WO1991002057A1 - Polypeptide production in fungi - Google Patents

Abstract

PAI-2, the main inhibitor of urokinase-type plasminogen activator, is produced with good yield in fungi, especially yeasts such as Saccharomyces cerevisiae, by rDNA techniques. Preferably, the PAI-2 is produced intracellularly as a soluble, correctly-folded product, even though in nature it is secreted.

Description

POLYPEPTIDE PRODUCTION IN FUNGI

The present invention relates to the production of polypeptides, specifically plasiriinogen activator inhibitor 2 (PAI-2), in fungi. PAI-2 is a naturally occurring inhibitor of serine proteases and, more specifically, of plasminogen activators of the urokinase-type (u-PA) and of the tissue-type (t-PA) (see

Astedt et al , 1987 for review). It is a member of a group of structurally and functionally related proteins known as the

SERPIN superfamily (Carrell and Travis, 1985), which includes αl-antitrypsin and plasminogen activator inhibitor 1. Such protease inhibitors act by mimicking the protease's natural substrate and forming a 1:1 covalent inactive complex which is subsequently cleared from the body. The primary determinant of SERPIN specificity is an amino acid at the reactive centre which is analogous to the amino acid immediately amino terminal to the peptide bond cleaved in the natural substrate. In the case of PAI-2 this amino acid is arginine, which is also present at the position at which u-PA and t-PA cleave their natural substrate plasminogen. It is clear , however , that other amino acids in the inhibitor are also important for determining specificity.

PAI-2 has been isolated from placenta (Kawano et al, 1968; Holmburg et al, 1978; Astedt et al, 1985), human monocytes (Golder and Stephens, 1983) and the human monocyte-like histiocytic lymphoma cell line U937 (Vassalli et al, 1984; Kruithof et al , 1986). It is found as an unglycosylated 47kD molecule in placental extracts (Astedt et al , 1985) and in a high molecular weight form (58-70kD) in which the protein is glycosylated to a variable degree (Wohlwend et al , 1987; Ye et al , 1988). The latter is secreted from human monocytes and U937 cells (Wohlwend et al , 1987; Genton et al , 1987) and is the predominant form in pregnancy plasma (Lecander and Astedt, 1986). Although the unglycosylated form is detected in culture medium from cultured U937 cells, it is not clear whether this is due to secretion or as a result of cell lysis (Ye et al , 1988).

Stephens et al (WO 86/01212) disclosed minactivin, a plasminoge activator inhibitor isolated from human monocyte cultures. This protein is PAI-2 and therefore identical to the molecule previously isolated from placenta (Kawano et al , 1968; Holmberg et al , 1978) except that, as the molecular mass was estimated to be 60-70kD, it was probably the glycosylated form of the protein.

Antalis et al (EP 238 275) disclose the production of minactivin by recombinant DNA technology. A minactivin (PAI-2) cDNA coding sequence was introduced into an Escherichia coli expression vector which directed the expression of active minactivin in - E. coll . Webb et al (EP 278 696) used recombinant" DNA technology to produce PAI-2 though they advocated removing 22 amino acids from the N-terminus of the protein to ensure maximal biological activity. These amino acids are, however, present in the active natural molecule and do not constitute a cleavable signal peptide (Ye et al, 1988). The effect of removal of these 22 amino acids from the N-terminus on the activity of PAI-2 has not been examined.

However, PAI-2 is secreted in the normal cells which produce it and therefore, although it lacks a cleavable leader sequence, it must have a sequence which directs secretion. This sequence would have been thought to be effective in fungal cells, at least to the extent of lodging the PAI-2 in the membrane fraction. For example, Livi et al (1988) found that interleukin-1β, which similarly has an internal secretion signal, is at least partially secreted in yeast. Surprisingly, however, it is found that PAI-2 is expressed as an intracellular protein. Moreover, it is obtainable from the soluble fraction when the cells are lysed, unlike when it is expressed in E. coli ( EP-A-238 275 ) . Thus , the undesirable pattern of fungal glycosylation is avoided and the protein can be recovered relatively easily. We have also found the yield to be surprisingly high. In essence, therefore, the present invention provides the production of plasminogen activator inhibitor 2 in fungi such as Saccharomyces cerevlslae. A PAI-2 cDNA or other coding sequence is operationally linked to an effective transcription promoter and transcription terminator in a plasmid which can be maintained in the yeast cells. The PAI-2 protein has been found to be expressed as an intracellular, unglycosylated protein which can relatively easily be recovered from cell extracts by simple purification steps.

Suitable fungal cells include the genera Plchla , Saccharomyces, Kl uyveromyces , Candida, Torul opsis , Hansen ul a ,

Schlzosaccharomyces, Citeromyces, Pachysolen, Debaromyces, Metschunikowia , Rhodosporidium, Leucosporldlum, Botryoascus, Sporidlobolus, Endomycopsis, and the like. Preferred genera are those selected from the group consisting of Plchla , Saccharomyces, Kluyveromyces, Yarrowia and Hansenula , because the ability to manipulate the DNA of these yeasts has, at present, been more highly developed than for the other genera mentioned above.

Examples of Saccharomyces are Saccharomyces cerevlslae, Saccharomyces itallcus and Saccharomyces rouxii .

Examples of Kluyveromyces are Kluyveromyces fragilis and Kluyveromyces lactis . Examples of Hansenula are Hansenula polymorpha , Hansenula anomala and Hansenula capsulata .

Yarrowia llpolytica is an example of a suitable Yarrowia species.

Saccharomyces cerevlslae and Schlzosaccharomyces pombe are particularly preferred. Filamentous fungi such as Asperglllus nlger are also suitable.

Fungal cells can be transformed by:

(a) digestion of the cell walls to produce spheroplasts;

(b) mixing the spheroplasts with transforming DNA (derived from a variety of sources and containing both native and non- native DNA sequences);

(c) regenerating the transformed cells.

The regenerated cells are then screened for the incorporation of the transforming DNA

It has been demonstrated that fungal cells of the genera Pichia, Saccharomyces, Kluyveromyces, Yarrowia and Hansenula can be transformed by enzymatic digestion of the cells walls to give spheroplasts; the spheroplasts are then mixed with the transforming DNA and incubated in the presence of calcium ions and polyethylene glycol, then transformed spheroplasts are regenerated in regeneration medium.

Methods for the transformation of S. cerevlslae are taught generally in EP 251 744, EP 258 067 and WO 90/01063, all of which are incorporated herein by reference.

By "PAI-2" we mean the polypeptide or polypeptides isolated from placentas and the other sources given above and minor variations of such polypeptide(s) which (a) have 80% homology (preferably 85%, 90%, 95% or 99%) with any such polypeptides for the respective corresponding regions of the two molecules, the regions being at least 350 (preferably at least 400) amino acids long, (b) have an arginine residue at the said position of the reactive centre and (c) have a second order rate constant for urokinase-type plasminogen activator inhibition of at least 10⁵ M^-1s^-1 and generally up to 2 × 10⁷ M^-1s^-1, for example about 2 × 10⁶ M^-1s^-1, as measured in the method of Thorsen et al (1988) Eur. J. Blochem. 175, 33-39 (see Table 1 especially). The PAI-2 is unglycosylated. The PAI-2 may be expressed as a fusion with other polypeptides or may be conjugated to other polypeptides or pharmacologically active compounds by known techniques, except that it is preferred for any fusion polypeptide not to be secreted.

Suitable promoters for S. cerevlslae include those associated with the phosphoglycerate kinase (PGK) gene, GAL1 or GAL10 genes, CYC1 , acid phosphatase ( PHO5), TRP1 , ADH1 , ADH2, the genes for glyceraldehyde-3-phosphate dehydrogenase, hexokinase, pyruvate decarboxylase, phosphofructokinase, glucose-6-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, triosephosphate isomerase, phosphoglucose isomerase, glucokinase, α-mating factor pheromone, α-mating factor pheromone, the promoter of EP-A-214 638, hybrid promoters involving hybrids of parts of 5' regulatory regions with parts of other 5' regulatory regions or with upstream activation sites (e.g. the promoter of EP-A-258 067) and the PRB1 promoter. Preferably, the promoter is inducible. The preferred promoters are the promoters of EP-A-258 067 and PRB1 .

The PRBl promoter is described in our co-pending application GB 8927480.7 (our ref. DELF/P7375GB) which is incorporated herein by reference. Example 2 below reproduces the relevant parts thereof for the purposes of this application.

The transcription termination signal is preferably the 3' flanking sequence of a eukaryotic gene which contains proper signals for transcription termination and polyadenylation in fungi. Suitable 3' flanking sequences may, for example, be those of the gene naturally linked to the expression control sequence used, i.e. may correspond to the promoter or, in the case of a hybrid promoter, the downstream part thereof. Alternatively, they may be different. Preferably, the termination signal is that of the S. cerevlslae PGK or ADH1 genes. Preferably, the DNA construct according to the present invention is provided at both ends with synthetic oligonucleotide linkers which allow insertion and cloning of the construct in a cloning vector. The fungal expression control sequence (i.e. effective in fungi), the DNA sequence coding for PAI-2 and the fungal transcription termination signals (i.e. effective in fungi) are operably linked to each other, i.e. they are juxtaposed in such a manner that their normal functions are maintained. Thus, the array is such that the expression control sequence effects proper expression of PAI-2 and the transcription termination signals effect proper termination of transcription and polyadenylation. The junction of these sequences is preferably effected by means of synthetic oligonucleotide linkers which may carry the recognition sequence of an endonuclease. The DNA constructs according to the invention may be prepared by methods known in the art, for example by linking a eukaryotic expression control sequence, a DNA sequence coding for PAI-2 and a DNA sequence containing eukaryotic transcription termination signals in such a way that proper expression of the coding sequence is effected in a yeast host.

The DNA sequence coding for PAI-2 can be isolated from the sources indicated above, or a complementary double-stranded PAI-2 DNA (PAI-2 ds cDNA) is produced from PAI-2 mRNA, or a synthetic gene coding for the amino acid sequence of PAI-2 is produced by means of chemical and enzymatic processes.

Genomic PAI-2 DNA and PAI-2 ds cDNA are obtained, for example, according to methods that are known per se. For example, genomic PAI-2 DNA is obtained from a placenta gene bank that contains the PAI-2 gene by cloning the placenta DNA fragments in a microorganism and identifying clones that contain PAI-2 DNA, for example by colony hybridisation using a radioactively labelled PAI-2 DNA-specific oligonucleotide that contains at least 15, and preferably from 15 to 30, nucleotides. The resulting DNA fragments as a rule contain in addition to the PAI-2 gene other undesired DNA constituents that can be removed by treatment with suitable exo- or endo-nucleases. Double-stranded PAI-2 cDNA can be produced by isolating mRNA from suitable placenta or monocyte cell lines, or U937 which are preferably induced to synthesise PAI-2, enriching the PAI-2 mRNA in the resulting mRNA mixture in a manner known per se, using this mRNA as a template for the preparation of single-stranded cDNA with RNA-dependent DNA polymerase, synthesising from this, with the aid of a DNA-dependent DNA polymerase, ds cDNA, and cloning the latter into a suitable vector. Clones that contain PAI-2 cDNA are identified, for example in the manner described above, by colony hybridisation using a radioactively labelled, PAI-2 DNA-specific oligonucleotide.

The PAI-2 coding sequence can also be produced by chemical synthesis. The process is characterised in that segments of the coding and of the complementary strand of said gene are chemically synthesised and resulting segments are linked enzymatically into a linear coding sequence for PAI-2.

The chemical synthesis of DNA is well-known in the art and makes use of conventional techniques. Appropriate techniques have been compiled by S.A. Narang [ Tetrahedron 39, 3 (1983)]. In particular, the methods described in EP-A-146 785 may be used and are herein incorporated by reference. According to the present invention there is further provided a hybrid vector having one or multiple DNA inserts each comprising a fungal expression control sequence, a DNA segment consisting of a DNA sequence coding for PAI-2 which DNA segment is under transcriptional control of said expression control sequence, and a DNA sequence containing eukaryotic transcription termination signals.

The hybrid vectors according to the invention are hybrid plasmids or linear DNA vectors and are selected depending on the host organism envisaged for transformation.

The invention relates also especially to hybrid plasmids which apart from the expression control sequence, the above DNA segment and the sequence containing transcription termination signals contain additional DNA sequences which are inessential or less important for the function of the promoter, i.e. for the expression of the PAI-2 gene, but which perform important functions, for example in the propagation of the cells transformed with said hybrid plasmids. The additional DNA sequences may be derived from prokaryotic and/or eukaryotic cells and may include chromosomal and/or extra-chromosomal DNA sequences. For example, the additional DNA sequences may stem from (or consist of) plasmid DNA, such as bacterial or eukaryotic plasmid DNA, viral DNA and/or chromosomal DNA, such as bacterial, yeast or higher eukaryotic chromosomal DNA. Preferred hybrid plasmids contain additional DNA sequences derived from bacterial plasmids, especially Escherichia coli plasmid pBR322 or related plasmids, bacteriophage, yeast 2μ plasmid, and/or yeast chromosomal DNA.

In the preferred hybrid plasmids according to the invention, the additional DNA sequences carry a yeast replication origin and a selective genetic marker for yeast. Hybrid plasmids containing a yeast replication origin, e.g. an autonomously replicating segment (ars), are extrachromosomally maintained within the yeast cells after transformation and are autonomously replicated upon mitosis. Hybrid plasmids containing sequences homologous to yeast 2μ plasmid DNA can be used as well. These hybrid plasmids may be integrated by recombination into 2μ plasmids already present within the cell or may replicate autonomously. The integration vectors of EP-A-251 744 or the "disintegration" vectors of EP-A-286 424 may be used.

As to the selective gene marker for yeast, any marker gene can be used which facilitates the selection for transformants due to the phenotypic expression of the marker. Suitable markers for yeast are particularly those expressing antibiotic resistance or, in the case of auxotrophic yeast mutants, genes which complement host lesions. Corresponding genes confer, for example, resistance to the antibiotic cycloheximide or provide for prototrophy in an auxotrophic yeast mutant, for example the URA1 , URA3 , ARG4 , LEU2 , HIS 4 , HIS3, TRP5 or TRP1 gene.

Advantageously, the additional DNA sequences which are present in the hybrid plasmids according to the invention also include a replication origin and a selective genetic marker for a bacterial host, especially Escherlchla coli . There are useful features which are associated with the presence of an E. coli replication origin and an E. coli marker in a yeast hybrid plasmid. Firstly, large amounts of hybrid plasmid DNA can be obtained by growth and amplification in E. coli and, secondly, the construction of hybrid plasmids is conveniently done in E. coli making use of the whole repertoire of cloning technology based on E. coli . E. coli plasmids, such as pBR322 and the like, contain both E. coli replication origin and E. coli genetic markers conferring resistance to antibiotics, for example tetracycline and ampicillin, and are advantageously employed as part of the yeast hybrid vectors.

The additional DNA sequences which contain, for example, replication origin and genetic markers for yeast and a bacterial host (see above) are hereinafter referred to as "vector DNA" which, together with the above DNA construct, containing inter alia the expression control sequence and the PAI-2 gene, is forming a hybrid plasmid according to the invention. The hybrid vectors according to the invention may contain one or multiple DNA inserts each comprising inter alia the expression control sequence and the DNA sequence encoding PAI-2. If the hybrid vectors contain multiple DNA inserts, preferably 2 to 4 DNA inserts, these can be present in a tandem array or at different locations of the hybrid vector. Preferred hybrid vectors contain one DNA insert or DNA inserts in a tandem array. The DNA inserts are especially head to tail arranged.

The hybrid plasmids according to the invention are prepared by methods known in the art. The process for the preparation of the hybrid vectors comprises introducing one or multiple DNA constructs containing a fungal expression control sequence, a DNA segment consisting of a DNA sequence coding for PAI-2 which DNA segment is under transcriptional control of said expression control sequence, and a DNA sequence containing fungal transcription termination signals, as such or introducing the components of said DNA constructs successively in the predetermined order into a vector DNA.

The construction of the hybrid plasmids according to the invention is performed applying conventional ligation techniques. The components of the plasmids are linked through common restriction sites and/or by means of synthetic linker molecules and/or by blunt end ligation. Another aspect of the invention involves fungal host organisms transformed with a hybrid vector having one or multiple DNA inserts each comprising a fungal expression control sequence, a DNA segment consisting of a second DNA sequence coding for PAI-2 which DNA segment is under transcriptional control of said expression control sequence, and a DNA sequence containing fungal transcription termination signals, and mutants thereof.

The host is transformed with a hybrid plasmid having one or multiple DNA inserts each comprising the elements set out above.

The transformation of the host cells is accomplished by methods known in the art. For example, the transformation of yeast with the hybrid vectors may be accomplished according to the method described by Hinnen et al [ Proc. Natl . Acad. Sci . USA 75, 1929 (1978)]. This method can be divided into three steps:

(1) Removal of the yeast cell wall or parts thereof using various preparations of glucosidases, such as snail gut juices (e.g. Glusulase^R or Helicase^R) or enzyme mixtures obtained from microorganisms (e.g. Zymolyase^R) in osmotically stabilized solutions (e.g. 1M sorbitol).

(2) Treatment of the "naked" yeast cells (spheroplasts) with the DNA vector in the presence of PEG (polyethylene-glycol) and Ca²⁺ ions. (3) Regeneration of the cell wall and selection of the transformed cells in a solid layer of agar. This regeneration is conveniently done by embedding the spheroplasts into agar. For example, molten agar (about 50°C) is mixed with the spheroplasts. Upon cooling the solution to yeast growth temperatures (about 30°C), a solid layer is obtained. This agar layer is to prevent rapid diffusion and loss of essential macromolecules from the spheroplasts and thereby facilitates regeneration of the cell wall. However, cell wall regeneration may also be obtained (although at lower efficiency) by plating the spheroplasts onto the surface of preformed agar layers.

Preferably, the regeneration agar is prepared in a way to allow regeneration and selection of transformed cells at the same time. Since yeast genes coding for enzymes of amino acid biosynthetic pathways are generally used as selective markers (-supra ) , the regeneration is preferably performed in yeast minimal medium agar. If very high efficiencies of regeneration are required the following two step procedure is advantageous:

(1) regeneration of the cell wall in a rich complex medium, and

(2) selection of the transformed cells by replica plating the cell layer onto selective agar plates.

When tjie DNA vector is a linear DNA vector used for transforming eukaryotic host cells, transformation is preferably done in the presence of a second vector containing a selective marker for yeast. This cotransformation allows enrichment for those host cells which have taken up DNA that cannot be directly selected for. Since competent cells take up any type of DNA a high percentage of cells transformed with a selective vector will also harbour any additional DNA (such as the above linear DNA vector). The transformed host cells can be improved in production of PAI-2 by mutation and selection using methods known in the art. The mutation can be effected, for example, by U.V. irradiation or suitable chemical reagents. Strains which are deficient in protease A and B are particularly preferred; such strains are generally available.

The host cell may be fermented to express PAI-2 in known ways. The PAI-2 may be purified by known techniques, for example separating off the cells, lysing them, collecting the supernatant, concentrating it and chromatographically separating the PAI-2. Separation by copper chelate chromatography is particularly advantageous and forms a further aspect of the invention.

The PAI-2 (in some cases labelled with radioactive or other labels) may be useful for locating and defining the boundaries of tumours in vitro or in vivo and treating inflammation and tumours, for example colorectal, breast, prostatic, pancreatic and renal cell carcinomas or bladder cancer. Aspects of the invention will now be described by way of example and with reference to the accompanying figures.

Figure 1 shows the PAI-2 DNA and amino acid sequences used.

Figure 2 shows the construction of plasmid pDBP4 ; A=AflII, B=BamH1, Bg=BglII, E=EcoRI , E=HindIII and P=PstI. Only the AflII sites present in the PAI-2 sequence are indicated.

Figure 3 shows an SDS reducing polyacrylamide gel of yeast protein extracts, stained with Coomassie brilliant blue; Lane 1 = S. cerevlslae transformed with pDBP4, Lane 2 = untransformed S. cerevisiae.

Figure 4 shows the construction of plasmid pDBPS1.

Figure 5 is a plasmid map of pAYE333.

Figure 6 describes the two nucleotide substitutions which introduce a HindIII recognition site close to the PRB1 translation initiation codon.

Figures 7 to 10 are respective plasmid maps of pAYE334, pAYE335, pDBP5 and pDBP6.

Figure 11 describes the construction of pDBP1 and pDBP2. Figure 12 is a plasmid map of pDBP7. Only the BglII and AflII sites present in the PAI-2 sequence are shown.

EXAMPLE 1 : PRODUCTION OF PAI-2 WITH A HYBRID PROMOTER

Standard recombinant DNA procedures are as described by Maniatis et al (1982) and the second edition thereof (Sambrook et al, 1989) unless otherwise stated. Construction and analysis of M13 recombinant clones was as described by Messing (1983) and Sanger et al (1977).

Strains - E.coli strains TGI [Delta (lac-pro), supE, thi, hsdD5/F'traD36, proA⁺B⁺, lacI^q, lacZ Delta M15] and DH5α [F-, ɸ80dlacZ Delta Ml5, Delta (lacZYA-argF)U169, recA1, endA1 hsdR17 (r_k - m_k ⁺), supE44, lambda-, thil, gyrA, relA1] were used for propagation of M13 phage and plasmids, respectively, and strain Y1090 [Delta lacU169, proA, Delta Ion, araD139, strA, supF, (trpC22::Tn10) (pMC9)] was used for screening and propagation of the lambda

gtll library. S.cerevisiae strains DB1 (a, leu2), DS569 (a, leu2, pra1) and DM477 (α, leu2, trp1, ura3, pra1, prb1) were used for expression of PAI-2.

A lambda gt11 cDNA library constructed from mRNA isolated from phorbol-12-myristate-13-acetate stimulated cells of the human monocyte-like histiocytic lymphoma cell line U937 (obtained from Clontech Laboratories Inc.) was used as a source of PAI-2 cDNA. The library was screened using radioactively labelled oligonucleotide probes corresponding to the DNA sequence encoding the N-terminus (oligo 1) and complementary to the DNA sequence encoding the C-terminal end (amino acids 400-410) of the PAI-2 protein, respectively (Ye et al . , 1987).

Oligo 1

5'-ATG GAG GAT CTT TGT GTG GCA AAC ACA CTC TTT-3'

Oligo 2

5'-GCC GAA AAA TAA AAT GCA CTT GGT TAT CTT ATG-3'

From the putative positive clones was selected one clone (lambda gtll-186) which appeared to contain the entire PAI-2 coding region. This was confirmed by sequence analysis of the DNA insert in this clone (Fig. 1) following transfer to M13mp19

(Norrander et al , 1983) to form pDBPl (Fig. 2).

To facilitate insertion into expression vectors, restriction enzyme recognition sites were created at the 5' and 3' ends of the PAI-2 gerie. A BglII site was created at the 5' end of the gene using the oligonucleotide primer 5 ' -TGCCACACAAAGATCTTCCATTGTTTCAATCT-3 '

to create a mutation in the third position of the second codon as shown below: -

An A fill site was created at the 3 end of the gene using the oligonucleotide primer

5'-CAGAAGCAGCACGCTTAGTCTTAAGGTGAGGAAATCTGCC-3 '

to create mutations in the third position of the last codon (proline) and in the first base after the stop codon as shown below:-

These two oligonucleόtides were annealed to single stranded pDBP1 and then used in an in vitro mutagenesis procedure (Amersham International plc) carried out according to the manufacturer's recommendations. A clone derived from this procedure and with the correct changes was designated pDBP2 (Fig. 2).

Oligonucleotide linkers were then used to position restriction sites at either end of the gene which are suitable for insertion of the gene into an expression vector. The linker positioned at the 5 end of the gene was

These two linkers were ligated with the BglII-AflII PAI-2 gene fragment from pDBP2 into HindIII + BamHI digested M13mp19 to form pDBP3 (Fig. 2). The BamHl fragment containing the gene was isolated from this plasmid and ligated into pKV50 (GB-A-2 196 635) at BglII to form pDBP4 (Fig. 2). The plasmid ρKV50 contains part of the S. cerevisiae 2 μm plasmid, the S. cerevisiae LEU2 gene as a selectable marker, a hybrid promoter which is injduced in the presence of galactose, and the S. cerevisiae PGK gene transcription terminator.

The plasmid pDBP4 was introduced into S. cerevisiae NY4 (leu2-, pral-) by transformation (Beggs, 1978) and transformants were selected on a minimal medium lacking leucine. Purified transformants were grown in 10ml YEPGal (1% yeast extract, 2% peptone, 2% galactose) shake flask cultures at 30°C, lysed and assayed for the presence of PAI-2 by Western blotting using polyclonal anti-PAI-2 antibody (American Diagnostica, Inc.). A protein of approximately 47kD, which reacted with the anti-PAI-2 antibody, was detected in the soluble fraction of the cells and represented over 5% of the soluble yeast protein (Figure 3). Small quantities of PAI-2 were also found in the culture supernatant but this was likely to be due to cell lysis since equivalent amounts of the intracellular protein enolase were also present.

Purification of recombinant PAI-2 from yeast - Yeast cells were harvested from 6 litres of YEP Gal shake flask culture by centrifugation at 2000xg for 15 min, washed by resuspension in 1 litre of 20mM Na phosphate pH7.0 and repelleted by a second centrifugation. The cells were then resuspended to 50% (w/v) in 200mM Na phosphate pH7.0 and broken with glass beads in a Biospec bead beater (Bartlesville, Oklahoma, USA) in accordance with the manufacturer's instructions. The beads were removed by filtration through a glass sinter funnel and washed with an equal volume of breakage buffer. The filtrate was centrifuged at 15,000xg for 20 min and then the supernatant was adjusted to ρH7.0 with IM HCl and to IM NaCl with solid NaCl. The slight turbidity generated by these additions was removed by vacuumassisted filtration through three layers of Whatman No. 1 filte paper. The filtrate was then loaded onto a 200ml (25 × 8cm) Fast Flow Chelate column (Pharmacia), previously charged with CuSO₄ and equilibrated with 20mM Na phosphate pH7.0, 1M NaCl. The column was extensively washed with the same high salt buffer followed by two column volumes of 20mM Na phosphate pH7.0. PAI-2 activity was eluted with lOmM imidazole, 20mM Na phosphate.

The eluate from the chelate column was adjusted to pH7.6 with 1M HCl and loaded directly onto a 50ml (10 × 2.6cm) Hiload Q Sepharose column (Pharmacia) equilibrated with 20mM Tris-HCl buffer pH7.6. After washing with the same buffer to restore the A₂₈₀ baseline, the column was eluted with a 0-250mM gradient of NaCl.

The anti-urokinase activity eluted in a broad peak towards the end of the gradient. The front of this peak, pool A, consisted of >95% monomeric PAI-2, as determined by densitometric scanning of a Coomassie stained Phast gel (Pharmacia) whereas pool B had a high proportion of dimeric PAI-2. This dimer could be dissociated by reduction. All characterisation work was done on the monomeric material from pool A. Overall, there was a recovery of 33% of the initial uPA inhibitory activity for 2.6% of the initial protein. We have found this purification procedure to be very reproducible. The final material was >95% pure as determined by densitometric scanning of gels and by HPLC analysis.

The purified PAI-2 inhibited uPA with similar reaction kinetics to natural PAI-2.

EXAMPLE 2 : PRODUCTION OF PAI-2 WITH THE PRBl PROMOTER

The structural gene, PRB1 , for the Saccharomyces cerevisiae vacuolar endoprotease protease B has been isolated by Moehle et al . (1987) Genetics 115, 255-263, on two prbl complementing plasmids called MK4 and FP8. When the yeast Saccharomyces cerevisiae is grown on glucose in shake flask culture, very little protease B activity is detected until the cells have catabolised the glucose and are utilising the ethanol accumulated during growth (Saheki, T. and Holzer, H. (1975) Biochem. Biophys. Acta 384, 203-214; Jones et al . (1986) UCLA Symp. Mol. Cell Biol. New Ser. 33, 505-518). This is believed to be a consequence of a transcriptional control mechanism which represses mRNA accumulation until the glucose has been exhausted and the culture enters the diauxic plateau (Moehle et al . (1987) Genetics 115, 255-263). Studies with protease B (prb1- ) deficient mutants implicate protease B in the protein degradation that occurs when vegetative cells are starved of nitrogen and carbon (Wolf, D. and Ehmann, C. (1979) Eur. J. Biochem 98, 375-384; Zubenko, G. and Jones, E. (1981) Genetics 97, 45-64).

The DNA sequence of the PRB1 gene has been reported, as has 150bp of the PRB1 promoter (Moehle et al . (1987) Mol. Cell. Biol. 7, 4390-4399). A more extensive DNA sequence of the PRB1 promoter is also available as an entry in the Genbank database, release 60, accession number M18097, locus YSCPRB1.

The whole of the PRB1 promoter may be used, or a smaller portion thereof, as may readily be determined. For example, the roughly lkbp sequence extending upstream from the start codon to the SnaB1 site is effective.

The transcription termination signal can be the 3' flanking sequence of a eukaryotic gene which contains proper signals for transcription termination and polyadenylation. Suitable 3' flanking sequences may, for example, be those of the gene naturally linked to the expression control sequence. Alternatively, they may be different. Preferably, the termination signal is that of the Saccharomyces cerevisiae ADH1 gene.

The 1.435kbp HindIII-EcoRI DNA fragment containing the protease B promoter was cloned into by the polylinker of the M13 bacteriophage mpl8 (Yanisch-Perron et al . (1985) Gene 33, 103-119), generating plasmid pAYE333 (Figure 5). Plasmid pAYE333 was linearised by partial digestion with SnaB1 and the double standard oligonucleotide linker 3 inserted by ligation.

This generates a NotI restriction site at the 5' end of the protease B promoter. The promoter element was further modified by site directed mutagenesis (oligonucleotide directed in vitro mutagenesis system-Version 2, Amersham) according to the manufacturer's instructions. Mutagenesis with the oligonucleotide

5' -CGCCAATAAAAAAACAAGCTTAACCTAATTC-3 ' introduces a HindIII restriction site close to the ATG translation initiation codon (Figure 6).

Plasmid pAAH5 (Goodey et al . 1987: In Yeast Biotechnology, 401-429, Edited by Berry, D.R., Russell, I. and Stewart, G.G. Published by Allen and Unwin) was linearised by partially digesting with BamHl. The 5' protruding ends were blunt-ended with T4 DNA polymerase and ligated with the double-stranded oligonucleotide Linker 3. A recombinant plasmid pAYE334 (Figure 7) was selected in which a NotI restriction site had replaced the BamHl site at the 3' end of the ADH1 terminator.

The 0.8kbp NotI-HindIII modified protease B promoter sequence was placed upstream of the 0.45kbp HindIII-NotI ADH1 transcription terminator on a pAT153 based plasmid (Twigg and Sherrat (1980) Nature 283, 216-218) to generate pAYE335 (Figure 8).

The large 6.38kbp HindIII-BamHl fragment from the yeast E. coli shuttle vector pJDB207 (Beggs, J.D. 1981 Molecular Genetics in Yeast, Alfred Benzon Symposium 16, 383-395) was treated with the Klenow fragment of E. coli DNA polymerase to create flush ends and ligated with the double stranded oligonucleotide Linker 3 to generate plasmid pDBP5 (Figure 9). The 1.25kbp NotI Protease B promoter/ADH1 terminator cassette from plasmid pAYE335 (Figure 8) was introduced into the unique NotI site of plasmid pDBP5 generating pDBP6 (Figure 10).

Two double stranded oligonucleotide linkers were used to allow the insertion of a human PAI-2 cDΝA into the expression plasmid pDBP6.

Linkers 5 and 6 were ligated with the 1.34kbp BglII-AflII PAI-2 cDNA from pDBP2 (Figure 11) into HindIII linearised pDBP6 generating plasmid pDBP7 (Figure 12).

Stable maintenance of pDBP7 by the S. cerevisiae strains DB1[cir°] and DS569[cir°] can not be accomplished until trans acting functions present on the native 2μ plasmid (Futcher, A.B., (1988) Yeast 4 , 27-40) have been introduced. This was achieved by co-transforming DB1[cir°] and DS569[cir°] with pSAC3 (Chinery, S.A. and Hinchliffe, E. (1989) Current Genetics 16, 21-25, EP286424) and pJDB207 and selecting for transformants on minimal media lacking leucine. Curing of DB1[cir°] pJDB207 and DS569[cir°] pJDB207 of the pJDB207 plasmid results in a cir⁺ derivatives of the original strains.

DB1[cir⁺] and DS569[cir⁺] were re-transformed to leucine prototrophy with plasmid pDBP7 and transformants selected on minimal media lacking leucine. DS569[cir⁺], DS569[cir⁺] pDBP7; DB1[cir⁺] and DB1[cir⁺] pDBP7 were grown for 72 hours in 10ml complex media (1% (w/v) yeast extract; 2% (w/v) bactopeptone and 2% (w/v) sucrose) in shake flask culture at 200rpm, 30°C. Cells were harvested by centrifugation, lysed and the soluble protein extracts analysed by SDS-polyacrylamide gel electrophoresis. An extra protein band of approximately 47Kd is only observed in the soluble protein extracts of those cultures containing the PAI-2 expressing plasmid pDBP7. This extra protein band is of the predicted molecular weight for full length human PAI-2 and was shown by Western blot analysis to react with anti-PAI-2 antibodies, thereby verifying its identity as PAI-2.

Comparative Example

In order to direct PAI-2 to the yeast secretory pathway, an expression plasmid was made in which DNA encoding an N-terminal signal sequence was placed upstream of DNA encoding PAI-2.

Four oligonucleotides were synthesised and annealed to form Linker 7.

This encodes the HSA/α-factor fusion leader described in WO90/01063 followed by the start of the PAI-2 coding sequence up to the BglII site. This linker was ligated with the BglI -BamHl fragment of pDBP3, encoding the remainder of PAI-2, into the BglII site of pKV50 to form pDBPS1 (Figure 4). This plasmid was introduced into S. cerevisiae NY4 by transformation, selecting for complementation of the leu2 mutation. Transformants were grown for 72 hours in YEPGal and the culture supernatant was analysed by Western blotting using anti-PAI-2 antibody. This revealed the presence of immunoreactive material of high molecular weight in the form of a smear characteristic of a glycosylated protein. Some material of a lower molecular weight was also present. This result indicated that, by provision of an N-terminal signal sequence, PAI-2 is directed to the secretory pathway where it was glycosylated in the manner associated with yeast. REFERENCES

Astedt et al (1985) Thromb. Haemost. 53, 122-125.

Astedt et al fl987) Fibrinolysis 1 , 203-208.

Baldari et al (1987) EMBO J. 6 , 229-234.

Barr et al (1988) J. Biol. Chem. 263, 16471-16478.

Bunn et al (1989) Fibrinolysis 3 , , Supp 1., 22.

Carrell and Travis (1985) Trends Biochem. Sci. 10, 20-24.

Genton et al (1987) J. Cell. Biol. 104, 705-712.

Golder and Stephens (1983) Eur. J. Biochem. 136, 517-522.

Holmberg et al (1978) Biochem. Biophys. Acta 544, 128-137

Kawano et al (1968) Nature 217, 253-254.

Kruithof et al (1986) J. Biol. Chem. 261, 11207-11213.

Lecander and Astedt (1986) Br. J. Haematol. 62, 221-228.

Lingappa et al (1979) Nature 281, 117-121.

Livi et al (1988) Yeast 4, S446

Maniatis et al (1982) Molecular Cloning: A laboratory manual.

Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

Meek et al (1982) J. Biol. Chem. 257 , 12245-12251.

Messing (1983) Methods Enzymol. 101, 20-78.

Norrander et al (1983) Gene 26, 101-106.

Sambrook et al (1984 Molecular Cloning : A laboratory manual

( 2nd Edn.) Cold Spring Harbor Laboratory, New York .

Sanger et al (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.

Tabe et al (1984) J. Mol. Biol. 180 , 645-666.

Vassalli et al (1984) J. Exp. Med. 159, 1653-1668. Vlodavsky et al (1987) Proc. Natl. Acad. Sci. USA 84 , 2292-2296. Wohlwend et al (1987) J. Exp. Med. 165, 320-339.

Ye et al , (1987) J. Biol. Chem. 262, 3718-3725.

Ye et al (1988) J. Biol. Chem. 263, 4869-4875.

Claims

1. A process for preparing PAI-2 comprising the fermentation of a fungal cell which is transformed with a genetic construct to express PAI-2, wherein the genetic construct does not comprise an extraneous secretion leader sequence expressible as a fused polypeptide with the PAI-2 and wherein the PAI-2 is not secreted from the fungal cell.

2. A process according to Claim 1 wherein the fungal cell is a yeast.

3. A process according to Claim 2 wherein the fungal cell is Saccharomyces cerevisiae.

4. A process according to any other of the preceding claims wherein the PAI-2 is identical to a naturally-occurring PAI-2, but unglycosylated.

5. A genetic construct for expressing PAI-2 in a fungal cell, the construct comprising a 5 expression regulatory sequence, a coding sequejace for PAI-2 and a 3 regulatory sequence, wherein there is no fungally-effective secretion signal coding sequence interposed between the 5 regulatory sequence and the PAI-2 coding sequence.

6. A genetic construct according to Claim 5 wherein the coding sequence for PAI-2 encodes a naturally-occurring PAI-2.