Vector and method for making tissue factor pathway inhibitor (TFPI) analogues in yeast

Abstract

A method for making TFPI analogues lacking part of the C-terminal end of the native TFPI molecule is described by cultivation of a yeast strain transformed with an expression vector containing a DNA sequence encoding such TFPI analogues. The TFPI analogues will at least contain the two first Kunitz domains and lack the third Kunitz domain.

Description

FIELD OF THE INVENTION

The present invention relates to a method of producing tissue factor pathway inhibitor (TFPI) analogues in yeast.

BACKGROUND OF THE INVENTION

Blood coagulation is a complex process involving many activating and inactivating coagulation factors. Anticoagulant proteins are known to be important for regulation of the coagulation process and anticoagulants are thus important in the treatment of a variety of diseases, e.g. thrombosis, myocardial infarction, disseminated intravascular coagulation etc.

Thus heparin is used clinically to increase the activity of antithrombin III and heparin cofactor II. Antithrombin III is used for the inhibition of factor Xa and thrombin. Hirudin is used for the inhibition of thrombin and protein C may be used for the inhibition of factor V and factor VIII. Anticoagulant proteins may also be used in the treatment of cancer.

Coagulation can be initiated through the extrinsic pathway by the exposure of tissue factor (TF) to the circulating blood (Y. Nemerson, Blood 71 (1988) 1-8). Tissue factor is a protein cofactor for factor VII/VIIa and binding of tissue factor enhances the enzymatic activity of factor VIIa (FVIIa) towards its substrates factor IX and factor X.

Recently a new anticoagulant protein, the tissue factor pathway inhibitor (TFPI) has been isolated (G. J. Broze et al., Proc. Natl. Acad. Sci. 84 (1987) 1886-1890).

On a molar basis TFPI has been shown to be a potent inhibitor of TF/FVIIa-induced coagulation (R. A. Gramzinski et al., Blood 73 (1989) 983-989). TFPI binds and inhibits factor Xa (FXa) and the complex between TFPI and FXa inhibits TF/FVIIa (Rapaport, Blood 73 (1989) 359-365). TFPI is especially interesting as an anticoagulant/antimetastatic agent as many tumor cells express TF activity (T. Sakai et al., J. Biol. Chem. 264 (1989), 9980-9988) and because TFPI shows anti-Xa activity like antistatin which has antimetastatic properties.

TFPI has been recovered by Broze et al. (supra) from HepG2 hepatoma cells (Broze EP A 300988) and the gene for the protein has been cloned (Broze EP A 318451). A schematic diagram over the secondary structure of TFPI is shown in U.S. Pat. No. 5,312,736 (WO 91/01253). The amino acid sequence of TFPI with its natural 28 amino acid signal peptide (Sequence ID Number 1 and 2) is shown in FIG. 1 where the N-terminal amino acid Asp is given the number 1. The protein consists of 276 amino acid residues and has in addition to three inhibitor domains of the Kunitz type, three potential glycosylation sites at position Asn117, Asn167 and Asn228. The molecular weight shows that some of these sites are glycosylated. Furthermore, it has been shown that the second Kunitz domain binds FXa while the first Kunitz domain binds FVIIa/TF (T. J. Girard et al., Nature 338 (1989) 518-520). TFPI has also been isolated from HeLa cells (PCT/DK90/00016) and it was shown that HeLa TFPI binds heparin.

In U.S. Pat. No. 5,312,736 certain TFPI analogues are described retaining the TFPI activity as well as anti Xa activity although parts of the molecule have been deleted. Furthermore, these analogues show a much lower affinity for heparin than full length TFPI, making them more useful as therapeutic agents than the native molecule. The TFPI analogues will furthermore have a longer half life as compared with native TFPI which will further reduce the amount of active ingredients for the medical treatment.

These TFPI analogues are thus characterized in having TFPI activity but with no or low heparin binding capacity under physiological conditions (pH, ionic strength).

In the present context the term "low heparin binding capacity" is intended to mean a binding capacity of about 50%, more preferably of about 25% and most preferably less than about 10% of that of native TFPI at physiological pH and ionic strength.

It was thus shown that the heparin binding capacity is lost when the sequence from amino acid residue number 162 to amino acid residue number 276 is deleted from the TFPI molecule. It was therefore concluded that the heparin binding domain is situated in this part of the TFPI molecule and it was assumed that the heparin binding domain comprises at least a region from Arg246 to Lys265 near the C-terminal end of the TFPI molecule being rich in positively charged amino acid residues.

The present invention is based on the surprising fact that active TFPI analogues lacking C-terminal parts of the molecule are expressed in good yields in yeast. This is the more surprising because attempts to express native TFPI in yeast gave only neglible amounts.

BRIEF DESCRIPTION OF THE INVENTION

In its broadest aspect, the present invention is related to a method of making TFPI analogues in yeast, said TFPI analogues containing at least the first and second Kunitz domain and lacking part of the C-terminal end of the native TFPI molecule by cultivation of a yeast strain transformed with an expression vector containing a DNA-sequence encoding such TFPI analogues in a suitable nutrient medium under conditions which are conductive to the expression of the TFPI.

In a more narrow aspect the present invention is related to a method of making TFPI analogues in yeast, said TFPI analogues containing at least the first and second Kunitz domain and lacking the third Kunitz domain from amino acid Cys189 to amino acid Cys239 as well as a substantial part of the amino acid sequence from Lys240 to Met276. With "a substantial part" is meant from about 70% to 100%. Preferably the TFPI analogues will lack the sequence from at least Cys189 to Lys265. The TFPI analogues may furthermore lack part of the amino acid sequence from amino acid Cys147 to Trp188 and part of the N-terminal sequence of native TFPI such as the sequence from amino acid residue 1 to 24.

The TFPI analogues may also contain a Ser residue as the N-terminal residue for efficient cleavage of a signal peptide by a signal peptidase. Thus, the N-terminal in the TFPI molecule may be replaced by a Ser or an additional Ser may be inserted adjacent to the original N-terminal residue.

The TFPI analogues may furthermore contain a modification of the potential N-glycosylation triade Asn117, Gln118, Thr119 to avoid glycosylation. Such modification will include a mutation and/or a deletion of one or more of the amino acid residues Asn117, Gln118, Thr119; see copending patent application Ser. No. 08/026,146. Asn117 may thus be replaced by Gln which cannot be glycosylated.

The present invention is thus related to a method of expressing TFPI analogues in yeast, said TFPI analogues containing at least the amino acid sequence from Phe25 to Glu148 of the native TFPI molecule and lacking the third Kunitz domain from amino acid Cys189 to amino acid Cys239 and a substantial part of the amino acid sequence from Lys240 to Met276 of the native TFPI molecule.

In a more preferred embodiment, the present invention is related to a method of expressing TFPI analogues in yeast, said TFPI analogues containing the amino acid sequence from Asp1 to Glu148 of the native TFPI molecule and lacking the third Kunitz domain from Cys189 to Cys239 and furthermore lacking a substantial part of the amino acid sequence from Lys240 to Met276 of the native TFPI molecule.

In an even more preferred embodiment the present invention is related to a method of making TFPI analogues in yeast, said TFPI analogues lacking the amino acid sequence from Gln162 to Met276 of the native TFPI molecule in yeast.

In a further aspect, the present invention is related to a recombinant yeast expression vector which comprises DNA sequences encoding functions facilitating gene expression, including a promoter and a terminator, linked functionally to a DNA sequence encoding the TFPI analogues and capable of expressing the TFPI analogues in yeast.

In a still further aspect, the present invention is related to a yeast strain containing a recombinant yeast expression vector as defined above.

DETAILED DESCRIPTION OF THE INVENTION

The cDNA for the native TFPI has been cloned and sequenced (T. C. Wun et al., J. Biol. Chem. 263 (1988) 6001-6004). DNA sequences encoding the desired TFPI analogue may be constructed by altering TFPI cDNA by site-directed mutagenesis using synthetic oligonucleotides encoding the desired amino acid sequence in accordance with well-known procedures.

The DNA sequence encoding the TFPI analogue of the invention may also be prepared synthetically by established standard methods. Thus oligonucleotides may be synthesized, by phosphoamidite chemistry in an automatic DNA synthesizer, purified, annealed, ligated and cloned in suitable vectors.

The yeast expression vector according to the invention may be any vector which may conveniently be subjected to recombinant DNA procedures, and which is capable of expressing the gene for the TFPI analogues in yeast. Thus, the vector may be an autonomously replicating vector, i.e. a vector which exists as an extrachromosomal entity, e.g. a plasmid. Alternatively, the vector may be one which, when introduced into a host cell, is integrated into the host cell genome and replicated together with the chromosome(s) into which it has been integrated.

In the expression vector, the DNA sequence encoding the TFPI analogue will be operably connected to a suitable promoter sequence. The promoter may be any DNA sequence which shows transcriptional activity in the yeast host cell of choice and may be derived from genes encoding proteins either homologous or heterologous to the yeast host cell. Suitable yeast promoters include promoters from yeast glycolytic genes (R. A. Hitzeman et al., J.Biol.Chem. 255 (1980) 12073-12080; T. Alber and G. Kawasaki, J.Mol.Appl.Gen. 1 (1982) 419-434) or alcohol dehydrogenase genes (T. Young et al., in Genetic Engineering of Microorganisms for Chemicals (Hollaender et al., eds.), Plenum Press, New York, 1982 pp. 335-361) or other highly expressed genes. Specific examples are the TPI1 (T. Alber and G. Kawasaki, op. cit., U.S. Pat. No. 4,599,311) or the ILV5 (J. G. L. Petersen and S. Holmberg, Nucl. Acids Res. 14 (1986) 9631-9651) promoter.

The DNA sequence encoding the TFPI analogues should also be operably connected to a suitable terminator sequence which show transcription termination activity in yeast. Such terminator sequences may be derived from the 3' untranslated regions of yeast genes such as TPI1 (Alber and Kawasaki, op. cit.) and ILV5 (J. G. L. Petersen and S. Holmberg, op. cit.). The vector may further comprise elements such as polyadenylation signals, transcriptional enhancer sequences and translational enhancer sequences.

Within the present invention it is preferred to express TFPI analogues in yeast host cells that can secrete the analogues into the culture media. To direct the TFPI analogues into the secretory pathway of the yeast host cell, a secretory signal sequence is operably linked to the TFPI analogue DNA sequence. The secretory signal should preferably be cleaved in vivo, e.g. by a signal peptidase or by the yeast KEX2 protease (D. Julius et al., Cell 37 (1984) 1075-1089) during export of the fusion protein to allow for secretion of the TFPI analogue having the correct N-terminal amino acid. Suitable secretory signals include the alpha factor signal sequence (J. Kurjan and I. Herskowitz, Cell 30 (1982) 933-943, U.S. Pat. No. 4,546,082 and EP 116,201), the PH05 signal peptide (WO 86/00637), secretory signal sequences derived from the BAR1 gene (U.S. Pat. No. 4,613,572 and WO 87/002670), the SUC2 signal peptide (M. Carlson et al., Mol. Cell. Biol. 3 (1983) 439-447) and the human serum albumin prepropeptide (A. Dugaiczyk et al., Proc. Natl. Acad. Sci. USA, 79 (1982) 71-75). Alternatively, a secretory signal sequence may be synthesized according to the rules established, for example, by G. von Heinje (Nucl. Acids Res. 14 (1986) 4683-4690). Examples of synthetic secretory signal sequences are described in WO 89/02463 and WO 92/13065.

Suitable yeast vectors for use in the present invention include YRp7 (K. Struhl et al., Proc. Natl. Acad Sci. USA 76 (1987) 1045-1039), YEp13 (J. R. Broach et al., Gene 8 (1979) 121-133), POT vectors (U.S. Pat. No. 4,931,373), pJDB249 and pJDB219 (J. Beggs, Nature 275 (1978) 104-109) and derivatives thereof. Such vectors will generally include a selectable marker, which may be one of any number of genes that exhibit a dominant phenotype for which a phenotypic assay exists to enable transformants to be selected. Preferred selectable markers are those that complement host cell auxotrophy, provide antibiotic resistance or enable a cell to utilize specific carbon sources, and include LEU2 (Broach et al., op.cit.), URA3 (Botstein et al., Gene 8 (1979) 17), HIS3 (Struhl et al., op.cit.) or POT1 (U.S. Pat. No. 4,931,373).

Techniques for transforming yeast are well known in the literature, and have been described for instance by Beggs (op.cit.). The genotype of the host cell will generally contain a genetic defect that is complemented by the selectable marker present on the expression vector. Choice of a particular host and selectable marker is well within the level of ordinary skill in the art. To optimize production of heterologous proteins, it is preferred that the host strain carry a mutation, such as the yeast pep4 mutation (E. W. Jones, Genetics 85 (1977) 23-33), which results in reduced proteolytic activity.

The recombinant expression vector of the invention may further comprise a DNA sequence enabling the vector to replicate in the host cell in question. An example of such a sequence is the yeast 2-micron sequence.

The procedures used to ligate the DNA sequences coding for the TFPI analogue of the invention, the promoter and the terminator, respectively, and to insert them into suitable vectors containing the information necessary for replication, are well known to persons skilled in the art.

The yeast host cell may be any yeast species which is capable of producing the TFPI analogue. Examples of suitable yeast host cells include strains of Saccharomyces spp., Schizosaccharomyces spp. Kluyveromyces spp., Pichia spp. and Hansenula spp., in particular strains of Saccharomyces cerevisiae.

The transformed yeast cells are grown according to standard methods in a growth medium containing nutrients required for growth of the particular yeast host cells. A variety of suitable media are known in the art and generally include a carbon source, a nitrogen source, essential amino acids, vitamins, minerals and growth factors. The growth medium will generally select for cells containing the DNA construct by, for example, drug selection or deficiency in an essential nutrient which is complemented by the selectable marker on the DNA construct or co-transfected with the DNA construct.

Suitable growth conditions for yeast cells, for example, include culturing in a medium comprising a nitrogen source (e.g. yeast extract or nitrogen-containing salts), inorganic salts, vitamins and essential amino acid supplements as necessary at a temperature between 4.degree. C. and 37.degree. C., with 30.degree. C. being particularly preferred. The pH of the medium is preferably maintained at a pH greater than 2 and less than 8, more preferably pH 5-6.

The TFPI analogue will preferably be secreted to the growth medium and may be recovered from the medium by conventional procedures including separating the host cells from the medium by centrifugation or filtration, precipitating the proteinaceous components of the supernatant or filtrate by means of a salt, e.g. ammonium sulphate, followed by purification by a variety of chromatographic procedures, e.g. ion exchange chromatography, affinity chromatography, or the like.

The novel TFPI analogues may be used for the treatment of patients having coagulation disorders or cancer; see U.S. Pat. No. 5,312,736.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is further described with reference to the drawings in which

FIG. 1A and 1B together show a synthetic gene for human TFPI including the signal peptide and the corresponding amino acid sequence,

FIG. 2 shows DNA-sequences and corresponding amino acid sequences for the prepropeptide of human serum albumin (pp.sub.HSA) (Sequence ID Number 3 and 4) and the synthetic secretion signal 212spx3 (Sequence ID Number 5 and 6) fused to the N-terminal of the mature form of TFPI. Only the three N-terminal amino acids of TFPI are shown,

FIG. 3 shows a restriction site map of yeast POT-expression plasmid pLP-ppTFPI,

FIG. 4 shows a Western analysis of affinity-purified TFPI and TFPI.sub.1-161 secreted from yeast transformants. Lane 1, Molecular weight markers, lanes 2 and 5, full length TFPI.sub.1-276 (0.7 U per lane) secreted by yeast expressing pp.sub.HSA TFPI; lane 3, TFPI secreted from a transfected BHK cell line (0.7 U) (A. H. Pedersen et al., J.Biol.Chem. 265 (1990) 16786-16793); lanes 4 and 6, TFPI.sub.1-161 secreted by transformant expressing pp.sub.HSA TFPI.sub.1-161 (0.7 U). After electrophoresis and blotting onto nitrocellulose filter, immunodetection was performed with monoclonal antibodies, lanes 2-4 with an anti-C-terminal antibody (O. Nordfang et al., Biochemistry 30 (1991) 10371-10376) and lanes 5 and 6 with anti-N-terminal antibody (A. H. Pedersen et al., op.cit.),

FIG. 5 shows an SDS polyacrylamide gel of purified TFPI.sub.1-161. Lane 1, Molecular weight markers; Lane 2, TFPI.sub.1-161 secreted to the growth medium in pilot-scale fermentor by a yeast transformant expressing pp.sub.HSA TFPI.sub.1-161 after initial ion exchange chromatography (200 U); lanes 3 and 4, two preparations of TFPI.sub.1-161 upon further purification by gel filtration and cation exchange chromatography (200 U); lane 5, TFPI isolated from transfected BHK cell line (200 U). The proteins were stained with Coomassie Brilliant Blue R250.

The invention is further described in the following examples which are not in any ways intended to limit the scope or spirit of the invention as claimed.

Experimental Part

Materials and Methods

Standard recombinant DNA techniques were carried out as described (T. Maniatis et al., Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, 1982).

Synthetic oligonucleotides were synthesized by the phosphoramidite method using an Applied Biosystems DNA Synthesizer Model 380B.

Restriction endonucleases and T4 DNA ligase were obtained from New England Biolabs. Modified T7 DNA polymerase (Sequenase) was obtained from United States Biochemicals. Restriction endonucleases and other enzymes were used in accordance with the manufacturers recommendations. pBS+ (Stratagene) was used as cloning vector for construction of the synthetic TFPI gene by cloning of synthetic DNA fragments.

E. coli strains XL-1 Blue (Stratagene) and MC1061 (M. J. Casadaban and S. N. Cohen, J. Mol. Biol. 138 (1980) 179-207) were used as bacterial recipients for plasmid transformations and as hosts for propagation and preparation of plasmid DNA.

Strains of Saccharomyces cerevisiae used as hosts for expression of TFPI analogues were the two diploids E18 (MATa/MAT.alpha. .DELTA.tpi::LEU2/.DELTA.tpi::LEU2 leu2/leu2 +/his4 pep4-3/pep4-3) (U.S. Pat. No. 4,931,373) and YNG452 (MAT.alpha./MAT.alpha. ura3-52/ura3-52 leu2-.DELTA.2/leu2-.DELTA.2 +/his4 pep4-.DELTA.1/pep4-.DELTA.1). The latter was derived from strain JC482 (J. F. Cannon and K. Tatchell, Mol. Cell. Biol. 7 (1987) 2653-2663).

Yeast expression vectors used for expression of TFPI analogues in yeast were of the POT-type (US Patent No. 4,931,373) or the URA3-LEU2d-2.mu. plasmid pAB24 (P. J. Barr et al., in Proc. Alko Symp. on Industrial Yeast Genetics (Korkola and Nevalainen, eds.) Found. Biotech. Industr. Ferment. Res. 5 (1987) 139-148).

DNA sequences were determined by the dideoxy chain termination method (Sanger et al., Proc. Natl. Acad. Sci. 74 (1977) 5463-5467) using double stranded plasmid DNA as template and .sup.32 P- or .sup.35 S labelled primers and Sequenase.

Western blot analysis was carried out as described by J. Mikkelsen and J. Knudsen (Biochem. J. 248 (1987) 709-714).

Affinity purification of the TFPI analogues was carried out from culture supernatants by affinity chromatography using polyclonal anti-TFPI immunoglobulin G coupled to Sepharose (O. Nordfang et al., Biochemistry 30 (1991) 10371-10376).

TFPI activity was measured in a chromogenic microplate assay, modified after the method of Sandset et al., (Thromb. Res. 47 (1987), 389-400). Heat treated plasma pool was used as a standard. This standard is defined as containing 1 U/ml of TFPI activity. Standards and samples were diluted in buffer A (0.05M Tris-HCl, 0.1M NaCl, 0.1M Na-citrate, 0.02% NAN.sub.3, pH 8.0) containing 2 .mu.g/ml polybrene and 0.2% bovine serum albumin. FVIIa/TF/FX/CaCl.sub.2 combination reagent was prepared in buffer A and contained 1.6 ng/ml FVIIa (Novo Nordisk A/S), human tissue factor diluted 60 fold, 50 ng/ml FX (Sigma) and 18 mM CaCl.sub.2. The assay was performed in microplate strips at 37.degree. C. 50 .mu.l of samples and standards were pipetted into the strips and 100 .mu.l combination reagent was added to each well. After 10 minutes incubation, 25 .mu.l of FX (3.2 .mu.g/ml) was added to each well and after another 10 minutes 25 .mu.l of chromogenic substrate for FXa (S2222) was added 10 minutes after the addition of substrate. The reaction was stopped by addition of 50 .mu.l 1.0M citric acid pH 3.0. The microplate was read at 405 nm.

The inhibitory activity of TFPI analogues in the extrinsic pathway of coagulation was measured in PT clotting assay using human plasma and diluted human thromboplastin (O. Nordfang et al., op.cit.).

EXAMPLE 1

Construction of genes encoding TFPI analogues for secretion in yeast

In U.S. Pat. No. 5,312,736, a synthetic gene for human TFPI with its 28 amino acid signal peptide was described. The DNA sequence was derived from the published sequence of a cDNA coding for human TFPI (Wun et al., J. Biol. Chem. 263 (1988) 6001-6004). The synthetic gene was assembled by the step-wise cloning of synthetic restriction fragments into plasmid pBS(+). The resulting gene was contained on a 922 base pair (bp) SalI restriction fragment. The gene had 26 silent nucleotide substitutions in degenerate codons as compared to the cDNA resulting in fourteen unique restriction endonuclease sites in order to facilitate the introduction of mutations in TFPI as well as the in-frame insertion of new secretion signals at the N-terminal of mature TFPI. The DNA sequence of the 922 bp SalI fragment and the corresponding amino acid sequence of human TFPI (pre-form) is shown in FIG. 1.

Using standard DNA manipulation technology, the coding sequences for TFPI analogues were constructed from the synthetic TFPI gene by replacing portions of the TFPI gene with appropriate synthetic DNA fragment. The DNA fragments were annealed oligodeoxynucleotides synthesized by phosphoramidite chemistry. Resulting plasmids were propagated in E. coli and the nucleotide sequences verified by DNA sequencing. The exemplary TFPI analogues constructed in this manner were all characterized by lacking different portions of the C-terminal one third of the TFPI polypeptide, and thus retaining at least the first two Kunitz Domains. Some derivatives had an Asn to Gln substitution at the potential N-glycosylation site at position 117 (Wun et al. op.cit., T. J. Girard et al., Thromb. Res. 55 (1989) 37-50) by changing the codon AAT (Asn) to CAA (Gln)) in order to avoid the addition of N-linked oligosaccharides at this site during expression and secretion in yeast. The following TFPI polypeptides were expressed in yeast:

______________________________________Name Characteristics Polypeptide______________________________________TFPI full length Asp1--Met276TFPI-117Gln full length Asp1--Met276(117Gln)TFPI.sub.1-252 lacking the 24 Asp1--Leu252 C-terminal amino acidsTFPI.sub.1-161 2-domain Asp1-Thr161TFPI.sub.1-161 -117Gln 2-domain, lacking Asp1--Thr161(117Gln) N-glycosylation sitesTFPI.DELTA.3-117Gln 2-domain, internal Asp1--Gly150/ deletion of domain Phe243--Met276-- 3 retaining the (117Gln) C-terminal, lacking N-glycosylation sites______________________________________

The TFPI polypeptides above were expressed in yeast in secretable forms by replacing the coding sequence for the 28 amino acid signal peptide of TFPI (FIG. 1A) with synthetic DNA fragments encoding two different heterologous secretion signals, namely the human serum albumin prepropeptide (pp.sub.HSA) (A. Dugaiczyk et al., Proc. Natl. Acad. Sci. USA, 79 (1982) 71-75) and the synthetic secretion signal 212spx3 (WO 89/02463). The 212spx3 signal consists of the signal peptide of mouse salivary gland .alpha.-amylase with residues Ala3Va14 changed to Phe, and Va17 to Leu, followed by a synthetic leader sequence. Both signals provided at their C-termini a pair of basic amino acids fused to the N-terminal Asp1 of mature TFPI to allow for secretion of correctly processed TFPI by cleavage in vivo by the yeast KEX2 protease (D. Julius et al., Cell 37 (1984) 1075-1089).

The DNA sequences and translated amino acid sequences for the two secretion signals fused to TFPI are shown in FIG. 2.

EXAMPLE 2

Construction of yeast expression plasmids

In order to express the TFPI mutant forms in yeast, the genes were placed between a yeast promoter and a yeast terminator in yeast-E. coli vectors. In the present example, the strong constitutive promoters of the ILV5 (J. G. L. Petersen and S. Holmberg, Nucl. Acids Res. 14 (1986) 9631-9651) or TPI1 (T. Alber and G. Kawasaki, J. Mol. Appl. Genet. 1 (1982) 419-434) genes of S. cerevisiae were used. Transcription terminator sequences were derived from the same two genes. The ILV5 promoter fragment employed was from position -346 to -1 of the published sequence of ILV5 with a BamHI site added at the upstream end via a synthetic linker. The ATG translation start codon of the different signal-TFPI fusion genes were inserted at the putative start codon for the ILV5 gene. The ILV5 terminator fragment used was from position +1172 to +1823 of the published sequence with a SalI site at the upstream position and a poly-linker sequence downstream. The TPI promoter was from position -11 to -379 of the published sequence except that the SphI site in this fragment had been deleted. At the upstream end was added a SphI site, and at position -10 was added en EcoRI site, which could be conveniently used for the insertion of DNA fragments encoding 212spx3-TFPI fusions (see FIG. 2). The TPI1 terminator was from the XbaI site in the 3' end of the TPI1 coding region and 0.7 kb downstream where a BamHI site was inserted.

The TFPI expression cassettes thus constructed were inserted in yeast vectors. In the present example, we used the POT-type (pRPOT, similar to CPOT, U.S. Pat. No. 4,931,373) characterized by carrying the Schizosaccharomyces pombe triosephosphate isomerase gene (POT) (P. R. Russell, Gene 40 (1985) 125-130) as selectable marker for the purpose of plasmid stabilization in tpi1 mutants of S. cerevisiae. The plasmids contained most of the yeast 2-micron plasmid for replication in yeast, and pUC13 sequences for selection and maintenance in E. coli. The POT vector contained in some cases also the defective LEU2d gene derived from plasmid PJDB219 (J. Beggs, Nature 275 (1978) 104-109). This marker was not used for selection in this example. A representative plasmid construct is shown in FIG. 3.

Some TFPI analogues were also expressed from plasmid pAB24 (P. J. Barr et al., in Proc. Alko Symp. on Industr. Yeast Genet. Kornola and Nevalaiken, eds. Found. Biotech. Industr. Ferment. Res. 5 (1982 ) 139-148 ).

EXAMPLE 3

Production of TFPI analogues in yeast

The TFPI expression plasmids with the POT gene as selectable marker were introduced into the diploid S. cerevisiae strain E18. The host strain grows poorly on media with glucose as the carbon source due to the chromosomal deletion of the gene for triose phosphate isomerase. For transformation, the strain was grown in medium with glycerol and lactate as the carbon sources and spheroblasts prepared according to J. Beggs (op.cit.). Transformants were selected in top-agar on minimal plates containing sorbitol and with glucose as the carbon source. In expression studies, reisolated transformants were grown in rich glucose medium (YPD) (F. Sherman, G. R. Fink and J. B. Hicks, Methods in Yeast Genetics, A Laboratory Manual. Cold Spring Harbor Laboratory, 1986) in shake flasks at 30.degree. C. to stationary phase. Following centrifugation, cell pellets and media were stored at -20.degree. C. until further analysis. Expression plasmids based on vector pAB24 were transformed into strain YNG452 selecting for uracil prototrophy on synthetic medium lacking uracil by the alkali cation transformation procedure (H. Ito et al., J. Bacteriol. 153 (1983) 163- 168). Transformants of YNG452 were propagated in synthetic media without uracil.

The activity of TFPI was measured in media supernatants by the chromogenic FXa/TF/FVIIa-dependent two-stage assay. The results are summarized in Table I.

TABLE I______________________________________Levels of secreted TFPI from yeast transformants of strain E18TFPI activity in the culture medium after three days of growthat 30.degree. C. was determined relative to absorbance of the culture at600 nm. Values are the mean of 2-10 independent experiments. Secreted TFPI Secretion TFPI/ activityPromoter signal TFPI analogue (U/ml .multidot. A.sub.600)______________________________________ILV5 pp.sub.HSA TFPI 0.002TPI1 212spx3 TFPI 0.008TPI1 212spx3 TFPI.sub.1-252 0.004ILV5 212spx3 TFPI-117Gln 0.007ILV5 pp.sub.HSA TFPI.sub.1-161 0.45TPI1 212spx3 TFPI.sub.1-161 4.9TPI1 212spx3 TFPI.sub.1-161 -117Gln 4.0TPI1 212spx3 TFPI.DELTA.3-117Gln 0.10 -- -- placebo <0.002______________________________________

As seen in Table I, plasmids for full length TFPI gave only low levels of activity in the culture medium with both secretion sequences. Full length TFPI with a substitution of one of the three potential sites for N-linked glycosylation (Asn117 to Gln) also showed very low levels of secretion.

Expressing of a TFPI variant, TFPI.sub.1-252 lacking the C-terminal basic region of TFPI gave rise to about the same low levels of secreted activity as observed with full length TFPI.

Expression of TFPI.DELTA.3-117Gln, which lacks the third Kunitz domain and potential N-glycosylation sites gave more than 10 times higher yields compared with full length TFPI.

About a 200-fold increase in secreted activity relative to the full length protein was observed when TFPI.sub.1-161 was expressed in yeast using the ILV5 promoter and the prepropeptide of HSA. With the TPI1 promoter and signal sequence 212spx3, a further 10-fold increase was obtained. Substitution of the single potential N-glycosylation site in TFPI.sub.1-161 (resulting in TFPI.sub.1-161 -117Gln) did not affect the level of secreted activity.

From these results it can be concluded that full length and near to full length TFPI containing the third Kunitz domain and a substantial part of the sequence from amino acid Lys240 to Met276 of native TFPI were expressed only poorly in yeast as measured by secreted activity, while surprisingly TFPI analogues lacking the third Kunitz domain and a substantial part of the sequence Lys240 to Met276 were secreted at significantly higher levels.

EXAMPLE 4

Anticoagulant properties of TFPI analogues from yeast

In order to measure the anticoagulant activities of TFPI and TFPI.sub.1-161 secreted from yeast, the polypeptides were partly purified by affinity-chromatography using a polyclonal antibody towards TFPI coupled to Sepharose and their anticoagulant activity determined in traditional prothrombin time (PT) coagulation assays. Anticoagulant units were normalized to chromogenic TFPI activity which was assumed to be similar to the different TFPI forms. In this assay, TFPI in human plasma used as standard was defined to have a relative anticoagulant activity of 1. The results are shown in Table II. A relative anticoagulant activity of 0.18 was determined for full length TFPI produced in yeast. A comparable activity was found for TFPI with the Asn117 to Gln substitution. These activities are 5-9 times lower than for human, high-activity TFPI produced in BHK cells, but similar to C-terminally fragmented TFPI prepared as described by O. Nordfang et al. (Biochemistry 30 (1991) 10371-10376) and may be due to partial C-terminal fragmentation. In contrast, about 60-fold lower anticoagulant activities were observed for the two-domain polypeptides TFPI.sub.1-161 and TFPI.sub.1-161 -117Gln as compared to intact TFPI from BHK cells.

TABLE II______________________________________Anticoagulant activity of TFPI variantsTFPI polypeptides secreted from yeast transformants werepurified by affinity chromatography and their anticoagulantactivity measured in a PT clotting assay relative toFXa/TF/FVIIa-dependent TFPI activity. TFPI Anticoagulant Relative activity activity anticoagulantTFPI molecule (U/ml) (units/ml) activity______________________________________TFPI 8.3 1.5 0.18TFPI-117Gln 14.0 4.3 0.31TFPI.sub.1-161 102.0 2.8 0.027TFPI.sub.1-161 -117Gln 99.0 2.6 0.026From BHK cells:TFPI 10.0 15.8 1.6Heterogenous C-term. 100.0 10.9 0.11______________________________________

EXAMPLE 5

Western analysis of secreted and intracellular TFPI and TFPI.sub.1-161

The analysis of secreted TFPI and TFPI analogues in Table I depended on activity measurements of the culture medium, and did not necessarily reflect the molar amounts of polypeptides being secreted. Thus, the low levels of full length TFPI could be a result of mainly inactive product being secreted, e.g. as a result of proteolytic degradation or improper folding. Therefore, the expression of TFPI and TFPI.sub.1-161 was analyzed by Western blotting using different antibodies towards TFPI (FIG. 4). The samples of supernatant media or cell extracts were partially purified by anti-TFPI affinity chromatography before SDS polyacrylamide gel electrophoresis and Western blot analysis. Expression of full length TFPI (lanes 2 and 5) gave rise to a predominant protein band with an apparent molecular mass of 50 kDa. The polypeptide reacted both with antibodies against the N- and C-terminal regions suggesting that the polypeptide was full length. The lower mobility as compared to TFPI expressed in transfected BHK cells (lane 3) suggested that more extensively glycosylation takes place in yeast. TFPI.sub.1-161 reacted as expected only with the anti-N antibody (lanes 4 and 6). The major immunoreactive species appeared as a broad band of about 25 kDa.

Comparable staining intensities were obtained in Westerns with TFPI from yeast or BHK cells, and with TFPI.sub.1-161 from yeast (FIG. 4). Since the same amounts of activity were analyzed (0.7 U per lane) it appears as the three preparations have similar specific activities.

A similar analysis by Western blotting was carried out with cell extracts of transformants expressing TFPI and TFPI.sub.1-161 fused to the prepropeptide of HSA. The cell extract of the full length TFPI transformant upon partial immuno-affinity purification showed 2-3 bands of 34-39 kDa which could represent full length, underglycosylated TFPI, while the TFPI.sub.1-161 transformant showed a major band of 25 kDa and a minor band of 22 kDa, probably representing glycosylated and unglycosylated TFPI.sub.1-161, respectively.

Pilot-scale purification of TFPI.sub.1-161

In order to obtain a large quantity of homogeneously pure TFPI.sub.1-161, a yeast transformant expressing the analogue fused to the prepropeptide of HSA was propagated in a pilot-scale fermentor to high cell density. The resulting fermentation medium after removal of the cells by centrifugation contained about 650 U/ml of TFPI activity. The purification scheme for TFPI.sub.1-161 from this batch is summarized in Table III. At each step, fractions with high TFPI activity and low protein content were pooled for further purification.

TABLE III______________________________________Purification of TFPI.sub.1-161Yeast transformants of strain YNG452 expressing TFP.sub.1-161fused to the prepropeptide of HSA was grown in a 2000 lfermentor and secreted TFPI.sub.1-161 purified from thecleared supernatant medium. Total Volume TFPI activity YieldPurification step (1) (10.sup.6 U) %______________________________________Fermentation medium 1050 680 100Cation exchange (pH 3.0) 73 92 13FF-Q anion exchange 7.8 73 11(pH 9.3)Freeze-drying and gel 1.2 78 11filtration on Superdex 75(pH 3.4)S-Sepharose cation exchange 1.4 57 8(pH 3.4)Precipitation by -- 49 7 -isopropanol, dissolutionin H.sub.2 O, freeze-drying______________________________________

About 5.times.10.sup.7 U were isolated corresponding to a purification yield of about 7%. The rather low yield was mainly caused by loss of material during the first step of purification involving cation exchange chromatography.

SDS-PAGE analysis of the purified TFPI.sub.1-161 showed a single protein band of 25 kDa after staining with Coomasie Brilliant Blue (FIG. 5, lanes 3 and 4). After the initial cation and anion exchange chromatography several additional polypeptides were present (lane 2). Among these, the major 25 kDa polypeptide and the 14 kDa polypeptide reacted with antibodies against the N-terminal region of TFPI (data not shown). This result suggested that the 14 kDa polypeptide is a proteolytic break-down product of TFPI.sub.1-161, possibly consisting of the first Kunitz domain.

As was the case for TFPI.sub.1-161 produced in COS-7 cells (copending patent application Ser. No. 07/828,90), TFPI.sub.1-161 produced in yeast differed from full length TFPI in its binding characteristics towards heparin. Whereas full length TFPI binds strongly to heparin-Sepharose with about 1M NaCl required for effective elution, TFPI.sub.1-161 did not bind to heparin under physiological conditions (0.1M NaCl, 50 mM Tris-HCl, 0.1% BSA, pH 7.4).

Based on weighing of the freeze-dried preparation of purified TFPI.sub.1-161 from yeast, a specific activity of 23,000 U/mg in the chromogenic FXa/TF/FVIIa-dependent assay was determined. The activity agrees well with the estimated 30,000 and 10,000 U/mg reported for TFPI isolated from transfected BHK cells (A. H. Pedersen et al., J. Biol. Chem. 265 (1990) 16786-16793) and from the HepG2 hepatoma cell line (Broze, G. J., Jr. et al., Thromb. Res 48 (1987) 253-259).

The anticoagulant activity of purified TFPI.sub.1-161 was further analyzed by its ability to inhibit the activity of tissue factor (TF) in direct PT clotting assay where TFPI.sub.1-161 was added to the plasma without preincubation with TF. It was found that a concentration of 65 .mu.g/ml of TFPI.sub.1-161 inhibited 90% of the TF activity. This inhibition was obtained for both high and low concentrations of the TF preparation (thromboplastin). With full length TFPI from transfected BHK cells only 0.5 .mu.g/ml was required for 90% inhibition of TF activity.

__________________________________________________________________________SEQUENCE LISTING(1) GENERAL INFORMATION:(iii) NUMBER OF SEQUENCES: 6(2) INFORMATION FOR SEQ ID NO:1:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 928 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: double(D) TOPOLOGY: linear(ii) MOLECULE TYPE: DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO (vi) ORIGINAL SOURCE:(A) ORGANISM: Synthetic(ix) FEATURE:(A) NAME/KEY: CDS(B) LOCATION: 8..919(ix) FEATURE:(A) NAME/KEY: sigpeptide(B) LOCATION: 8..91(ix) FEATURE:(A) NAME/KEY: matpeptide(B) LOCATION: 92..919(xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:GTCGACCATGATT TACACAATGAAGAAAGTACATGCACTTTGGGCTAGC49MetIleTyrThrMetLysLysValHisAlaLeuTrpAlaSer28-25-20-15GTATGCC TGCTGCTTAATCTTGCCCCTGCCCCTCTTAATGCTGATTCT97ValCysLeuLeuLeuAsnLeuAlaProAlaProLeuAsnAlaAspSer10-51GAGGAA GATGAAGAACACACAATTATCACAGATACGGAGCTCCCACCA145GluGluAspGluGluHisThrIleIleThrAspThrGluLeuProPro51015CTGAAACTTAT GCATTCATTTTGTGCATTCAAGGCGGATGATGGGCCC193LeuLysLeuMetHisSerPheCysAlaPheLysAlaAspAspGlyPro202530TGTAAAGCAATCATGAAAA GATTTTTCTTCAATATTTTCACTCGACAG241CysLysAlaIleMetLysArgPhePhePheAsnIlePheThrArgGln35404550TGCGAAGAATTTATA TATGGGGGATGTGAAGGAAATCAGAATCGATTT289CysGluGluPheIleTyrGlyGlyCysGluGlyAsnGlnAsnArgPhe556065GAAAGTCTGGAAGAG TGCAAAAAAATGTGTACAAGAGATAATGCAAAC337GluSerLeuGluGluCysLysLysMetCysThrArgAspAsnAlaAsn707580AGGATTATAAAGACAAC ACTGCAGCAAGAAAAGCCAGATTTCTGCTTT385ArgIleIleLysThrThrLeuGlnGlnGluLysProAspPheCysPhe859095TTGGAAGAGGATCCTGGAATAT GTCGAGGTTATATTACCAGGTATTTT433LeuGluGluAspProGlyIleCysArgGlyTyrIleThrArgTyrPhe100105110TATAACAATCAGACAAAACAGTGTGAAAGG TTCAAGTATGGTGGATGC481TyrAsnAsnGlnThrLysGlnCysGluArgPheLysTyrGlyGlyCys115120125130CTGGGCAATATGAACAATTTTGAGACA CTCGAGGAATGCAAGAACATT529LeuGlyAsnMetAsnAsnPheGluThrLeuGluGluCysLysAsnIle135140145TGTGAAGATGGTCCGAATGGTTTCCA GGTGGATAATTATGGTACCCAG577CysGluAspGlyProAsnGlyPheGlnValAspAsnTyrGlyThrGln150155160CTCAATGCTGTTAACAACTCCCTGACTC CGCAATCAACCAAGGTTCCC625LeuAsnAlaValAsnAsnSerLeuThrProGlnSerThrLysValPro165170175AGCCTTTTTGAATTCCACGGTCCCTCATGGTGT CTCACTCCAGCAGAT673SerLeuPheGluPheHisGlyProSerTrpCysLeuThrProAlaAsp180185190AGAGGATTGTGTCGTGCCAATGAGAACAGATTCTACTACAAT TCAGTC721ArgGlyLeuCysArgAlaAsnGluAsnArgPheTyrTyrAsnSerVal195200205210ATTGGGAAATGCCGCCCATTTAAGTACTCCGGATGTGG GGGAAATGAA769IleGlyLysCysArgProPheLysTyrSerGlyCysGlyGlyAsnGlu215220225AACAATTTTACTAGTAAACAAGAATGTCTGAGGGCAT GCAAAAAAGGT817AsnAsnPheThrSerLysGlnGluCysLeuArgAlaCysLysLysGly230235240TTCATCCAAAGAATATCAAAAGGAGGCCTAATTAAAACC AAAAGAAAA865PheIleGlnArgIleSerLysGlyGlyLeuIleLysThrLysArgLys245250255AGAAAGAAGCAGAGAGTGAAAATAGCATATGAAGAAATTTTTGTT AAA913ArgLysLysGlnArgValLysIleAlaTyrGluGluIlePheValLys260265270AATATGTGAGTCGAC9 28AsnMet275(2) INFORMATION FOR SEQ ID NO:2:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 304 amino acids(B) TYPE: amino acid(D) TOPOLOGY: linear(ii) MOLECULE TYPE: protein(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:MetIleTyrThrMetLysLysValHisAlaLeuTrpAlaSerValCys28 -25-20- 15LeuLeuLeuAsnLeuAlaProAlaProLeuAsnAlaAspSerGluGlu10-51AspGluGluHisThrIleIleT hrAspThrGluLeuProProLeuLys5101520LeuMetHisSerPheCysAlaPheLysAlaAspAspGlyProCysLys25 3035AlaIleMetLysArgPhePhePheAsnIlePheThrArgGlnCysGlu404550GluPheIleTyrGlyGlyCysGluGlyAsnGlnAsn ArgPheGluSer556065LeuGluGluCysLysLysMetCysThrArgAspAsnAlaAsnArgIle707580IleLysThrTh rLeuGlnGlnGluLysProAspPheCysPheLeuGlu859095100GluAspProGlyIleCysArgGlyTyrIleThrArgTyrPheTyrAsn10 5110115AsnGlnThrLysGlnCysGluArgPheLysTyrGlyGlyCysLeuGly120125130AsnMetAsnAsnPheGluThrLeuG luGluCysLysAsnIleCysGlu135140145AspGlyProAsnGlyPheGlnValAspAsnTyrGlyThrGlnLeuAsn150155160 AlaValAsnAsnSerLeuThrProGlnSerThrLysValProSerLeu165170175180PheGluPheHisGlyProSerTrpCysLeuThrProAlaAspArgGly 185190195LeuCysArgAlaAsnGluAsnArgPheTyrTyrAsnSerValIleGly200205210LysCysArgProP heLysTyrSerGlyCysGlyGlyAsnGluAsnAsn215220225PheThrSerLysGlnGluCysLeuArgAlaCysLysLysGlyPheIle230235 240GlnArgIleSerLysGlyGlyLeuIleLysThrLysArgLysArgLys245250255260LysGlnArgValLysIleAlaTyrGluGluIlePheVal LysAsnMet265270275(2) INFORMATION FOR SEQ ID NO:3:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 81 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: double(D) TOPOLOGY: linear(ii) MOLECULE TYPE: DNA(iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO(vi) ORIGINAL SOURCE:(A) ORGANISM: Synthetic(ix) FEATURE:(A) NAME/KEY: CDS(B) LOCATION: 1..81(ix) FEATURE:(A) NAME/KEY: miscfeature(B) LOCATION: 1..54(ix) FEATURE:(A) NAME/KEY: miscfeature(B) LOCATION: 55..72(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:ATGA AGTGGGTTACTTTCATCTCTTTGTTGTTCTTGTTCTCTTCTGCT48MetLysTrpValThrPheIleSerLeuLeuPheLeuPheSerSerAla151015TAC TCTAGAGGTGTTTTCAGGAGGGATTCTGAG81TyrSerArgGlyValPheArgArgAspSerGlu2025(2) INFORMATION FOR SEQ ID NO:4:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 27 amino acids (B) TYPE: amino acid(D) TOPOLOGY: linear(ii) MOLECULE TYPE: protein(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4:MetLysTrpValThrPheIleSerLeuLeuPheLeuPheSerSerAla151015TyrSerArg GlyValPheArgArgAspSerGlu2025(2) INFORMATION FOR SEQ ID NO:5:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 231 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: double(D) TOPOLOGY: linear(ii) MOLECULE TYPE: DNA(iii) HYPOTHETICAL: NO( iv) ANTI-SENSE: NO(vi) ORIGINAL SOURCE:(A) ORGANISM: Synthetic(ix) FEATURE:(A) NAME/KEY: CDS(B) LOCATION: 76..231(ix) FEATURE:(A) NAME/KEY: miscfeature(B) LOCATION: 76..222(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5:GAATTCATTCAAGAATAGTTCAAACAAGAAGATTACAAACTATCAATTTC ATACACAATA60TAAACGATTAAAAGAATGAAGGCTGTTTTCTTGGTTTTGTCCTTGATCGGA111MetLysAlaValPheLeuValLeuSerLeuIleGly15 10TTCTGCTGGGCCCAACCAGTCACTGGCGATGAATCATCTGTTGAGATT159PheCysTrpAlaGlnProValThrGlyAspGluSerSerValGluIle1520 25CCGGAAGAGTCTCTGATCATCGCTGAAAACACCACTTTGGCTAACGTC207ProGluGluSerLeuIleIleAlaGluAsnThrThrLeuAlaAsnVal303540GCCATGGCTAAGAGAGATTCTGAG231AlaMetAlaLysArgAspSerGlu4550(2) INFORMATION FOR SEQ ID NO:6:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 52 amino acids(B) TYPE: amino acid (D) TOPOLOGY: linear(ii) MOLECULE TYPE: protein(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6:MetLysAlaValPheLeuValLeuSerLeuIleGlyPheCysTrpAla151015GlnProValThrGlyAspGlu SerSerValGluIleProGluGluSer202530LeuIleIleAlaGluAsnThrThrLeuAlaAsnValAlaMetAlaLys3540 45ArgAspSerGlu50

Claims

1. A recombinant vector capable of directing the expression of a polypeptide analog of Tissue Factor Pathway Inhibitor (TFPI) in yeast, said analog having the amino acid sequence of native TFPI from Asp.sup.1 to Thr.sup.161 ;

said vector comprising a promoter, a transcription initiation signal, and a terminator operably associated with a nucleic acid sequence which encodes a translation product consisting of a secretion-directing signal peptide concatenated to said TFPI analog.

2. A vector according to claim 1 wherein the encoded signal peptide is selected from the group consisting of the human serum albumin prepropeptide (pp.sub.HSA) and the synthetic signal sequence 212spx3.

3. A recombinant vector capable of directing the expression of a polypeptide analog of Tissue Factor Pathway Inhibitor (TFPI) in yeast,

said vector comprising a promoter, a transcription initiation signal, and a terminator operably associated with a nucleic acid sequence which encodes a translation product consisting of a secretion-directing signal peptide concatenated to said TFPI analog;

the amino acid sequence of said analog differing from the sequence of naturally occurring TFPI by the deletion of the native TFPI sequence from Gln.sup.162 to Met.sup.276 and by one or more of the modifications selected from the group consisting of

the deletion of one or more additional regions of the native TFPI sequence,

the mutation or deletion of one or more of the residues in the Asn.sup.117 -Gln.sup.118 -Thr.sup.119 triad such that the polypeptide is incapable of being glycosylated at this site in a yeast expression system, and

the addition of a Ser residue at the N-terminal of the analog polypeptide; provided that said analog comprises at least the first and second Kunitz domains of native TFPI.

4. A vector according to claim 3 wherein the encoded signal peptide is selected from the group consisting of the human serum albumin prepropeptide (pp.sub.HSA) and the synthetic sequence 212spx3.

5. A vector according to claim 3 wherein the residue of said analog corresponding to Asn.sup.117 of native TFPI has been replaced by another amino acid residue.

6. A vector according to claim 5 wherein said residue corresponding to Asn.sup.117 of native TFPI has been replaced by Gln.

7. A yeast cell transformed with the expression vector of any one of claims 1 to 6.

8. A method of making a TFPI analog, comprising

culturing the yeast cell of claim 7 under conditions suitable for the expression and secretion of said analog; and

recovering said TFPI analog.