AU676132B2

AU676132B2 - Method of using bar1 for secreting foreign proteins

Info

Publication number: AU676132B2
Application number: AU74003/91A
Authority: AU
Inventors: Vivian L. Mackay
Original assignee: Zymogenetics Inc
Current assignee: Zymogenetics Inc
Priority date: 1985-10-25
Filing date: 1991-04-02
Publication date: 1997-03-06
Anticipated expiration: 2006-10-20
Also published as: DK320287D0; JP2523562B2; SK776486A3; UA41863C2; SK279041B6; CA1316133C; FI872801A0; EP0243465A1; HUT43624A; CZ284251B6; IE63822B1; AU6543286A; CN86107554A; JPS63501614A; IE862804L; DK320287A; WO1987002670A1; HU206897B; AU7400391A; FI872801A

Description

COMMONWEALTH OF AUSTRALIA PATENTS ACT 1952 COMPLETE SPECIFICATION (Original) FOR OFFICE USE Class Int. Class Application Number: Lodged: Complete Specification Lodged: Accepted: Published: Priority: Related Art: Name of Applicant: ZYMOGENETICS, INC.

Address of Applicant: 4225 Roosevelt Way, N.E., 6" Seattle, Washington 98105, UNITED STATES OF AMERICA Actual Inventor(s): Vivian L. MacKAY Address for Service: DAVIES COLLISON, Patent Attorneys, 1 Little Collins Street, Melbourne, 3000.

Complete Specification for the invention entitled: "Method of Using Barl for Secreting Foreign Proteins" The following statement is a full description of this invention, including the best method of performing it known to us: -1- -la- S" METHOD OF USING BAR1 FOR SECRETING FOREIGN PROTEINS *ee *f C* eggo 9 Ce« The present invention is directed to novel DNA constructs containing at least the translated signal 5 peptide portion of the Saccharomyces cerevisiae B&1 gene and at least one structural gene foreign to a host cell transformed with said construct. Transformation of host organisms by such constructs will result in expression of a primary translation product "*10 comprising the structural protein encoded by the foreign'gene fused to the signal peptide of BAR~ so that the protein is processed through the host cell secretory pathway and may be secreted from the host cell into the culture medium or the periplasmic space.

Various procaryotic and eucaryotic microorganisms have been utilized as hosts for production of heterologous polypeptides, polypeptides which are not naturally produced by the host, by way of recombinant DNA methodology. Various eucaryotic fungal species are of particular interest, including Saccharomyces cerevisiae, Schizosaccharomyces pombe, Asperqillus and Neurospora. In particular, much work has been done in -2the budding yeast S. cerevisiae. Yeast cells, when transformed with a suitable DNA construct, such as a plasmid, have been made to express heterologous genes contained in the plasmid. However, a major limitation to this technology is that, in many cases, the protein products are not secreted into the medium by the host cells and it is thus necessary to disrupt the cells and purify the desired protein from the various contaminating cellular components without denaturing .10 or inactivating it. Thus, it is desirable to be able to direct the transformed cells to secrete the heterologous product which would simplify purification of that product. Additionally, it may be desirable for some proteins to enter a host cell secretory pathway to facilitate proper processing, i.e., disulfide bond formation.

cerevisiae is known to secrete some of its naturally produced proteins, although knowledge of the process is quite limited compared to what is known 20 about secretion of proteins from bacteria and mammalian cells. It appears that most of the secreted yeast proteins are enzymes which remain in the periplasmic space, although the enzymes invertase and acid phosphatase may also be incorporated into the cell wall. The proteins which are known to be secreted into the culture medium by S. cerevisiae include the mating pheromones (a-factor and a-factor), killer toxin, and the protein accounting for Barrier activity (hereinafter "Barrier"). Secretion through the cell wall into the medium is also referred to as "export".

g. cerevisiae mating type a cells produce a-factor which is secreted into the culture medium, whereas A cells produce two secreted polypeptides, A-factor and Barrier. The gene for a-factor has been cloned, sequenced, and analyzed (See Kurjan and Herskowitz, -3- Cell 30: 933-943 (1982)). The signal peptide (a short peptide sequence believed to direct the cell to secrete an attached protein), leader sequence (comprising a precursor polypeptide that is cleaved from mature a-factor) and non-translated gene sequences (including promoter and regulatory regions) from the a-factor gene may be used to direct the secretion of foreign proteins produced in yeast.

(Brake et al., Proc. Nati Acad. Sci. USA 81:4642-4646, 1984) Expression of the a-factor gene S: is regulated by the Tcal gene product and processing of the a-factor precursor into the mature protein appears to require at least two steps, believed to be under control of the STE13 and KE genes.

In contrast to a-factor, Barrier appears to be glycosylated, based on its ability to bind concanavalin A. Barrier is produced by a cells and expression appears to be under the control of the MATa2 gene. With the exception of possible signal r 20 peptide cleavage, no processing of the Barrier precursor has heretofore been demonstrated and the STE13 and KEX2 genes are not believed to be involved in the expression of Barrier.

Due to the above differences between the control of expression and processing of a-factor and Barrier, one cannot determine A priori which of these yeast genes is better suited to direct the secretion of a particular foreign protein. Because the two proteins appear to be processed through different secretory pathways, it would be desirable to exploit the characteristics of the Barrier secretion uystem.

It is therefore an object of the present invention to provide DNA constructs containing a segment of the P:\O1i'ERMS\74003-91.CLM 15/1096 -4- EBAR gene which codes for at least the signal peptide and also containing a foreign structural gene which results in expression in a microbial host of a heterologous protein which is processed through the secretory pathway of the host cell.

It is a further object of the present invention to provide a method for expressing foreign genes in a microbial host which results in a heterologous protein or a portion thereof encoded by sich a gene being processed through the secretory pathway of the host cell.

In a first aspect, the present invention provides a DNA construct comprising a portion of the Saccharomyces cerevisiae BAR1 gene comprising the signal peptide coding sequence thereof, and at least one structural gene foreign to a fungal host cell transformed with said 15 construct, said construct including a processing site at the junction of said BAR1 gene portion and said foreign structural gene, and a promoter which controls the expression, in said fungal host cell transformed with said construct, of a fusion polypeptide or protein *r resulting from said portion of the BAR1 gene and said structural gene.

In another aspect, this invention also provides a method for producing a heterologous protein in a transformed fungal cell and directing said protein into the secretory pathway of the cell comprising the steps of: transforming a fungal host cell with a DNA construct comprising a portion of the S. cerevisiae BAR1 gene comprising at least the signal peptide coding sequence thereof, and a structural gene encoding said heterologous protein, said construct including a processing site at the junction of said BAR1 gene portion and said structural gene, and a promoter which controls the expression in said fungal cell of a fusion protein comprising said signal peptide and said heterologous protein; growing said fungal cell from step under growth conditions -7 1) suitable to select for the production of said fusion protein.

organis- of fus ion prote-in- corprinirg the- SARI sjinal poptd-e and the foreign prot-an.- In the accompanying drawings: FIG. 1 shows the nucleotide sequence of the BAR1 gene and the derived amino acid sequence of the primary translation product. The ATa2 binding site is underlined, the putative signal peptide cleavage site S" is indicated with an arrow, and potential glycosylation sites are marked with asterisks.

1 0 FIG. 2 is a diagram of the plasmid pZV9; S* FIG. 3 illustrates the construction of plasmid p254; FIG. 4 illustrates the construction of the plasmids pZV30, pZV31, pZV32 and pZV33; FIGS. 5A and 5B illustrate the construction of the 15 plasmid pZVSO; FIG. 6 illustrates the construction of the plasmid 115; FIG. 7 illustrates the construction of the plasmid pZV49; FIG. 8 illustrates the construction of the plasmid pZV134 comprising the TP promoter; FIG. 9 illustrates the subcloning of a portion of the al gene; FIG. 10 illustrates the construction of plasmid /u 1 7 i if -6- FIG. 11 illustrates the construction of plasmids comprising the TP11 promoter and BARI-Mlal fusions; FIG. 12 illustrates the construction of plasmid pSW22; FIG. 13 illustrates the construction of plasmids comprising BARl-MFal fusions; FIG. 14 illustrates the construction of pZV100; l .10 FIG. 15 illustrates the construction of plasmid SpZV102; *0* 00 FIG. 16 illustrates the construction of the plasmid 0 3.0 pSW96; FIG. 17 illustrates the construction of the plasmid pSW97; and 0 FIG. 18 illustrates the construction of plasmids pSW98 and pSW99. A indicates the mutation at codon As used herein the term "DNA construct" means any DNA molecule, including a plasmid, which has been modified by human intervention in a manner such that the nucleotide sequences in the molecule are not identical to a sequence which is produced naturally. The term "DNA construct" also includes clones of DNA molecules which have been so modified. The terms "expression vector" and "expression plasmid" are defined as a DNA construct which includes a site of transcription initiation and at least one structural gene coding for a protein of interest which is to be expressed in a host organism. An expression vector will usually also contain appropriate regions such as a promoter and terminator which direct the expression of the protein -7of interest in the host organism and an origin of replication. Expression vectors according to the present invention will also usually. contain a selectable marker, such as a gene for antibiotic resistance or a nutritional marker.

The term "DNA construct" will also be considered to include portions of the expression vector integrated into the host chromosome.

The term "plasmid" will have its commonly accepted 10 meaning, an autonomously replicating DNA construct, usually close-looped.

*e* *0 The term "signal peptide" refers to that portion of a primary translation product which directs that product into the secretory pathway of the cell which produces 15 it. The signal peptide is usually cleaved f.-om the remainder of the nascent polypeptide by a signal peptidase during this process. A signal peptide is characterized by the presence of a core of hydrophobic amino acids, occurs at the amino terminus of the primary translation product, and is generally from about 17 to 25 amino acides in length. Signal peptidase cleavage sites have been characterized by von Heinje (Eur. J. Biochem. 133:17, 1983). As used herein, the term "signal peptide" may also refer to functional portions of the naturally occurring signal peptide.

The present invention provides a method for directing transformed cells to direct heterologous proteins through a secretory pathway accomplished by transforming a host with a DNA construct containing a foreign gene linked to the yeast PA-R gene or a portion thereof which codes at least for the BAR1 -8signal peptide. The proteins so processed may be secreted into the periplasmic space or the culture medium. The yeast BAR1 gene codes for Barrier activity which is believed to be a glycosylated protein secreted by cerevisiae a cells. The secreted Barrier allows mating type A cells to overcome the Gl arrest induced by a-factor. It is believed that Barrier may be a protease. (See Manney, J.

Bacteriol., 155: 291-301, 3983). Transcription of the 10 BAR1 gene is stimulated by a-factor. Barrier, or an analogous activity, is not detected in a or a/a cells and the BAR1 gene is not transcribed in these cell types.

The sequence of 2750 base pairs encompassing the BAR1 gene is shown in Figure 1, together with the derived amino acid sequence of the primary translation product. The ATG translation initiation site of BA1 is at position 681 of the approximately 2.75 kb fragment shown in Figure 1 which was subcloned from a 20 fragment obtained from a yeast genomic library (Nasmyth and Tatchell, CeUll 1: 753-764, 1980). An open reading frame starts with the ATG codon at +1 and extends 1761 bp in the 3' direction. The sequence of S" the first 24 amino acids of the BR1 primary translation product appears to be similar to sequences of known yeast and mammalian signal peptides. Thus, the alanine at position 24 may be used as a cleavage site, as in yeast invertase and acid phosphatase.

Cleavage could also occur after amino acid 23. At least nine potential asparagine-linked glycosylation sites exist in the primary translation product, although the extent of glycosylation of the mature secreted Barrier is not yet known. The promoter and regulatory regions of the BM1 gene are located within a region of approximately 680 bp on the 5' side of the -9translation initiation codon. Full promoter function and response to a-factor stimulation have been localized to the ATG-proximal approximately 680 bp of the 5' untranslated region.

The DNA constructs of the present invention will preferably encode a cleavage site at the junction of the Barrier and foreign protein portions. A preferred such site is a KEX2 cleavage site, a sequence of amino acids which is recognized and cleaved by the product 10 of the S. cerevisiae KEX2 gene (Julius ej A. Cell 32:1075-1080, 1984). A EX2 site is characterized by a pair of basic amino acids, such as lysine and arginine. It is preferred that the sequence of the EX2 site be Lys-Arg or Arg-Arg. The BAR1 primary translation product contains two such pairs located in the structural region: Arg-Arg at positions 177-178 and Lys-Lys at positions 404-405. As noted above, the KEX2 gene is not involved in the processing of the Barrier precursor protein, suggesting that the 20 potential processing sites are blocked by protein conformation or glycosylation and further suggesting that Barrier may normally be processed through a pathway different than that used by such KEX-processed proteins as a-factor. However, applicants have found that by including a KEX2 processing site in a fusion protein comprising a portion of the BAB gene primary translation product comprising the signal peptide thereof, together with a protein of interest, the fusion protein is cleaved at the KE2 site, resulting in secretion of the protein of interest. It has also been found that by reducing the efficiency of signal peptidase cleavage of the Barrier portion of a fusion protein comprising a EXi cleavage site, enhanced levels of the protein of interest are exported. The E1IE cleavage site may be provided by the BARW sequence or the gene of interest, or may be introduced into the fusion by linker addition, site-specific mutagenesis, etc.

Thus, according to the present invention, a portion of the BAR1 gene comprising the ATG initiation codon and signal peptide coding sequence thereof may be joined to a foreign gene of interest and transformed into a eukaryotic host cell. The resultant fusion gene will include a processing site, preferably a XEX2 cleavage 10 site, at the junction of the BAR1 and foreign sequences. Such a construct may also comprise regulatory regions and the promoter from the 5' non-coding region of the BAR1 gene, or may comprise regulatory regions and/or promoters from other genes. In addition to the promoter from the BAR1 gene, other promoters which may be used include the promoters from the g. cerevisiae alcohol dehydrogenase I or alcohol '"A4 dehydrogenase II genes, the genes of the A. cerevisiae '2 glycolytic pathway, such as the TPI1 promoter, and 20 corresponding genes from other species, including the fission yeast Schizosaccharomvces pombe (Russell and Hall, Z. Biol. Chem. 258: 143-149, 1983 and Russell, Nature 301: 167-169, 1983). The P. cerevisiae alcohol dehydrogenase I gene has been described by Ammerer (Methods in Enzymolocv 1I1: 192-201, 1983). The alcohol dehydrogenase II gene has been described by Russell et al. Biol. Chem. 258: 2674-2682, 1983).

Glycolytic genes of 2. cerevisiae have been described by Kawasaki (Ph. D. Thesis, University of Washington, 1979) Hitzeman et al. Biol. Chem. 12073- 12080), 1980), Kawasaki and Fraenkel (Bio:hemn.

Biophvs. es. Comm. =i9: 1107-1112, 1982) and Alber and Kawasaki Mol. ApnL. Genet. 1: 419-434, 1982).

-11eS 9.

6060 6 6690 *560

S

@6e 66 5* In a preferred embodiment, the signal peptide coding sequence of the BARi gene is altered to reduce the efficiency of signal peptidase cleavage of the Barrier portion of a fusion protein comprising a EEX2 cleavage site. This may be accomplished by site-specific mutagenesis of potential cleavage sites, preferably those sites at the amino acid 23-24 (of the Barrier protein sequence) juncture or at the amino acid 24-25 juncture.

10 Methods used to form DNA constructs according to the present invention involve conventional techniques.

The structural BAR1 gene, or a portion thereof, and the structural gene to be expressed will preferably be under control of a single promoter. Methods of ligation of DNA fragments are amply described and are well within the ability of those with ordinary skill in the art to perform. The DNA coding sequence of the protein to be expressed may be essentially that of any protein, particularly proteins of commercial 20 importance, such as interferons, insulin, proinsulin,* a-l-antitrypsin, growth factors, and tissue plasminogen activator.

4 96SS S S 05 6 *561 0 After preparation of the DNA construct containing the BAR1 gene or a portion thereof and the structural gene to be expressed, the construct will be transformed into the host organism under transforming conditions.

Techniques for transforming prokaryotes and eukaryotes (including mammalian cells) are known in the literature.

Preferably, the host organism will be a strain of the budding yeast Saccharomyces cerevisiae; however, other fungi, including the fission yeast Schizosaccharomvces -12too* too* d:96 0604 0 pornbe and the filamentous fungi Aspercrillus nidulans and Neurospora spp. may also be used.

The following examples are giveni by way of example and not by way of limitation. Unless otherwise indicated, standard molecular- biology methods were used throughout. Restriction endonucleases were obtained from Bethesda Research Laboratories, New England BinLabs and Boehringer-Mannheim Biochemicals and were used as directed by the manufacturer, generally with the 10 addition of pancreatic R.Nase (10 ug/ml) to digests.

T4 DNA ligase was obtained from Bethesda Research Laboratories or Boehringer-Mannheim. and was used as directed. M13 and pUC host strains and vectors were obtained from Bethesda Research Laboratories.

M13 cloning was done as described by Messing, Methods in nzynolocry jQ0, 20-77 (1983). DNA polymerase I (Klenow fragment) was used as described in Haniatis et al., Molecular Cloning: Laboratory Manual, Cold Spring Harbor Laboratory (1982). coli cultures were transformed by the method of Bolivar et al.0 Gene 2_1 95-113 (1977). cerevisiae cultures were transformed by the method of Beggs, Nature 275, 104-109 (1978) as modified by MacKay, Methods inl Enzvmnolocrv IQ01, 325 (1983). S. pombe was transformed as described by Russell, Natu7re 20i, 167-169 (1983). The mating pheromone a-factor was prepared by the method of Duntze et Al., Eur. BichPe,. 357-365, 1973, as modified by Manney pt Al. g. Cell Dil 96*0 1592- 1600, 1983 or was purchased from Sigma Chemical Co.

Oligonucleotides were synthesized on an Applied Biosystems Model 380A DNA synthesizer and purified by polyacrylamide gel electrophoresis on denaturing gels.

:.@of 6,60 *0 fe St -13- LHETHQJ2 Assay for Barrier Activity

S

0

S

0595 p.

4.

o The assay used for detection of Barrier production by transformed yeast cells relies on the ability of Barrier to reverse the inhibition of growth of sensitive A cells exposed -to a-factor. The test strain is one that is abnormally sensitive to a-factor, such as strain RC629 (MATa barl), since it produces no Barrier activity. A lawn is prepared using such a strain in a soft agar overlay on an agar plate. A sufficient quantity of a-factor (0.05 0.1 unit, as assayed by 10 Manney, ibid.) is added to the overlay to inhibit growth of the cells. Transformants to be screened for Barrier production are then spotted onto the lawn.

Secretion of Barrier by the transformed cells reverses the a-factor growth inhibition immediately surrounding 15 the spot, thereby allowing the sensitive cells to recover. The recovered cells are observed as a fringe of growth around the normally smooth edge of the colony of transformed cells. The presence of this fringe indicates that the plasmid in the transformed strain directs the expression and secretion of Barrier.

sbs S S is aw S S 5, 5 IRI and IRC Assays IRI and IRC assays were performed using commercial kits obtained from Novo Industri, Bagsvaerd, Denmark.

Guinea pig anti-porcine insulin and guinea pig antihuman C-peptide antibodies are supplied with the kits.

-14- ASSAY FOR IRI p1 sample in NaFAM (0.04 M4 phosphate buffer pH 7.4 containing bovine serum albumin) pl antibody (stock diluted 1:30) 16-24 hours at V C 50 l 125I-insulin (diluted 1:100) 2 hours at 4 C pl, It Staphylococcus aureus in NaFAM min. at 0' C 3.4Wash twice with 1% BSA/TNEN* 0 Centrifuge and count pellets The different steps can be conveniently done in microtiter dishes, until the pellets are transferred to the scintillation vials.

ASSAY FOR IRC s* 50 pl sample in NaFAM 641 0 A1 antibody (stock diluted 1:50) 16-24 hours at 46 C pl125 A "C peptide (stock diluted 2-4 hours at 46 C pl 1% Aureus in NaFAM 45 min. 00 C Wash twice 1% BSA/TNEN* Centrifuge and count pellets *TI;EX is 20 mM Trio pH 100 MM NaCl 1 mM EDTA ASSAY FOR ALPHA-FACTOR ACTIVITY The assay used for detection of a-factor export by transformed yeast cells employs the ability of a-factor to inhibit the growth of sensitive A cells.

The test strain (such as S. cerevisiae strain RC629 (MATa barl)) contains a mutation in the BAR1 gene that prevents the production of Barrier activity and renders the cells super-sensitive to a-factor. A lawn of the test strain is made in a soft agar overlay over a plate of standard yeast selective synthetic medium 10 medium lacking leucine). Transformants to be screened for a-factor export are spotted onto the lawn and incubated at 30'C. Secretion of a-factor by the transformants will cause growth inhibition in the lawn immediately surrounding the colony. The halo of growth inhibition in the lawn of test cells indicates that the colony is exporting active a-factor. A comparison of halo size enables one to estimate the relative quantities of a-factor exported by each transformant.

EXAMPLE 1 Expression of Proinsulin in S. cerevisiae Using The DBAR Gene A recombinant plasmid pool comprising the entire yeast genome was constructed (Nasmyth and Tatchell, Cgll 19: 753-764, 1980) using the shuttle vector YEpl3 (Broach et kl., Gene k: 121-133, 1979). Yeast DNA fragments, produced via a partial Sau 3A digestion, were inserted into Bam HI-digested YEpl3. The plasmid pool was used to transform 2. cerevisiae strain XP635-10C (MATa leu2-3 leu2-112 barl-1 gUi: ATCC #20679) and transformants were selected for leucine prototrophy and -16growth on a concentration of a-factor that is inhibitory to the A barl cells. Resultant colonies were then screened for the ability to secrete Barrier activity. Two colonies were found which carried both leucine independence and the ability to secrete Barrier. These colonies carried the plasmids designated pBAR2 and pBAR3.

Plasmid DNA isolated from those two transformants was o.u used to transform E. coli strain RRI (ATCC #31343).

Transformants were selected for ampicillin resistance.

Poe lasmids pBAR2 and pBAR3 were purified from the coli transformants and characterized by restriction endonuclease digestion and electrophoresis on agarose or acrylamide gels. Plasmid pBAR2 was shown to contain an insert of approximately 9.2 kilobases. E.

coli RRI transformed with plasmid pBAR2 has been deposited with ATCC under accession number 39410.

Subcloning showed that the pBAR3 plasmid insert comprised a portion of the insert of pBAR2, but oriented in the vector in the opposite direction.

Further subcloning and screening for Barrier secretion localized the functional BAR1 gene sequence to a region of approximately 2.75 kb. This fragment comprises the coding sequence, nontranslated transcribed sequences, promoter, regulatory regions, transcription terminator, and flanking chromosomal sequences.

The plasmid pBAR2 was digested with restriction endonucleases Hind III and Xho I and a fragment of approximately 3 kb was purified by electrophoresis through an agarose gel. This fragment was inserted into plasmid pUC13 which had been digested with Hind III and Sal I. The resulting recombinant plasmid, desig- -17nated pZV9 (FIG. can be used to transform E. coli, but lacks the necessary origin of replication and selectable marker for a yeast vector. The plasmid pZV9, in a transformant strain of E. gcol RRI has been deposited with ATCC under accession no. 53283.

For the BARI gene to be used to direct the secretion of proinsulin, fragments of the BARI gene comprising the 5' regulatory region and a portion of the coding sequence were used. The fusion of the BAR1 and proinsulin gene fragments was made in the proper reading frame and at a point in the BARl sequence at which the resulting fusion polypeptide may be cleaved, preferably in vivo. Several sites in the fA1. gene are potential cleavage sites; the Arg-Arg at position 177-178 was selected as the test site of fusion with proinsulin. Accordingly, the 5' regulatory sequences and approximately 800 bp of the coding, sequence of BAR were purified from plasmid pZV9 as a 1.9 kb Hind III-Sal I fragment.

Referring to FIG. 3, there is shown a method for subcloning human preproinsulin cDNA. A human preproinsulin cDNA (preBCA clone), p27, is produced by G-tailing Pst I digested pBR327 and inserting C-tailed DNA, made by reverse transcribing total RNA from human pancreas. Plasmid pBR327 is described by Soberon _t A1., Gene 2: 287-305 (1980) and the sequence of human preproinsulin is reported by Bell g Nature 122: 525-527 (1979). The complete translated sequence was cut out as a Nco I-Hga I fragment. The protruding ends were filled in with DNA-polymerase I (Klenow fragment), and synthetic Eco RI linkers (GGAATTCC) and Xba I linkers (CTCTAGAG) were attached simultaneously.

The fragment was subcloned into pUC13 (Vieira and Messing, Gene 19: 259-268, 1982; and Messing, Meth. 1 n -18- Enzvmoloy 191: 20-77, 1983) which had been cut with Eco RI and Xba I. Because the addition of an Eco RI linker at the 5' end restores the Nco I site (CCATGG) at the initiation codon, plasmids were screened for the presence of Eco RI, Nco I and Xba I sites flanking a 340 bp insert. A plasmid having these properties is termed p47, shown in FIG. 3. A proinsulin (BCA) fragment with a blunt 5' end was generated by primer repair synthesis (Lawn ee al, Nuc. Acids Res. 1: 10 6103-6114, 1981) of plasmid p47. Subsequent eas digestions with Xba I yielded a 270 bp fragment which s" was inserted into pUC12. *(Vieira and Messing, ibid., and Messing, ibid). The vector was prepared by cutting with Hind III, blunting the ends with DNA polymerase I (Klenow fragment), cutting with Xba I, and gel purifying. The resultant vector fragment, comprising a blunt end and a Xba I sticky end, was ligated to the above described BCA fragment. As mature BCA starts with the amino acid phenylalanine (codon TTT), blunt end ligation of the two fragments regenerates a Hind III site at the junction. Plasmids were screened first for the recovery of the Hind III site and then by sequencing across the junction using an M13 sequencing primer. Plasmid p254 had the correct sequence.

Referring to FIG. 4, plasmid p254 was digested with Hind III and Eco RI and the ca. 270 bp proinsulin fragment was gel purified. The fragment ends were blunted using DNA polymerase I (Xlenow fragment) and deoxynucleotide triphosphates. Sal I linker sequences (GGTCGACC) were treated with T4 polynucleotide kinase and 3 2 P-ATP and were ligated to the blunted proinsulin fragment. Digestion with Sal I and Bam HI, followed by electrophoresis on a 1.5% agarose gel, -19yielded a proinsulin fragment with Sal I and Bam HI cohesive ends.

The proinsulin fragment and the 1.9 kb BAR, fragment were ligated together into pUC13 which had been digested with Hind III and Bam HI. This construct was used to transform coli K12 (JM83).

Transformed cells were screened for ampicillin resistance and production of white colonies. Further screening by restriction endonuclease digestion using Hind III, Bam HI, and Sal I identified a plasmid (pZV27) containing a Hind III-Bam HI fragment of the proper size and a single Sal I site.

In order to link the first amino acid of proinsulin to the Arg-Arg potential processing site of the BAE1 gene product, the intervening material in the BAR1proinsulin fusion was deleted. A synthetic oligonucleotide was used to direct the looping-out of this extraneous material in the following manner. Referring to FIG. 4, plasmid pZV27 was digested with Hind .20 III and Bam HI and the ca. 2.2 kb BARk-proinsulin fusion fragment was gel purified. This fragment was then inserted into the replicative form of the phage vector M13mpll (Messing, Meth. in Enzymolov 19Af: 20-77, 1983) which had been digested with Hind III and Bam HI. This recombinant DNA was used to transfect coli K12 (JM103) (Messing, ibid.). White plaques were picked and the replicative forms of the recombinant phage were screened for the correct restriction patterns by double enzyme digestions using Hind III Sal I and Sal I Bamr HI. A construct showing the desired pattern is known as mpll-ZV29. The oligonucleotide primer (sequence: 3' GGATCTTCTAAACACTTG was labelled using y-32P-ATP and T4 polynucleotide kinase. 7.5 pmol of kinased primer was then combined with 60 ng of M13 sec:.encing primer (Bethesda Research Laboratories, Inc.) This mixture was annealed to 2 jig of single stranded znpll-ZV29 and the second strand was extended using T4 DNA ligase and DNA polyrnerase I (Kienow fragment), as described 1 r oligonucleotidedirected mutagenesis (two primer method) by Zoller li al. (Manual for Advanced Technicnies in Molecular Cloning Course, Cold Spring Harbor Laboratory, 1983).

DNA prepared in this way was used to transfect Z. cQ.2J @650K12 (JElO3) and plaques were screened using the Q1: 0l kinased oligomer as probe (Zoller g- Aj.t 1 ibid.) Plaques so identified were used for preparation of phage replicative form (RF) DNA (Messing, ibid.).

Restriction enzyme digestion of RF DNA identified two clones having the proper Xba I restriction pattern 4*0*:15 (fragments of 7.5 kb, 0.81 kb, and 0.65 kb) and lacking a Sal I restriction site (which was present in Otto the deleted region of the BLU-proinsulin fusion).

RF DN~A from these two clones was digested with Hind III and Barn HI, and the 1.9 kb fusion fragment from each was gel purified. These fragments were ligated to ptJCl3 and YEp13 (Broach S~ Al,, gA 8:121-133, 1979) vecto,-s which had been digested with Hind III and Barn HI. pUC/DflM-proinsulin hybrid plasmids for subsequent sequencing were used to transform Z. g~o K12 (3TMB3). Two of these plasmids were designated pZV32 and pZV33. YEp3-derived recombinants were used to transform X. g~oU RR1 (Nasmyth and Reed, E= )-At Acad, MA 27:2119-2123, 1980). Two of these plasmids were designated pZV3O and pZV3. (Fig 4).

Sequencing of pZV32 and pZV33 was done by the method of Maxam, and Gilbert (Meth. in Enyml f&:57, 1980). The BAB.I-proinsulin fusion was sequenced from -21the Bgl II site located approximately 190 bp to the side of the junction and from the Sau 961 site located about 140 bp to the 3' side of the junction (in the proinsulin gene). Data from these experiments confirmed that the desired fusion between the AR1 and proinsulin genes had been constructed.

o, 8. cerevisiae strain XP635-10C was transformed with plasmids pZV30 and pZV31. One liter cultures were grown in standard yeast synthetic medium lacking *10 leucine. After 34 hrs., a-factor was added to a 10 ml aliquot of each culture. After an additional 11 hours, cultures were harvested by centrifugation.

Cell pellets and supernatants were tested for the presence of insulin or insulin-like material. Results of two such assays on-the supernatant of a culture transformed with plasmid pZV31 showed 3 pmole IRI material per ml of culture medium, and 5.8 pmol IRC material per ml of culture medium. IRI is correctly folded insulin, proinsulin, or degradation products thereof. IRC is free C-peptide, incorrectly folded proinsulin, or degradation products thereof.

EXAMP .2 Expression of Proinsulin in 2. cerevisiae Using The Alcohol Dehydrogenase I Promoter, BAR1 Gene, and Triose Phosphate Isomerase Terminator The p. cerevisiae alcohol dehydrogenase I promoter (hereinafter ADHI promoter; also known as ACI promoter) was tested for use in directing expression of foreign polypeptides in conjunction with BAR1 sequences. A plasmid comprising these sequences was constructed.

-22- The plasmid pZV50 (FIG. 5B) comprises the g.

cerevisiae AI promoter, the BARI-proinsulin fusion described above, and the terminator region of the S.

cerevisla_ triose phosphate isomerase gene (Alber and Kawasaki, a. Molec. APP, Genet. 1: 419-434, 1982). It was constructed in the following manner. Referring to FIG. 5A, plasmid pAH5 (Ammerer, ibid.) was digested with Hind III and Baa HI and the 1.5 kb ADJI promoter fragment was gel purified. This 10 fragment, together with the Hind III-Eco RI polylinker

S*

fragment from pUC13, was inserted into Eco RI, Bam HI-digested pBR327, using T4 DNA ligase. The resultant plasmid, designated pAM5, was digested with Sph I and Xba I, and the approximately 0.4 kb ADHI promoter fragment was purified on a 2% agarose gel.

Plasmid pZV9 was digested with Xba I, and the approximately 2 kb BARI fragment, containing the entire BAR coding region, was similarly gel purified.

These two fragments, ADHI promoter and LAR sequence, 20 were ligated to Xba I, Sph I-digested YEpl3 to produce plasmid pZV24. Digestion of pZV24 with Sph I and Bgl II, followed by gel purification, yielded an ADHI promoter-BARl fusion of approximately 800 bp, which contains the ATG translation start codon, but lacks the codons for the Arg-Arg potential processing site.

Plasmid pZV33, containing the BAR-proinsulin fusion, was digested with Bgl II and Xba I, and the fusion fragment (ca. 500 bp), which includes the Arg-Arg codons, was purified.

Referring to FIG. 6, The L terminator was obtained from plasmid pFGl (Alber and Kawasaki, ibid.). pFG1 was digested with Eco RI, the linearized plasmid ends were blunted using DNA polymerase I (Klenow fragment), and Bam HI linker sequences (CGGATCCA) were added.

The fragment was digested with Bam HI and religated to -23produce plasmid p136. The 700 bp TPI1 terminator was purified from p136 as a Xba I-Barn HI fragment. This fragment was inserted .Jnto Xba Barn -'-digested YEp13, which was then cut with Hind III, ends blunted using DNA polyrnerase I (Kienow fragment), and religated to produce plasmid p270. The terminator was purified from p270 as a Xba I-Barn HI fragment, and was inserted into Xba 1, Barn HI-digested ptJCl3 to yield plasmid m11l5.

jo Referring to Figure 5B, the terminator was removed from-plasmid m115 by digestion with Xba I and Sst I, followed by gel purification. The three fragments: ADHI-BAR1 fusion, B~Al-proinsulin fusion, and TP11 terminator, were inserted into plasmid pUCiB (Norrander et al., Gene 21: 101-106, 1983) which had besn digested with Sph I and Sst 1. This DNA was used to transform ~.coli X12 (JI48j). Selection for ainpicillin resistance and screening for production of seen white colonies identified a plasmid (pZV45) containing the desired inserts. Plasmid pZV45 was subsequently digested with Sph I and Barn HI, and the APIiBAU-L proinsulin-TPI terminator sequence was gel purified, This fragment was inserted into YEpl3 which had been digested with Sph I and Barn HI, to produce the cerevisiae expression vector cerevisiae strain XP635-10C was transformed with plasmid pZV5O and cultured and assayed as described in Example 1 above. No IRI material was f ound in the medium, and IRC material was less than 0. 5 pmol per ml. Cells extracted with 0.1% Nonidet P-40 showed 1 prnol IRC material per ml of cell extract.

-24- EXAMPLE 3 Expression of Proinsulin in Schizosaccharomycgs Rombe Using the BARl Gene and a. Rombe Alcohol Dehydrogenase Promoter This example demonstrates the use of portions of the BARI gene to direct the secretion of foreign polypeptides expressed in a transformed Schizo-~ a'saccharomyces pombe host. A plasmid was constructed 5which combines the p. ombe alcohol dehydrogenase (ADH) gene promoter with the BARI-proinxsulin gene fusion.

goo: 0 0 The S. Rombe ADH promoter was obtained from a library of DNA fragments derived from S. pombe strain 972h- (ATCC 24843), which had been cloned into YEpI3 as 16014described by Russell and Hall P~l Chm :04 143-149, 3.983) The promoter sequence was purified 6 from the library as a 0.75 kb Sph I-Eco RI fragment.

This fragment and the Eco RI-Hind III polylinker fragment of pUC12 were ligated into YEpl3 which had been digested with Sph I and Hind III. The resulting plasmid is known as pEVP-ll.

Ref~erring to FIG. 7 for constructing the 2. pombe expression vector, the ADH promoter was purified from pEVP-ll as a Sph I-Xba I fragment. Plasmid pZV33 was digested with Xba I and Bgl 11 and the ca. 340 bp pARI fragment, which includes the ATO initiation codon, was purified. pZV33 was digested with Bgl II and Sst 1, and the PARI-proinsulin fusion sequence was pur:ified. The three fragments were combined with Sph I, Sst I-digested pUCl8 to produce plasmid pZV46. As pUCIS is not effective for transformation of a. p-gtthe plasmid was subjected to two double enzyme digestions. An ADH promoter-BARi fusion fragment was purified from a Hind III Bgl II digest, and a BAB1proinsulin fusion sequence was purified from a Bgl II Xba I digest. These fragments were inserted into Hind III, Xba I-digested YEpl3 to produce 1. pombe expression vector pZV49.

One liter cultures of transformed pombe strain 318-4h (ATCC #20680) were grown 36 hours at 30'C in t standard yeast synthetic medium (-leu D) containing 200 mg/l aspartic acid and 100 mg/1 each of histidine, adenine, and uracil. Cultures were harvested by centrifugation and the supernatants assayed by IRI and IRC assays. Two samples from pZV49-transformed cells contained 1.6 pmol/ml IRI material and 0.5 pmol/ml IRC-reactive material, respectively. A control sample from a culture transformed with YEpl3 contained no detectable IRC-reactive material.

0 6* e EXAMPLE 4 Export Of Alpha-Factor Using BAR1 Signal Peptide The BAR1 signal peptide was tested for its ability to direct the export of a-factor from a yeast transformant. Several plasmids containing DNA fragments coding for fusion proteins with varying lengths of the BAR1 protein and 1 or 4 copies of mature a-factor were constructed. These plasmids were transformed into an a/a diploid host strain and the transformants assayed for a-factor production by the halo assay.

Plasmids pSW94, pSW95, pSW96, and pSW97 comprise the S. cerevisiae triose phosphate isomerase (TPIl) promoter, a 355 bp or 767 bp fragment of the BAR1 gene (comprising 114 or 251 codons of the 5' end of the BAR coding sequence, respectively) and either one or -26four copies of the alpha-factor (Ial) coding sequence. These constructs are described in Table 1.

TABLE 1 Plasmid BAR1 Fragment a-factor pSW94 355 bp 4 copies S* pSW95 767 bp 4 copies pSW96 355 bp 1 copy pSW97 767 bp 1 copy Plasmid pM220 (also known as pM210) was used as the source of the TPT promoter fragment. E. coa RRI transformed with pM220 has been deposited with ATCC under accession number 39853. Plasmid pM220 was digested with Eco RI and the 0.9 kb fragment comprising the TPI1 promoter was isolated by agarose gel electrophoresis and the ends were blunted with DNA polymerase I (Klenow fragment). Kinased Xba I linkers were added to the fragment, which was then digested with Bgl II and Xba I. This modified TP1 promoter fragment was then ligated into the 3.4 kb Bg' II-Xba I vector fragment of pDR1107 to produce pZV18. Plasmid pDR1107 was constructed by subcloning the 900 bp Bgl II-Eco RI TP promoter fragment of pM220 into pIC7 (Marsh, Erfle and Wykes, Gene J2: 481-485, 1984) to generate plasmid pDR1101. Plasmid pDR1101 was digested with Hind III and Sph I to isolate the 700 bp partial TP11 promoter fragment. Plasmid pDR1100, comprising the 800 bp Xba I-Bam HI TPI terminator fragment of pM220 subcloned into pUC18, was cut with Hind III and Sph I. The 700 bp partial TPII promoter was ligated into the linearized pDROO00 to produce pDR1107.

-27- The Eco RI site at the 3' end of the TP~ promoter in pZVll8 was then destroyed. The plasmid was digested with Hind III and Eco RI and the 0.9 kb fragment was isolated and ligated to a synthetic linker constructed by annealing oligonucleotis ZC708 (5'AATTGCTCGAGT 3 and ZC709 (3'CGAGCTCAGATC The linker addition eliminates the Eco RI site at the 3' terminus of the TP11 promoter fragment and adds XhoI and Xba I sites.

This fragment was then joined to Hind III-Xba I cut pUC13. The resultant plasmid was designated pZV134 (Figure 8).

Cloning of the yeast mating pheromone a-factor (iEal) gene has been described by Kurjan and Herskowitz (ibid). The gene was isolated in this laboratory in a 15 similar manner from a yeast genomic library of partial Sau 3A fragments cloned into the Bam HI site of YEpl3 (Nasmyth and Tatchell, Cell 1: 753-764, 1980). From this library, a plasmid was isolated which expressed a-factor in a diploid strain of yeast homozygous for the mata2-34 mutation (Manney, e t Cell. Biol., 9: 1592, 1983). The clone contained an insert •overlapping with the KFal gene characterized by Kurjan *0 "and Herskowitz. This plasmid, known as pZA2, was digested with Eco RI and the 1.7 kb fragment 25 containing fcal was isolated and ligated into Eco RI cut pUC13. The resultant plasmid, designated p192, was cleaved with Eco RI and the resultant 1.7 kb MFal fragment was isolated and digested with Mbo II. The 550 bp Mbo II-Eco RI fragment was isolated and ligated to kinased Sal I linkers. The linkered fragment was cut with Sal I. The resulting 0.3 kb Sal I fragment was ligated into Sal I cut pUC4 (Vieira and Messing, Gene, t2: 259-268, 1982) to produce a plasmid designated p489 (Figure 9).

I I) -28- A gene fusion comprising portions of the BAR1 (114 codons) and MFal coding sequence was then constructed.

Plasmid pZV24 (Example 2) was digested with Sph I and Bgl II and the 0.8 kb ADH1 promoter-BARl fragment was isolated. Plasmid p489 was cleaved with Bam HI and the 0.3 kb MFal fragment was isolated. These two fragments were joined in a three part ligation to Sph I+Bam HI cut YEpl3. The resultant plasmid was designated pZV69 (Figure 11).

A second fusion gene encoding 251 amino acids of Barrier joined to a portion of the alpha-factor percursor was constructed. Plasmid pZV16, containing the 767 bp Xba I-Sal I BARI fragnant from pZV9 (Example 1) ligated into Xba I+Sal I cut pUC13, was linearized by digestion with Sal I. This 4.0 kb 06* fragment was joined with the 0.3 kb Sal I fragment from p489 encoding the four copies of a-factor. A plasmid having the BARl-MFal fusion in the correct orientation was designated as pZV71. The BARI-Mal fusion from pZV71 was then joined to the ADHI promoter. Plasmid pZV71 was digested with Xba I and Pst I and the 1.07 kb fragment was isolated. The ADH1 promoter was isolated as a 0.42 kb Sph I-Xba I fragment from pZV24. These two fragments were joined, 25 in a three part ligation, to Sph I+Pst I cut pUC18.

The resulting plasmid, pZV73, was digested with Sph I and Bar HI and the 1.5 kb fragment comprising the expression unit was isolated and ligated into the Sph I+Bam HI cut YEpl3 to form pZV75 (Figure For ease of manipulation, the BARl-MFal fusion units from pZV69 and pZV75 were subcloned with the T£U promoter into pUC18 (Figure 11). Plasmid pZV69 was digested with Eco RI and Bam HI and the 0.55 kb fragment containing the fusion was isolated. The 0.9 -29kb TPI promoter fragment was isolated from pZV118 by digestion with Hind III and Eco RI. A three part ligation was done by using the .55 kb BAR1-MIal fragment, the 0.9 kb JPI1 promoter fragment and pUC19 cut with Hind III and Bam HI. The resultant plasmid was designated pSW59. Plasmid pZV75 was digested with Eco RI and Bam HI .to isolate the 954 bp BAR~-NSal fusion fragment. This BARl-MFal fragment was ligated in a three part ligation with the 0.9 kb Hind III+Eco RI TP11 promoter fragment and pUC18 cut with Hind III and Bam HI to generate plasmid In construction of expression plasmids, the source of the 5' 116 bp of the BAR1 coding sequence was pSW22, which was constructed in the following manner (Figure 12). The BAR coding region found in pSW22 originated from pZV9. Plasmid pZV9 (Example 1) was cut with Sal I and Barn HI to isolate the 1.3 kb BARI fragment. This fragment was subcloned into pUC13 cut with Sal I and Bam HI to generate the plasmid designated pZV17. Plasmid pZV17 was digested with Eco RI to remove the 3'-most 0.5 kb of the BAR1 coding region. The vector-BARi fragment was religated to create the plasmid designated pJH66. Plasmid pJH66 was linearized with Eco RI and blunt-ended with Klenow 5 25 fragment. Kinased Bam HI linkers CCGGATCCGG were added and excess linkers were removed by digestion with Bam HI before religation. The resultant plasmid, pSW8, was cut with Sal I and Bam HI to isolate the 824 bp fragment encoding amino acids 252 through 525 of B1. This AR1 fragment was fused to a fragment encoding the C-termina portion of substance P (Munro and Pelham, ~IM 1: 3087-3093, 1984). Plasmid pPM2, containing the synthetic oligonucleotide sequence encoding the dimer form of substance P in M13mp8, was obtained from Munro and Pelham. Plasmid pPM2 was linearized by digestion with Bam HI and Sal I and ligated with the 824 bp BAR1 fragment. The resultant plasmid pSW14 was digested with Sal I and Sma I to isolate the 871 bp BA 1substance P fragment. Plasmid pZV16 (Figure 10) was cut with Xba I and Sal I to isolate the 767 bp coding sequence of BARi. This fragment was ligated with the 871 bp BARi-substance P fragment in a three part ligation with pUC18 cut with Xba I and Sma I.

The resultant plasmid was designated pSW15. Plasmid was digested with Xba I and Sma I to isolate the 1.64 kb BARl-substance P fragment. The ADH1 promoter was obtained from pRL02-, comprising the 0.54 kb Sph I-Eco RI fragment containing the ADH1 promoter and 116 45 bp of the BAR1 5' coding region from pZV24 in pUC18.

Plasmid pRL029 was digested with Sph I and Xba I to isolate the 0.42 kb AD promoter fragment. The TPl2 terminator (Alber and Kawasaki, A. ol. Ap1. Gen. i: 419-434, 1982) was provided as a 0.7 kb Xba I+Eco RI fragment in pUC18. The linearized fragment containing the TPIl terminator and pUC18 with a Klenow filled Xba I end and an Sph I end was ligated with the 0.42 kb ADH1 promoter fragment and the 1.64 kb BARI-substance P fragment in a three part ligation to produce plasmid pSW22.

Plasmid pSW94 was then constructed (Figure 13). The 2.3 kb fragment containing the BARl-substance P fusion and the terminator was isolated from plasmid pSW22 as an Xba I-Sst I fragment. The Hind III-Xba I TPI1 promoter fragment isolated from pZV134 was joined tc the BARl-substance P-El terminator fragment in a three part ligation with Hind III+SstI cut pUC18. The resultant plasmid, pSW81, was cleaved with Hind III and Eco RI to isolate the 1.02 kb fragment containing the TPIl promoter and the 5' 116 bp of BAR1. Plasmid -31pSW59 was cut with Eco RI and Barn HI to isolate the 0.55 kb PARl-Uta1 fusion fragment. This fragment was then ligated in a three part ligation with the promoter-BARI fragment from pSW8l and YEpl3 linearized with Hind III and Barn HI resulting in plasmid pSW94.

The construction of pSW95 is illustrated in Figure 13.

Plasmid pSW6O was cut with Eco RI and Barn HI to isolate the 954 bp BARI-Mcal fusion fragment. Plasmid pSW8l was cut with Hind III and Eco RI to isolate the 1.02 kb TP1I promoter-BARI fragment which was joined with the DAR-Mal fusion fragment in a three part ligation into Hind III+Baii HI cut YEpl3. The resultant plasmid was designated TP11 promoter-B3ARI-Mj~al fi*.sion constructs containing only one copy of a-factor originated from lAl-Mral fusions (encoding fou, copies of a-factor) that tane the =2l rromoter and Mal prepro sequence (see Figures 24, 15, and 16). Plasmid pZVl6 was digested with Eco RI and Sal 1. The isolated 651 bp BARi fragment was ligated with a kinased Hind IZI-Eco *e..e.RI &&RJL specific adaptor (produced by annealing ~*.oligonucleotides ZC566: 51AGCTTTAACAAACGATGGCACTGGTCACTTAG 3 and ZC567:, AATTCTAAGTGACCAGTGCCATCGTTTGTTAA) into pUCl3 cut with Hind III and Sal 1. The resultant plasmid, pZV96, was digested with Hind III and Sal I to isolate .the 684 bp ARI fragment. Plasmid pM220 provided the =PJ promoter fused to the Ka1 prepro sequence.

Plasmid pM22O was digested with Bgl 11 and Hind III to isolate the 1.2 kb promoter-M~al prepro fragment.

The 3' portion of the coding region was obtained by cutting pZV9 with Sal I and Barn HI to isolate the 1.3 kb DARI fragment. The 684 bp Hind III-Sal I PAU.

fragment, the 1.2 kd, Bgl 11-Hind III TM -32promoter-MFol prepro fragment and the 1.3 kcb Sal I-Ban HI BARI fragment were joined with YEp13 linearized with Barn HI in a four part ligation. The construct with the desired orientation of proriter and Macl-L&I.2 fusion was designated pZV100 (Fig'. 14).

For ease of manipulation, the truncated Mal prepro sequence-BARI fusion fragment from pZVIOO was subcloned into pUC13 as a 1.6 kb Pst I-Barn HI fr~agment. The resultant plasmid, pZVl0l, was cleaved with Pst I and Eco RI to isolate the 270 bp Mial prepro-BARI fragment. Plasmid pZV69 was digested with Eco RI and Barn HI to isolate the 0.55 kcb BAR-ZLal fusion fragment (encoding four copies of a-factor).

This fragment and the 270 bp Xfal prepro-DM.~ fragmdnt were ligated in a three part ligation into pUCl3 cut with Pst I and Barn HI. The resultant plasmid was lose@ designated pZVlO2 (Figure 15) :00 An expression unit comprising the TP1 promoter, a portion of BARI, and a single copy of the a-factor coding sequence was then constructed (Figure 16).

4 Plasmid pZV102 was cut with Pst I and Barn HI to isolate the 0.82 kb Mjal prepro-AlJ, fragment. A 1 kb Hind III-Pst I fragment comprising the ~JP1 promoter and the truncated Mcl prepro sequence from pM220 was joined to the 0.82 kb JL~al prepro-BAR1 fragment isolated from pZV1O2 in a three part ligation with YEpl3 cut with Hind III and Barn HI. The resultant plasmid was designated pZVlO5. Plasmid pZV1O5 was cleaved with Hind III to isolate the 1.2 kb =B1 promoter-Mal prepro fragment. Plasmid pZV1O2 was digested with Hind III to isolate the vector fragment containing the terminal a-factor copy. This 2.-8 kb vector-Mal fragment was ligated to the 1. 2 kb promoter-Mal prepro fr, gment- The plasmid with the -33correct orientation and a single copy of M#al coding sequence was designated pSW61. Plasmid pSW61 was linearized by a partial digestion with Hind III.

Plasmid pZV102 was digested with Hind III to isolate the 0.3 kb BARl-M al fragment. This fragment was ligated into the linearized pSW61. The plasmid with the insert in the correct orientation at the Hind III site 264 bp 3' to the MEal start codon was designated Plasmid F i, s cleaved with Eco RI and Barn HI to isolate the lI ;p BARl-Mal fragment. Plasmid pSW81 (Figure 13) was digested with Hind III and Eco RI to isolate the 1.02 kb TP1 promoter-BARl fragment.

This fragment was joined to the BAB" ,ial fragment in a three part ligation with YEpl3 linearized with Hind SiIs III and Bam HI. The resultant plasmid, pSW96, contains the TPI1 promoter and 356 bp of the 5' coding sequence of BAR~ fused to one copy of the a-factor coding sequence.

The second BARl-MFal construct containing 767 bp of BAR1 fused to one copy of the MEal coding sequence was made using pZV75 as the source of the BAR1 fragment (Figure 17). Plasmid pZV75 was digested with Eco RI and Sam HI to isolate the 954 bp D&ARl-K al fragment.

Plasmid pZV101, containing the Eal prepro sequence 25 fused to 6AR was cut with Pet I and Eco RI to isolate the 0.27 kb MFal prepro-ARl fragment. This fragment was joined to the 954 bp BARI-MFal fragment in a three part ligation with pUC13 linearized with S" Pst I and Bam HI. The resultant plasmid, pZV104, was cleaved with Hind III to isolate the 0.70 kb BAR1-Mal fragment. This fragment was ligated to pSW61 which was linearized by partial digestion with Hind III.

The plasmid with the insert in the correct orientation at the Hind III site 264 bp 3' to the start codon of Kral was designated pSW74. Plasmid pSW74 was cut with -34- Eco RI and Bam HI to isolate the 738 bp BARI-&Fal fragment. Plasmid pSW81 was cut with Hind III and Eco RI to isolate the 1.02 kb TI1 promoter-BAR1 fragment.

This fragment was joined to the 738 bp BAl -Mal fragment in a three part ligation with Hind III+Bam HI cut YEpl3. The resultant plasmid, pSW97, contains the TPI1 promoter and 767 bp of the 5' end of BAR1 fused to the single copy of the a-factor coding sequence.

The a/a diploid S. cerevisiae strain XP733 (MATa leu2-3 leu2-112 barl-1 cal2/MATa Ieu2-3 leu2-112 barl- ga12) was transformed with plasmids pSW73, pSW94 and pSW95. Plasmid pSW73 comprises the TP11 promoter, MFal signal peptide and prepro sequence and the coding region for the four copies of a-factor in YEpl3. The transformants were spotted onto a lawn of cerevisiae RC629 cells in a soft agar overlay over a plate of selective media and incubated overnight at 30'C. A comparison of halo sizes using pSW73 and the control shows that pSW94 exports approximately 15% as much a-factor as pSW73.

The constructs containing BAR& fused tc one copy of the MFal coding sequence were assayed tor export of o-factor in the same manner. Plasmid pSW67, comprising the TPI promoter, Moal signal peptide, prepro and the coding region for one copy of a-factor in YEpl3 was used as a control for plasmids pSW96 and pSW97. A comparison between halo sizes indicated that pSW96 directs secretion of approximately 30 40% as much a-factor as pSW61 and pSW97 directs secretion of approximately 10 15% as much a-factor as pSW67.

EXAMPLE MUTATION OF THE BAR1 SIGNAL PEPTIDE CLEAVAGE SITE As described above, it has been found that altering the signal peptide cleavage site of the Barrier precursor could be expected to facilitate processing and export of Barrier-containing fusion proteins through the ~EX2 pathway. Potential cleavage sites are between amino acids 23 and 24 and between amino acids 24 and 25. Thus, the DNA sequence coding for the BAPI primary translation product was mutated to encode a proline residue at position 25. Plasmids pSW98 and pSW99 are YEpl3-based plasmids comprising the S. cerevisiae TPI1 promoter, a 355 bp or 767 bp fragment of the BDAB gene, including the mutated signal peptide cleavage site, and one copy of the a-factor coding sequence.

The signal peptide mutation was introduced by standard Jn vitro mutagenesis methods (Zoller et Al., Manual i Advanced Techniques in Molecular Cloning Course, Cold Spring Harbor Laboratory, 1983) using a phage M13 template and a synthetic mutagenic oligonucleotide 5'A3' 0 (sequence 5

ATTACTGCTCCTACAAACGAT

3 The phage template pSW54 was constructed by ligating the 0.54 kb

S**

Sph I-Eco RI fragment of pSW22 with Sph I-Eco RI digested M13mpl9. Following in vitL mutagenesis, potentially mutagenized plaques were screened by 5 plaque hybridization with 32 P-labeled mutagenic oligonucleotide and were sequenced to confirm the presence of the mutation. The replicative form of one of the confirmed mutagenized phage, mZC634-7, was digested with Sph I and Eco RI and the 0.54 kb fragment was isolated and ligated with Sph I+Eco RI cut pUC18. The resulting plasmid, pSW66 (Figure 18), was digested with Hind III and Xba I to remove the -36- ADHI promoter, and the frangm-nt comprising the vector and BA~R sequences was ligated with the 0.9 kb Hind III-Xba I TEPI promoter fragment of pZV134. This plasmid, with the TP promoter and 119 bp of the end of BAR including the signal peptide cleavage mutation, was designated pSW82.

Referring to Figure 18, plasmid pSW82 was digested with Hind III and Eco RI and with Bgl II and Eco RI and the resulting 1.02 kb fragments were isolated.

The Hind III-Eco RI fragment of pSW82 was ligated with the Eco RI-Bam HI fragment of pSW74 and Hind III+Bam HI digested YEpl3 to form pSW99. The Bgl II-Eco RI fragment of pSW82 was ligated with the 0.30 kb Eco RI-Bam HI fragment of pSW70 and Bam HI digested YEpl3, to form pSW98. Plasmid pSW98 includes the T2U promoter, 355 bp of the 5' end of the mutagenized BARI sequence and a single copy of the a-factor coding sequence. Plasmid pSW99 contains the identical expression unit except for having 767 bp of the mutagenized BAR sequence.

An analysis by the halo assay showed that the cleavage site mutation enhanced a-factor export when using the plasmids encoding one copy of a-factor. Transformants containing pSW98 exported about 50% more a-factor than those containing pSW96, the wild-type control.

Claims

400. 4 0909 *4 C 4 -37- THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS: 1. A DNA construct comprising a portion of the Saccharomyces cerevisiae BAR1 gene comprising the signal peptide coding sequence thereof, and at least one structural gene foreign to a fungal host cell transformed with said construct, said construct including a processing site at the junction of said BAR1 gene portion and said foreign structural gene, and a promoter which controls the expression, in said fungal host cell transformed with said construct, of a fusion polypeptide or protein resulting from said portion of the BAR1 gene and said structural gene. 2. A construct according to claim 1 wherein said fusion polypeptide or protein comprises a KEX2 processing site. 3. A construct according to claim 2 wherein said signal peptide coding sequence of said BAR1 gene is altered to reduce the efficiency of signal peptidase cleavage of the fusion polypeptide or protein. 4. A construct according to claim 1 wherein said promoter is a yeast glycolytic pathway gene promoter. A construct according to claim 1 wherein said promoter is selected from the group consisting of the S. cerevisiae BAR1 promoter, S. cerevisiae alcohol dehydrogenase I promoter, and Schizosaccharomyces pombe alcohol dehydrogenase promoter. 6. A construct according to claim 5 wherein said construct is G, F1 'A P:\OPIlIRJMS\7400391.CLM 15/10/96 -38- selected from the group consisting of pZV30, pZV31, pZV49, and pZV50 (as herein described). 7. A construct according to claim 1 further comprising the transcription terminator region of the Saccharomyces cerevisiae triose phosphate isomerase gene. 8. A construct according to claim 1 wherein said portion of said BAR1 gene further comprises the 680 base pair sequence of the 5'-untranslated region adjacent to the tranclation initiation site. A transformed fungal cell containing a construct according to any of claims 1-8. C C A cell according to claim 9 wherein said fungus is Saccharomyces cerevisiae. 11. A cell according to claim 9 wherein said fungus is Schizosaccharomyces pombe. 12. A cell acco2ding to claim 9 wherein said fungus is Aspergillus or Neurospora. 13. A method for producing a heterologous protein in a transformed fungal cell and directing said protein into the secretory pathway of the cell comprising the steps of: transforming a fungal host cell with a DNA construct -j comprising a portion of the S. cerevisiae BAR1 gene comprising at leas the signal peptide coding sequence P:kOPELRlMS\74003-9Z.CLM -15/10/96 -39- thereof, and a structural gene encoding said heterologous protein, said construct including a processing site at the junction of said BAR1 gene portion and said structural gene, and a promoter which controls the expression in said fungal cell of a fusion protein comprisiny said signal peptide and said heterologous protein; growing said fungal cell from step under growth conditions suitable to select for the production of said fusion protein. S. 14. A method according to claim 13 wherein said fungus is Saccharomyces cerevisiae. 15. A method according to claim 13 wherein said fungus is Schizosaccharomyces pombe. 16. A method according to claim 13 wherein said fungus is Aspergillus or Neurospora. 17. A method according to claim 13 wherein said fusion protein comprises a KEX2 processing site. 18. A method according to claim 17 wherein said signal peptide coding sequence of said BAR1 gene is altered to reduce the efficiency of signal peptidase cleavage of the polypeptide or protein. 19. A method according to claim 13 wherein said promoter is a yeast glycolytic pathway gene promoter. I':\OI'RlMS74003-91.CLM( 15/1096 A method according to claim 13 wherein said promoter is selected from the group consisting of the S. cerevisiae BAR1 promoter, S. cerevisiae alcohol dehydrogenase I promotel, and Schizosaccharomyces pombe alcohol dehydrogenase promoter. 21. A method according to claim 20 wherein said DNA construct is selected from the group consisting of pZV30, pZV31, pZV49, and pZV50 (as herein described). 22. A method according to claim 13 wherein said heterologous protein is exported from said fungal cell. 23. A method according to claim 13 wherein said protein comprises proinsulin or insulin. 24. A protein produced according to the method of claim 13. 25. A protein produced according to the method of claim 13 comprising the amino acid sequence of human proinsulin or insulin. P:\OI'1RJMS\74003-9I.CLM 151196 41 26. A DNA construct according to claim 1, a method according to claim 13 or a protein produced thereby substantially as hereinbefore described with reference to the Examples and/or drawings. Dated this 15th day of October 1996. Zymo~enetics, Inc. By its Patent Attorneys, Davies Collison Cave S S 0**t S S S S 1,Y)