EP0208706A1 - Dna sequence useful for the production and secretion from yeast of peptides and proteins - Google Patents

Abstract

Composition, in particular the nucleotide sequence of the DNA gene encoding the yeast premelibiase and the amino acid sequence contained in the premelibiase protein. The invention describes the DNA sequence of the premelibiase gene, in combination with the amino acid sequence of the secreted melibiase, and thus defines the amino acid sequence of the pre-region and the nucleotide sequence encoding this sequence. The premelibiase pre-region can be used to obtain peptides and proteins other than melibiase from genetically modified yeasts by means of biological engineering techniques, since the fusions carried out between its nucleotide sequence and the coding sequence of foreign peptides and proteins ensure the secretion of these genetic products in yeasts, when they are contained in appropriate gene expression vehicles. The nucleotide sequence of the promotion region of the premelibiase gene is also described, this sequence being able to be regulated by protein products of the GAL4 and GAL80 genes. The promotion region can be integrated into a vector structure with a structural gene and optionally comprising the signal sequence defined above, in order to be able to regulate the expression of a protein in a transformed yeast host.

Description

DNA SEQUENCE USEFUL FOR THE PRODUCTION AND SECRETION FROM YEAST OF PEPTIDES AND PROTEINS.

This invention relates to recombinant DNA technology. In particular, it relates to sequences of DNA which in one aspect function to promote expression of a gene and, in another aspect, function to signal secretion of the protein coded by the expressed gene. Such sequences are valuable in manipulating yeast cells.

BACKGROUND OF THE INVENTION

Secretion, a process by which particular products of gene expression are exported from the cell to the culture medium, is a natural property of living cells. For cells that have been genetically modified to express foreign peptides and proteins, the feature of secretion of these products is desirable since the products are produced in a partially purified and readily accessible state by virtue of their extracellular location.

Protein and polypeptides produced in cells are destined either to remain within the cell or to be secreted from it. The mechanism by which the destinies of these two categorically different types of proteins is determined has been experimentally shown to reside in a functional "signal" peptide coded for by a nucleotide sequence i.e. a signal sequence which in almost all known cases precedes the actual sequence of the gene coding for the protein to be secreted. According to the evidence, immature proteins translated from the complementary nucleotide sequence and bound for secretion initially bear a precursorial sequence of amino acids or "signal" which is recognized by receptors on the membraneous secretory pathway within the cell thereby permitting the immature protein to pass into the secretory pathway. Once so engulfed, the signal sequence of the precursor protein is proteolytically cleaved to provide the active protein which may be released from the cell, after subsequent modification of its structure, e.g. glycosylation.

By deciphering and utilizing such specific signal sequences it is possible to qenetically engineer cells capable of producing protein for routine secretion from the cell. Typically, retrieval of protein from genetically enqineered cells has required lysis of the engineered cells and purification of the desired protein, since the vector does not contain the appropriate nucleotide sequence to confer secretory properties to the proteins translated from the sequence. This process results in destruction of the cell whereas, according to one aspect of the present invention as described below, the cell may be sustained in a viable condition by supplying sufficient growth medium while continuing to secrete the desired protein. Moreover, the protein may be extracted directly from the nutrient broth within which the engineered cells are incubated, on a continuous basis according to known chemical procedures.

As described, secretion of products expressed in yeast from genetically-modified expression vehicles is, in principle, assured by providing the foreign gene inserted in the vehicle with a nucleotide sequence encoding a signal peptide appropriate to yeast. This signal peptide should be contiguous with, and at the N-terminal end of the peptide comprising the gene product of interest.

Expression per se of foreign peptides and proteins in qenetically engineered yeasts, without secretion of the molecules so expressed, has been attempted on many occasions. In some cases these genes are expressed in the yeasts, but often they are not, or at least not at required levels. In order that a particular foreign gene be expressed, certain of its features must be changed. This may be done by fusing part of a natural gene of the organism to part of the foreign gene, thus reducing the differences between the unexpressed foreign gene and the expressed natural gene.

One of the most common features to be changed is the region upstream of the coding sequence of the gene. This region typically includes sequences necessary for the initiation of transcription, as well as sequences which enhance and/or repress i.e. regulate, the level of transcription, often in combination with particular proteins present in the cell, together with other sequences which affect the transcription and translation of the gene. This region is commonly called the 'promoter', although as stated it includes more than the sequence which is strictly defined as that region which promotes transcription.

When it is desired that the product of a given foreign gene be secreted from yeast, another feature which is commonly added or replaced is the signal or leader sequence. This signal sequence is a short polypeptide stretch which is typically encoded in a sequence at the beginning of the translated portion of the gene.

Thus, when it is desired that the product of a given foreign gene be expressed and secreted from yeast, the promoter and/or signal sequence as described above may be added, or used to replace analogous foreign sequences with sequences which are functional in yeast.

An attempt to obtain secretion in yeast of foreign proteins without the use of yeast signal peptides was described by Hitzeman et al (1). The proteins of interest, human interferons, carried, in their genes, their own signal sequences appropriate to secretion in human cells; when expressed in yeast, the interferons were expressed efficiently but secreted poorly. Recent reports have examined the utility of using yeast signal peptides to obtain secretion of foreign gene products from yeast. Emr et al (2) made gene fusions between nucleotide sequences coding for the leader region of the secreted yeast peptide, alpha factor, and nucleotide sequences coding for invertase, a yeast protein normally secreted into the periplasmic space, i.e. the region between the yeast cell membrane and cell wall. While this resulted in invertase secretion to the periplasmic space, it is not clear that this was due to the alpha factor leader, since the invertase gene in the fusion included a segment of the invertase signal peptide sequence. Indeed, since alpha factor is secreted into the medium, and invertase into the periplasm, the localisation of invertase in the experiments of Emr et al may be due to the invertase component of the gene fusion. Brake et al (3) constructed gene fusions between the alpha factor leader region and a synthetic DNA sequence encoding human epidermal growth factor (EGF) and found that this fusion resulted in efficient secretion from yeast of mature EGF, provided a limited region of the sequence of the alpha factor leader-coding region was altered, by in vitro mutagenesis. Bitter et al (4) constructed fusions between the alpha factor leader region and each of the proteins ß-endorphin and a "consensus" human interferon; they found secretion of these foreign proteins but also some degradation due to proteolysis during the secretion events.

However, in the case of interferon, at least 95% of the secreted protein was recovered intact from yeast culture fluids. It seems clear that, at least in the case of alpha factor, yeast secretion siqnals contain information that facilitates secretion from yeast of foreign peptides and proteins, albeit peptides and proteins that are secretory gene products in their natural hosts.

Melibiase, or α-galactosidase (E.C. 3.2.1.22), a product of the growth of the yeast S. carlsbergensis ( S . uvarum), and of a few strains of the yeast S. cerevisiae, is a naturally secreted protein. A DNA fragment obtained from one of these strains by cloning into a yeast plasmid, has been shown by

Post-Beittenmiller et al (5) to confer the melibiase-positive phenotype on otherwise melibiase-negative yeast, following introduction into the cells by transformation. SUMMARY OF THE INVENTION

It is an object of the present invention to provide the amino-acid sequence of the precursor premelibiase and the pre-region thereof, and the nucleotide sequence encoding each of these entities and to provide a vector construct containing the nucleotide sequence coding for the pre-region.

It is a further object of the present invention to provide processes wherein signal sequences capable of signalling secretion of melibiase may be combined with other genes or gene parts and introduced via cloning vectors into yeast cell hosts for secretion of proteins therefrom by means of such genes or gene parts.

It is a further object of the present invention to provide a regulatable promoter region which is capable of promoting transcription of the DNA region coding for the protein melibiase gene native to Saccharomyces carlsbergensis.

It is a further object of the present invention to provide processes wherein the promoter regions described above may be combined with a signal sequence and a polypeptide coding region via cloning vectors and introduced into yeast cell hosts to obtain secretion of the polypeptide encoded by the coding region.

Thus, the present invention provides in a first aspect, the signal sequence of the melibiase gene native to yeast, preferably derived from Saccharomyces carlsbergensis. In a second aspect, the present invention provides the regulatable promoter region of the premelibiase gene native to yeast, preferably derived from Saccharomyces carlsbergensis or a synthetic or semi-synthetic equivalent thereof.

In a third aspect, there is provided a coding region which codes for premelibiase and melibiase native to yeast, preferably derived from Saccharomyces carlsbergensis.

In a fourth aspect, there are provided vector constructs comprising at least one of the promoter region and the signal sequence for use in transforming a host yeast cell. In its preferred form, both the promoter region and the signal sequence are integrated on the vector construct and operatively associated with a coding region either native or foreign to the host yeast cell to be transformed. Further, where each of the promoter region, signal sequence and coding region originate in the premelibiase region native to S. carlsbergensis the vector construct so formed is used to transform a foreign host.

BRIEF REFERENCE TO THE DRAWINGS

FIGURE 1 represents the nucleotide sequence of a portion of plasmid pMP550 from Hind III to the first Bam HI site, including the 5' flank of the premelibiase gene including the promoter, the premelibiase gene sequence, showing in the lower line the amino-acid sequence of pre-melibiase and in the upper line the corresponding nucleotide sequence, and extending into the 3' untranslated sequence to the first Bam HI restriction site; FIGURE 2 represents schematically a section of DNA segment shown in Figure 1;

FIGURE 2a provides the nucleotide sequence of the natural melibiase secretion signal represented in Figure 2;

FIGURE 2b provides the nucleotide sequence of a modified melibiase secretion signal in which the natural sequence depicted in Figure 2a is coupled with a synthetic linker sequence;

FIGURE 3 illustrates, via plasmid maps, a scheme for creating plasmid p4;

FIGURE 4 illustrates, via plasmid maps, a scheme for creating plasmid p5;

FIGURE 4a provides the nucleotide sequence of a segment of plasmid p5 shown in Figure 4; and

FIGURE 5 illustrates, via plasmid maps, a scheme for creating each of plasmids p6 and p7.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The signal sequence preferred herein acts as the signal sequence for melibiase secretion and is obtained from S . carlsbergensis by methods described hereinafter. The particular amino acid sequence coded by the signal sequence is reproduced below:

met-phe-ala-phe-tyr-phe-leu-thr- ala-cys-ile-ser-leu-lys-gly-val- phe-gly Analooously, an example of the nucleotide sequence corresponding to the above amino acid sequence is reproduced below:

5' ATG TTT GCT TTC TAC TTT CTC ACC GCA TGC ATC AGT TTC AAG GGC GTT TTT GGG 3'

The first six nucleotides 3' of reference numeral 30 shown in Figure 1 represent a Hind III site on pMP550. Reference numeral 25 identifies the first Bam HI site downstream of Hind III site 30. Within this 5' Hind III - Bam HI 3' region are located the promoter region and signal sequence of the melibiase gene as well as the melibiase coding region. The entire, separated and isolated amino acid sequence of the premelibiase coding redion is shown in the lower lines startinq at reference numeral 10 and ending at reference numeral 1 2 . The corresponding nucleotide sequence is shown in the lines above this particular amino acid sequence. In Fig. 1, a represents adenine nucleotide, t represents thymine nucleotide, g represents guanine nucleotide and c represents cytosine nucleotide. The amino-acid corresponding to each codon has been entered using the normal triplet abbreviations therefor. The structural gene for premelibiase is the sequence starting at reference numeral 10 with the atg codon equivalent to the initial methionine of premelibiase and ends with the tct codon at reference numeral 12 equivalent to serine, and immediately preceding the tga translation-stop sequence. This sequence consists of 1,413 nucleotides as shown in Figure 1. Before reference numeral 10 and beyond reference numeral 12 , translation of amino acids does not occur, and the single row represents a nucleotide sequence. In the naturally occurring material, atg starting codon 16 is preceded by a promoter DNA sequence (included in the sequence between reference numerals 20 and 10) to which reference is made hereinafter. The nucleotide sequence, contained within that for premelibiase, which encodes secreted melibiase, is that sequence illustrated in Fig. 1 commencing at reference numeral 14 with the codon gtg, equivalent to valine at the N-terminus of the secreted melibiase, and continuing to the end 12 with the tct codon, equivalent to serine. This sequence is 1,359 nucleotides in length. The entire protein illustrated in Fig. 1 has a molecular weight of 52,101 (between reference numerals 10 - 12) .

The nucleotide sequence from reference numerals 10-14 indicated on Fig. 1 encodes the pre-region, or signal peptide, of premelibiase. This sequence starts with the atg codon equivalent to the initial methionine 16 of premelibiase at reference numeral 10 , and ends with the ggg codon at reference numeral 14, equivalent to the qlycine residue 18 adjacent to the start of the secreted melibiase. This sequence consists of 54 nucleotides.

This signal sequence for melibiase has been found to have an unexpectedly large ability to direct the secretion from yeast cells of gene products coded by regions associated with the MEL sequence. Whilst it is not intended that the invention should in any way be limited or bound by any particular theory or mode of operation, it is postulated that the MEL signal has some ability to direct gene products produced under its influence into superior secretion pathways or via other modes of efficient operation.

Preferred promoters which are useful either in connection with the signal sequence described above or independently thereof are those which are capable of promoting transcription, in a regulatable fashion, of the premelibiase gene native to S. carlsbergensis. Preferably, the promoter region is identified in and derived from this yeast according to procedures described herein although it may be synthesized to provide a nucleotide sequence analogous to the naturally occurring one.

Selection of this particular promoter as one component in a vector construct provides distinct advantages. The premelibiase promoter region regulates transcription in response to certain proteins, including the products of the GAL4 and GAL80 genes. Specifically an increase in the level of expression of the GAL4 gene within the host containing the promoter region has been demonstrated to enhance the level of expression of a structural gene over which the promoter exerts its transcription-promoting function. Also, it has been demonstrated that where normal GAL80 product production within the host containing the promoter region is disrupted, expression of genes associated with the promoter region is enhanced and further that expression of such genes is no longer dependent on the presence of galactose in the medium as is the case in the presence of the GAL80 protein.

The promoter sequence is shown in Figure 1, between reference numerals 2 0 and 10.

With the knowledge of the full and complete sequences of the signal sequence and the promoter region, respectively of the melibiase gene, these sequences can either be identified and isolated from yeast cells, in whole or in part, and/or produced by synthetic methods.

Then, the promoter region, for example, may be integrated into a vector and operatively linked either with a polypeptide coding region per se, where only expression is desired or with a signal sequence for secretion which in turn is operatively linked to and in the same reading frame as a coding region which encodes a desired protein. In this latter construct, the signal sequence is preferably as particularly described herein.

Examples of host yeasts in which this process may be used include S . carlsbe rgensis , S. ce revisiae , Saccharomycopsis lipolytica (also known as Candida lipolytica or Yarrowia lipolytica) etc. It will be understood that S. carlsbergensis is also known as S. uvarum.

Examples of genes useful in the constructs according to the invention are genes foreign to the aforementioned yeasts, include human, animal and bacterial genes, such as those for human proteins and peptides such as interferons, insulins, hormones, growth factors, enzymes etc., bacterial and fungal proteins such as cellulases etc., plant proteins, etc., peptide flavors and enhancers, antigens, etc.

Suitable expression vehicles or plasmids for introducing the constructs of the invention are those which contain genes or portions thereof appropriate to their maintenance and selection in the chosen host yeast and in E. coli, and yeast sequences appropriate to expression in the inserted foreign sequence, namely a strong yeast promoter, to ensure efficient transcription of the inserted nucleotide sequence, and a yeast sequence downstream of the insert to ensure adequate transcription termination. Specific examples of such vehicles are known and commercially available - - see for example reference (6) listed herein. In addition, yeast sequences appropriate to the expression of the product, other than the promoter and signal sequence described here, may include a transcription terminator etc. as well as such modifications as are required to create the desired construct in a functional form.

The techniques for preparing and replicating the appropriate vector plasmids, and introducing them into the host yeast cells, are known and are within the general skill of the art in this field.

Embodiments of the invention are further described in the following specific examples. Example 1 - Nucleotide sequencing of promoter region, signal sequence, coding region and flanking regions of melibiase and premelibiase

As shown by Post-Beittenmiller et al (5), plasmid PMP550, which is plasmid Yep24 (7 and 26) containing an additional fragment of yeast DNA, confers the melibiase-positive phenotype on otherwise melibiase-negative yeast when introduced into them by transformation, implying that the DNA insert contained in pMP550, an approximately 2.9 kilobase EcoRI to Bam HI fragment from a melibiase - positive yeast, contains the structural gene for premelibiase. Plasmid pMP550, obtained from J.E. Hopper of Pennsylvania State University, was prepared for sequencing of the DNA insert as follows: two μg pMP550, and 0.2 μg of the replicative form (RF) of phage M13 mpl8 (Norrander et al, 8) were each digested with EcoRI and Bam HI (New England Biolabs) and electrophoresed in 1 percent w/v low melting-point agarose. Gel slices containing the approximately 2.9 kilobase pMP550 fragment and the approximately 7.2 kilobase M13mp18 fragment were melted at 65°C to release the DNA, and a few μl of each mixed together and ligated at 12°C for 3h in 70mM Tris. HC1 buffer (pH 7.5), 7mM MgCl₂, 70 μM ATP, and 0.9 units T4 DNA ligase, in 20 μ l . The ligated DNA was used to transform E. coli strain JM109 (9) which had been made competent by the method of Hanahan (10) . Six of the plaques which appeared from the transformation were selected and the DNA isolated and checked for the presence of the pMP550 DNA fragment. All were of the appropriate structure; one was selected for sequencing, which was done as follows: RF DNA was prepared from E. coli JM 109 transfected with the selected phaqe, and 20 μg digested to completion with Eco RI and Pst I

(New England Biolabs), ethanol precipitated, and dissolved in

66mM Tris HCl buffer (pH 8.0) containing 0.66mM Mg Cl₂.

Following exactly the method of Henikoff (11), the DNA was digested with exonuclease III (PL-Pharmacia) and, at intervals, samples removed and treated with nuclease SI (BRL) E. coli DNA polymerase I (Klenow fragment) (Boehrinqer Mannheim), and T4 DNA ligase (Boehringer Mannheim), to generate a series of circular

DNA molecules with a decreasing size of the pMP550 insert sequence adjoined to the Ml3mpl8 sequence. Following recovery of the ligated DNA's by transfection of E. coli JM 109, and anelectrophoretic analysis of the size of the pMP550 DNA remaining in each, single-stranded DNA was prepared from selected phage and sequenced by the dideoxy technique (12), as detailed in the

BRL manual (13). The sequences obtained were combined using the program described by Larson and Messing (14), and further analysed using the programs of Devereux et al (15). Two thousand eight hundred and sixty-four nucleotides comprising the entire Eco RI to Bam HI fragment initially obtained from pMP550, were sequenced. Figure 1 shows this sequence, extending from the complete Eco RI site 28 through to the complete Bam HI site

25. Figure 1 also shows the predicted amino acid sequence of premelibiase, by translation of the long open reading frame. The calculated molecular weiqht of premelibiase, from the nucleotide sequence, is 52,101. Based on the known N-terminal amino acid sequence of secreted melibiase (see Example 2), the amino acids from methionine represented by reference numeral 16, to glycine represented by the reference numeral 18 comprise the melibiase signal peptide. Therefore, the calculated molecular wt. of secreted melibiase is 50,104. This agrees well with the molecular weight of melibiase that was purified from culture supernatants of yeast transformed with plasmid pMP550

(Example 3) and deglycosylated, i.e. approximately 50,000.

The sequence of the remainder of the S. carlsbergensis flank was determined as follows: two μg pMP550 and 2 μg of M13mpl8 were each digested with Xmal and Sphl (New England Biolabs) and electrophoresed in 0.8 percent w/v low melting-point agarose. Gel slices containing the approximately 1.8 kilobase pMP550 fragment and the approximately 7.2 kilobase M13mp18 fraqment were melted and a few μl of each mixed, ligated and used to transform JM109 essentially as described above. One plaque was selected for further work after verification of the presence of the required insert.

A series of deletion clones were made extending through the insert essentially according to the method of Dale et al (1985). Briefly, 5 μg of single-stranded DNA were hybridized with a primer (RD29), and then digested with Hind III overnight at 50°C. The 3' end of the single-stranded DNA was deleted to varying degrees with T4 DNA polymerase (New England Biolabs), tailed with terminal transferase (IBI) and dATP (Sigma), annealed to RD29 primer and ligated. Following transfection of

JM109, and an electrophoretic analysis of the size of the pMP550

DNA remaining in each of several phage clones, single-stranded

DNA was prepared and sequenced as described above. A further nine-hundred and twenty-nine nucleotides of sequence from pMP550 were determined by this method. The two methods therefore yielded the complete sequence of the S. carlsbergensis-derived

DNA in pMP550.

Example 2 - Amino Acid Sequencing of Melibiase

2.1 Purification of Melibiase

2.1.1 Preparation of a Yeast that Secretes Melibiase

Plasmid pMP550 was introduced into the melibiase-negative yeast strain S . cerevisiae 284 (α,leu 2-3, leu 2-112, ura 3-52, ade 1, mel°), supplied by J. E. Hopper, Pennsylvania State University, by transformation, according to the method described by Ito et al (16). The transformed cells were spread onto agar plates that select for the growth of cells independent of exogenous uracil and incubated for 2-3 days at 30°C. Composition per litre of this medium, designated selective medium I, was: glucose, 20g; agar, 17g; yeast nitrogen base without amino acids, 6.7g; succinic acid, 5.8g; dipotassium hydrogen phosphate, 8.7g, adenine, arginine, histidine, isoleucine, leucine, methionine, tryptophan and tyrosine, 0.02g each, lysine, 0.03g; phenylalanine, 0.05g; threonine, 0.10g and valine, 0.15g; pH was 4.6-4.8. Colonies that appeared on selective medium I were transferred to fresh selective medium I, allowed to grow at 30°C, then transferred to agar plates containing selective medium II, that is selective medium I with 20g/L lactic acid, 30g/L glycerol, and 20g/L galactose, in the place of 20g/L glucose. This was to allow induction by galactose of melibiase synthesis (5) . The colonies were then used to inoculate liquid medium consisting of selective medium II without agar. Following cell growth at

30°C, the cultures were assayed for secreted melibiase by the following procedure: 500 μl culture was centrifuged to pellet the cells, then 10 μ l of the supernatant fluid added to 190 μl pH 4 buffer (30mM citric acid, 38mM dibasic sodium phosphate) containing 12.5mM p-nitrophenyl α -D-galactoside (p-NPG; Sigma

Chemical Co.) . After 10 min at 30°C, 800 μl 0.1 M sodium carbonate was added and the yellow color measured spectrophotometrically at 407 nm. All uracil-independent transformants tested produced secreted melibiase, while S. cerevisiae strain 284 itself produced no secreted melibiase.

One transformant, designated PMP550. Sc284.1, was selected for melibiase purification.

2.1.2 - Purification of Melibiase

Ten litres of selective medium II without agar was inoculated with yeast transformant pMP550. Sc284.1 and incubated at 30°C with shaking until a maximum optical density at 660 nm had been reached. The culture was centrifuged at 3000xg, to remove the cells, and the supernatant liquid concentrated to 100 ml in an Amicon ultrafiltration unit fitted with a PM-10 membrane. The concentrate was clarified by centrifugation at

40,000xg for 30 min. then freeze-driied. The material was then dissolved in 25 ml of 50mM Tris. HCl buffer (pH 7.5) (buffer) and dialysed against buffer (3 changes at 2.5 litres per chanqe). The material was filtered through a 0.45 micron membrane, to give a solution containing 16.6 mg protein and 3534 units of melibiase (17). Melibiase activity was then separated from the bulk of other proteins on a Pharmacia FPLC system fitted with a Mono Q anion exchange column (HR5/5). The melibiase, which remained bound in the column, was eluted by application of a linear sodium chloride gradient, 125 to 245 mM in buffer. One ml fractions were collected, assayed, and the active fractions pooled; 7.61 mg protein was recovered, containing 3437 units of melibiase. The material was lyophilized, dissolved in 6.8 ml buffer, and chromatographed, in

2 aliquots, on a Biogel A-1.5M column (1.5 x 90.0 cm) .

Fractions containing the highest melibiase activity were pooled and contained 3.8 mg protein and 2302 units of melibiase. When electrophoresed in an SDS-polyacrylamide gel according to

Laemmli (18) this material contained 2 bands, representing 98% of the protein. One band was diffuse, encompassing the size range 75 to 100 kdaltons, midpoint 82 kdaltons, the second sharp and with a calculated molecular weight of 70 kdaltons. In order to determine the molecular weight of the melibiase polypeptide chain, and to determine if it was pure, covalently-linked carbohydrate, known to be part of the melibiase secreted from S. carlsbergensis (Lazo et al, 19; Lazo et al, 20) was enzymatically removed from the sample with endo-ß-N-acetylglucosaminidase H (endo H) from Streptomyces griseus, as follows; 40 μq of protein was incubated with 5 milliunits of endo H (Boehringer Mannheim Canada) at 37°C for 25h in a 40 μl reaction containing 0.05 M citric acid and 0.1 M disodium hydrogen phosphate. The sample was lyophilized then analysed by SDS-polyacrylamide gel electrophoresis. A single protein band, with a calculated molecular weight of 50 kdaltons, was observed.

2.2. Amino Acid Sequencing

Pure melibiase, without endo H treatment to remove carbohydrate side-chains, was subjected to an analysis of the first 10 N-terminal amino acid residues, by derivatisation with phenylthiohydantoin. The sequence, determined by the amino acid sequencing facility in the Department of Biochemistry, University of Toronto, was: gly or N-terminus a la- X-pro- ser- tyr-a sn-gly- leu-gly- leu or val The reason for the amhiquity at the first 2 residues is not understood. Nevertheless, the identity of the sequence of the remaining 8 residues with those predicted by the nucleotide sequencing of the DNA insert in plasmid pMP550 (Figure 1) leads to the conclusion that the amino acid sequence of secreted melibiase, starting at the N-terminus, is as follows:

val- s er-pro- ser- tyr- a sn- g ly- leu- g ly- l eu- - -

2.3 Structure of the Melibiase Signal Sequence

Based on the amino acid sequence of the N-terminus of secreted melibiase, and the amino acid sequence predicted by the nucleotide sequencing for premelibiase, the melibiase signal sequence, starting at the N-terminus, is defined as follows:

met-phe-ala-phe-tyr-phe-leu-thr-ala-cys-ile-ser- leu-lys-gly-val-phe-gly

Example 3 - Use of an expression cassette containing the melibiase promoter and signal sequence to effect production from Saccharomyces cerevisiae of a novel extracellular protein

3.1 Construction of the cassette

The expression cassette is illustrated schematically in Figure 2. Its essential element are (i) flanking restriction sites, in this example Bam HI, which allow for its ready insertion into appropriate E. coli/S. cerevisiae shuttle vectors; (ii) the DNA sequences for the melibiase promoter 32

(which include upstream activation sequences) and signal peptide

36 in a form such that a coding sequence other than that for melibiase (shown as 38) can be placed in operative association with them. In this example, this is accomplished by creating a

Bgl II site adjacent to the sequence encoding the melibiase signal peptide. This allows a gene fusion to be made between the codons that specify the signal peptide and these that specify the novel protein. The operative association then is promoter: signal peptide codons: novel protein codons. The segment designated 34 in Figure 2 includes additional sequences such as those specifying transcription, termination, cleavage and polyadenylation. The expression cassette was constructed as described below and as shown in Figure 3 to which reference is now made, using conventional recombinant DNA techniques.

Plasmid pMP550 (Figure 3) is plasmid pBR322 (Bolivar et al; 22) containing a 4.3 kilobase DNA fragment 40 which includes the melibiase gene and its flanking DNA sequences (Post.

Beittenmiller et al; 5, and Sumner-Smith et al; 23). The nucleotide sequence of fragment 40 is specified in Figure 1, as bounded by the Hind III site 30 and Bam HI site 25. It should be noted that the first 346 nucleotides (reference numerals 30 to 20 in Figure 1) are derived from plasmid pBR322. This is because of the cloning strategy used, and is incidental to the function of the expression cassette derived from the 4.3 kilobase DNA fragment. Plasmid pMP550 was restricted with Hind

III and Bam HI and the 4.3 kilobase fragment 40 cloned into the plasmid pUC8 (Vieira and Messing; 24) between its Hind III and

Bam HI sites. The resulting plasmid, pi, was restricted with

Hind III, the DNA ends repaired by filling-in using the Klenow fragment of Ε . coli DNA polymerase I, and ligated, by the method of Lathe et al (25), to a synthetic oligonucleotide 42 of the sequence 5' CCGGATCCGG 3'.

3' GGCCTAGGCC 5' The product of this ligation, plasmid p2, consisted of pUC8 containing the 4.3 kilobase fragment 40 from pMP550 now bounded by Bam HI sites. Plasmid p2 was restricted with Sphl which cuts uniquely within the DNA sequence that encodes the melibiase signal peptide. The linear ized plasmid was then ligated to a synthetic oligonucleotide 42 of the sequence

5' CATCAGTTTGAAGGGCGTTTTTGGGAAGATCTCTGCATG 3' 3' GTACGTAGTCAAACTTCCCGCAAAAACCCTTCTAGAGAC 5'

by the method of Lathe et al (25) . Recombinant plasmids were sequenced across the position of the inserted oligonucleotide and plasmid p3, which contained the oligonucleotide in the orientation shown in Figure 2, was selected. Plasmid p3 was then restricted with Bam HI and the 4.3 kb fragment cloned into the Bam HI site of the shuttle plasmid Yep 24(26), generating plasmid p4. Plasmid p4 contains the 4.3 kilobase fragment originally from pMP550 but the DNA sequence encoding the melibiase signal peptide was now modified, as specified in Figure 2. The 4.3 kilobase Bam HI fragment 40 in p4 (Figure 3) constitutes the expression cassette which is an example of a preferred embodiment of the present invention. 3.2 Use of the expression cassette to engineer production of a novel extracellular protein in yeast cultures

A DNA sequence known to encode a celluase ( an endoqlucanase or endo-1, 4-β-D-glucanase, EC 3.2.1.4) in the soil bacterium Cellulomonas fimi ATCC 484 was used to demonstrate the utility of the expression cassette in Saccharomyces cerevisiae. As depicted in Figure 4, the endoqlucanase codinq sequence is contained in a 2.4 kilobase Bam HI fraqment 46 from the plasmid pEC2.2 (Figure 4; Skipper et al; 27) provided by Dr. Robert C. Miller, University of British Columbia. The coding sequence is known to be missing the codons for the first 76 amino-terminal amino acids of the bacterial protein, including the 31 amino acid bacterial signal peptide (28). Previous experiments had shown that Saccharomyces cer-gy-isiae cells transformed with a plasmid containing the 2.4 kilobase Bam HI fraqment from pEC2.2 in an operative association with the yeast alcohol dehydroqenase 1 promoter produced a very low level of extracellular endoglucanase activity (Skipper et al; 27) . The addition of an operative signal peptide provided by codons specifying the amino terminus of the yeast preprotoxin protein resulted in a greatly enhanced level of extracellular endoglucanase activity (Skipper et al; 27).

The 2.4 kilobase Bam HI fragment 46 from pEC2.2 was inserted into plasmid p4 (Figure 4) by cohesive-end ligation at the Bgl II site of this plasmid. Recombinant plasmids conta ining the endoglucanase insert were restricted with Bgl II, to identify those containing the insert in the appropriate orientation, and several of those sequenced across the melibiase/endoqlucanase junction to identify those containing the predicted qene fusion. One of these plasmids, p5, was selected for further analysis. It contained the endoqlucanase coding sequence fused to the sequence encoding the melibiase siqnal peptide, in the correct translational reading frame (see

Figure 4a) . One additional codon (AAG , lysine) was present between the terminal codon (GGG, qlycine) of the melibiase signal peptide and the 77th codon (ATC, isoleucine) of endoglucanase, as expected from the construction (Figure 5).

The 6.7 kilobase Bam HI fragment 48 in p5, containing 4.3 kilobases of the expression cassette and the 2.4 kilobase endoglucanase codinq sequence, comprised the operative melibiaserendoqlucanase qene fusion, as shown in Figure 4a.

Plasmid p5 is Yep24 containinq the melibiase: endouiucanase gene fusion. Since Yep24 carries the URA3 gene, it can be used to transform ura3 yeast strains to uracil-independence. To extend the range of recipient

Saccharomyces cerevisiae strains for assessment of endoglucanase expression from the melibiase elements the 6.7 kilobase Bam HI fragment 48 from p5 was cloned into each of 2 further shuttle vect ors , Yep21 and Yep 24.TRP1 ARS , generat ing plasmids p6 and p7, respectively (Figure 5). Plasmid Yep21 (26) carries the yeast LEU2 gene for transformation of leu2 strains to leucine-independence. Plasmid Yep24.TRP1 ARS was constructed from plasmids Yep24 (26) and Yrp7 (26), as illustrated in Figure 5. It is useful for transferring trp1 yeast to tryptophan-independence, and ura3 yeast to uracil-independence.

To test for endoglucanase expression in yeast,

Saccharomyces cerevisiae strain 20B12 (α, trpl, pep4.3; Jones,

29) was transformed with plasmid p7 to tryptophan-independence and assayed for extracellular endoglucanase activity. Two assays were used: a qualitative assay based on the direct staining of agar plates (containing the substrate carboxymethylcellulose) on which the transformants had been grown, and a qualitative assay based on measurement of carboxymethylcellulose hydrolysis to reducing-sugar equivalents. In the plate assay, transformants were grown at

30°C on agar plates containing *C tryptophan medium (30), 1.2 peitrent agar, 0.9 percent carboxymethylcellulose (high viscocity, Sigma Chemical Co.), 2 percent glucose, and 50 mM sodium phosphate buffer; final pH was 6.8 to 7.0. The yeast colonies were washed off with water, then the plates were stained with Congo Red (2 mg/ml) for a few minutes and then rinsed with 1M sodium chloride. Under these conditions, endoglucanase activity is indicated by a colourless zone, since β-1,4 glucans of seven or fewer glucose residues do not bind

Congo Red (Wood, 31) . All transformants produced extracellular endoglucanase by this assay. In the quantitative assay, transformants were grown in *tryptophan medium containing 2 percent glucose and 50mM sodium phosphate; (pH 6.8 to 7.0). The cultures were centrifuged to removed the cells, the supernatant fluids dialysed against water, and the dialysates lyophilized to dryness. Following reconstitution in water the samples were assayed for Production of reducing-sugar equivalent from carobxymethy lcellulose (low viscocity, Siqma Chemical Co.) by the method of Miller et al (32) , as described by Skipper et al

(27) . All transformants produced extracellular endoglucanase by this assay. A typical transformant produced 6.4 units endoglucanase oer ml culture, or 2.32 units oer 10⁸ cells, where 1 unit produces 1 μmole reducing suqar per minute at 37°C.

Example 4 - Use of different configurations of yeast chromosomal genes to effect altered regulation of the expression of a novel protein from the expression cassette

In common with the galactose pathway genes GALl, GAL2 , GAL7 , and GAL10 expression of the MEL-1 (melibiase) gene is induced by galactose and repressed by glucose. This control by the carbon source appears to be mediated by the products of the GAL4 and GAL80 genes. GAL4 produces a product which is required for transcription of all the genes. The action of GAL4 is inhibited by the GAL80 product, which is thought to bind to it or to its site of action on the DNA in the absence of galactose. In the presence of galactose the binding of GAL80 product is prevented and all the qenes are turned on. The site o f b i nding o f the GAL4 produc t is in the 5 ' -flanking reg ion of each targ et gene , a reg ion call UAS or upstream activation sequence (see Sumner-Smith et a l; 23 , for br ief review) . Manipulations of the GAL4 and GAL80 genes are therefore expected to modify the regulation of expression of proteins dependent on the melibiase promoter, of which extracellular endoglucanase is used as an example here.

Three Saccharomyces cerevisiae strains: Sc284, Sc295 and Sc300 (Table 1) were used to examine the effects of GAL4 and GAL80 modifications on endoglucanase production from plasmids containing the operative melibiase: endoglucanase fusion described above. These strains were provided by Dr. J.E. Hopper, Pennsylvania State University. Strain Sc284 (leu 2-3, leu 2-112, ura 3-52, ade 1, mel°) is wild type with respect to GAL4 and GAL80. Strain Sc295 (ura 3-52, ade 1, gal80-deletion) lacks the GAL80 gene, so that galactose is unnecessary for full expression of the melibiase promoter. Strain Sc300 (leu 2-3, leu 2-112, ade 1, gal80-deletion, gall-ga17-gal10-deletion, ADH1:GAL4) lacks GAL80 and in addition has 2 modifications intended to increase the availability of the GAL4 protein to the melibiase promoter. First, the GAL1, GAL7 and GAL10 gene cluster is deleted. Second, an additional copy of the GAL4 gene is present, expressed from the strong promoter of the yeast alcohol dehydrogenase 1 gene; this is estimated to increase the synthesis of GAL4 protein by a factor of 150 over cells containing the normal single chromosomal GAL4 gene. The phenotype of strain Sc300 with respect to melibiase promoter function is not known in any detail, but it could be reasonably predicted to be both constitutive promoter function, due to the GAL80 deletion, and an elevated maximal function, due to the increased availability of GAL4 product.

The 3 yeast strains were each transformed with an operative expression cassette/endoglucanase gene fusion, contained in an E . coli/S. cerevisiae shuttle vecto r appropriate to the select ion of t ransfo rmants (Table 1) . Transformants were then grown in liquid media containing carbon sources either non-inducing, inducing or repressing for the melibase promoter in a normal GAL4 , GAL80 strain, then assayed for extracellular endoglucanase activity. The results are illustrated in Table 1 and discussed below:

1. Transformant 113.1 (strain Sc284) . The absolute amount of extracellular endoglucanase produced by this transformant was low, making it difficult to assess the regulatory effect of the various combinations of carbon sources with any accuraccy. However the response of endoglucanase expression to the carbonsources was consistent with that expected for a strain carrying wild-type GAL4 and GAL80 genes: induction by galactose, repression by glucose.

2. Transformant 149.1 (strain Sc295). Endoglucanase expression was maximal in the absence of galactose, consistent with the expected effect of the GAL80 deletion, and greatly reduced by 2 percent glucose, consistent with normal glucose repression. The presence of galactose in fact partially repressed endoglucanase synthesis, an effect that has also been seen in GAL80 deletion strains expressing extracellular melibiase (unpublished obserations). 3. Transformant 142.11 (strain Sc300) . Endoglucanase expression in this strain was the highest seen in any yeast recipient. Unexpectedly, galactose was required for maximal endoglucanase expression, and glucose was only partially effective as a represser of its synthesis. Both these unpredicted effects are presumably due to the increased availability of the GAL4 product.

These experiments demonstrated clearly that manipulations of genes known to affect melibiase promoter function affected expression of extracellular endoglucanase, so providing additional evidence of the utility of the expression cassette in producing this and other novel protein from S. cerevisiae.

The following plasmids were deposited with ATCC before the filing date hereof, i.e. on December 16, 1985 having been designated as defined below and having been allotted the following accession number. All are contained within a suitable

S. coli host.

Plasmid Designation ATCC Accession No. p4 NAS-1 53360 p5 NAS-2 53361 p6 NAS-3 53362 p7 NAS-4 53363

Cultures containing these constructs are also currently maintained in a permanently viable state at the laboratories of

Allelix Inc., 6850 Goreway Drive, Mississauga, Ontario, Canada and are identified by the designations used herein.

r

(l) Table 1

(2 ) NI (non-inducing) in 2 percent lactic acid, 3 percent glycerol, 0.05 percent glucose; I (inducinq) is NI plus 2 percent qalactose; R (repressing) is I plus 2 percent glucose.

(3 ) In each case, the carbon sources were using in selective media, either *C uracil (33) for transformants 113.1 and 142.11, or *C leucine (34) for transformant 142.11. Final pH was 6.8 to 7.0.

(4) Cultures were centrifuged and the supernatant fluids dialysed against water, lyophilized, reconstituted in water, and assayed for production of reducing-sugar equivalents from carboxymethylcellulose, as described (Skipper et al; 27). One unit endoglucanase produces 1 μmole reducing-sugar per minute at 37°C.

TABLE OF REFERENCES CITED IN DISCLOSURE

1) Hitzeman, R.A., Leung, J.W., Perry, L.J., Kohr, W.J., Levine, H.L., Goeddel, D.V. (1983). Science 219, 620-625

2) Emr, S.A. Schekman, R., Flessel, M.C., Thorner, J. (1983) Proc. Natl. Acad. Sci. USA 80, 7080-7084.

3) Brake, A.J., Merryweather, J.P., Coit, A.G., Heberlein, U.A., Masiarz, F.R., Mullenbach, G.T., Urdea, M.S., Valenzuela, P., Bare, P.J. (1984). Proc. Natl. Acad. Sci. USA 81, 4642-4646.

4) Bitter, G.A., Chen, K.K. Banks, A.R., Lai, P-H., (1984) . Proc. Natl. Acad. Sci USA 81, 5330-5334.

5) Post-Beittenmiller, M., Hamilton, R.W., Hopper, J.E.

(1984). Mol. Cell Biol. 4, 1238-1245.

6) Goff, C.G., Moir, D.T., Kohno, T., Gravius, T.C., Smith, R.A., Yamasaki, E., Taunton-Rigby, A. (1984). Gene 27, 35-46.

7) Botstein, D., Falco, S.C., Stewart, S.E., Brennan, M., Scherer, S., Stinchcomb, D.T., Struhl, K., Davis, R.W.

(1979) Gene 8, 17-24. 8) Norrander, J., Kempe, T., Messing, J. (1983) . Gene 26, 101-106.

9) Messing, J. (unpublished).

10) Hanahan, D. (1983). J. Mol. Biol. 116, 557-580.

11) Henikoff, S. (1984) . Gene 28, 351-359.

12) Sanger, F. , Nicklen, S., Coulson, A.R. (1977). Proc. Natl. Acad. Sci USA 74, 5463-5467.

13) BRL M13mp7 cloning/dideoxy sequencing user manual. Bethesda Research Laboratories, Inc., Gaithersberg, Maryland 20877.

14) Larson, R., Messing, J. (1982). Nucl. Acids Res. 10, 39-49.

15) Devereux, J., Haeberli, P., Smithies, O. (1984). Nucl. Acids Res. 12, 387-395.

16) Ito, H., Fukada, K., Murata, K., Kimura, A. (1983). J. Bacteriol. 153, 163-168.

17) 1 unit of melibiase hydrolyses 1 μmole per minute of p-n itropheny l - α - D-galactos ide a t 30 °C . 18) Laemmli, U.K. (1970). Nature 227, 680-685.

19) Lazo, P.S., Ochoa, A.G., Gascon, S. (1977). Eur. J. Biochem. 77, 375-382.

20) Lazo, P.S. Ochoa, A.G., Garcon, S. (1978). Arch. Biochem. Biophys. 191, 316-325.

21) Dale, R.M.K., McClure, B .A. , Houchins, J.P. Plasmid 13, 31-40. (Manuscript received before publication).

22) Bolivar, F. , Rodriquez, R.L., Grene, P.J., Betlach, M.C. , Heyneker, H.L. , Boyer, H.W., Crosa, J.H., Falkow, S. (1977) . Gene 2, 95-113.

23) Sumner-Smith, M. , Bozzato, R.P., Skipper, N., Davies,

R.W. , Hopper, J.E. , (1985). Gene 36, 333-360.

24) Vieiera, J. , Messing, J. (1982) . Gene 19, 259-268.

25) Lathe, R. , Keny, M.P. , Skory, S. , Lecocq, J.P. (1984) . DNA 3, 173-182.

26) American Type Culture Collection, Rockville, Maryland. 27) Skipper, N., Sutherland, M., Davies, R.W., Kilburn, D.,

Miller, R.C, Warren, A., Wong , R. (1985). Science 230, 958-960.

28) Personal communication from R.C. Miller and R. Wong, University of British Columbia.

29) Jones, E. (1978) . Genetics 85, 23-33.

30) *C tryptophan medium contained, in grams per litre, yeast nitrogen base, 6.7; adenine, arginine, isoleucine, tyrosine, uracil and leucine 0.02 each; lysine, 0.03; phenylalanine, 0.05; threonine, 0.10; valine, 0.15; Na₂HPO₄, 3.0; KH₂PO₄, 3.0;

KH₂PO₄, 1.5.

31) Wood, P.J. (1980). Carbohydrate Research 85, 271-287.

32) Miller, G.L., Blum, R. , Glennon, W.E., Burton, A.L.

(1960). Anal. Biochem. 2, 127-132.

33) * uracil medium contained, in grams per litre, yeast nitrogen base, 6.7; adenine, arginine, isoleucine, tyrosine, tryptophan, and leucine, 0.02 each; lysine, 0.03; phenylalanine, 0.05; threonine, 0.10; valine, 0.15; Na₂HPO₄ , 3.0; KH₂PO₄, 1.5. 34) *leucine medium contained, in grams per litre, yeast nitrogen base, 6.7; adenine, arginine, isoleucine, tyrosine, tryptophan, and uracil, 0.02 each; lysine,

.0.03; phenylalanine, 0.05; threonine, 0.10; valine,

0.15; Na₂HPO₄, 3.0; KH₂PO₄, 1.5.

Claims

1. A DNA construct capable of expression in a yeast cell transformed therewith, said construct comprising

the promoter region of the melibiase gene and

a polypeptide coding region operatively associated therewith.

2. The construct according to claim 1 wherein the polypeptide coding region encodes other than melibiase.

3. The construct according to claim 2 wherein the polypeptide coding region is derived from the Cellulomonas fimi endoglucanase gene.

4. The construct according to claim 2 wherein the coding region is followed by and is operatively associated with a transcription-terminating region.

5. The construct according to claim 2 wherein the promoter reglon is derived from S. carlbergensis.

6. The construct according to claim 2 wherein the DNA sequence of the promoter region is substantially as shown in Figure 1 of the accompanying drawings.

7. A method for altering the expression characteristics of a yeast cell which comprises introducing into said yeast cell a DNA segment comprising the promoter region of the melibiase gene operatively linked to a polypeptide coding region.

8. The method according to claim 7 wherein the polypeptide coding region of said DNA segment encodes other than melibiase.

9. The method according to claim 8 wherein said yeast cell is selected from S . cerevisiae and S. carlsbergensis.

10. The method according to claim 8 wherein the promoter region has the nucleotide sequence substantially as shown in Figure 1 of the accompanying drawings.

11. The method according to claim 9 wherein the polypeptide coding region is derived from the Cellulomonas fimi endoglucanase gene.

12. A yeast cell containing the promoter region of the melibiase gene operatively associated with and exerting a transcription-controlling function over a coding region which codes for a polypeptide other than melibiase.

13. A yeast cell according to claim 12 which is selected from the species Saccharomyces cerevisiae and Saccharomyces carlsbergensis.

14. A yeast cell according to claim 12 wherein the polypeptide coding region is derived from the Cellulomonas fimi endoglucanase gene.

15. A DNA construct which comprises

a promoter region operative in a yeast cell

a signal sequence coding for a signal peptide that is functional in a yeast cell, operatively associated with the promoter region, and

a polypeptide coding region operatively linked to said signal sequence, wherein at least one of said signal sequence end said promoter region is a melibiase gene segment.

16. The construct according to claim 15 wherein said signal sequence is the signal sequence of the melibiase gene.

17. The construct according to claim 16 wherein the signal sequence codes for the amino acid sequence substantially as follows:

(N-terminus) met-phe-ala-phe-tyr-phe-leu-thr-ala-cys- ile-ser-leu-lys-gly-val-phe-gly.

18. The construct according to claim 16 wherein the signal sequence has a nucleotide sequence substantially as follows:

(5') ATG TTT GCT TTC TAC TTT CTC ACC GCA TGC ATC AGT TTG AAG GGC GTT TTT GGG (3')

19. The construct according to claim 16 wherein the signal sequence is derived from S. carlsbergensis.

20. The construct according to claim 15 wherein the promoter region is the melibiase promoter region.

21. The construct according to claim 20 whe rein the promoter region has the nucleotide sequence substantially as shown in Figure 1 of the accompanying drawings.

22. The construct according to claim 16 wherein the promoter region is the melibiase promoter region.

23. The construct according to claim 22 wherein the polypeptide coding region encodes other than melibiase.

24 The construct according, to claim 23 wherein the polypeptide coding region is derived from the Cellulomonas fimi endoglucanase gene.

25. The construct according to claim 16 wherein the promoter region is the melibiase promoter region and has the nucleotide sequence substantially as shown in Figure 1 of the accompanying drawings.

26. The construct according to claim 15 wherein the polypeptide coding region is followed by and is operatively associated with a transcription-terminating region.

27. A method for altering the expression and secretion characteristics of a yeast cell which comprises introducing into said yeast cell a construct as defined in any one of claims 15 -26 with the proviso that when both said promoter region and said signal sequence are segments of the melibiase gene then either the polypeptide coding region encodes other than melibiase or the construct is foreign to the yeast cell.

28. The method according to claim 27 wherein the yeast cell is of the species Saccharomyces cerevisiae.

29. The method according to claim 27 wherein the yeast cell is of the species Saccharomyces carlsbergensis.

30. A yeast cell containing the signal sequence of the melibiase gene operatively linked with a polypeptide coding region which codes for other than melibiase.

31. A yeast cell according to claim 30 wherein the promoter region is the promoter region of the melibiase gene.

32. A yeast cell according to claim 31 wherein the polypeptide coding region is derived from Cellulomonas fimi and encodes endoglucanase.

33. A yeast cell according to claim 30 which is of the species S . carlsbergensis.

34. A yeast cell according to claim 30 which is of the species S. cerevisiae.

35. A process for obtaining secretion of a polypeptide from a yeast cell which comprises growing, in a growth-promoting medium, a yeast cell transformed by a DNA construct comprising

a promoter region operative in a yeast cell

a signal sequence coding for a signal peptide that isfunctional in a yeast cell and is operatively associated with the promoter region, and

a polypeptide coding region encoding said polypeptide and operatively linked with said signal sequence, wherein at least one of said signal sequence and said promoter region is a melibiase gene segment.

36. The process according to claim 35 wherein the polypeptide coding region encodes other than melibiase.

37. The process according to claim 36 wherein said promoter region is the melibiase promoter region.

38. The process according to claim 37 wherein said signal sequence is the melibiase signal sequence.

39. The process according to claim 38 wherein the polypeptide coding region is derived from Cellulomonas fimi and encodes endoglucanase.

40. The process according to claim 37 wherein transcription of the signal sequence and ceding region is regulated by changing conditions in the growth medium which affect the activity of the GAL4 and/or GAL80 genes and their products.

41. The signal sequence of the melibiase gene native to yeast.

42. The signal sequence according to claim 41 which is native to S. carlsbergensis.

43. The signal sequence according to cla im 42 having a DNA sequence which codes for the amino acid sequence substantially as def ined below:

(N-terminus) met-phe-ala-phe-tyr-phe-leu-thr-ala-cys- ile-ser-leu-lys-gly-val-phe-gly .

44. The signal sequence according to claim 43 having the DNA sequence substantially as defined below:

(5') ATG TTT GCT TTC TAC TTT CTC ACC GCA TGC ATC AGT TTG AAG GGC GTT TTT GGG (3')

45. A DNA construct comprising a DNA segment containing the melibiase secretion signal sequence and defining restriction sites flanking said signal sequence for incorporation of additional DNA components into the construct.

46. The construct according to claim 45 wherein said DNA segment has the nucleotide sequence substantially as shown in Figure 4A.

47. The promoter region of the melibiase gene native to yeast.

48. The promoter region according to claim 47 which is native to S. carlsbergensis.

49. The promoter region according to claim 47 having the

DNA sequence substantially as shown in Figure 1 of the accompanying drawings.

50. A DNA construct comprising a DNA segment containing the melibiase promoter and defining a restriction site downstream of said promoter for incorporation of an additional DNA component into said vector.