IE912600A1 - Gamma-carboxylase and methods of use - Google Patents

Abstract

Protein compositions having gamma-carboxylase activity enriched at least 20,000-fold as compared to liver microsomes are provided. Also provided are DNA molecules encoding gamma-carboxylase. The protein compositions and DNA molecules are useful within methods of producing gamma-carboxylated vitamin K-dependent proteins.

Description

Technical Field The present invention relates generally to isolated proteins and methods of making proteins. More specifically, the invention relates to isolated gammacarboxylase, DNA sequences encoding gamma-carboxylase, and methods of using the DNA sequences in the production of gamma-carboxylated proteins.

Background of the Invention Gamma-carboxyglutamic acid (abbreviated gla) 15 is an amino, acid found in certain calcium-binding proteins. These proteins include factor VII, factor IX, factor X, prothrombin, protein C and protein S, plasma proteins that are components of the coagulation system; protein Z, also found in plasma; pulmonary surfactant20 associated proteins (Rannels et al., Proc. Natl. Acad. Sci. USA 84; 5952-5956, 1987) ; and the bone proteins osteocalcin (also known as bone gla-protein) and matrix gla-protein. Proteins containing this amino acid are variously referred to as vitamin K-dependent proteins, gla proteins, or gamma-carboxylated proteins. The plasma vitamin K-dependent proteins are dependent on glamediated binding to calcium and membrane phospholipids for their biological activity. Gla-containing proteins show a high degree of amino-terminal amino acid sequence homology.

Gamma-carboxyglutamic acid is formed by the post-translational modification of specific glutamic acid residues in a reaction requiring vitamin K hydroquinone (KH2) as a cofactor. The carboxylation reaction requires the oxidation of KH2 to vitamin K epoxide (KO) , however i carboxylation and KO formation are not strictly coupled. The reaction is believed to proceed by removal of a * hydrogen ion from the γ-carbon of glutamic acid, followed by the addition of CO2 · KH2 is regenerated by the action of reductases. Gamma-carboxylase is the subject of a recent review by Vermeer (Biochem. J. 266: 625-636, 1990).

In recent years a number of vitamin K-dependent proteins have been produced by genetic engineering. Production levels of' active protein are, however, limited by the ability of host cells to properly carboxylate glutamic acid residues in the nascent proteins. It has not heretofore been possible to produce biologically active vitamin K-dependent proteins in prokaryotic or lower eukaryotic (e.g. fungal) host cells, necessitating a reliance on higher eukaryotic (e.g. mammalian) cell lines, which are generally more difficult and expensive to culture. Even so, production levels of biologically active vitamin K-dependent proteins have been quite low when compared to levels of other proteins produced using genetically engineered mammalian cell lines. For example, Hagen et al. (U.S. Patent No. 4,784,950) reported the expression of recombinant human factor VII at up to 200 ng/ml in a baby hamster kidney cell line; Kaufmann et al. (J. Biol. Chem. 261: 9622-9628, 1986) produced 100 Mg/ml of factor IX using the CHO cell line, but only 1% of the material was biologically active; and Walls et al. (Gene 81: 139-149, 1989) produced protein C at up to 25 Mg/ml following amplification. In contrast, other proteins are typically produced in biologically active form at levels of 100-200 pq/ml. Further studies have shown that secretion of vitamin K-dependent proteins by cultured mammalian cells is reduced up to 20-fold if vitamin K is omitted from the culture media, and factor IX and protein C are poorly secreted if their propeptides are deleted. Together, these experimental results suggest that vitamin K-dependent carboxylation is required for efficient production of tliese proteins, and that the inability of cell lines to efficiently carboxylate may limit expression levels. These data further indicate that improperly carboxylase preparation. processed (i.e. uncarboxylated or undercarboxylated) vitamin K-dependent proteins are not efficiently secreted and may be unstable in the host cells.

Previous attempts to isolate gamma-carboxylase have not resulted in useful levels of purification. Canfield et al. (Arch. Biochem. Biophys. 202: 515-524, 1980) reported a · 100- to 150-fold enrichment of activity in a rat liver microsomal Girardot (J. Biol. Chem. 257: 15008-15011, 1982) reported a 400-fold purification of a carboxylasesubstrate complex from rat liver microsomes. Van Haarlem et al. (Biochem. J. 245: 251-255, 1987) prepared a crude microsomal fraction enriched for carboxlase activity from bovine aortae. This preparation also exhibited vitamin K reductase activity. Hubbard et al. (Proc. Natl. Acad. Sci. USA 86: 6893-6897, 1989) reported the purification of a 77 kDa carboxylase from bovine liver. This protein was later determined to be BiP (also known as GRP78) (Furie, 1990 Gordon Research Conference, Hemostasis and Thrombosis; Walsh, 1990 Gordon Research Conference, Hemostasis and Thrombosis).

There remains a need in the art for purified preparations of gamma-carboxylase to permit the characterization of this enzyme. The isolated enzyme may be used to direct the in vitro carboxylation of vitamin Independent proteins. Carboxylase-encoding DNA molecules may be used in producing biologically active vitamin independent proteins in lower eukaryotes and prokaryotes. In addition, DNA molecules encoding the enzyme may be used within genetic engineering processes to increase production of vitamin K-dependent proteins in cultured mammalian cell lines. The present invention provides such enzyme preparations and DNA molecules, as well as methods for producing vitamin K-dependent proteins that employ the enzyme preparations or DNA molecules. In addition, the present invention provides other, related advantages.

Summary of the Invention In one aspect, the present invention provides protein compositions having gamma-carboxylase activity enriched at least 20,000-fold as compared to liver microsomes. In one embodiment, the protein is selected from the group consisting of bovine, human and rat gammacarboxylases. In another embodiment, the protein is liver gamma-carboxylase. ' In yet another embodiment, the gammacarboxylase activity of the protein composition is enriched 100,000-fold as compared to liver microsomes.

In another aspect, protein compositions as described above are provided wherein the protein is affixed to a solid support. These compositions are useful, for example, in carboxylating vitamin K-dependent proteins in vitro.

In yet another aspect, the present invention provides DNA molecules encoding gamma-carboxylase, including bovine, human and rat gamma-carboxylases. In one embodiment, the gamma-carboxylase is liver gamma20 carboxylase. In another embodiment, the DNA molecule is a cDNA molecule.

A related aspect of the present invention provides a cultured cell transfected or transformed to express a DNA sequence encoding gamma-carboxylase. Within certain embodiments, the cultured cell is a eukaryotic cell, such as a yeast cell or a mammalian cell. In another embodiment, the cultured cell is further transfected or transformed to express a DNA sequence encoding a vitamin K-dependent protein, such as factor VII, factor IX, factor X, prothrombin, protein C, activated protein C, protein S, protein Z, osteocalcin, matrix gla-protein or pulmonary surfactant-associated proteins. Within another embodiment, the vitamin Kdependent protein is a human protein. The cultured cells may be used within methods of producing vitamin Kdependent proteins by culturing the cells and isolating the vitamin K-dependent protein. Within these methods, » the cells may be cultured in the presence of 0.1-10 Mg/ml vitamin K.

In another aspect, the present invention provides methods of producing a vitamin K-dependent protein comprising the steps of (a) introducing into the germline of a non-human animal a first DNA sequence encoding gamma-carboxylase and a second DNA sequence encoding a vitamin K-dependent protein, (b) growing the animal and (c) isolating from the animal the vitamin K10 dependent protein encoded by the second DNA sequence. In one embodiment, the vitamin K-dependent protein is selected from the group consisting of factor VII, factor IX, factor X, prothrombin, protein C, activated protein C, protein S, protein Z, osteocalcin, matrix gla-protein and pulmonary surfactant-associated proteins. In another embodiment, the vitamin K-dependent protein is a human protein.

A related aspect of the present invention provides non-human animals into which have been introduced a cloned DNA molecule encoding gamma-carboxylase, wherein the DNA molecule is expressed in the animals. The animals may further have introduced into them a second DNA molecule encoding a vitamin K-dependent protein, wherein the second DNA molecule is also expressed. In one embodiment, the vitamin K-dependent protein is selected from the group consisting of factor VII, factor IX, factor X, prothrombin, protein C, activated protein C, protein S, protein Z, osteocalcin, matrix gla-protein and pulmonary surfactant-associated proteins. In another embodiment, the vitamin K-dependent protein is a human protein.

These and other aspects of the invention will become evident upon reference to the accompanying detailed description and drawings. i Brief Description of the Drawings Figure 1 shows a polyacrylamide gel of fractions from a gamma-carboxylase purification procedure. Molecular weight markers (200, 93, 68 and 43 kD) are indicated.

Figure 2 illustrates the construction of leaderdeleted IX-pD5. Symbols used are 0-1, the adenovirus 5 ΟΙ map unit sequence; E, the SV4 0 enhancer; MLP, the adenovirus 2 major late promoter; Ll-3, the adenovirus 2 tripartite leader; 5', 5' splice site; 3', 3' splice site; FIX, factor IX cDNA; and pA, the SV40 early polyadenylation signal.

Figure 3 shows three samples of purified bovine gamma-carboxylase (lanes 1-3). Lane 4, molecular weight markers (205, Ί17, 106, 80 and 50 kd) ; lane 5, molecular weight markers as shown.

Figure 4 shows the results of a Western blot of purified bovine gamma-carboxylase probed with anitsera to bovine gamma-carboxylase. Positions of molecular weight markers are shown.

Figure 5 shows a polyacrylamide gel of proteins purified from G3 and D30 microsomes using antibody affinity chromatrography, cation exchange chromatography and lentil lectin chromatography. Lane 1, molecular weight markers (200, 93, 68, 43 and 26 kD.) ; lane 2, D30 microsomes; lane 3, G3 microsomes; lane 4, markers.

Detailed Description of the Invention Prior to setting forth the invention, it may be 30 useful to define certain terms to be used hereinafter.

Transfection or transformation: The process of stably and hereditably altering the genotype of a recipient cell or microorganism by the introduction of purified DNA. This is typically detected by a change in the phenotype gf the recipient organism. The term ’’transformation is generally applied to microorganisms, 4 while transfection” is used to describe this process in cells derived from multicellular organisms.

Cultured cell: A cell capable of being grown in liquid or solid media over a number of generations. In the case of cells derived from multicellular organisms, a cultured cell is a cell isolated from the organism as a single cell, a tissue, or a portion of a tissue.

As noted above, carboxylation of specific 10 glutamic acid residues is necessary for the biological activity of certain proteins. The present invention provides protein compositions that are highly enriched for gamma-carboxylase activity, as well as DNA molecules encoding gamma carboxylase. These compositions and molecules are useful within methods of making vitamin Independent proteins.

Within the present invention, gamma-carboxylase was isolated from microsomes prepared from rat and bovine liver and from a cultured human cell line producing recombinant factor IX. . Gamma-carboxylase may also be isolated from other tissues or cells known to produce the protein. The isolated carboxylase may be used for in vitro carboxylation of vitamin K-dependent proteins. The carboxylase also provides a useful tool for the isolation of genomic and cDNA clones encoding it. These clones are used to transfect or transform cultured eukaryotic or prokaryotic cells to produce host cells suitable for the high level production of vitamin K-dependent proteins, including vitamin K-dependent blood coagulation proteins and vitamin K-dependent bone proteins.

A preferred source of gamma-carboxylase is liver microsomes, such as microsomes prepared from rodent (e.g. rat or mouse), bovine or human liver, although liver microsomes from other species may also be used. The production of (biologically active, recombinant human vitamin K-dependent proteins in host cells derived from a variety of mammalian species (Hagen et al., U.S. Patent *E 912600 . No. 4,784,950; Foster et al., U.S. Patent No. 4,959,318) demonstrates that gamma-carboxylase is active across species lines, thus a gamma-carboxylase derived from one mammalian species will be useful in carboxylating proteins from a second mammalian species, although the syngeneic carboxylase may be preferred. Other suitable sources include microsomes from cell lines known to be capable of producing active vitamin K-dependent proteins, such as the transformed human kidney 293 cell line (available from American Type Culture Collection under accession number CRL 1573) . It will be understood by those skilled in the art that, in principle, nearly any tissue or cell type is a candidate source of gamma-carboxylase. Gammacarboxylase . activity has been detected in most tissues examined, and carboxylase activity may be detected in a candidate tissue or cell line by assaying the ability of that tissue or cell line to incorporate radiolabeled CO2 into protein in a carboxylation reaction as described herein. Typically, microsomes are prepared from 1-2 g of tissue and incubated in a carboxylase reaction with NaH14CC>3. The reaction is run in the presence and absence of vitamin K hydroquinone, and the reaction products are analyzed by gel electrophoresis and autoradiography. The presence of one or more labeled protein bands specific to the vitamin K-containing reaction is indicative of gammacarboxylase in the tested tissue.

Microsomes are prepared from cells or tissues according to procedures known in the art. One suitable method is that described by Swanson and Suttie (Biochemistry 21: 6011-6018, 1982), which is incorporated herein by reference. Briefly, the cells or tissue are homogenized in a slightly acidic to slightly basic buffer containing sucrose and a protease inhibitor. Microsomes are harvested from this preparation by ultracentrifugation. Preparation of microsomes is also disclosed by Rannels et al. (ibid) and Soute et al. (Thromb. Haemostasis 57: 77-81, 1987) .

K A protein extract is then prepared from the microsomes. A suitable extract may be prepared by disrupting the membranes in a tissue homogenizer, solubilizing in detergent, removing insoluble membrane material, and precipitating detergent-solubilized proteins by adjusting the (RR^)2SO4 concentration of the solution to about 60% of saturation. The precipitated proteins are recovered by centrifugation.

Within the present invention, several different 10 methods were used to fractionate the microsome extracts and purify gamma-carboxylase. Certain of these methods take advantage of the ability of gamma-carboxylase to bind to a vitamin K-dependent protein such as factor IX, prothrombin,, or protein C.

In a preferred embodiment, gamma-carboxylase is isolated from extracts of microsomes from vitamin Ktreated animals by affinity chromatography on a propeptide of a vitamin K-dependent protein. To maximize recovery, it is preferred to incubate the extract and propeptide for at least 16 hours. Bound proteins are eluted from the propeptide and electrophoresed. Comparison to the pattern of proteins bound to a non-specific peptide is used to identify gamma-carboxylase. Propeptides of vitamin Kdependent proteins may be synthesized on the basis of known amino acid sequence data. See, for example, the disclosures of Kurachi et al. (Proc. Natl. Acad. Sci. USA 79: 6461-6464, 1982), Foster et al. (Proc. Natl. Acad. Sci. USA 82: 4673-4677, 1985), Hagen et al. (Proc. Natl. Acad. Sci. USA 83: 2412-2416, 1986) and Leytus et al.

(Biochemistry 25: 5098-5102, 1986), which are incorporated herein by reference. Propeptides may be synthesized by the solid phase method of Merrifield (Fed. Proc. 21: 412, 1962; J. Amer. Chem. Soc. 85: 2149, 1963) or by use of an automated peptide synthesizer. To isolate gamma35 carboxylase, a < propeptide is covalently bound to a suitable solid support, such as a derivatized agarose gel (e.g. cyanogen bromide-activated agarose gels or activated thiol agarose gels). A preferred such support is activated thiol Sepharose 4B™, available from Pharmacia, Piscataway, NJ. Methods for coupling peptides to solid supports are known in the art, and specific instructions are generally provided by the supplier. When using an activated thiol support, the coupled propeptide and gammacarboxylase are eluted with a reducing agent such as dithiothreitol. In the alternative, a complete proprotein (e.g. pro-factor IX or pro-prothrombin) is attached to the support and bound ganuna-carboxylase is eluted using an isolated propeptide of a vitamin K-dependent protein. Chromatography can be carried out in either batch processing or on a column. This step typically provides about a 500-fold enrichment of gamma-carboxylase activity.

An alternative embodiment utilizes antibody affinity chromatography. An antibody, preferably a monoclonal antibody, to a vitamin K-dependent protein is coupled to a solid support (e.g. activated thiol Sepharose or CNBr-activated Sepharose). Antibodies to vitamin K20 dependent proteins are disclosed by, for example, Wakabayashi et al. (J. Biol. Chem. 261: 11097-11105, 1986), Kaufman et al. (ibid.) and Busby et al. (Nature 316: 271-273, 1985). In one embodiment, a microsome extract, prepared from warfarin-treated animals in a manner similar to that described above, is applied to the immobilized antibody. Gamma-carboxylase is eluted from the antibody using a propeptide of a vitamin K-dependent protein or by incubating the antibody-enzyme complex in a carboxylation reaction mix. In the alternative, calcium30 dependent antibodies are used, and the carboxylase is eluted by altering the calcium concentration. The eluant, containing gamma-carboxylase, is recovered.

Antibody affinity chromatography is also used to identify gamma-carboxylase. Cultured cells are transfected to (express a vitamin K-dependent protein lacking its propeptide. An extract of these cells is exposed to the immobilized antibody. Bound protein is * eluted and electrophoresed. The protein pattern is compared to a parallel preparation made from cells transfected to express the corresponding native vitamin independent protein (i.e. including the propeptide). Gamma5 carboxylase is found specifically in the preparation from cells expressing the native protein.

Further purification is achieved using ion exchange chromatography and/or lectin affinity chromatography. Ion exchange chromatography may be carried out using either anion exchange or cation exchange media. Bovine gamma-carboxylase has been found to bind to cation exchange media, while human and rat gammacarboxylases have been found to bind to anion exchange media. Suitable anion exchange media include derivatized, cross-linked agaroses and dextrans, including DEAE and QAE derivatives. Particularly preferred anion exchange media include DEAE Sepharose CL-6B™ and Q-Sepharose™, available from Pharmacia. Suitable cation exchange media include carboxymethyl and sulfopropyl derivatized resins.

Suitable lectins include mannose-specific lectins, with lentil lectin particularly preferred. Additional purification may be obtained by using lectins with other specificities to remove contaminating glycoproteins. Lectins are coupled to support matrices according to conventional methods (e.g. CNBr activation), or lectin matrices may be obtained from commercial suppliers.

In a typical anion exchange chromatography step, a partially purified carboxylase preparation is dialyzed in a neutral to slightly basic buffer containing phospholipids and a non-ionic detergent. Preferred buffers include 20 mM Tricine pH 8.5 or 20 mM Bis-Tris pH 6.8 containing 1 mg/ml phosphatidyl choline type VE, 50 Mg/ml phosphatidyl serine, 50 M NaCl), preferably about 0.5 M NaCl. The eluant, which contains the carboxylase, is recovered.

It has been found by the inventor that gamma5 carboxylase contains core glycosylation, allowing it to bind to mannose-specific lectins. However, gammacarboxylase lacks more complex carbohydrate additions. Based on these observations, lectins of different specificities may be used in purifying gamma-carboxylase.

Lectin affinity chromatography is typically carried out using lentil lectin bound to a derivatized agarose support that has been washed in phosphate buffered saline. A partially purified carboxylase preparation (e.g. purified by propeptide affinity chromatography or propeptide affinity chromatography followed by anion exchange chromatography) is combined with the washed lectin resin. It is preferred to allow the mixture to incubate for at least about 16 hours. The resin is then washed to remove unbound material, and the bound carboxylase is eluted with a suitable sugar, such as mannose or α-methyl mannoside. The typical level of enrichment at this point is about 20,000 x over the initial microsome preparation. The level of enrichment can be increased to about 100,000 x by using a second lectin to remove contaminating glycoproteins. Preferred second lectins include Ervthrina cristagalli lectin, Tetraqonolobus purpureas lectin, Pseudomonas aeruginosa PA-I lectin and Limulus polvphemus lectin, which do not bind gamma-carboxylase. These and other lectins are available from commerical suppliers (e.g. Sigma Chemical Co., St. Louis, MO). Lectins are bound to solid supports using conventional chemical methods as generally described above.

Cation exchange chromatography of a partially purified gamma-carboxylase preparation is carried out using a neutral to slightly acidic buffer containing phospholipids and a nonionic detergent.

» An alternative purification protocol employs, in order, propeptide or antibody affinity chromatography, lectin affinity chromatography and ion exchange chromatography.

Throughout purification, it is preferred to perform all manipulations involving gamma-carboxylase in the cold (e.g. 4’C). and in the presence of a 1:1 ratio of detergent to phospholipid.

Purification of gamma-carboxylase is monitored 10 by assaying the biological activity of samples using the carboxylation assay described in more detail below. Protein samples may be electrophoresed on non-denaturing polyacrylamide gels in the presence of phosphatidyl choline and CHAPS, and protein bands cut from the gels and tested for activity in the carboxylation assay. Total protein concentration of samples is monitored according to routine procedures, such as methods utilizing bicinchoninic acid (e.g. BCA™ protein assay reagent, available from Pierce Chemical Co., Rockford, IL) or by gel scanning.

The above-described methods were used to identify proteins that specifically bound to a prothrombin propeptide. Within one aspect of the invention, further purification of rat gamma-carboxylase by lentil lectin affinity chromatography was used to identify proteins having molecular weights of approximately 90 and 150 kDa on denaturing gels. Gamma-carboxylase is identified from these candidate proteins, for example by amino acid sequence analysis. Picomole quantities of protein are sequenced essentially as described by Aebersold et al.

(Proc. Natl. Acad. Sci. USA 84: 6970-6974, 1987), incorporated herein by reference. Peptide fragments are generated by gel electrophoresis of the protein and transfer to nitrocellulose, followed by proteolytic digestion. The Resulting enzymatic cleavage fragments are separated by HPLC and sequenced on an automated gas-phase sequenator. The amino acid sequences are compared to carboxylase. Although most . known sequences, such as the sequences contained in the GenBank™ database of the U.S. Department of Health and Human Services or the Protein Sequence Database of the National Biomedical Research Foundation. In the alternative, gamma-carboxylase is identified by preparing peptides corresponding to the amino acid sequences of the candidate proteins, preparing antibodies to the peptides, and using the antibodies to identify the gamma-carboxylase by immunodepletion of biological activity.

Gamma-carboxylase purified as described above may be used to generate polyclonal or monoclonal antibodies according to conventional procedures. These antibodies are useful for large-scale affinity purification of the protein.

Information obtained from the isolated protein is then used to clone DNA molecules encoding gammaStandard cloning methods are employed, tissues are believed to produce gammacarboxylase, for cloning it is preferred to use a library derived from a tissue known to produce substantial amounts of vitamin K-dependent proteins or known to contain gammacarboxylase activity, such as liver, kideny, testis, heart, bone or lung. Liver tissue is a particularly preferred source.

For example, a DNA molecule encoding gammacarboxylase can be isolated using oligonucleotide probes designed from the amino acid sequence of the protein. It is preferred to use an amino acid sequence having a low level of degeneracy. These probes are used to screen genomic or cDNA libraries according to known procedures. In general, it is preferred to use a cDNA library.

In the alternative, oligonucleotide primers designed on the basis of the amino acid sequence are used within the polymerase chain reaction cloning method (disclosed by Mullis et al., U.S. Patent No. 4,683,195 and Mullis, U.S. Patent No. 4,683,202, which are incorporated herein by reference).

. The primers will generally comprise a restriction site at the 5' end to facilitate subsequent cloning of the fragment and about 14 to 30, preferably about 18-20, nucleotides corresponding to the predicted DNA sequence. For highly degenerate sequences, shorter primers (e.g. those comprising about 14 nucleotides corresponding to the- DNA sequence) are preferred. Primer synthesis is simplified by substituting inosine at positions of four-fold degeneracy. When working with primers of high degeneracy, the initial PCR reactions will generally be run at a low stringency (low annealing temperature). In this event, it is advantageous to repeat the PCR procedure at slightly higher stringencies to reduce the background (non-specific sequences). The PCR reaction products are separated by gel electrophoresis, and bands of the predicted size for the fragment of interest are sequenced to confirm their identity. PCRgenerated clones are used as probes to screen cDNA or genomic libraries or to generate longer clones by the RACE (rapid amplification of cDNA ends) procedure of Frohman et al. fProc. Natl. Acad. Sci. USA 85: 8998-9002, 1988). The RACE procedure can be run using combinations of PCRgenerated primers, PCR-generated primers with mixed oligonucleotide primers or oligo(dT) or mixed with oligonucleotide primers designed from amino acid sequence having low degeneracy at the nucleotide level. PCR methods are disclosed by Innis et al., eds. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc., San Diego, 1990 and by Erlich., ed. PCR Technology: Principles and Applications for DNA Amplification. Stockton Press, NY, 1989.

In a third alternative, antibodies raised against purified gamma-carboxylase or fragments of gammacarboxylase are used to screen expression libraries, such as cDNA librari’es prepared in the Agtll vector (ATCC 37194).

. The identity of cloned sequences may be confirmed in an expression cloning system, such as the Agtll system of Young and Davis (Proc. Natl. Acad. Sci. USA 80: 1194-1198, 1983). Using standard immunization protocols, antibodies (preferably monoclonal antibodies) are prepared against peptides synthesized from the gammacarboxylase sequence; These antibodies are used to screen the expression library. Methods for fusing lymphocytes and immortalized cells and generating monoclonal antibodies from the resultant hybridomas are disclosed by Kohler and Milstein (Nature 256: 495-497, 1975; Eur. J.

Immunol. 6: 511-519, 1976). Alternatively, the antibodies are used to purify larger quantities of carboxylase, and the purified protein is used to generate polyclonal antisera, which are in turn used to screen the library.

The present invention enables the production of vitamin K-dependent proteins in a wide variety of cultured host cells, including prokaryotic (e.g. bacterial), lower eukaryotic (e.g. fungal) and higher eukaryotic (e.g. mammalian) cells. -Methods for transforming or transfecting these cell types to express exogenous DNA sequences are known in the art and are disclosed, for example, by Itakura et al. (U.S. Patent No. 4,704,362), Goeddel et al. (U.S. Patent No. 4,766,075), Hinnen et al.

(Proc. Natl. Acad. Sci. USA 75: 1929-1933, 1978), Murray et al. (U.S. Patent No. 4,801,542), Upshall et al., (U.S. Patent No. 4,935,349), Hagen et al. (U.S. Patent No. 4,784,950) and Axel et al. (U.S. Patent No. 4,399,216), which are incorporated herein by reference.

A DNA molecule encoding gamma-carboxylase is expressed in a cultured cell to provide or enhance the ability to gamma-carboxylate proteins. Cells expressing cloned gamma-carboxylase sequences are thus particularly useful in making vitamin K-dependent proteins, including factor VII, factor IX, factor X, prothrombin, protein C, activated protein C, protein S, protein Z, osteocalcin, . matrix gla-protein and lung surfactant-associated proteins.

DNA sequences encoding plasma vitamin Independent proteins and the expression of these DNA sequences are known in the art. See, for example, the disclosures of Kurachi et al. (Proc. Natl. Acad. Sci. USA 79: 6461-6464, 1982), Foster et al. (European Patent Office Publication EP 266,190; U.S. Patent No. 4,959,318), Bang et al. (U.S. Patent No. 4,775,624), Hagen et al. (ibid), Degan et al. (Biochemistry 22: 2087-2097, 1983), Leytus et al. (Biochemistry 25: 5098-5102, 1986) and Wydro et al. (European Patent Office Publication EP 255,771), which are incorporated herein by reference. A cDNA encoding rat matrix gla-protein is disclosed by Price et al. (Proc. Natl. Acad. Sci. USA 84_: 8335-8339, 1987), incorporated herein by reference. A cDNA encoding rat bone gla-protein is disclosed by Pan and Price (Proc. Natl. Acad. Sci. USA 82: 6109-6113, 1985), incorporated herein by reference.

Within the present invention, cells are transfected or transformed to express a DNA sequence encoding gamma-carboxylase and a DNA sequence encoding a vitamin K-dependent protein. The DNA sequences are introduced into the host cells on separate expression vectors or on the same expression vector. Cells expressing the introduced sequences are then selected. To facilitate selection, one or more selectable markers are provided on the expression vector(s). It is preferred to first obtain cells expressing the cloned gamma-carboxylase sequence, then transfect or transform these cells to express a vitamin K-dependent protein.

Expression vectors for use in transfection and transformation include an expression unit wherein the DNA sequence to be expressed is operably linked to transcriptional ! promoter and terminator sequences. Expression vectors may further include such elements as enhancers, polyadenylation signals, splice sites, selectable markers, and one or more origins of replication. Selection of these elements and construction of expression vectors skill in the art.

A preferred mammalian cells, such is within the level of ordinary group of host cells is cultured as the 293 (ATCC CRL 1573), BHK (ATCC CRL 1632 and .CRL 10314) and CHO (ATCC CCL 61) cell lines. Within a preferred embodiment, a cultured mammalian cell line is transfected with an expression 10 vector containing an expression unit for gamma-carboxylase and an expression unit for a selectable marker such as neomycin or hygromycin resistance. The selectable marker may also be provided on a separate vector. Transfectants are selected on the basis of antibiotic resistance. 15 Resistant cells are then screened for increased production of gamma-carboxylase, for example by assaying extracts of transfected and untransfected cells. Cells producing gamma-carboxylase are transfected with a second expression vector containing expression units for a vitamin K20 dependent protein and a ‘second selectable marker. It is preferred that the second selectable marker be an amplifiable selectable marker, such as a gene encoding dihydrofolate reductase. Use of an amplifiable selectable marker allows for amplification of the exogenous DNA, 25 leading to enhanced expression of the vitamin K-dependent protein. DNA sequences encoding secreted vitamin Kdependent proteins, such as coagulation proteins, will generally include a secretory signal sequence. For intracellular proteins, such as bone and matrix gla30 proteins, a secretory signal sequence may be omitted. In an alternative embodiment, the DNA sequence encoding gamma-carboxylase and a vitamin K-dependent protein are placed on the same expression vector under the control of a single promoter to produce a polycistronic message. 35 Transfected cells are selected on the basis of a selectable marker and are screened for gamma-carboxylase activity as described above or by measuring the amount of ,E 912600 gamma-carboxylase protein using a gla-dependent antibody (Wakabayashi et al., ibid). Preferred promoters for use in cultured mammalian cells include the SV40 promoter (Subramani et al., Mol. Cell. Biol. 1: 854-864, 1981), the adenovirus 2 major late promoter (Kaufman and Sharp, Mol. Cell. Biol. 2’ 1304-1319, 1982) and the mouse metallothionein-I gene promoter (Palmiter et al., U.S. Patent No. 4,579,821).

Media for culturing a variety of prokaryotic and 10 eukaryotic host cells are known in the art. These media will generally contain a carbon source, a nitrogen source, vitamins (including vitamin K) and minerals, as well as growth factors and other nutrients required by the particular host cell. In many instances, the media will further contain a selective agent to insure stability of the exogenous DNA sequence(s) . In the case of cultured mammalian cells, many of the required growth factors and other nutrients may be provided as serum, such as fetal bovine serum, although a variety of serum-free media formulations are known in the art. Within a preferred embodiment, cultured mammalian cells transfected to express a DNA sequence encoding gamma-carboxylase are cultured in a medium containing 0.1-10 Mg/ml vitamin K, preferably about 5 pg/ml vitamin K.

Gamma-carboxylase may also be expressed in nonhuman transgenic animals, particularly transgenic warmblooded animals. Methods for producing transgenic animals, including mice, rats, rabbits, sheep and pigs, are known in the art and are disclosed, for example, by Hammer et al. (Nature 315: 680-683, 1985), Palmiter et al.

(Science 222: 809-814, 1983), Brinster et al. (Proc. Natl. Acad. Sci. USA 82: 4438-4442, 1985), Palmiter and Brinster (Cell 41: 343-345, 1985) and U.S. Patent No. 4,736,866, which are incorporated herein by reference. Briefly, an expression unit,ι including a DNA sequence to be expressed together with appropriately positioned expression control sequences, is introduced into pronuclei of fertilized . eggs. Introduction of DNA is commonly done by microinjection. Integration of the injected DNA is detected by blot analysis of DNA from tissue samples, typically samples of tail tissue. It is preferred that the introduced DNA be incorporated into the germ line of the animal so that it is passed on to the animal's progeny. Although systemic expression of gammacarboxylase and vitamin K-dependent proteins is generally preferred, proteins showing toxicity to the host animal are best expressed in a regulated and/or tissue-specific manner. Tissue-specific expression is achieved through the use of a tissue-specific promoter, such as an insulin gene promoter. Use of an inducible promoter, such as a metallothionein gene promoter (Palmiter et al., 1983, ibid), allows regulated expression of the transgene. In this way, animals are made transgenic for gammacarboxylase and a vitamin K-dependent protein. The animals are grown and the vitamin K-dependent protein is isolated from cells or body fluids (e.g. milk, blood or urine) of the animal.

Gamma-carboxylase purified as disclosed above or prepared by genetic engineering methods (e.g. cell culture or transgenic animals) is also useful within in vitro carboxylation systems. One such system is disclosed by Vermeer et al. (International Patent Application Publication WO 87/04719). Briefly, the purified enzyme is affixed to a solid support (e.g. derivatized dextran or agarose beads), and an un- or undercarboxylated vitamin Kdependent protein is passed over the support. The reaction is carried out in the presence of vitamin Khydroquinone. Such a system is particularly useful for activating vitamin K-dependent proteins produced in prokaryotic or lower eukaryotic host cells. If the protein lacks a propeptide, exogenous propeptide is added to the reaction (mixture to drive carboxylation (Knobloch and Suttie, J. Biol. Chem. 262: 15334-15337, 1987).

. Proteins containing propeptides are proteolytically processed after carboxylation to remove the propeptide.

The following examples are offered by way of illustration, not by way of limitation.

Examples Vitamin K was obtained from Sigma Chemical Co., St. Louis, MO. and was reduced by the method of Sadowski et al. (J. Biol. Chem. 251: 2770-2776, 1976). PSN antibiotic mix was obtained from GIBCO BRL Life Technologies, Inc. (Gaithersburg, MD) . Phospholipids and lectins were obtained from Sigma.

Gamma carboxylation was assayed by measuring the incorporation of radiolabeled CO2 into the synthetic peptide substrate NBoc-L-glu-L-glu-L-leu-methylester (EEL) (obtained from Bachem Bioscience, Philadelphia, PA) using essentially the method of Vermeer (ibid.). The reaction mixture contained 950 μΐ of 3.8 M (NH4)2SO4, 150 μΐ 1% CHAPS, 150 μΐ 10 mg/ml phosphatidyl choline type III, 1% Na cholate, 75 μΐ 0.2 M dithiothreitol, 750 μΐ 1 mM EEL and 150 μΐ NaH14CO3 (50 mCi/mmol). One hundred twenty-eight microliters of this mixture was combined with a 100 μΐ test sample plus 5 μg/ml vitamin K hydroquinone. Control reactions lacked vitamin K. The reaction was allowed to proceed for one hour in the dark at room temperature. The reaction was stopped by the addition of 1 ml of io% trichloroacetic acid, placed on ice for 30 minutes, and centrifuged 15 minutes at 4’C. One milliliter of the resulting supernatant was boiled for five minutes and combined with 10 ml of scintillant (BioSafe II, obtained from Research Products International Corp., Prospect, IL) and counted. Incorporation of label was compared to a control reaction lacking vitamin K. 293 cells were metabolically labeled as follows.

The 293 cells (were seeded into 125 mm plates to a confluency of 10% and were grown in Dulbecco's Modified Eagle medium (Hazelton Biologies, Lexena, KA) supplemented IE 91260° ‘ with 1% G418 + 1% PSN (GIBCO BRL, Gaithersburg, MD) + 10% dialyzed, fetal bovine serum (HyClone Laboratories, Inc., Logan, UT) for two days to 60-70% confluency. The media was removed from one set of 293 cell cultures by aspiration and the cells were washed with 10 ml of PBS (Sigma, St. Louis, MO) that had been prewarmed to 37*C. The PBS was removed’ and 25 ml of prewarmed pulse media (100 ml Dulbeccozs Modified Eagle medium -cys/-met [Hazelton Biologies], 1 ml 100 mM sodium pyruvate [Irvine Scientific, Santa Ana, CA], 1 ml of 200 mM L-glutamine [Hazelton Biologies], 1 ml of lOOx PSN, 10 ml of dialyzed, fetal bovine serum) containing 20 pci/ml of Neg-072 Expre35S35S [35S] Protein Labeling Mix (NEN Research Products, Willmington, DE) was added to each plate. The cells were incubated at 37*C, and the media was changed every 10-12 hours for two days. One day after the first set of cultures was processed, the remaining duplicate cultures were labeled as described above and grown at 37*C with media changes every 10-12 hours for one day. After labeling, the media from each culture was removed and stored on ice. The plates were washed with 2.5 ml lx Versene (GIBCO BRL) and the Versene wash was combined with the spent media. After the Versene wash, an additional 2.5 ml of lx Versene was added to each plate, and the plates were incubated at room temperature for 10-15 minutes. After incubation, the cells were harvested and added to the spent media.

EXAMPLE 1 Peptides corresponding to the propeptide of human prothrombin and a portion of the human GM-CSF receptor (Table 1) were synthesized using an Applied Biosystems Model 431A peptide synthesizer (Applied Biosystems, Inc., Foster City, CA).

/ TABLE 1 pro-prothrombin (Sequence ID Number 1) CGGHVFLAPQQARSLLQRVRR CSF (Sequence ID Number 2) CGGKDKLNDNHEVEDEY Thirty-eight mg of pro-prothrombin (pro-PT) peptide (Sequence ID Number 1) or 44 mg of CSF peptide (Sequence ID Number 2) were dissolved in 7 ml or 6 ml, respectively, of 0.1 M ammonium acetate pH 4.5. The solutions were each combined with 10 ml (-2.5 g) of activated thiol Sepharose 4B™ (Pharmacia, Piscataway, NJ) that had been sequentially washed in phosphate buffered saline (PBS) and 0.1 M ammonium acetate pH 4.5. The mixtures were placed on a rocker overnight at room temperature. The mixtures were then filtered on a scintered glass funnel, and the eluants were saved for determination of the coupling efficiency (as described by Ware et al., J. Biol. Chem. 264: 11401-11406, 1989). The resins were each rinsed with 250 ml 0.1 M ammonium acetate pH 4.5. The rinsed resins were each combined with 10 ml of 0.1 M ammonium acetate pH 4.5 containing 4 mM β- mercaptoethanol, rocked for 20 minutes at room temperature, and filtered again. The resins were then rinsed with 250 ml 0.1 M ammonium acetate pH 4.5, then 30 with 200 ml PBS, and were stored at 4*C until ready for use. Just before use, the resins were recovered by filtration.

Rat liver microsomes, prepared essentially as described by Swanson and Suttie (ibid.) (approximately 800 mg), were suspended in 40 ml of 0.025 M imidazole pH 7.2 containing 0.25 M sucrose, 0.5 M KCl, 0.2% Triton X-100. Membranes were disrupted with 8 strokes of a tissue . homogenizer (Dura Grind stainless steel Dounce Tissue Grinder, Wheaton Scientific, Millville, NJ), and the solution was placed on ice for 30 minutes. The chilled solution was centrifuged at 150,000 x g for 1 hour at 4’C, and the supernatant was recovered. To the supernatant (40 ml volume) was added 60 ml saturated (NH4)2SO4 over approximately 30 minutes, and the mixture was stirred for an additional 20 minutes at 4’C. The solution was centrifuged at 12,000 x g for 20 minutes at 4’C.

The (NH4)2SO4 pellet was resuspended in 4 ml buffer A (20 mM sodium phosphate pH 7.4 containing 1 mg/ml phosphatidyl choline type VE, 50 Mg/ml phosphatidyl serine, 50 Mg/ml phosphatidyl ethanolamine, 0.1% 3-[(3cholamidopropyl)-dimethylammonio]-l-propanesulfonate [CHAPS], 0.15 M NaCl, 15% glycerol) to a final volume of approximately 15 ml. The resulting solution was dialyzed against 0.5 1 of buffer A at 4’C for 5 hours, then the buffer was replaced with 0.5 1 of fresh buffer A and dialysis was continued at 4’C overnight. The final volume of the dialysate was approximately 20 ml.

The dialysed microsome extract was combined with ml of prewashed activated thiol Sepharose 4B, and the mixture was rocked for 3.5 hours at 4’C. The mixture was then poured into a column, and the eluant was recovered.

The column was rinsed with additional buffer A until most of the protein was eluted as determined by a reduction in eluant viscosity. The eluant fractions were combined (total volume = 30 ml), and two-thirds of this preparation was combined with 10 ml of the pro-PT Sepharose. The remaining one-third of the microsome preparation was combined with 5 ml of the CSF Sepharose, and the mixtures were rocked overnight at 4’C, then centrifuged at 2,000 x g for 5 minutes at 4’C. The resins were recovered^ rinsed with 5 ml of buffer A adjusted to 0.5 M NaCl, and centrifuged. The rinse procedure was repeated 4 times, and the resin was resuspended in 5 ml (CSF) or 10 ml (proPT) buffer A adjusted to 50 mM dithiothreitol and placed . on a rocker for 24 hours at 4’C. The preparations were again centrifuged at 2,000 x g for 5 minutes at 4’C, and the supernatants were recovered and assayed for gammacarboxylase activity and protein content (by BCA).

Results, shown in Table 2, indicated that 10 times as much activity bound to the pro-PT resin as bound to the CSF resin.

TABLE 2 Protein Total Fraction Volume Concentration CPM Microsome suspension 40 ml 20 mg/ml 4 X 10 Microsome supernatant 40 ml 15 mg/ml 3 X 10 Dialyzed microsome extract 20 ml 19 mg/ml 2 X 10 Pre-resin eluant 20 ml 11 mg/ml 2 X 10 CSF resin 5 ml N.D. 1 X 10 Pro-PT resin 10 ml N.D. 2 X 10 CSF eluant 5 ml N.D. 3 X 10 Pro-PT eluant N.D. = not determined 10 ml N.D. 6 X 10 The eluants from the pro-PT and CSF resins were electrophoresed on an 8% SDS-polyacrylamide gel. Figure 1 is a photograph of the gel showing equal volumes of the pro-PT eluant (lane 1) and the CSF eluant (lane 2).

EXAMPLE 2 Approximately 800 mg of rat liver microsomes (prepared as described in Example 1) were suspended in 40 ml of 0.025 M imidazole pH 7.2 containing 0.25 M sucrose, 0.5 M KC1, 0.2% Triton X-100. Membranes were disrupted with 10 strokes of a tissue homogenizer, and the solution was placed on ice for 30 minutes. The chilled solution was centrifuged at 150,000 x g for 1 hour at 4’C, and the supernatant was /recovered. To the supernatant was added 65 ml saturated (NH4)2SO4 over approximately 30 minutes, and the mixture was stirred 30 minutes at 4'C. The » solution was centrifuged at 12,000 x g for 20 minutes at 4*C.

The (NH4)2SO4 pellet was resuspended in buffer A to a final volume of approximately 20 ml. The resulting solution was dialyzed against 0.5 liters of buffer A at 4‘C for 5 hours, then the buffer was replaced with 0.5 liters of fresh buffer A, and dialysis was continued at 4’C overnight.

Ten milliliters of activated thiol Sepharose 4B 10 was washed in an equal volume of buffer A, and the mixture was rocked at 4‘C for 3 hours. The mixture was then centrifuged at 2,000 x g for 5 minutes at 4’C, the resin was recovered, and the wash procedure was repeated four times. The washed resin was combined with the dialyzed microsomal preparation and placed on a rocker for 3 hours at 4’C. The preparation was then centrifuged at 2,000 x g for 5 minutes at 4’C, and the supernatant was recovered.

The supernatant fraction was then combined with activated thiol Sepharose that had been coupled to the synthetic prothrombin propeptide as described above. The mixture was placed on a rocker at 4’C overnight, then centrifuged at 2,000 x g for 5 minutes at 4’C. The resin was recovered, rinsed with 5 ml of buffer A adjusted to 0.5 M NaCl, and centrifuged. The rinse procedure was repeated 3 more times, and the washed resin was mixed with 5 ml of buffer A adjusted to 50 mM dithiothreitol and placed on a rocker for 1 hour at 4’C. The preparation was again centrifuged at 2,000 x g for 5 minutes at 4 °C, and the supernatant, containing gamma-carboxylase purified about 200-fold over the initial microsome preparation, was recovered.

Purification was monitored by assaying for gamma-carboxylase activity. Of 4 x 108 cpm of activity in the initial microsome suspension, 2 x 106 cpm was recovered in the/pro-PT eluant.

The partially purified carboxylase was further purified by lentil lectin chromatography. Three ‘ milliliters of lentil lectin Sepharose 4B (Pharmacia) was washed in 4 changes (10 ml each) of PBS. The washed resin was combined with 5 ml of the pro-PT eluant and placed on a rocker at 4’C overnight. The resin was washed 6 times (10 ml each) in buffer A adjusted to 0.5 M NaCl as described above, and the supernatant was recovered. The resin was combined with 5 ml of buffer A adjusted to 0.5 M NaCl, 0.5% CHAPS and 5 mg/ml phosphatidyl choline, and containing 0.5 M mannose, and the mixture was placed on a rocker overnight at 4 °C. The resin was pelleted by centrifugation as described above, and the supernatant was recovered. Fractions were assayed for gamma-carboxylase activity (Table 3) . The two supernatant fractions were electrophoresed on a polyacrylamide gel. Figure 1 shows the lentil lectin flow-through (lane 3) and the lentil lectin eluant (lane 4) . Equal amounts of carboxylase activity were loaded in lanes 1 (pro-PT eluant) and 4. Comparison of the lanes revealed the presence of bands of approximately 90 kD and 150 kD that were specific to the pro-PT and lentil lectin preparations (lane 4) . At this point, gamma-carboxylase activity was enriched 10,000- to 20,000-fold over the initial microsome preparation. fraction pro-PT eluant Lentil lectin resin Lentil lectin flow-through Mannose elution Volume Total com TABLE 3 ml 2.5 X 10 ml 1.2 x 10 ml 3 X 104 ml 3 X 105 The partially purified material from the proprothrombin purification could also be further purified by anion exchange chromatography. In a test of purification conditions, 5 Al of the propeptide resin eluant was dialyzed against two changes of 20 mM Tricine (Sigma Chemical Co., St. Louis, MO) pH 8.5 containing 0.1% CHAPS •Ε 912600 (Sigma) , 1 mg/ml phosphatidyl choline type VE, 50 /ig/ml phosphatidyl ethanolamine and 50 Mg/ml phosphatidyl serine. One milliliter of the dialyzed solution was then combined with 0.5 ml of DEAE Sepharose (Pharmacia) that had been equilibrated in the same Tricine buffer or Tricine buffer containing 0.2M, 0.5 M or 0.8 M NaCl. The mixtures were placed on a rocker overnight at 4*C, then centrifuged at 2,000 x g for 5 minutes. The supernatants were recovered and assayed for carboxylase activity. The resins were then rinsed 4 times in 1 ml of Tricine buffer or Tricine buffer containing the appropriate concentration of NaCl with centrifugation at 2,000 x g for 5 minutes. Each resin sample was recovered and assayed for gammacarboxylase activity. Results are shown in Table 4.

TABLE 4 Resin Sample 0 M NaCl 0.2 M NaCl 0.5 M NaCl 0.8 M NaCl Total com 2 χ 105 1 χ 105 0 0 Supernatant 0 M NaCl 2 x 104 0.2 M NaCl 8 X 104 0.5 M NaCl 2 χ 105 0.8 M NaCl 2 X 105 In a parallel set of experiments, material from the pro-PT purification step was dialyzed against 2 changes of 20 mM Bis-Tris (Sigma) pH 6.8 containing 0.1% CHAPS, 1 mg/ml phosphatidyl choline type VE, 50 Mg/ml phosphatidyl ethanolamine and 50 fig/ml phosphatidyl serine. Four hundred microliters of the dialyzed solution was combined with 0.5 ml of SP Sephadex™ (Pharmacia) or CM Sepharose™ (Pharmacia) that had been equilibrated in the same Bis-Tris buffer, and the mixtures were rocked . overnight at 4’C. Essentially all of the gammacarboxylase activity was found in the eluant.

EXAMPLE 3 Eighty milligrams of liver microsomes prepared from warfarin-treated rats was disrupted with 8 strokes of a tissue homogenizer and placed on ice for 15 minutes. The microsome preparation was then combined with 5 ml of CNBr-activated Sepharose to which was coupled 4 mg of mouse anti-proinsulin monoclonal antibody that had been washed 4 times (10 ml) in .025 M imidazole pH 7.2, 0.25 M sucrose. The mixture was rocked for 3 hours at 4’C, then centrifuged at 2,000 x g for 5 minutes at 4'C. The eluant was retained. The resin was washed with 5 ml of imidazole-sucrose buffer, centrifuged, and the eluant was recovered and combined with the first eluant.

The eluant pool was combined with 10 ml (3 g) of CNBr-activated Sepharose coupled to anti-factor X polyclonal antibody. The mixture was rocked overnight at 4’C, then washed 6' times with 10 ml of imidazole-sucrose buffer containing 0.5 M KC1. The resin was recovered and incubated in a gamma-carboxylase reaction with cold NaHCC>3 (10 ml total reaction volume) to release the enzyme. After one hour, the reaction mixture was centrifuged at 2,000 x g for 5 minutes. The eluant was recovered, and the resin was washed with 4 ml of 0.025 M imidazole pH 7.2, 0.5 M KC1. The wash was recovered and combined with the eluant.

The combined eluants were combined with 1 ml of washed lentil lectin Sepharose 4B, and the mixture was rocked overnight at 4’C. The mixture was centrifuged at 2,000 x g for 5 minutes at 4’C. The resin was washed four times in 3 ml of buffer A containing 0.5 M NaCl. The washed resin was recovered and combined with 2.5 ml of buffer A adjusted to 0.5 M NaCl, 5 mg/ml phosphatidyl choline type VE, 0.5% CHAPS and containing 0.5 M mannose. This mixture was rocked overnight at 4’C, then centrifuged < at 2,000 x g for 5 minutes at 4*C. The eluant fraction was recovered. Recovered fractions were assayed for gamma-carboxylase activity. Results indicated that about 70% to 90% of the carboxylase activity bound to the resin, and about 30% of the initial activity was recovered in the final eluant.

EXAMPLE 4 Bovine liver microsomes were prepared as 10 described by Swanson and Suttie (ibid.). Ten tubes (each containing 200 mg of protein) of microsomes were pooled in pairs, and each pool was suspended in 4 ml of SI (0.25 M sucrose, 0.025 M imidazole pH 7.2) containing 0.2% Triton X-100. The.microsomes were disrupted with 10 strokes of a tissue homogenizer, pooled, then placed on ice for 30 minutes. One milliliter was removed for subsequent assay. The mixture was centrifuged at 150,000 x g for 1 hour at 4’C. The supernatant was recovered and 1 ml was saved for assay. To the remainder of the supernatant was added 60 ml of saturated (NH4)2SO4 over 30 minutes at 4'C, and the mixture was stirred for an additional 30 minutes at 4*C. The precipitate was harvested by centrifugation and removal of the liquid fraction. The precipitate was resuspended in 5 ml of buffer A' (buffer A lacking phosphatidyl ethanolamine and phosphatidyl serine) and dialyzed twice in 600 ml of buffer A'. One milliliter was removed for assay. The dialysate (-40 ml) was combined with 20 ml activated thiol Sepharose (ATS) and rocked for three hours. The mixture was poured into a column and the flow-through was collected. One milliliter was removed for assay.

Ten milliliters of the flow-through was combined with 5 ml of CSF Sepharose (Example 1) , and 20 ml was combined with 10 ml of pro-PT Sepharose (prepared as in Example 1) . Thfe mixtures were rocked overnight at 4’C, then centrifuged at 2,000 x g for five minutes at 4’C. The resins were washed five times in 10 ml of buffer A' t adjusted to 0.5 M NaCl, then rocked at 4*C in buffer A' containing 0.5 M NaCl and 50 mM dithiothreitol for between 5 and 24 hours. The mixtures were centrifuged and the supernatants were retained.

Samples removed during the purification procedure were assayed for gamma-carboxylase activity and are shown in total protein Table 5. content (by BCA). Results 10 Fraction Volume (Pill... TABLE 5 Protein fma/ml) Activity (cpm/ml) Specific Activity (cpm/ma) TX Suspension 50 40 5 x 106 105 15 TX Sup. 45 41 4 x 106 105 (NH4)2SO4 ppt (dialyzed) 40 30 4 x 106 105 ATS Sup. 40 23 3 x 106 105 proPT eluant 10 -0.1 1 x 106 -107 20 CSF eluant 5 ~ -0.1 ND - ND= Not determined Five ml each of the proPT and CSF eluants were combined with 5 ml of lentil lectin Sepharose 4B that had been prewashed in buffer A'. The mixtures were rocked overnight at 4*C, then washed six times with 10 ml of buffer A' adjusted to 0.5 M NaCl. The washed resins were each mixed with 5 ml of buffer A' adjusted to 0.5 M NaCl, 0.5 M mannose, 0.5% CHAPS and 5 mg/ml phosphatidyl choline type VE. The mixtures were rocked overnight at 4*C and spun at 2,000 x g for 5 minutes. The supernatants were taken for gamma-carboxylase assay and gel electrophoresis.

’ EXAMPLE 5 Gamma-carboxylase was purified from a 293 cell line transfected to express a factor IX molecule lacking the propeptide (amino acids -18 to -1) . An expression vector for a propeptide-deleted human factor IX protein was constructed as shown in Figure 2. A pUC118 plasmid containing a human factor IX cDNA insert was digested with Eco RV and Hind III, and the 3.3 kb fragment, comprising the vector and 5' factor IX sequences, was recovered. The same plasmid was also digested with Hae III and Hind III, and the 1.3 kb fragment, comprising the 3' portion of the factor IX cDNA, was recovered. The two fragments were joined by ligation with oligonucleotides ZC2575 (5* ACA TTC AGC ACT GAG TAG AT 3'; Sequence ID Number 7) and ZC2576 (5' ATC TAC TCA GTG CTG AAT GT 3'; Sequence ID Number 8) . The DNA was transformed into E. coli strain HB101. Positive colonies were selected by ampicillin resistance. One hundred twenty-one positive colonies were replated and screened with 32P-labeled ZC2575 (Sequence ID Number 7). Twelve positive colonies were selected and screened by digestion with Hae III + Eco RI, Hae III, and Eco RV + Pvu II.

Two positive clones were selected on the basis of restriction analysis and digested with Bam HI to isolate the propeptide-deleted inserts. The inserts were ligated to Bam HI-digested pD5 (disclosed by Foster et al., U.S. Patent No. 4,959,318, which is incorporated herein by reference), a mammalian cell expression vector comprising the adenovirus 5 replication origin (0-1 map units), SV40 enhancer, adenovirus 2 major late promoter and tripartite leader, a set of splice signals, the SV40 late polyadenylation signal, and a unique Bam HI cloning site. The resulting plasmids were digested with Bam HI, and the factor IX inserts were sequenced. A positive plasmid clone wa^ designated leader-deleted IX-pD5.

The propeptide-deleted factor IX expression vector was co-transfected into 293 cell lines by the * calcium phosphate method with a plasmid (pD5neo) containing a neomycin resistance gene. The transfected cells were cultured in DMEM supplemented with lx PSN mix (GIBCO BRL) and 10% dialyzed fetal calf serum (FCS).

Randomly selected colonies were screened for factor IX production by enzyme-linked immunosorbent assay (ELISA). Positive colonies.' were selected and screened by radioimmune precipitation with polyclonal antisera to factor IX. This cell line was designated G3.

In a similar manner, a pD5 vector containing a native factor IX cDNA insert was co-transfected into 293 cells with pD5neo. The cells were cultured in DMEM supplemented with lx PSN mix, 10% FCS and 1% G418 for 36 hours, then· split 1:10 into five plates. Six colonies were chosen and plated into 6-well dishes in DMEM containing 10% FCS, 1% G418 and 1% PSN mix. After seven days, the cultures were split, half into DMEM containing 10% FCS, 1% G418, 5 gg/ml vitamin K and 1% PSN mix, and half into DMEM containing 10% FCS, 1% G418 and 1% PSN mix.

The six colonies growri in the vitamin K-supplemented medium were screened for factor IX production by ELISA. The ELISA-positive colonies grown in vitamin Ksupplemented media were screened by radioimmune precipitation (RIP) with polyclonal antisera to factor IX.

The highest factor IX producing cell line as determined by ELISA and RIP was designated D30. D30 cells that had been cultured in DMEM containing 10% dialyzed FCS, 1% G418 and 1% PSN were used for the microsomal preparations.

The D30 and G3 cell lines were cultured in DMEM containing 10% dialyzed FCS, 1% G418 and 1% PSN in 150 mm diameter plates. Forty plates of each unlabeled cell line were combined with one or two plates of metabolically labeled cells. The cells were harvested by aspirating off the medium and rinsing each plate twice with 2.5 ml of Versene (GIBCO B^L) to remove the cells. The rinses from each set of plates were pooled and centrifuged at 2,000 x g for five minutes at 4’C. The cell pellets (-2 x 109 cells) were each resuspended in 200 ml of PBS, centrifuged at 2,000 x g for five minutes at 4*C and resuspended in 10 ml of cold SI containing 2 mM PMSF. The resuspended cells were placed on ice and sonicated with four 15-second pulses with 30 seconds between pulses, then disrupted with 7 strokes of a tissue homogenizer. These preparations were centrifuged at. ‘4,000 x g for 15 minutes at 4’C. The supernatants were removed and centrifuged at 24,600 x g for 1 hour at 4*C. The pellets (microsomes) were each resuspended in 5 ml of SI containing 5% CHAPS. The resuspended microsomes were combined with 3 ml of antiproinsulin Sepharose (Example 3) , and the mixtures were rocked for 3 hours at 4’C. The mixtures were centrifuged at 2,000 x g for five minutes at 4’C. The supernatants were recovered and placed on ice. The pellets were rinsed with 5 ml of SI, mixed, centrifuged as above, and the supernatants were recovered and combined with the first supernatants.

The microsomes were then combined with antifactor IX resin (prepared by coupling 5 mg of ESN4 antibody [American Diagnostica, New York, NY] with 5 ml of CNBr-activated Sepharose). The resin was washed with 5 ml of 3 M KSCN, rocked 1 hour at 4’C, washed four times with 5 ml of SI, then combined with the microsome preparation and rocked overnight at 4’C. The resin was then washed four times with 10 ml of SI containing 0.5 M KCl.

The carboxylase was eluted from the resin by three different procedures. In the first, the resin was incubated with 5 ml of 0.2 M glycine pH 10 containing 50% ethylene glycol on a rocker for one hour at 4’C. The mixture was then rocked one hour at 4*C and spun at 2,000 x g for five minutes at 4’C. The supernatant, containing the factor IX and the bound carboxylase, was recovered. This elution method inactivates the carboxylase but is suitable for physical characterization procedures, such as analysis by gel electrophoresis. In the second elution procedure, the resin was incubated in 5 ml of SI containing 0.1% CHAPS, 1 mg/ml phosphatidyl choline type III and 100 μΜ pro-PT peptide. The mixture was rocked overnight at 4’C, centrifuged, and the supernatant was recovered and assayed for carboxylase activity. Fifty percent of the carboxylase activity was found in the eluant. In the third elution procedure, the carboxylase was eluted from the resin by incubation in a gammacarboxylase reaction mixture with cold NaHC03 to release the enzyme, as described above.

Two to five ml of the carboxylase reaction eluate was mixed with 1 ml of lentil lectin Sepharose 4B and rocked overnight at 4’C. The mixture was centrifuged for five minutes at 2,000 x g at 4’C. Gamma-carboxylase was eluted .from the resin by rocking it overnight at 4’C with 2.5 ml of 20 mM Tricine pH 8.5, 0.5% CHAPS, 1 mg/ml phosphatidyl choline type VE, 50 Mg/ml phosphatidyl serine, 50 Mg/ml phosphatidyl ethanolamine, 0.5 M NaCl and 0.5 M mannose. The supernatant was recovered.

In a similar preparation, unlabeled gammacarboxylase was purified by chromatography on ESN4Sepharose. The carboxylase was eluted from the resin with SI containing 100 mM CaCl2, 0.1 % CHAPS and 1 mg/ml phosphatidyl choline type III. Assay of fractions taken during the purification indicated that about 50-70% of the carboxylase activity was eluted from the resin (Table 6).

TABLE 6 D3O G3 Protein Activity Protein Activity 5 Fraction (mg/ml) coro/ml (mg/ml) com/ml Cell Extract 10 5 x 105 10 5 x io5 Microsome pellet 10 5 x 105 10 5 x 105 Preresin sup. ' 5 5 x 105 5 5 x 105 ESN4 sup. 5 2 x 105 5 2 x 105 10 CaCl2 eluant C.01 4 X 104 <.01 ND ESN4 resin — 1 X 105 — 1 X 104 ND = not determined EXAMPLE 6 Rat ’ microsomes (250,000 cpm total carboxylase activity) were electrophoresed on a native polyacrylamide gel. The microsomes (-0.5 ml) were combined with an equal volume of 2x sample buffer (0.125 M Tris pH 6.8, 20% glycerol, 0.0125% bromphenol blue) at 4*C. The sample was loaded onto an 8% polyacrylamide slab gel containing 0.375 M Tris, 1% PC/CHAPS solution (prepared by removing the chloroform from 10 ml of phosphatidyl choline type VE and resuspending the waxy residue in 20 ml of 100 mg/ml CHAPS), 0.1% ammonium persulfate and 0.025% TEMED. The gel was run at 4*C, 40 volts in running buffer containing 3 g/1 Tris base, 14.4 g/1 glycine and 1% PC/CHAPS solution until the dye front ran off the bottom of the gel.

The gel lane containing the microsomal preparation was cut out and sliced into 21 pieces. Each piece was incubated in the carboxylase assay for one hour at room temperature. One slice exhibited gammacarboxylase activity 7000 cpm above the background level.

EXAMPLE 7 To determine whether the carboxylase could be purified away from more extensively glycosylated contaminants, the partially purified rat pro-PT eluant was * subjected to lectin chromatography on lectins that bind complex oligosaccharide side-chains prior to lentil lectin chromatography. These lectins include Ervthrina cristagalli lectin, which binds /J-D-gal (1-4) -DglcNAc moieties, Pseudomonas aeruginosa PA-I lectin, which binds D-gal moieties, Limulus Polyphemus lectin, which binds sialic acid residues and Tetragonolobus purpureas lectin, which binds α-L-fuc moieties.

Partially purified rat carboxylase as a pro-PT 10 eluant was prepared essentially as described in Example 2. One milliliter of the eluant containing 5 χ 106 cpm was added to 0.5 ml of either Ervthrina cristagalli lectin coupled to CNBr-activated Sepharose 4B, Pseudomonas aeruginosa PA-I lectin coupled to CNBr-activated Sepharose 15 4B, Limulus Polyphemus lectin coupled to CNBr-activated Sepharose 4B or Tetragonolobus purpureas lectin resin. The mixtures were rocked overnight at 4’C. The mixtures were centrifuged at 2,000 x g for five minutes at 4’C, and the supernatants were added to 0.5 ml lentil lectin20 Sepharose 4B resin. The pellets were rinsed four times (0.2 ml each) in buffer A (Example 1) and the washes were combined with their respective supernatants. A control containing untreated pro-PT eluant was added to 0.5 ml lentil lectin-Sepharose 4B. The mixtures were rocked 25 overnight at 4’C followed by centrifugation at 2,000 x g for five minutes at 4’C recover the resin, and the pellets were rinsed with 1 ml buffer A containing 0.5 M NaCl. The rinse procedure was repeated three times. The carboxylase was eluted from the resins with 1 ml of buffer A adjusted 30 to 0.5 M NaCl, 0.5% CHAPS, 5 mg/ml phosphatidyl choline containing 0.5 M mannose. The mixtures were rocked overnight at 4’C. After incubation, the mixtures were centrifuged at 2,000 x g for five minutes at 4’C and the supernatants were removed to fresh tubes.

The eiuants were tested in carboxylase assays and were run on an 8% reducing polyacrylamide gel. The eluant was added to lx SDS sample buffer (50 mM Tris-HCl, < pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromphenol blue, 10% glycerol) . The mixtures were boiled for five minutes and the mixtures were loaded onto an 8% polyacrylamide gel. The results of the carboxylase assay showed that the carboxylase does not bind to the Erythrina cristagalli lectin, Pseudomonas aeruginosa PA-I lectin, Limulus Polyphemus ‘ lectin or Tetragonolobus purpureas lectin. An autoradiograph of the gel electrophoresis showed that these lectins removed major contaminants from the carboxylase preparation.

A partially purified rat pro-PT eluant was sequentially chromatographed using Tetragonolobus purpureas lectin resin and lentil lectin-Sepharose 4B. Partially purified rat carboxylase was prepared essentially as described in Example 1 using the pro-PT peptide. Ten milliliters of pro-PT eluant containing 4 x 106 cpm was dialyzed into 20 mM Tricine, pH 8.5, o.l% CHAPS, 1 mg/ml phosphatidyl choline type VE, 50 Mg/ml phosphatidyl ethanolamine and 50 Mg/ml phosphatidyl serine. The dialyzed - eluant was added to 6 ml of prewashed Tetragonolobus purpureas lectin resin. The sample was rocked overnight at 4’C followed by centrifugation at 2,000 x g for five minutes at 4’C. The supernatant was removed into 5 ml of prewashed lentil lectin-Sepharose 4B, the pellet was rinsed with the dialysis buffer, and the second supernatant was combined with the supernatant-lectin mixture. The supernatantlectin mixture was rocked overnight at 4’C, and the mixture was centrifuged at 2,000 x g for five minutes at 4’C. The supernatant was discarded and the pellet was washed six times with 10 ml of buffer A adjusted to 0.5 M NaCl. The carboxylase was eluted from the lectin resin with buffer A adjusted to 0.5 M NaCl, 0.5% CHAPS, 5 mg/ml phosphatidyl choline containing 0.5 M mannose overnight at 4’C. The samplie was centrifuged at 2,000 x g for five minutes, and the supernatant was subjected to a • carboxylase assay and was TCA precipitated and run on a polyacrylamide gel as described above.

EXAMPLE 8 Bovine liver microsomes, prepared essentially as described by Harbeck et al. (Thrombosis Res, 56: 317-323, 1989), were resuspended to 30 mg/ml in 70 ml of 20 mM Tris pH 7.3, 0.1 M NaCl and centrifuged at 105,000 x g for sixty minutes at 4’C. The supernatant was discarded, and the pellets were resuspended in 70 ml of 20 mM Tris pH 7.3, 0.1 M NaCl and solubilized with 0.5% CHAPS. The solution was incubated at 4 °C for 30 minutes, then centrifuged at 150,000 x g for 60 minutes at 4‘C. The supernatant · was discarded, and the pellet was resuspended in 20 mM Tris pH 7.3, 1 M NaCl, 1% CHAPS. The suspension was incubated 30 minutes at 4’C, then combined with antiprothrombin resin (prepared by coupling CNBr-activated Sepharose-4B and affinity purified rabbit antisera to bovine prothrombin). The mixture was incubated for 16 hours at 4’C, then loaded into a column (35 ml volume), and the resin was washed with five column volumes of 20 mM Tris pH 7.3, 0.1 M NaCl, 0.5% CHAPS, 0.5% phosphatidyl choline (type III-E), 5 mM dithiothreitol (buffer C), then with five column volumes of buffer C containing 1 mM ATP and 5 mM MgCl2 at 20’C. Aliquots of unbound material and of anti-prothrombin resin were assayed as described above. In general, 50% of the carboxylase activity could be bound to the resin.

Carboxylase was eluted from the resin with buffer C containing 100 μΜ of human factor X propeptide (consisting of the -18 to -1 sequence of human factor X (Leytus et al., ibid.; incorporated herein by reference)). Elution was performed in several batches, with 3-hour incubations at 20’C for each batch. Aliquots of post35 elution anti-prdthrombin resin and of propeptide eluant were assayed, and approximiately 70% of the carboxylase * was found to be eluted from the column by the propeptide (Table 7).

The propeptide eluant (200 ml) was adsorbed onto 0.5 ml of S-Sepharose (Pharmacia) at 4*C for 16-20 hours, and the resin was washed with ten volumes of 50 mM Tris pH 7.4, 100 mM NaCl, 0.25% phosphatidyl choline type VE, 0.25% CHAPS. The resin was then rocked for 30 minutes in 0.5 ml of 50 mM Tris pH 7.4, 200 mM NaCl, 0.25% phosphatidyl choline VE, 0.25% CHAPS, then incubated for an additional 30 minutes in 1 ml of the same buffer adjusted to 0.5 M NaCl. The resin was pelleted by low speed centrifugation. Both adsorption of propeptide eluant and salt elution were quantitative, giving a recovery of;100% for this step (Table 7).

Material eluted from S-Sepharose was then adsorbed onto 200 μΐ lentil lectin Sepharose (Sigma Chemical Co.). The mixture was adjusted to 5 mM MnCl2, 5 mM CaCl2, and incubated at 4*C for sixteen hours. The resin was washed with a 100-fold excess (100 volumes) of 50 mM Tris pH 7.4, 100 mM NaCl, 0.25% phosphatidyl choline VE, 0.25% CHAPS, and carboxylase was eluted in 1 ml of the same buffer containing 0.5 M a-methyl-mannoside and 10 mM EDTA. Overall recovery at this step was 50% (Table 7) . Neither prothrombin nor any other vitamin K-dependent protein could be detected in the eluant, indicating that adsorption of the carboxylase to lentil lectin occurs via a direct interaction.

TABLE 7 5 Sample Activity .(cpm) Protein Imgl* Specific Activity (cDm/mo) Fold Purifi- cation 10 Solubilized microsomes 6xl07 5500 1 x 104 1 detergent- extracted microsomes 9 X 107 2000 5 χ 104 5 15 Propeptide eluant 2 X 107 .16 1 x 108 104 20 S-Sepharose' eluant 1.6 χ 107 .08 2 x 108 2 X 104 lentil lectin eluant 4 X 106 .002 2 X 109 2 X 105 *determined by scanning Coomassie blue or silver stained gels using a BSA standard The eluant from the lentil lectin Sepharose was electrophoresed, and the gel was stained with Coomassie blue. A single 90 kD band was observed (Figure 3). When the gel was silver stained or when radio-iodinated protein was analyzed, this 90 kD band was the major band, although several minor protein bands were also present.

The propeptide-eluted bovine carboxylase (6 x 106 cpm activity in 100 ml buffer C containing 100 μΜ propeptide) was batch adsorbed onto l ml S-Sepharose for 16 hours at 4*C. Bound material was washed with 10 volumes of 50 mM Tris pH 7.4, 100 mM NaCl, 0.25% phosphatidyl choline VE, 0.25% CHAPS, and a small aliquot of the resin wa«s assayed for carboxylase activity. The remaining resin was mixed with an equal volume of Freund's adjuvant (obtained from ICN Biochemicals, Costa Mesa, CA) . and injected intraperitoneally into five Balb/c mice. The mice were boosted at 2-week intervals with carboxylase prepared in the same manner, except incomplete Freund's adjuvant was used.

Following the third injection of antigen, serum was collected and tested using an activity immunocapture assay. Serum from a non-immunized mouse was used as a control. Increasing amounts (0-20 μΐ) of sera were incubated with bovine microsomes (1.5 mg in 50 μΐ) for 8 hours at 4’C followed by the addition of 50 μΐ of a 1:1 mixture of protein A Sepharose (Sigma Chemical Co.) in 50 mM Tris pH 7.4, 100 mM NaCl. Samples were rocked at 4*C for 12 hours, and the resins were then washed 1000-fold with 50 mM Tris pH 7.4, 100 mM NaCl, 0.25% phosphatidyl choline VE, 0.25% CHAPS. The resins were then incubated in 128 μΐ of the carboxylase reaction for four hours at 20’C, precipitated with 10% TCA (1 ml), then counted after removal of 14CO2 by boiling. Test samples were run to show that carboxylase activity is linear over this period.

Samples were run in duplicate and an average was taken for each. With sera isolated from carboxylase-injected mice a linear increase in carboxylase activity occured with an increase in antisera. With the highest amount of antisera used (20 μΐ), 20% of the carboxylase activity bound to the resin.

Western blot (Towbin et al., Proc. Natl. Acad. Sci. USA 76: 4350-4358, 1979; U.S. Patent No. 4,452,901) analysis of carboxylase was carred out in parallel with either non-immune or anti-carboxylase antisera. Lentil lectin-eluted carboxylase (105 cpm of activity) was electrophoresed on an 8% denaturing gel and transferred to nitrocellulose (Biotrace NT, Gelman Sciences, Ann Arbor, MI) at 60 volts over 24 hours at 4’C in 1.4% glycine, 25 mM Tris pH 8.8, 20% methanol and .005% SDS.

Nitrocellulose wias washed in Western buffer A (50 mM Tris pH 7.4, 5 mM EDTA, 0.05% NP-40, 150 mM NaCl and 0.25% gelatin), then incubated in 20 ml Western buffer A supplemented with a 1:1000 dilution of sera. After 16 hours rocking at 4’C, the nitrocellulose was washed in Western buffer A (three times, 100 ml each), then incubated in 20 ml Western buffer A containing 125Ilabeled rabbit anti-mouse IgG (2 χ 106 cpm) for one hour at 4’C. The nitrocellulose was washed three times in 100 ml of Western buffer B (50 mM Tris pH 7.4, 5 mM EDTA, 0.05% NP-40, 1 M NaCl, 0.4% N-laurylsarcosine, 0.25% gelatin), air dried and exposed to film. The only immunoreactive protein observed with the anti-carboxylase antisera was a 90 kD band (Figure 4) . No reactivity was observed with control sera (data not shown).

The starting microsomal preparations and SSepharose eluants were also analyzed by Western blot. Egual amounts of carboxylase activity from each stage of purification were analyzed, and equal amounts of carboxylase reactivity were observed. These results indicate that there was no significant loss of activity during purification.

Example 9 Microsomal preparations were prepared from 80 150-mm plates (ca. 2 χ 109 cells) of D30 and G3 cells (Example 5) . The cells were taken off the plates with Versene (2.5 ml per 150 mm plate to rinse; then 2.5 ml to take off cells) and concentrated by centrifugation at 2000 x g for 5 minutes. The cells were rinsed in cold PBS (total volume 400 ml) and spun at 2000 x g for 5 minutes. The cells were resuspended in SIP (0.25 M sucrose, 0.25 M imidazole pH 7.2, 2 mM PMSF) to a final volume of 20 ml and sonicated in four 15-second bursts with 30 seconds between bursts. The cells were then disrupted with seven strokes of a tissue homogenizer and centrifuged for 15 minutes at 4 000 x g, 4’C. Microsomes were then prepared from the supernatants by spinning at 45,000 RPM for 1 hour at 4’C in a Beckman ultracentrifuge. The resulting supernatants were discarded, and the pellets were and the •Ε 912600 . disrupted in 10 ml SIP using 7 strokes of a tissue homogenizer, then adjusted to 0.2% CHAPS. After 30 minutes on ice the samples were combined with an antifactor IX resin (prepared from polyclonal anti-factor IX IgG, purified over a protein A-Sepharose column, then coupled to CNBr-activated Sepharose at 5 mg/ml).

The anti-factor IX resin was rocked at 4’C overnight, then washed with 0.1 M NaCl in 0.05 M Tris pH 7.4, then eluted with 0.05 M Tris pH 7.4, 0.1 M NaCl, 0.25% phosphatidyl choline, 0.25% CHAPS plus 100 μα propeptide (either the -18 to -1 sequence of human prothrombin or of human factor X) . Elution was effected by rocking the resin in this buffer overnight at 4’C, then collecting the supernatant and assaying it.

The propeptide eluants were then incubated with 0.2-0.5 ml Q-Sepharose (Pharmacia) overnight at 4’C. The resin was rinsed in 0.05 M Tris pH 7.4, 0.1 M NaCl, 0.25% phosphatidyl choline, 0.25% CHAPS, then incubated for 30 seconds at 4’C with the same buffer adjusted to 0.2 M NaCl. The buffer was then adjusted to 0.5 M NaCl, incubated for another 30 seconds, and the eluant was collected and assayed. Protein content was determined by gel scanning. Results are shown in Table 8.

TABLE 8 Sample microsomes Activity (£ΕΒ>) Protein (mg) * Specific Activity (cpm/mg) Fold Purifi- cation D3 0 1 x 106 100 l x 104 1 10 G3 3 χ 105 100 3 X 103 propeptide eluant D30 3 χ 105 . 02 1.5 x 107 1500 G3 2 χ 104 .02 1 X 106 15 O-Sepharose eluant D30 2 χ 105 .01 2 χ 107 2000 G3 1 x 104 .01 1 X 106 Eluants from microsomes prepared from metabollically labeled 1530 and G3 cells were further fractionated on lentil lectin resin. Purified fractions were gel electrophoresed and exposed for autoradiography. As shown in Figure 5, the lentil lectin eluant from D30 cells contains an approximately 65kD band not present in the G3 cell sample.

Whole cell extracts were prepared from D30 cells. Eighty 150 mm plates of cells were harvested with Versene, concentrated, washed and disrupted as described above. 0.5 ml of CHAPS was added, and the lysate was placed on ice for 30 minutes. The lysate was then centrifuged at 4000 x g for 15 minutes, and the was retained. The Sepharose-coupled ESN4, resulting extract and carboxylase was was supernatant adsorbed to eluted using 0.05 M Tris pH 7.4, 0.1 M NaCl, 0.25% phosphatidyl choline, 0.25% CHAPS, 100 mM CaCl2 (ESN4 is Ca++-dependent). Elution was performed for 1 hour at 4’C. The eluant was then adsorbed onto concanavalin A Sepharose » (3 00 μΐ) by incubating at 4’C overnight. The resin was washed with 0.05 M Tris pH 7.4, 0.1 M NaCl, 0.25% phosphatidyl choline, 0.25% CHAPS, a small aliquot of resin was assayed to monitor the amount of activity, and the rest of the resin was injected intraperitoneally into Balb/c mice. Approximately 3-5 x 105 cpm of activity was used per mouse per boost.

A sample of D30 whole cell extract was run out for Western blot analysis. The sample was boiled in SDS sample buffer for five minutes, centrifuged one minute at room temperature in a Beckman Microfuge 12 at a setting of 4, and electrophoresed on an 8% discontinuous denaturing polyacrylamide gel at 40 volts for 15 hours at room temperature. Protein was transferred to nitrocellulose at 60 volts for 24 hours. The nitrocellulose was removed and washed three times, 20 minutes per wash, in 100 ml Western buffer A at 4’C, rocking. The gel was stained to verify transfer. The washed nitrocellulose was placed in 20 ml fresh Western buffer A, and 20 μΐ of anti-carboxylase 0 antiserum (Example 8) was added. The blot was rocked for 18 hours at 4’C, then washed three times as above. The blot was placed in 20 ml fresh Western buffer A containing 10 μΐ (0.5 μg; 2.0 x 106 cpm) iodinated rabbit anti-mouse IgG (Organon Teknika Corp., West Chester, PA; iodinated with 125i obtained from Amersham, Arlington Heights, IL) and rocked for one hour at 4*C. The nitrocellulose was then washed three times as above, air dried, wrapped in saran and exposed to X-ray film.

EXAMPLE 10 Total RNA was prepared from bovine liver using guanidine isothiocyanate (Chirgwin et al. Biochemistry .18: 52-94, 1979) and CsCl centrifugation (Gilsin et al.

Biochemistry 13: 2633-2637, 1974). Poly(A)+ RNA was selected from the total RNA using oligod(T) cellulose chromatography (Aviv and Leder, Proc. Natl. Acad. Sci. USA 69: 1408, 1972).

« First strand cDNA was synthesized from one time poly d(T)-selected bovine liver poly(A)+ RNA in two separate reactions. One reaction, containing radiolabeled dATP, was used to assess the quality of first strand synthesis. The second reaction was carried out in the absence of radiolabeled dATP and was used, in part, to assess the quality of second strand synthesis. Superscript reverse transcriptase (GIBCO BRL) was used specifically as described below. A 2.5x reaction mix was prepared at room temperature by mixing, in order, 8 μΐ of 5x reverse transcriptase buffer (GIBCO BRL; 250 mM TrisHCl, pH 8.3, 375 mM KCl, and 15 mM MgCl2) , 2.0 μΐ 200 mM dithiothreitol (made fresh or stored in aliquots at -70°C) and 2.0 μ! of a deoxynucleotide triphosphate solution containing 10 mM each of dATP, dGTP, dTTP and 5-methyl dCTP (Pharmacia). The reaction mix was aliquoted into two tubes of 6 μΐ each. To the first tube, 1.0 μΐ of 10 MCi/μΙ a32P-dATP (Amersham) was added and 1.0 μΐ of water was added to the second reaction tube. Fourteen microliters of a solution containing 10 pg of bovine liver poly(A)+ RNA diluted in 14 μΐ of 5 mM Tris-HCl, pH 7.4, 50 μΜ EDTA was mixed with 2 μΐ of 1 μq/μl first strand primer, ZC3747 (GAC AGA GCA CAG AAT TCA CTA CTC GAG TTT TTT TTT TTT TTT; Sequence ID Number 10), and the primer was annealed to the RNA by heating the mixture to 65°C for 4 minutes, followed by chilling in ice water. Eight microliters of the RNAprimer mixture was added to each of the two reaction tubes followed by 5 μg of 200 U/μΙ Superscript reverse transcriptase (GIBCO BRL). The reactions were mixed gently, and the tubes were incubated at 45 °C for 30 minutes. After incubation, 80 μΐ of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA was added to each tube, the samples were vortexed and centrifuged briefly. Two microliters of each reaction was removed to determine total counts and TCA precipitable counts (incorporated counts). Two microliters of each sample were analyzed by alkaline gel electrophoresis to assess the quality of first strand - synthesis. The remainder of each sample was ethanol precipitated. The nucleic acids were pelleted by centrifugation, washed with 80% ethanol and air dried for ten minutes. The first strand synthesis yielded 1.0 ng of liver cDNA or a 20% conversion of RNA into DNA.

Second strand cDNA synthesis was performed on the RNA-DNA hybrid, from the first strand reactions under conditions which encouraged first strand priming of second strand synthesis resulting in DNA hairpin formation. The nucleic acid pellets containing the first strand cDNA were resuspended in 71 μΐ of water. To assess the quality of second strand synthesis, a32P-dATP was added to the unlabeled first strand cDNA. To encourage formation of the hairpin structure, all reagents except the enzymes were brought to room temperature, and the reaction mixtures were set up at room temperature. (Alternatively, the reagents can be on ice and the reaction mixture set up at room temperature and allowed to equilibrate at room temperature for a short time prior to incubation at 16’C.) Two reaction tubes were' set up for each synthesis. One reaction tube contained the unlabeled first strand cDNA and the other reaction tube contained the radiolabeled first strand cDNA. To each reaction tube, 10 μΐ of 5x second strand buffer (100 mM Tris, pH 7.4, 450 mM KC1, 23 mM MgCl2, 50 mM (NH4)2(SO4), 3 μΐ of beta-NAD and 1 μΐ of a deoxynucleotide triphosphate solution containing 10 mM each of dATP, dGTP, dTTP and dCTP (Pharmacia) , 1 μΐ α32ΡdATP or 1 μΐ of water (the radiolabeled dATP was added to the tube containing the unlabeled first strand cDNA), 0.6 μΐ of 7 U/μΙ £. coli DNA ligase (New England Biolabs, Beverly, MA), 3.1 μΐ of 8 U/μΙ E. coli DNA polymerase I (Amersham), and 1 μΐ of 2 U/μΙ of RNase H (GIBCO BRL). The reactions were incubated at 16 *C for 2 hours. After incubation, 3 μΐ was taken from each reaction tube to determine total' and TCA precipitable counts. Two microliters of each sample were analyzed by alkaline gel electrophoresis to assess the quality of second strand ‘ synthesis by the presence of a band of approximately twice unit length. To the remainder of each sample, 2 pi of 2.5 lg/μΐ oyster glycogen, 5 pi of 0.5 M EDTA and 200 pi of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA were added, the samples were phenol-chloroform extracted, and isopropanol precipitated. The nucleic acids were pelleted by centrifugation, washed with 80% ethanol and air dried. The yield of double stranded cDNA in each of the reactions was approximately 2 pg.

The single-stranded DNA in the hairpin structure was clipped using mung bean nuclease. The double stranded cDNA samples were resuspended and combined in 30 pi of water. Five microliters of lOx mung bean buffer (0.3 M NaOAC, pH 4-6, 3 M NaCl, 10 mM ZnSO4) , 5 pi of 10 mM dithiothreitol, 5 pi of 50% glycerol, and 5 pi of 10 ϋ/μΐ mung bean nuclease (Promega Corp, Madison, WI) were added and the reactions were incubated at 30’C for 30 minutes. After incubation, 50 pi of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA was added to each tube, and 2 pi of each sample was subjected to alkaline gel electrophoresis to assess the cleavage of the second strand product into unit length cDNA. One hundred microliters of 1 M Tris-HCl, pH 7.4 were added to each sample, and the samples were twice extracted with phenol-chloroform. Following the final phenol-chloroform extraction, the DNA was isopropanol precipitated. The DNA was pelleted by centrifugation, washed with 80% ethanol and air dried. Approximately 2 pg of DNA was obtained from each reaction.

The cDNA was blunt-ended with T4 DNA polymerase after the cDNA pellet was resuspended in 30 pi of water. Five microliters of ΙΟχ T4 DNA polymerase buffer (330 mM Tris-acetate, pH 7.9, 670 mM KAc, 100 mM MgAc, 1 mg/ml gelatin), 5 pi of 1 mM dNTP, 5 pi 50 mM dithiothreitol, 5 pi of 1 U/μΙ T4 DNA polymerase (Boehringer-Mannheim) were added to each trube. After an incubation at 15 *C for 1 hour, 150 pi of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA was added and the sample was phenol-chloroform extracted . followed by isopropanol precipitation. The cDNA was pelleted by centrifugation, washed with 80% ethanol, air dried and resuspended in 4 til water. Eco RI adapters (Invitrogen, Cat. # N409-20) were ligated to the blunted cDNA.

The first strand primer encoded an Xho I cloning site to allow the cDNA to be directionally cloned into an expression vector. The cDNA was digested with Xho I followed by phenol-chloroform extraction and isopropanol precipitation. After digestion, the cDNA was electrophoresed in a 0.8% low melt agarose gel, and the cDNA over 2.0 kb was electroeluted using an Elutrap (Schleicher and Shuell, Keene, NH) . The electroeluted cDNA in 500. pi of buffer was isopropanol precipitated and the cDNA was pelleted by centrifugation. The cDNA pellet was washed with 80% ethanol.

The double-stranded, linkered, and Xho I-cut cDNA was resuspended in 34 pi of distilled water. Five microliters of 10X kinase buffer (500 mM Tris, pH 7.8, 100 mM MgCl2, 1 mM EDTA), 5 Ml of 10 mM ATP, 1 pi 200 mM dithiothreitol and 5 Ml of T4 polynucleotide kinase (1 U/μΙ, GIBCO BRL) were added to the DNA. The reaction mixture was gently mixed, and incubated at 37C for one hour. After incubation, the mixture was phenol-chloroform extracted, chloroform extracted and isopropanol precipitated using 5 pg of mussel glycogen (Boehringer Mannheim, Indianapolis, IN) as a carrier.

One hundred and twenty nanograms (1.5 pi) of the cDNA was combined with 0.5 pi of distilled water, 2 pi of Lambda Uni-ZAP (1 pg/pl, Stratagene Cloning Systems, La Jolla, CA) , which had been digested with Xho I and Eco RI and treated with calf intestinal phosphatase, and 2 pi of 3X ligation mix (7 pi 10X ligation buffer (500 mM Tris, pH 7.8, 100 mM MgCl2, 10 mM ATP, 500 Mg/ml BSA), 7 Ml 100 mM DTT, 7 Ml T4 UNA ligase (1 U/pi, Boehringer Mannheim, Indianapolis, IN)). The ligation mixture was incubated for nine hours at room temperature. Following the • ligation, the DNA was packaged into phage heads using Gigapack Plus II (Stratagene Cloning Systems) according to the manufacturer supplied protocol. The packaging reaction was plated out onto PLKF' host cells (Stratagene Cloning Systems) on 59 plates at approximately 250,000 plaque forming units/plate for a yield of approximately 15 χ 106 independent plaques. The plates were overlaid with TM buffer (10 mM Tris, pH 7.8, 10 mM MgSC>4) and allowed to elute for six hours at room temperature. The liquid lysate was removed, pooled and stored at 4’C. Fifty millilers of chloroform was added to the lysate to prevent bacterial growth. As a test of quality, the library was screened with a radiolabeled human plasminogen DNA fragment as a probe. Positive clones were seen at a frequence of 0.53% and greater than 50% of the positives were full length when DNA prepared from the clones was analyzed.

EXAMPLE 11 Nucleotide sequences encoding gamma-carboxylase were obtained using polymerase chain reactions (PCR) and oligonucleotides designed from amino acid sequences determined by amino acid microsequencing of the partially purified material described in Example 8. Partially purified S-Sepharose eluant (Example 8) was concentrated either using a Centricon concentrator (Amicon, Danvers, MA) or by extracting the phospholipids from the SSepharose eluant and eluting the protein with methanol/chloroform as described by Wessel and Flugge (Anal. Biochem. 138: 141-143, 1984, which is incorporated by reference herein). The concentrated material was electrophoresed in an 8% SDS-polyacrylamide gel and transferred to nitrocellulose essentially as described in Example 8. The nitrocellulose was stained with amido black (Sigma, s£. Louis, MO) essentially as described by Aebersold et al. (Proc. Natl. Acad. Sci. USA 84: 69706974, 1987) and Schaffner and Weismann (Anal. Biochem. 56/. •Ε 912600 * 502-514, 1973), which are incorporated herein by reference, and the 90 Kd band was cut out. The band was rinsed with distilled water and stored at -20’C. The procedure was repeated four times to collect a sufficient sample for microsequencing. The 90 kD bands on nitrocellulose were pooled and then was microsequenced by Harvard Microsequen'cing (Cambridge, MA) . The amino terminal sequences of three peptides were determined as shown in Table 9.

TABLE 9 Peptide 1 (Sequence ID Number 3) F T L L A Ρ T S P G D Τ Τ Ρ [K] Peptide 2 (Sequence ID Number 4) GRDPALPTLLNPK Peptide 3 (Sequence ID Number 5) DD [R] GPSGQGQGQGQFLIQQVT Peptide 4 (Sequence ID Number 6) FLWDEGFHQLVIQR where residues enclosed in brackets [] indicate 25 probable/reasonable residues at the designated position Families of degenerate oligonucleotides, ZC4135 (Table 10; Sequence ID Number 11) and ZC4136 (Table 10; Sequence ID Number 12) were designed to correspond to the terminal portions of the amino acid sequences of Peptide 1 (Table 9; Sequence ID Number 3) and in addition, Eco RI restriction sites were added to the 5' termini of the oligonucleotides to facilitate subcloning. The oligonucleotides- were synthesized on an Applied Biosystems 394 RNA/DNA Synthesizer (Applied Biosystems, Foster City, CA) . The antisense oligonucleotide family, ZC4135 (Table ; Sequence ID Number 11), had a 128-fold degeneracy. The sense oligonucleotide family, ZC4136 (Table 10; Sequence ID Number 12), had a 512-fold degeneracy.

TABLE 10 Degenerate Oligonucleotide Primer Families ZC4135 (Sequence ID Number 11) ATA GAA TTC TTG GGG GTG GTG TC ZC4136 (Sequence ID Number 12) ATT AGA ATT CTT CAC GCT GCT CGC ZC4138 (Sequence ID Number 13) A A C CC TTT CCG ACG AGG CCG GG ZC4204 (Sequence ID Number 14) A A c c c T T TAT AGA ATT CGT GAC TTG TTG GAT I ΙΕ 9^26θθ TABLE 10 continued ZC4205 (Sequence ID C ATA AGA ATT CGA TGA ZC4217 (Sequence ID A C T ATA AGA ATT-CGG GCC ZC4241 (Sequence ID AAA A C T T TGG TGG AAG CCC TCG Number 15) AAA CCC G G TCG TGG TCC Number 16) AAA CCC C TT T T GAG GGG GCA Number 17) TCC CA First strand bovine cDNA was prepared and used as the template cDNA for the PCR reactions. First strand cDNA was synthesized from one time poly d (T)-selected bovine liver poly(A)+ RNA in two separate reactions. One reaction, containing radiolabeled dATP, was used to assess the quality of first strand synthesis. The second reaction was carried out in the absence of radiolabeled dATP and was used, in part, to assess the quality of second strand synthesis. Superscript reverse transcriptase (GIBCO BRL) was used specifically as described below. A 2.5x reaction mix was prepared at room temperature by mixing, in order, 8 μΐ of 5x reverse transcriptase butter (GIBCO BRL; 250 mM Tris-HCl, pH 8.3, 375 mM KC1, and 15 mM MgCl2), 2.0 pi 200 mM dithiothreitol (made fresh or stored in aliquots at -70*C) and 2.0 pi of a deoxynucleotide triphosphate solution containing 10 mM each of dATP, dGTP, dTTP and 5-methyl dCTP (Pharmacia). The reaction mix was aliquoted into two tubes of 6 μΐ each. To the first tube, 1.0 μΐ of 10 /xCi/gl a32P-dATP (Amersham) was added and 1.0 μΐ of water was added to the second reaction tube. Fourteen microliters of a solution containing 10 gg of Bovine liver poly(A)+ RNA diluted in 14 /ll of 5 mM Tris-HCl pH 7.4, 50 μΜ EDTA was mixed with 2 μΐ of 1 μ9/μ1 first strand primer, ZC2938 (GAC AGA GCA CAG AAT TCA CTA GTG AGC TCT TTT TTT TTT TTT TT; Seguence ID Number 9) , and the primer was annealed to the RNA by heating the mixture to 65’C for 4 minutes, followed by chilling in ice water. Eight microliters of the RNAprimer mixture was added to each of the two reaction tubes followed by 5 μς of 200 U/μΙ Superscript reverse transcriptase (GIBCO BRL). The reactions were mixed gently, and the tubes were incubated at 45’C for 3 0 minutes. After incubation, 80 μΐ of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA were added to each tube, the samples were vortexed and centrifuged briefly. Two microliters of each reaction was removed to determine total counts and TCA precipitable counts (incorporated counts). Two microliters of each sample was analyzed by alkaline gel electrophoresis to assess the quality of first strand synthesis. Five microliters of 0.5 M EDTA and 10 μΐ of 5 Μ KOH were added to each tube, and the reactions were placed at 65’C for five minutes. Following the incubation, the reaction was terminated by the addition of 35 μΐ of 8 M NH4AC and 135 μΐ of isopropanol. The mixture was chilled on ice for 15 minutes and centrifuged for 20 minutes at 10,000 rpm. The pellet was resuspended in 5 μΐ of distilled water. A yield of 1.5 μg of cDNA was obtained.

A PCR reaction mixture was prepared which contained 50 ng fef first strand bovine cDNA, 5 μΐ of 2.5mM dNTPs (containing all four deoxynucleotide triphosphates, Cetus, Emeryville, CA) , 5 μΐ of lOx PCR buffer (Perkins35 . Elmer Cetus, Norwalk, CN) and 5 μΐ of each oligonucleotide (ZC4135 and ZC4136? Table 10; Sequence ID Nos. 11 and 12) at a starting concentration of 20 pmole/μΐ. The total volume was 50 μΐ. One microliter of Taq I polymerase (Perkins-Elmer Cetus) was added, and the PCR reaction was carried out under the conditions shown in Table 11.

TABLE 11 Precvcle - 3 cycles 94“C for 1 minute 32*C for 1 minute 72*C for 1.5 minute Amplification cycle - 35 cycles 94 *C for 1 minute 55 *C for 1 minute 20 72*C for 1 .5 minute After the last cycle of the Amplification cycle, the reaction products were separated by electrophoresis in a Nusieve gel (FMC Bioproducts, Rockland, ME). The DNA of approximately 65 base pairs was isolated by melting the gel at 65’C in 5 ml of 10 mM Tris pH 8, 1 mM EDTA and 5 ml of phenol. The mixture was spun at 4000 rpm for five minutes at room temperature (Beckman GPR centrifuge, Beckman Instruments, Carlsbad, CA) and the aqueous layer was transferred to a fresh tube. The amplified DNA in the aqueous layer was ethanol precipitated and pelleted by centrifugation.

The PCR reaction was repeated on the pellet by first resuspending the pellet in 30 μΐ of distilled water.

PCR reagents w&re added as described above, and the reaction mixture was brought to a final volume of 50 μΐ. The PCR cycles were repeated as described in Table 11 • except that the precycle was omitted and the amplification cycle was run twice. The resulting reaction products were separated on a gel as described above. Several bands of approximately 65 base pairs in length were seen including a unique band not seen in control reactions using each oligonucleotide primer family independently. The region around 65 base pairs' was cut from the gel, and the DNA was extracted as described above.

The PCR reaction was repeated a third time using 10 the second round reaction products. The reactions were run as described in Table 11 except that the precycle was omitted. The reaction products were gel purified as described above and the resulting DNA was digested with Eco RI. The Eco RI-digested DNA was phenol extracted and 15 ethanol precipitated.

The amplified DNA was ligated into pUC118 or pUC119 which had been linearized by digestion with Eco RI digested and treated with calf alkaline phosphatase to prevent recircularization. The ligation mixture was 20 transformed into DH10B competent £. coli cells (GIBCO BRL) using a Biorad Gene Pulser (Biorad Richmond, CA) , according to the high efficiency electro-transformation protocol specified by the manufacturer.

Plasmid DNA was prepared from selected and the DNA is sequenced for positive of the gamma-carboxylase insert. In addition, Southern analysis was carried out on aliquots of the plasmid DNA using the degenerate oligonucleotide ZC4138 (Table 10; Sequence ID Number 13), corresponding to 30 an internal amino acid sequence of Peptide 1 (Table 9; Sequence ID Number 3) flanked by the oligonucleotide primers ZC4135 and ZC4136 (Table 10; Sequence ID Nos. 11 and 12).

Nucleotide sequences were also obtained using 35 families of degenerate oligonucleotide probes which were designed to correspond to the terminal portions of the amino acid sequences of Peptide 3 (Table 9; Sequence ID transformants, identification , Number 5) and contain Eco RI restriction sites at the 5' termini of the oligonucleotides to facilitate subcloning. The oligonucleotide family ZC4204 (Table 10; Sequence ID Number 14) corresponds to the amino terminus of Peptide 3 (Table 9; Sequence ID Number 5). The oligonucleotide family ZC4205 (Table 10; Sequence ID Number 15) corresponds to the oarboxyterminus of Peptide 3 (Table 9; Sequence ID Number 5), The oligonucleotide family ZC4217 (Table 10; Sequence ID Number 16) corresponds to internal amino terminal residues of Peptide 3 (Table 9; Sequence ID Number 5).

Bovine first strand cDNA, prepared as described in Example 10, was used as a template in two PCR reactions. . I-n the first reaction oligonucleotide family ZC4205 (Table 10; Sequence ID Number 15) was combined with oligonucleotide family ZC4204 (Table 10; Sequence ID Number 14). In the second reaction, oligonucleotide family ZC4205 (Table 10; Sequence ID Number 15) was combined with oligonucleotide family ZC4217 (Table 10; Sequence ID Number 16) . The PCR reactions were carried out essentially as described above using the protocol described above and the conditions described in Table 12 except that the PCR products were gel purified using a crush and soak method. The PCR products were isolated from the gel by cutting the appropriate band from the gel, crushing the band with a pasteur pipette in TE (10 mM Tris, pH 8.0, 1 mM EDTA) and allowing the crushed gel to soak in the TE overnight at room temperature. After the overnight incubation, the gel-TE mixture is phenol extracted and the aqueous layer containing the DNA is ethanol precipitated.

TABLE 12 Precvcle - 4 cycles 94*C for 1 minute 36’C for 1 minute 72*C for 1 .5 minute First Amplification cycle - 25 cycles 94’C for 1 minute 60“C for 1 minute 72*C for 1.5 minute Second Amplification cycle - 25 cycles 94’C for 1 minute 60’C for 1 minute 72’C for 1.5 minute The sequences of the PCR products are used to generate oligonucleotide probes that are used to probe a bovine liver cDNA library, prepared as described in 25 Example 10 and probed according to standard techniques essentially as described by Sambrook et al. (Molecular Cloning, A Laboratory Manual, Second Edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989), which is incorporated herein by reference). 30 Oligonucleotide probes designed from the PCR product sequences are also used in a RACE protocol essentially as described by Frohman et al. (Proc. Natl. Acad. Sci. USA 85: 8998-9002, 1988, which is incorporated herein by reference) to obtain gamma-carboxylase sequences.

The am’ino acid sequence of Peptide 4 (Table 9; Sequence ID Number 6) contains a sequence from which a family of degenerate oligonucleotides may be designed • containing a degeneracy of 64 over 20 base pairs. The family of degenerate oligonucleotides, ZC4241 (Sequence ID Number 17) is used as a probe of the bovine lambda gtll cDNA library described in Example 10. The library is screened using standard techniques described Sambrook et al. (ibid.). Positive clones are plaque purified and analyzed by restriction and sequence analysis.

EXAMPLE 12 Monoclonal antibodies were prepared essentially as described by Hart (U.S. Patent Application, Serial No. 07/139,960, which is incorporated herein by reference).

Briefly, partially purified bovine gammacarboxylase· was bound to S-Sepharose, as described in Example 8 or to sulfo-propyl Sepharose. The bound material was mixed with Complete Freund's Adjuvant (ICN Biochemicals, Costa Mesa, CA.), and 150 pi of the solution per mouse was injected intraperitoneally into seven-weekold Balb/c female mice (Simonsen Labs, Gilroy, CA).

Booster injection material was prepared by absorbing partially purified gamma-carboxylase to sulfo-propyl or SSepharose (Pharmacia), and the bound material was mixed with Incomplete Freund's Adjuvant. Each mouse was boosted with 150 μΐ of booster injection material at approximately two week intervals for the first two to three months and then at approximately monthly intervals thereafter. Mice were identified as expressing anti-carboxylase antibodies after the first three injections as described in Example 8.

Two three- to four-week-old Balb/c mice, which were shown to express anti-carboxylase antibodies, were sacrificed and the spleens and lymph nodes were removed from the immunized mice and minced with scissors on top of fine-mesh stainless steel screens. The minced tissues were washed through the fine screens into petri dishes with 10 ml RPMI 1640 (GIBCO BRL) . The remainder of the minced tissues were pressed through the screens with a . spatula and the screens were washed with 5 ml RPMI 1640. To remove any remaining cell material, the bottom of the screens were scraped and the material was added to the petri dishes.

The strained tissues were transferred into 50 ml centrifuge tubes, and the petri dishes were washed with 10 ml RPMI 1640 to remove any remaining material. The cell suspensions were centrifuged for 10 minutes at 2 00 x g. The supernatants were discarded and the pellets were resuspended in 4 ml RPMI 1640. After resuspension 1 ml fetal calf serum (BioCell, Carson, CA) was added to each tube.

The red blood cell contaminants were lysed by adding 18 . ml sterile distilled water to the cell suspensions. The mixtures were swirled quickly and 5 ml 4.25% NaCl was added to each tube. The cell suspensions were centrifuged for 10 minutes at 200 x g. The supernatants were discarded, and the pellets resuspended in 10 ml RPMI 1640.

To remove any remaining tissue material, the suspensions were filtered through two layers of sterile gauze into 50 ml tubes. The centrifuge tubes and gauze were rinsed with an additional 10 ml RPMI 1640. Dilutions of the resultant cell suspensions were counted with a hemacytometer to determine the yield of lymphocytes. The prepared cells were kept at room temperature for approximately one hour until ready for use.

Thymus glands obtained from baby mice are the source of the thymocytes which act as the feeder layer for the cell fusions. Thymocytes were prepared from thymus glands obtained from two three- to four-week old Balb/c mice. The thymus glands were rinsed with NS-1 medium (Table 13) and minced on a fine-mesh stainless steel screen with scissors. The minced tissues were rinsed through the screen with 10 ml NS-l medium into a petri dish. The thymus tissue was pressed through the screen with a spatula into the petri dish, and the screen was . washed with 10 ml NS-1 medium. The bottom of the screen was scraped to remove any adhered material and the material was pooled in the petri dish. The cells were transferred to a 50 ml centrifuge tube through two layers of sterile gauze. The petri dish and gauze were rinsed with an additional 10 ml NS-1 medium. The cells were centrifuged for 10 minutes at 200 x g. The supernatants were discarded and the pellets resuspended in 10 ml NS-1 medium. Dilutions of the cell suspensions were counted using a hemacytometer. The yield should be about 400 million cells. The cells were stored at room temperature until ready for use.

TABLE 13 NS-1 Medium For a 500 ml solution: ml ml 5 ml ml ml gm mM non-essential amino acids (GIBCO BRL, Lawrence, MA) 100 mM sodium pyruvate (Irvine, Santa Ana, CA) 200 mM L-glutamine (GIBCO BRL) lOOx Penicillin/Streptomycin/Neomycin (GIBCO BRL) inactivated fetal calf serum (Hyclone, Logan, UT) NaHCO3 lOOx HT Stock 38.5 mg thymidine 136.1 mg hypoxanthine (Sigma, St. Louis, MO.) » Table 13 continued IQOOx A Stock 17.6 ng aminopterin Sterile distilled water was added to the aminopterin to a volume of 50 ml. The aminopterin was dissolved by the drop-wise addition, of 1 N NaOH, and sterile distilled water was added to a final volume of 100 ml. The solution was sterilized by filtration through a 0.22 μιη filter and stored frozen at -20’C. 5Ox HAT ml 5 ml 45 ml 100X HT 1000X A stock distilled water Sterilize the solution by filtration through a 0.22 pm filter. Store frozen at -20'C.

Freezing Medium 7 ml NS-1 medium 2 ml fetal calf serum 1 ml DMSO Mix the ingredients and make fresh for each freezing.

The NS-1 mouse myeloma cell line was used for the fusion. To optimize the fusion procedure, the NS-1 line was cloned out to isolate a clone with a high fusion efficiency. The NS-1 cells were cloned out by limiting dilution into 96-well microtiter plates at an average of five and ten cells per well in NS-1 medium + 2.5 x 105 thymocytes/ml (as prepared above). The plates were incubated at 37’fc with 7% C02 for ten days.

On day ten, the cells were examined microscopically and screened for wells containing single . colonies. On the same day, 100 μΐ of fresh NS-1 medium containing 2.5 x 106 thymocytes was added to the cells. On the fourteenth day, eight of the most vigorously growing single colonies were chosen to expand for fusion.

The eight candidate colonies were transferred to individual 24-well plates containing 1.5-2 ml NS-1 medium + 2.5 x 106 thymocytes/well. The plates were incubated at 37 *C with 7% C02 and the cells were split at appropriate intervals to obtain a sufficient number of cells to transfer to flask culture. Ten million cells from each clone were inoculated into a 75 cm2 tissue culture flasks containing 50 ml NS-1. The flasks were incubated at 37’C, with 7% CO2, until the cells reached a density of at least 5 x 105 cells/ml. The cells were then harvested by centrifugation and diluted to a concentration of approximately 5 x 10® cells/ml with freezing medium (Table 13). The cells were divided into 1 ml aliquots and frozen stepwise, first at -80’C and then at -130’C.

The clones were assayed by quick thawing one vial of each clone in water held at 37’C. The cells were inoculated into flasks containing NS-1 medium to a concentration of 2 x 105 cells/ml. The cells were grown at 37’C with 7% CO2 · The cells were cut back to 2 x 105 cells/ml daily. One day before fusion, two 75 cm flasks containing 50 ml of cell cultere were set up. Each candidate NS-1 clone was mixed with immunized mouse spleen cells and fused as described below. The results of the fusion showed that one clone, designated clone F, had an increased fusion efficiency.

For fusion, 2.5 x 107 NS-1 clone F cells were quickly thawed, as described above, and were added to the prepared immunized mouse spleen and lymph node cells. The mixed cells were centrifuged for 10 minutes at 200 x g, and the supernatant was removed. The cell pellet was resuspended in ljbo μΐ RPMI 1640 and warmed in a 37’C water bath.

*E 912600 . One milliliter of a 50% polyethylene glycol (PEG) solution in RPMI 1640 was adjusted to within the range of pH 7.0 to pH 8.0 using 1% sodium bicarbonate. The PEG solution was added to the cell suspension over a period of one minute with gentle stirring. The solution was stirred for one additional minute. One milliliter of the NS-1 medium was · added over a period of 1 minute with gently stirring. An additional milliliter was added to the susupension over a period of one minute. Eight milliliters of NS-1 media was added over a period of two minutes with gentle stirring and the suspension was then pelleted by centrifugation at 125 x g for 10 minutes at room temperature. The supernatant was discarded, and the cells were .gently resuspended in 25 ml of NS-1 medium.

The cells were transferred into a 175 cm2 flask.

Four hundred million thymocytes (prepared as above) were added to the flask. The volume was adjusted to 160 ml with NS-1 medium, and the mixture was incubated at 37’C with 7% C02 for two to four hours.

After incubation, 3.2 ml of 50x HAT (Table 13) was added. The cell suspension was transferred to eight 96-well plates at 200 μΐ per well., and the plates were incubated at 37’C with 7% CO2. The plates were examined microscopically after three days to determine fusion efficiency with the expectation of approximately five hybridoma colonies per well. The cells were fed after seven days by replacing 100 μΐ of the medium with fresh NS-1 medium containing lx HAT and 2.5 x 106 thymocytes per ml. The hybridomas were tested between day nine and fourteen for the production of specific monoclonal antibodies.

The monoclonal antibodies were tested both for the ability to selectively bind 125I-gamma-carboxylase and for the ability to bind to material which exhibited carboxylase activity, as measured by C02 incorporation into a peptide substrate.

Iodination of gamma-carboxylase was performed using Iodobeads (Pierce, Rockford, IL) and the standard protocol supplied by the manufacturer. One hundred microliters of gamma-carboxylase was added to one Iodobead with 0.5 mCi of 125Iodine, and the mixture was placed on ice for 15 minutes. After incubation, 1.5 ml of TNC/P buffer (Table 13) was added, and the mixture was passed over a G25 column (Pharmacia). Eight fractions were collected with a first fraction volume of 1.5 ml and subsequent fractions volumes of 0.5 ml. All eight fractions were counted on the gamma counter.

TABLE 14 TCN/P 0.1 M 50mM 0.25% 0.25% NaCl Tris, pH7.4 Phosphatidyl Choline Type V E (Sigma) CHAPS (Sigma) 20 RIP buffer 10 ml 2 M Tris-base, pH 8 25 ml 4 M NaCl 5 ml Nonidet P-40 (Sigma) .0 gm Sodium deoxycholate 1.5 gm Sodium Iodine .0 gm BSA fraction V 1% Total volume brought to 1 liter and adjusted to pH 8.

TABLE 14 continued carboxylase reaction mix 950 μΐ 3.8 M Ammonium sulfate 750 Ml 10mM EEL (BOC-Glu-Glu-Leu-OME) (Bachem Bioscience, Philadelphia, PA) 150 Ml 1% CHAPS (C32H58N2O7S) (Sigma) 150 Ml 1% Phosphatidyl Choline Type HIE (Sigma) in 1% sodium cholate Ml 0.2 M DTT (Dithiothreitol) 150 Ml NaH14CC>3 ([5 mCi/2.5 ml] Amersham, Arlington Heights, IL) Ml Vitamin K hydroquinone Urea/SDS Six grams of Urea in a final volume of 10 ml of water was heated in a 37’C water bath until all the urea was dissolved. One milliliter of 20% SDS was added, and the solution was stored at room temperature.

The ability of the monoclonal antibodies to selectively bind 125I-gamma-carboxylase was tested using a radioimmunoprecipitation assay (RIPA). Further identification of the binding complex was made using gel separation. Five microliters of 125I-gamma-carboxylase (prepared as described above) was added to each well of a microtiter plate. Fifty microliters of TNC/P buffer was added to each well, and 1 M sodium iodine was added to each well to achieve a final concentration of 10 mM.

Finally, 50 Ml of each monoclonal hybridoma-conditioned medium was added. The plate was incubated shaking for one hour at 4’C. One microliter of 0.1 mg/ml rabbit antimouse IgG (Cappel Laboratories) which had been diluted in PBS was added to each well, and the plate was incubated shaking at 4’C fbr one hour.

Precleared Pansorbin cells (Calbiochem, La Jolla, CA) were prepared by centrifuging 1 ml of throughly * mixed Pansorbin for 40 seconds at maximum speed in a microfuge at 4’C. The supernatant was removed and the pellet was resuspended in 1 ml RIP buffer. The cells were incubated on ice for 20 minutes. After incubation the cells were pelleted and resuspended, as described above, two more times. The cells were stored at 4’C.

Following, the incubation, 25 pi of precleared Pansorbin Staphylococcus aureus cells in RIP buffer (Table 14) were added to each well. The plates were incubated at 4’C, on a shaker at high speed. The cells were pelleted by centrifuging the plate in a Beckman TJ-6 centrifuge with a TH-4 rotor (Beckman, Carlsbad, CA.) at 2000 rpm at 4’C for 5 minutes, and the supernatants were removed using a bent manifold (Fischer 21-169-10E, Fischer Scientific Group, Santa Clara, CA) . The pellets were resuspended in 150 pi RIP buffer per well. This procedure was repeated two more times.

After the final rinse, each resuspended pellet was transferred to a plastic tube and counted on a Packard Cobra Auto-Gamma counter (Downers Grove, IL).

The monoclonal antibodies were tested for the ability to bind a protein having carboxylase activity. Carboxylase activity was measured as a function of the incorporation of 14C-C02 in a carboxylase reaction. Fifty microliters of hybridoma-conditioned medium from each clone was added to the wells of a 96-well Linbro Titertek multiwell plate (Flow Laboratory). Fifty microliters of bovine microsomes were added to each well. After mixing, the plate was covered with a Costar Serocluster Plate Sealer (Costar, Cambridge, MA) and incubated with gentle shaking for one hour at 4’C.

Ten microliters per well of a 1:10 dilution of 1 mg/ml purified rabbit antimouse IgG (Cappel, West Chester, PA) was added to each well, and the plates were incubated at 4’C for one fiour. Following the incubation, 25 pi of TNC/P-precleared Pansorbin Staphlvococcus aureus cells was added to each well and the plates were incubated for one * hour at 4’C. Precleared cells were prepared as described above except that TNC/P buffer (Table 14) was substituted for RIP buffer. After incubation, the plates were spun in a Beckman Model TJ-6 centrifuge with a TH-4 rotor at 2,000 rpm for five minutes at 4’C. The supernatants were aspirated and the pellets were resuspended in 150 pi of TNC/P buffer. The plates were spun as described above and the supernatants were again aspirated. The pellets were resuspended in 150 pi of TNC/P buffer and spun as described above. The rinses were repeated a total of three times. Following the final rinse the pellet was resuspended in 100 pi per well TNC/P buffer and 133 pi carboxylase reaction mix (Table 14) was added to each sample.

The samples were transferred to microfuge tubes and were incubated from one hour to overnight at room temperature. After incubation, 1 ml of 10% trichloroacetic acid was added to each sample. After incubation, the samples were spun fifteen minutes at maximum speed in an Eppendorf microfuge at 4’C. One milliliter of supernatant from each sample was transferred to scintillation vials, and 1 ml of sterile water was added to each vial. The samples were boiled with one boiling chip each for five minutes, until only several microliters of liquid remained. The vials were cooled to room temperature and 10 ml of Bio-Safe II (Research Products International, Mt. Prospect, IL) was added to each vial. The samples were counted in a Beckman LS-1800 Liquid Scintillation counter.

Wells from the first fusion were screened on the first day for the ability to selectively bind 125I-gammacarboxylase using the RIPA assay described above. Wells found to produce antibodies capable of selectively binding 125I-gamma-carboxylase were screened on the second day to detect wells producing antibodies capable of binding a protein having carboxylase activity in the carboxylase activity assay described above. From the first fusion, ‘ two wells, designated 170.1.1 and 170.3.1, were found to be positive in both assays.

Wells from the second fusion were screened in parallel using both the RIPA assay and the carboxylase activity assay described above. Of the wells, two clones, designated 171.2.1 and 171.4.4, were shown to produce antibodies capable of both selectively binding 125I-gammacarboxylase and binding a protein having carboxylase activity. Approximately sixty wells were identified as producing antibodies capable of binding a protein having carboxylase activity but incapable of binding 125I-gammacarboxylase. Of these latter clones, two, designated 171.5.1 and 171.4.1, were chosen for further use.

The six clones 170.1.1, 170.3.1, 171.2.1, 171.4.1, 171.4.4, and 171.5.1 were serially diluted into 96-well microtiter plates to isolate clones arising from single cells. The clones arising from single cells were re-screened and the clones were expanded. The media from the monoclonal lines were passed over a Protein A column for purification essentially as described in Affinity Chromatography: Principles and Methods (Pharmacia Fine Chemicals, Uppsala, Sweden, 1983, which is incorporated herein by reference). Clone 170.3.1 cells were injected intraperitoneally into mice to obtain ascites fluid which was purified over a protein A column as described above to purify monoclonal antibodies for use in subsequent experiments.

SXAMBI&· 13.

The monoclonal antibodies described above were used individually to screen the bovine lambda gtll cDNA library prepared as described in Example 10 or were combined at a concentration of 1 /xg/ml each to screen the cDNA library. Polyclonal serum from mice immunized as generally described in Example 12 was also utilized to screen the bovine lambda gtll cDNA library.

. One to three million phage from the bovine lambda gtll cDNA library described in Example 10 were absorbed to a limiting number of PLKF' cells. The cells were then plated with £. coli Y1090 cells to obtain 50,000 plaques per 150 mm plate. Alternatively, one to three million phage were plated with PLKF' cells to obtain 50,000 plaques per 150 mm plate. The plates were grown at 37’C or 42 *C for 3.5 hours. Nitrocellulose filters, previously soaked in 10 mM IPTG, were placed on the lawn and the plates were incubated overnight at 37’C or 42’C. The filters were removed from the lawn and a second presoaked nitrocellulose filter was laid on the lawn and incubated at 37’C or 42’C for an additional six hours. The second filter was removed from the lawn.

The duplicate filters were washed two times in TBS (50 mM Tris, pH 8.0, 150 mM NaCl). Following the second wash, the filters were blocked for one hour in TBS + 3% BSA. The blocked filters were transfered to a solution containing 1-2 Mg/ml of antibody (170.3.1.1, 171.2.1, the monoclonal antibody pool or polyclonal serum) in TBS + 3% BSA. The filters were incubated for 4-20 hours at room temperature. Following incubation, excess antibody was removed with two washes with TBS, one wash with TBS + 0.1% NP-40, two additional washes with TBS and a final wash in TBS + 3% BSA. The filters were incubated in TBS + 3% BSA + 125I-Protein A (obtained from Amersham or iodinated using the Iodobead iodination protocol described in Example 13 using Protein A obtained from Sigma) for 4-20 hours at room temperature. c Excess label was removed by two washes in TBS, 1 wash with TBS + 0.1% NP-40 and three washes with TBS. The filters were dried and exposed to X-ray film. Positive plaques were picked for plaque purification.

Although specific embodiments of the invention have been described herein for purposes of illustration, various modifications can be made without deviating from the spirit and scope of the invention. Accordingly, the invention is not limited except as by the appended claims.

( SEQUENCE LISTING (1) GENERAL INFORMATION: (i) APPLICANT: Berkner, Kathy L (ii) TITLE OF INVENTION: GAMMA-CARBOXYLASE AND METHODS OF USE (iii) NUMBER OF SEQUENCES:. 17 (iv) CORRESPONDENCE ADDRESS: (A) ADDRESSEE: Seed and Berry (B) STREET: 6300 Columbia Center, 701 Fifth Avenue (C) CITY: Seattle (D) STATE: WA (E) COUNTRY: USA (F) ZIP: 98104-7092 (v) COMPUTER READABLE FORM: (A) MEDIUM TYPE: Floppy disk (B) COMPUTER: IBM PC compatible (C) OPERATING SYSTEM: PC-DOS/MS-DOS (D) SOFTWARE: Patentln Release #1.0, Version #1.25 (vi) CURRENT APPLICATION DATA: (A) APPLICATION NUMBER: (B) FILING DATE: (C) CLASSIFICATION: (vii) PRIOR APPLICATION DATA: (A) APPLICATION NUMBER: US 07/557,220 (B) FILING DATE: 23-JUL-1990 (viii) ATTORNEY/AGENT INFORMATION: (A) NAME: Maki, David J (B) REGISTRATION NUMBER: 31,392 (C) REFERENCE/DOCKET NUMBER: 990008.544PC (ix) TELECOMMUNICATION INFORMATION: (A) TELEPHONE: 206-622-4900 (B) TELEFAX: 206-682-6031 (2) INFORMATION FOR SEQ ID NO:1: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 21 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (v) FRAGMENT TYPE: (internal « (vi) ORIGINAL SOURCE: (A) ORGANISM: Homo sapiens (xi) SEQUENCE DESCRIPTION: SEQ ID NO:1: Cys Gly Gly His Val Phe Leu Ala Pro Gin Gin Ala Arg Ser Leu Leu 15 10 15 Gin Arg Val Arg Arg 20 (2) INFORMATION FOR SEQ ID N0:2: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 17 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (v) FRAGMENT TYPE: internal (vi) ORIGINAL SOURCE: (A) ORGANISM: Homo sapiens (xi) SEQUENCE DESCRIPTION: SEQ ID N0:2: Cys Gly Gly Lys Asp Lys Leu Asn Asp Asn His Glu Val Glu Asp Glu 15 10 15 Tyr (2) INFORMATION FOR SEQ ID NO:3: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 15 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (v) FRAGMENT TYPE: internal (xi) SEQUENCE DESCRIPTION: SEQ ID N0:3: Phe Thr Leu Leu Ala Pro Thr Ser Pro Gly Asp Thr Thr Pro Lys 15 10 15 (2) INFORMATION FOR SEQ ID N0:4: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 13 amino acids (B) TYPE: aminp acid (D) TOPOLOGY: ’linear (v) FRAGMENT TYPE: internal (xi) SEQUENCE DESCRIPTION: SEQ ID N0:4: Gly Arg Asp Pro Ala Leu Pro Thr Leu Leu Asn Pro Lys 1 5 10 (2) INFORMATION FOR SEQ ID NO:5: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 21 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (v) FRAGMENT TYPE: internal (xi) SEQUENCE DESCRIPTION: SEQ ID N0:5: Asp Asp Arg Gly Pro Ser Gly Gin Gly Gin Gly Gin Gly Gin Phe Leu 15 10 15 lie Gin Gin Val Thr 20 (2) INFORMATION FOR SEQ ID N0:6: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 14 amino acids (B) TYPE: amino acid (D) TOPOLOGY: linear (v) FRAGMENT TYPE: internal (xi) SEQUENCE DESCRIPTION: SEQ ID N0:6: Phe Leu Trp Asp Glu Gly Phe His Gin Leu Val He Gin Arg 1 5 10 (2) INFORMATION FOR SEQ ID NO:7: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 20 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC2575 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: ACATTCAGCA CTGAGTAGAT 20 (2) INFORMATION FOR SEQ ID N0:8: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 20 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS:-single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC2576 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:8: ATCTACTCAG TGCTGAATGT 20 (2) INFORMATION FOR SEQ ID N0:9: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 43 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC2938 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:9: GACAGAGCAC AGATTCACTA GTGAGCTCTT TTTTTTTTTT TTT 43 (2) INFORMATION FOR SEQ ID NO:10: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 42 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC3747 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:10: GACAGAGCAC AGAATTCACT ACTCGAGTTT TTTTTTTTTT TT ‘ (2) INFORMATION FOR SEQ ID NO:11: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 23 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: . (B) CLONE: ZC4135 (Xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: ATAGAATTCT TNGGNGTNGT RTC 23 (2) INFORMATION FOR SEQ ID NO:12: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4136 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: ATTAGAATTC TTYACNYTNY TNGC 24 (2) INFORMATION FOR SEQ ID NO:13: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 14 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4138 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:13: CCNACNWSNC CNGG ‘ (2) INFORMATION FOR SEQ ID NO:14: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4204 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:14: TATAGAATTC GTNACYTGYT GDAT 24 (2) INFORMATION FOR SEQ ID NO:15: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4205 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:15: ATAAGAATTC GAYGAYMGNG GNCC 24 (2) INFORMATION FOR SEQ ID NO:16: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4217 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:16: ATAAGAATTC GGNCCNWSNG GNCA ‘ (2) INFORMATION FOR SEQ ID NO:17: (i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 20 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (vii) IMMEDIATE SOURCE: (B) CLONE: ZC4241 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:17: TGRTGRAANC CYTCRTCCCA 20

Claims (27)

Claims What is claimed is:

1. A proteig composition having gamma-carboxylase activity enriched at least 20,000-fold as compared to liver microsomes.

2. The protein composition of claim 1 wherein said protein is selected from the group consisting of bovine, human and rat gamma-carboxylases.

3. The protein composition of claim 1 wherein said protein is liver gamma-carboxylase.

4. The protein composition of claim 1 wherein said protein is affixed to a solid support.

5. The protein composition of claim 1 wherein said activity is enriched 100,000 - fold as compared to liver microsomes.

6. A protein composition according to claim 1, wherein said protein comprises the amino acid sequence of Peptide 1, Peptide 2, Peptide 3 or Peptide 4 of Table 9 (Sequence ID Nos. 3, 4, 5 and 6).

7. An isolated DNA sequence encoding gammacarboxylase.

8. A DNA sequence according to claim 7 wherein said gamma-carboxylase comprises the amino acid sequence of Peptide 1, Peptide , 2, Peptide 3 or Peptide 4 of Table 9 (Sequence ID Nos. 3, 4, 5 and 6).

9. A DNA sequence according to claim 7 wherein said gamma-carboxylase is selected from the group consisting of bovine, human and rat gamma-carboxylases.

10. A DNA sequence according to claim 7 wherein said gamma-carboxylase is liver gamma-carboxylase.

11. A cDNA sequence according to any of claims 710.

12. A cultured cell transfected or transformed to express a DNA sequence according to any one of claims 7-10.

13. A cultured cell according to claim 12 wherein said gamma-carboxylase comprises the amino acid sequence of Peptide 1, Peptide 2, Peptide 3 or Peptide 4 of Table 9 (Sequence ID Nos 3, 4, 5 and 6).

14. A cultured cell according to claim 12 wherein said cell is a mammalian cell.

15. A cultured cell according to claim 12 wherein said cell is further transfected or transformed to express a DNA sequence encoding a vitamin K-dependent protein.

16. A method of producing a vitamin K-dependent protein, comprising the steps of: culturing a cell transfected or transformed to express a first DNA sequence encoding gamma-carboxylase and a second DNA sequence encoding a vitamin K-dependent protein; and isolating the vitamin K-dependent protein encoded by said second DNA sequence.

17. A method according to claim 16 wherein said gamma-carboxylase comprises the amino acid sequence of Peptide 1, Peptide 2 Peptide 3 or Peptide 4 of Table 9 (Sequence ID Nos. 3, 4, 5 and 6).

18. A method according to claim 16 wherein said cell is a cultured eukaryotic cell.

19. A method according to claim 16 wherein said cell is a cultured mammalian cell. cell is

20. A cultured method according in the presence of to claim 16 wherein 0.1 - 10 Mg/ml vitamin said K.

21. A method according to claim 16 wherein said gamma-carboxylase is selected from the group consisting of human, bovine and rat gamma-carboxylases.

22. A method according to claim 16 wherein said gamma-carboxylase is liver gamma-carboxylase.

23. A protein composition according to claim 1, substantially as hereinbefore described.

24. An isolated DNA sequence according to claim 7, substantially as hereinbefore described.

25. A cultured cell according to claim 12, substantially as hereinbefore described.

26. A method according to claim 16 of producing a vitamin K-dependent protein, substantially as hereinbefore described and exemplified.

27. A vitamin K-dependent protein, whenever produced by a method claimed in a preceding claim.