GB2172890A - Epidermal growth factor production - Google Patents

Abstract

A DNA sequence encoding a fusion protein comprising a carrier protein linked through a Lys residue to an EGF or EGF analogue, which EGF or EGF analogue does not contain an internal Lys residue and from which fusion protein the EGF or EGF analogue is cleavable by a Lys-specific protease, is incorporated in a vector which is capable, in a transformed host, of expressing the fusion protein. The EGF or EGF analogue is released from the fusion protein by treatment with a Lys-specific protease. The EGF or EGF analogue is useful for depilating animals, especially defleecing sheep. <IMAGE>

Description

SPECIFICATION Epidermal growth factor production The present invention relates to the production of epidermal growth factor (EGF).

Gene synthesis and expression in microorganisms is a recent, but established technology. However, it is generally observed that expression of small polypeptides, such as insulin chains, directly in E. coli is extremely inefficient. Primarily, this is because proteolysis by intracellular proteases is rapid. However, it is also because of the necessity to construct an efficient ribosome binding site and surrounding sequences within the gene that are compatible with the desired amino terminal sequence. The rules governing the efficiency of translation initiation are not completely understood. In addition, an initiator methionine codon is required and the N-terminal methionine may not be efficiently removed by the host protein synthetic machinery, as, for example, is the case with human growth hormone, particularly at high levels of expression.

As an alternative to the expression of a mature polypeptide, a synthetic gene may be fused to part of a gene for a host protein that can itself be expressed at high levels in a controllable manner (EP-A-0001930).

The resultant fusion protein may well be stabilised against proteolysis. A polypeptide of interest may then be released by specific chemical or enzymic cleavage.

EGF can cause defleecing of sheep (GB-A-2082186). However, substantial quantities of the growth factor are required for this purpose, at a dose of 3 to 5 mg per sheep. EGF can be extracted from adult male mouse submaxillary glands. This murine EGF (mEGF) is a 53 residue polypeptide of known primary structure.

However, the amounts of mEGF which can be extracted via this route are insufficient.

We have now found that a fusion protein containing a lysine (Lys) link to mEGF can be expressed in E. coli and cleaved by Lys-specific proteolysis to release the mEGF. mEGF lacks Lys, and is thus resistant to the specific proteases. This approach has general applicability and enables the production of much larger amounts than has previously been possible of any EGF or EGF analogue which does not contain an internal Lys residue. Several alternative fusion protein constructions which did not embody this approach failed to yield mEGF.

According to the present invention, there is provided a DNA sequence encoding a fusion protein comprising a carrier protein linked through a Lys residue to an EGF or EGF analogue, which EGF or EGF analogue does not contain an internal Lys residue and from which fusion protein the EGF or EGF analogue is cleavable by a Lys-specific protease.

The invention further provides a vector which incorporates this DNA sequence and which is capable, in a transformed host, of expressing the fusion protein. A host transformed with such a vector also forms part of the invention.

The invention also provides a process for the preparation of an EGF or EGF analogue which does not contain an internal Lys residue, which process comprises culturing such a transformed host such that the said fusion protein is expressed and treating the fusion protein with a Lys-specific protease to release the EGF or EGF analogue. The EGF or EGF analogue thus obtained can be used to depilate an animal, especially for defleecing sheep, by administration to the animal.

A DNA sequence capable of expressing a fusion protein containing a Lys link to an EGF or EGF analogue which does not contain Lys is provided in a vector, e.g. a plasmid. The codons for the Lys residue and the EGF or EGF analogue are provided in the same reading frame as the codons for the carrier protein. These codons are preceded by a promoter. A host is transformed with the vector. Generally, the host is a bacterial host such as E. coli. The fusion protein is expressed in the host.

We have found that a suitable DNA sequence comprises a promoter followed by a carrier protein gene linked through a Lys codon to the gene for the EGF or EGF analogue which is in turn immediately followed by a stop codon. Such a sequence may be produced by first constructing a synthetic gene in which the gene for the EGF or EGF analogue is immediately preceded by a Lys codon and immediately followed by a stop codon. This synthetic gene is ligated with a DNA sequence comprising a promoter for the carrier protein and the carrier protein gene. Preferably, the carrier protein gene and the beginning of the synthetic gene (the Lys codon end) are provided with matching sticky ends for this purpose. The ligation is arranged so that the codons of the synthetic gene are in the correct reading frame with respect of the promoter and carrier protein gene.

Alternatively, the fusion protein encoded by a DNA sequence may comprise two or more runs each of an EGF or EGF analogue. One run is linked to the next through a Lys residue so that treatment of the fusion protein with a Lys-specific protease releases each EGF or EGF analogue. For this embodiment, a synthetic gene is constructed in which DNA sequences coding for each EGF or EGF analogue are linked in tandem through Lys codons. Repeats of a DNA sequence of an EGF or EGF analogue may therefore be linked via Lys codons. Such a synthetic gene may then be ligated with a DNA sequence comprising a promoter for the carrier protein and the carrier protein gene as above.

It is important that the EGF or EGF analogue forming the fusion protein does not contain an internal lysine residue. Preferably the EGF is mEGF. EGF analogues are synthetic and natural derivatives of the EGF polypeptide family which contain a sequence of amino acids (or amino acid substitutes) effective in regulating hair growth and especially in defleecing sheep. EGF analogues may therefore be fragments of an EGF, for example mEGF1-45, mEGF1-47, mEGF1-48 or mEGF1-51. Also, an analogue may contain an amino acid residue, in the place of a naturally-occurring residue, which does not affect the ability of the analogue to regulate hair growth and in particular to act as a defleecing agent for sheep.

The synthetic gene may be constructed by standard methods. Oligodeoxynucleotides for the gene may be synthesized by a manual solid phase method, for example similar to that described by Sproat and Bannwarth (1983). These oligodeoxynucleotides except those yielding the 5'-OH ends of the assembled gene are then fully kinased with an excess of [832P]-ATP and polynucleotide kinase. Ligation experiments are conducted using different subsets of oligodeoxynucleotides, intermediates thus obtained are purified, for example by polyacrylamide gel electrophoresis, and then ligated together to form the synthetic gene in which the gene for the EGF or EGF analogue is sandwiched between a Lys codon and a stop codon. The methods used may be similar to those previously described by Smith eta/. (1982) and Edge etal. (1981).

Generally, the synthetic gene is provided with sticky ends to assist ligation into a suitable vector. A host is then transformed with the vector and colonies containing the synthetic gene selected.

The synthetic gene is ligated with a DNA sequence comprising a promoter for a carrier protein and the carrier protein gene. The sequence may be a separate fragment or part of a vector. The sequence may also incorporate an operator and/or attenuator depending upon the expression system which is used. We have used part of a Trp operon containing the trp promoter, attenuator and part of the TrpE gene. However, the synthetic gene may be placed on the control of any appropriate prokaryotic promoter. A promoter/operator such as a tac promoter-operator may be used.

The carrier protein must form a fusion protein which is not susceptible to proteolytic cleavage by endogenous proteases in the host in which the fusion protein is expressed. Desirably, most of the fusion protein is present in the host as inclusion bodies. The carrier protein may or may not be in itself a complete protein. As indicated above, we have used part of a Trp operon. Consequently, the fusion protein which has been expressed consists of part of the TrpE protein linked via a Lys residue to mEGF.

A host, prokaryotic or eukaryotic, capable of expressing the fusion protein may be prepared in any suitable manner. For example, an expression vector for the fusion protein can be produced by ligating the synthetic gene in the correct reading frame with the carrier protein gene which is itself part of a DNA fragment comprising a promoter, and providing this ligation product in a suitable vector. The vector is then used to transform a host. The host is cultured under such conditions as to ensure expression of the fusion protein.

We have found that the level of production of the fusion protein can be increased by increasing the copy number of a plasmid which expresses the fusion protein using a temperature-sensitive runaway-copynumber plasmid.

The EGF or EGF analogue is obtained by extracting the fusion protein from the host cells and digesting it with a Lys-specific protease. The EGF or EGF analogue is thus released and can be purified, for example by chromatography. A suitable Lys-specific protease is endoproteinase LysC which cleaves proteins at the carboxyl group of Lys (US-A-4414332).

We obtained the endoproteinase Lys C as follows. An overnight slope of Lysobacter enzymogenes (ATCC 27796) was inoculated into and cultured in 500ml shake flasks containing yeast extract, 0.25%, glucose 0.1%, tryptone-soya 1.0% and MgCI2,1.OmM in H2O with swirling at 25"C for 53 h. The supernatant was recovered from the fermentation product by centrifugation and endoproteinase LysC obtained as described (US-A-4414332). The final pH of the fermentation was S9, but pH was not specifically controlled.

The EGF or EGF analogue which is obtained can be used for depilating animais and in particular for defleecing sheep. The EGF or EGF analogue may be administered to the sheep by subcutaneous infusion or injection or by slow release infusion from an implanted capsule or per os. Preferably, sufficient of the EGF or EGF analogue is administered to reduce the mean staple plucking force to 6 N/ktex or below, for example 2-6 N/ktex. Methods for making this measurement are described by A.J. Gordon, Aust. J. of Exp. Agric. Anim.

Husb.2O4O49 and Moore etal., Search. 1212S129. Typically, 3-5 mg of the EGFor EGFanalogue is given to each sheep.

The following examples illustrate the invention. In the accompanying drawings: Figure 1 shows the sequence of the Lys-mEGF gene of Example 1; Figure 2 shows the construction of the Expression Vector pWRL500 in Example 2; Figure 3 shows the construction of the Expression Vector pWRL505 in ExampleS; and Figure 4shows the construction of the Expression Vector pEGFtactrp2 in Example 6 Example 1: Synthesis ofLys-mEGFgene and preparation of plasmid containing this gene We designed the Lys-mEGF gene shown in Figure 1. The gene codes for the amino acid sequence of mEGF and for a linking peptide, Leu-Lys. The linking protein allowed insertion of the gene into the Bg1 II site that includes the codon for lle-323 in a TrpE gene. This enabled expression of a fusion protein of 378 amino acid residues, from which mEGF, which lacks lysine residues, could be released by a lysine-specific protease, Endoproteinase LysC. A stop codon followed by an EcoRI restriction site to facilitate cloning into the BamHI EcoRI site of pAT1 53 was provided at the end of the gene.

Codons were selected on the basis of preference of use byE coli in highly expressed proteins (Grantham etna!., 1981; Gouy & Gautier 1982). A number of modifications were made to remove regions of undesirable complementarity and repeats that could interfere with correct ligation of the oligodeoxynucleotide segments. An average length of 21 residues was chosen for oligonucleotide synthesis, since these could be efficiently synthesized and purified and would provide good overlaps for effective hybridization before ligation.

The sixteen oligodeoxynucleotides EGF-3 to EGF-18 shown in Figure 1 were synthesized by a manual solid-phase method based on that described by Duckworth petal. (1981) and Gait eta!. (1982), but with the following modifications, similar to those described by Sproat & Bannwarth (1983). The 3'-terminal deoxynucleoside was linked through a 3'-O-succinamido moiety to 3-aminopropyl controlled-pore glass beads (240 A pore size). Protected dinucleotide triethyl ammonium salts were prepared according to Chattopadhyaya & Reese (1979). Mesitylene sulphonyl-3-nitro-1 2,4-triazole with N-methyl imidazole catalyst (Efimov et 1982) was used to activate protected mono- and di-deoxynucleotide phosphate diesters for each condensation step.Trichloroacetic acid (10%) in l,l,l-trich loroethane was used to remove the dimethoxytrityl protecting group at each cycle, for the minimum time required as judged by removal of coloured material (40-80 seconds). The final protected oligodeoxynucieotide was treated with 4nitrobenzaldoxime and 1,1 ,3,3-tetramethylguanidine in 50% dioxan to cleave 2-chlorophenyl protecting groups and to cleave the oligodeoxynucleotide from the glass support. Ammonia and acetic acid deprotection steps followed, as described (Duckworth etal., 1981).

The oligodeoxynucleotides were purified by ion-exchange hplc on a Partisil SAX column at 55"C in 30% formamide with gradients of 10-700 mM KH2PO4 and desalted on a Sephadex G25 column. Each oligonucleotide was pure as revealed by autoradiography of [32P]-phosphate labelled material separated by electrophoresis on 15% polyacylamide gels in 8M urea.

Oligodeoxynucleotides EGF-3 to EGF-1 S and EGF-1 8 were preparatively phosphorylated at their 5'-hydroxyl groups using polynucleotide kinase and an excess of [ 32p] ATP. Ligations were conducted of three sets of oligodeoxynucleotides: set E, containing 40 pmol EGF-17, 17.5 pmol each of kinased EGF-3, EGF-4, EGF-5 and EGF-18, and 20.4 pmol of kinased EGF-6; set F, containing 20.4 pmol kinased EGF-7 and EGF-11 and 17.5 pmol each kinased EGF-8, EGF-9 and EGF-10; set G, containing 20.4 pmol kinased EGF-12, 17.5 pmol each EGF-13, EGF-14 and EGF-15 and 40 pmol EGF-16.

The oligodeoxynucleotides were added to polypropylene tubes in a total 1 S18,ul H2O for each mixture.

The tubes were heated to 1000C for 2 mins. then allowed to cool slowly in a 2 litre water bath initially at 1000C in an insulating jacket, overnight. The tubes were cooled to 0 C, ligase buffer (15 coil each) and bovine serum albumin (0.1%) were added, followed by 1 ul T4 DNA ligase (BioLabs). Ligation was for 1 h. at 370C. The DNA products were precipitated with ethyl alcohol and the correctly sized DNA fragments were purified by denaturing 12% acrylamide gel electrophoresis. The products of the three separate ligations were mixed and annealed together by heating to 100 C and cooling slowly.Ligase buffer with 0.1% BSA was added and 1 ul T4 DNA was added. After 1 h. at 37"C the DNA was precipitated with ethanol.

This product was ligated into the pAT153 EcoRI-BamHI large fragment at 7 C overnight, and the final ligation mix used to transform E. coli K12 HB101. Ampicillin-resistant colonies were selected for tetracyline sensitivity, and selected colonies were studied by restriction enzyme mapping, using Pstl + BstEll and EcoRI + BamHI. The expected fragments of 2700 and 790 bp in the former and 3300 and 167 bp in the latter digests were identified in 11 out of 15 colonies tested. These colonies therefore contained the desired plasmid, designated pEGF6. Confirmation was obtained using Hpall, Taql and EcoRI mapping.The EGF genes from two colonies were completeiy sequenced by the Maxam-Gilbert procedure (Maxam & Gilbert, 1980). One isolate has the correct sequence, while the other contained a single base change that did not affect the amino acid sequence.

Example 2: Construction of Expression Vector for (part TrpE)-L ys-mEGF fusion protein The complete mEGF gene and downstream sequences were excised from pEGF6 using a BamHI-Pst I digest. Plasmid pBRtrp, that contained the genes for part of the Trp operon, was digested with EcoRI and Bglll and the fragment containing the trp promoter, attenuator and part of the TrpE gene was isolated. These two fragments were ligated into the EcoRI-Pst I large fragment of pAT153 to yield pWRLS00 that contained a gene for a (part TrpE)-Lys-mEGF fusion protein under the control of the Trp promoter-operator.The lengths of the fragments used to construct pWRLS00 were 2907 bp Pst-EcoRI from pAT153 1210 bp EcoRI-Bglll from pBRtrp 930 bp Pst-BamHI from pEGF6 The fragments were ligated together in a single mixture. The construction of pWRLS00 is shown in Figure 2.

Plasmid pBRtrp is a derivative of pBR322 with the 2.08 kb Hpal-Hind Ill fragment of the tryptophan operon of E. coR encoding the promoter, leader peptide and structural gene of TrpE located between the EcoRI and Hind Ill sites of pBR322.

Plasmid pWRLS00 was used to transform E. coli HB101. Colonies were selected for ampicillin and tetracycline resistance. Plasmids from 20 colonies were studied by EcoRI restriction mapping and plasmids containing the expected 3660,1220 and 167 bg fragments were further characterised using Pstl + Bst Ell digestion. pWRLS00 yielded the expected 790 and 4257 bp fragments.

Example 3: Expression of (part TrpE)-Lys-mEGF fusion protein The E. coli bearing pWRLS00 were induced to produce high levels, estimated at about 10% of total cell protein by inspection of Coomassie blue stained polyacrylamide gel electrophoretograms, of the fusion protein, Mr 42,085, following starvation for tryptophan and addition of indole acrylic acid.The structure of the linking region of the fusion protein and the DNA sequence there are as follows:

Endoproteinase Lys-C cleavager3EGF 11 TrpE (1-320)-lle-Glu-lle-Leu-Lys-Asn-Ser-Tyr ATT GAG ATC CTT AAG AAT TAT TAT TAA CTC TAG GAA TTC TTA AGA ATA Bglll/BamHI EcoRI fusion Example 4: Purification ofmEGF (i) using 2-litre fermentation vessels The harvested E coli (from 2400 A600 units of culture) were disrupted by using lysozyme/EDTA and freeze-thawing, following by DNAse treatment.The bulk of the fusion protein was present as inclusion bodies, as seen by immunocytochemical analysis by using protein A-colloidal gold to reveal rabbit anti-mEGF immunoglobulin G molecules bound to the fusion protein in fixed E. coD cells, and was in the pellet fraction after centrifugation at 40,000 xg for 1 h. at 20"C. The pellet was homogenized in 8M urea, 1 mM EDTA, 1 mM 2-mercaptoethanol, 50 mM NH4HCO3, pH 8.3, and the mixture was diluted S4old into 50 mM NH4HCO3.The opalescent mixture was centrifuged as above and the supernatant, containing the bulk of the fusion protein, was digested with Endoproteinase Lys C (0.75 U, Boehringer Mannheim GmbH) for 24h. at 37"C. A further 0.75 U of protease was added and the digestion continued for 3 days. The digest was dialyzed for 2h. against 1 mM NH4HCO3, 0.2 mM EDTA, pH 8, at 4"C, and then against 50 mM KCl, 0.1 M NaCI for 2h.

The mixture was centrifuged at 40,000 xg for 10 mins at 4 C, and the supernatant was concentrated to 15 ml in an Amicon ultrafiltration cell with a UM2 membrane. The solution was chromatographed on a column (2.5 cmx 80 cm) of Bio Gel P-10 in 50 mM HCI, 0.1 M NaCI and the mEGF peak, eluting after the total column volume (Savage & Cohen, 1972) was collected. The solution was adjusted to pH 5.6 with NH3 and concentrated almost to dryness. It was dissolved in 20 mM NH40Ac, adjusted to pH 5.6 with acetic acid, and chromatographed on DEAE cellulose in a gradient of pH 5.6 ammonium acetate buffer. The first peak eluted after application of the gradient was mEG F; a second peak contained a mixture of proteins very similar to mEGF and probably degradation products.A test for endotoxin (limulus assay) revealed very small amounts (3.1 ng/mg mEGF).

(ii) using 200 litre fermentation The harvested E. colt cells were processed in 1 Kg batches of packed cell pellet, stored at -200C until used.

The thawed cells were suspended with 1.5 1. H2O and were disrupted by addition of 50 mg phenylmethane sulphonyl fluoride in 10 ml ethanol, 50 ml 1 M Tris base, 5 ml 0.5M EDTA, pH 8, 1 ml 2-mercaptoethanol, 5 ml NP40 detergent, 2-3 ml 5M NaOH,to adjust to pH 8.5, and 1g lysozyme,followed by incubation for 2h. at 30"C. MgCI2 (5 ml, 1 M) and DNAse (50 mg) were added and incubation continued 2h. EDTA (10 ml, 0.5M, pH 8) was added.The mixture was centrifuged at 27,000 xg for 45 mins at 20"C. The pellet was homogenized wtih 1.25 litres of 50 mM NH4HCO3, 1 mM EDTA, 1 mM 2-mercaptoethanol (pH 8.3) and collected by repeat centrifugation. Urea (0.9919 pellet) was added and the mixture was warmed to 37"C and homogenized. The clear viscous solution was diluted 2-fold, cleared by centrifugation and further diluted to 5 litres with 50 mM NH4HCO3. 2-hydroxyethyl disulphide (2.75 ml) and Endoproteinase LysC (12 U) were added and the mixture was incubated 3 days at 370C. These conditions gave almost complete cleavage.

The digest was made 15% in acetonitrile and adjusted to pH 3.75. The bulk of denatured polypeptide material precipitated and was removed by settling and filtration, and mEGF was recovered from the supernatant by reverse-phase chromatography on Prep RP18 silica gel in a gradient of 15-50% acetonitrile in 0.1% trifluroacetic acid. The crude mEGFwas diluted, neutrilized and chromatographed on DEAE cellulose as above, and the mEGF peak was acidified and chromatographed on Bio Gel P-1 The final product was neutralised, dialyzed and freeze-dried. The yield was about 15% of the theoretical based on estimates of the fusion protein content in the bacterial pellet; approximately 200 mg of mEGF was obtained.

(iii) purity analysis Reverse-phase and ion-exchange hplc of the purified mEGF indicated a purity greater than 85%. Amino acid analysis and peptide mapping data were in agreement with the structure of mEGF.

Example 6: Construction of an alternative Expression Vector for (part TrpE)-Lys-mEGFfusion protein In order to increase levels of expression of the fusion protein, the (part TrpE)-Lys-mEGF gene was transferred from pWRL500 to a temperature-sensitive runaway-copy-number vector, pMMI (Wong etna/., 1982). This is shown in Figure 3. The 3.2 kb fragment derived by BamHI and Pstl digestion of plasmid pMMI, containing the origin of replication, was ligated with BamHI-Pstl digested pWRLS00.E. coll HB1 01 was transformed with the ligation product and ampicillin-resistant colonies were selected for growth at 30"C but not at 42"C. Colonies containing the correctly formed pWRL505 were identified by restriction mapping using BamHI and Pstl.

E. colitransformed with pWRL505 was grown at 30"C in glucose minimal media with casamino acids to A600 = 1. The temperature was raised to 37"C for 2h., which increased the copy number of the plasmid many-fold, and indoleacrylic acid was added to induce expression from the Trp promoter. After a further 6h.

at 30"C the expression of the (part TrpE)-Lys-mEGF attained levels of about 20% of total cell protein. The mEGF was purified as described in Example 4 (ii) and was of similar purity as described in Example 4(iii).

Example 6: Expression ofa part (TrpE)-Lys-mEGFfusionprotein under the contro!ofthe tac promoteroperator.

High level expression of proteins from the trp promoter requires starvation for tryptophan. Expression of proteins containing tryptophan, such as mEGF, is not expected to be optimal under these conditions. The tac promoter (de Boer petal. (1983)) can be induced by isopropyl- -D-thiogalactoside (IPTG) in amino acid rich media, and more reproducibly high expression yields may be expected. The tac promoter in ptac 12 (Amann eft at (1983)) therefore was used to construct a plasmid, pEGFtactrp2, capable of expression a fusion protein comprising mEGF linked through a Lys residue to part of the TrpE protein. The construction of pEGFtactrp2 is shown in Figure 4.

pWRL500 from Example 2 was subjected to a partial EcoRI digestion to reorientate the trp promoter and the fusion protein gene. The resulting plasmid is pEGFtrp1. This plasmid contains a recognition site for BstXI at base 80 of the TrpE coding sequence. The trp promoter of pEGFtrpl through to this BstXI recognition site of the TrpE gene was replaced by a sequence comprising the tac promoter from ptac 12 through to base 80 of the TrpE coding sequence as shown below. The Shine-Dalgarno sequences, ribosome-binding sequences, are labelled SD. The resulting plasmid, pEGFtactrp1, was subjected to a partial EcoRI digestion to reorientate the tac promoter and fusion protein gene to obtain pEGFtactrp2.

Sequence of the tac promoter construct used for expression of fusion proteins in pEGFtactrp 1 and 2 TTCCGACATCATAACG GTTCTG GCAAATATTCTGAAATGAGCTGTTGACAATTAATCATC 20 40 60 -35 MetA GG CTCGTATAATGTGTGGAATTGTGAGCG GATAACAATTTCACACAG GAAACAGTTATGA 80 100 120 -10 SD snLeuGlyProAsnLyslleArgGluxxx MetGln ..mTrpE ATTTG G G CCCGAACAAAATTAGAGAATAACAATG CAA 140 ~~~~~ SD E. coli K1 2 J M105 (Yanisch-Perron, etal. (1985)) was transformed with pEGFtactrp2.The strain was grown in a rich medium in a 2 litre fermenter at 300C and induced with IPTG at O.D.650 1.0 Densitometric analysis of Coomassie blue stained SDA-polyacrylamide gel electrophoretograms indicated that 38% of total cell protein was (part TGrpE)-Lys-mEGF fusion protein. The mEGF was obtained and purified according to Example 4(i) and was of a similar purity as described in Example 4 (iii).

References Amann petal. (1983) Gene 25, 167-178.

de Boer etal. (1983) Proc. Natl. Acad. Sci. USA 80,21-25.

Chattopadhyaya & Reese (1979) Tetrahedron Lett. 5059-5062.

Duckworth etna!. (1981) Nucl. Acids Res. 9,1691-1706.

Edge et aL (1981) Nature 292, 756-762.

Efimov et al. (1982) Tetrahedron Lett. 23,961-964.

Gait etal. (1982) Nucl. Acids Res. 10, 6243-6254.

Gouy & Gautier (1982) Nucl. Acids Res. 10,7055-7074.

Grantham petal. (1981) Nucl. Acids Res. 9, r43-r74.

Maxam & Gilbert (1980) Methods Enzymol. 65,499-560.

Savage & Cohen (1972) J. Biol. Chem. 247, 7609-7611.

Smith petal. (1982) Nucl. Acids Res. 10,4467-4482.

Sproat & Bannwarth (1983) Tetrahedron Lett. 24, 5771-5774.

Wong eta/. (1982) Proc. Natl. Acad. Sci. USA 79,3570-3574.

Yanisch-Perron petal. (1985) Gene 33,103-119.

Claims (21)

1. A DNA sequence encoding a fusion protein comprising a carrier protein linked through a Lys residue to an EGF or EGF analogue, which EGF or EGF analogue does not contain an internal Lys residue and from which fusion protein the EGF or EGF analogue is cleavable by a Lys-specific protease.

2. A DNA sequence according to claim 1 wherein the EGF or EGF analogue is mEGF, mEGF1-45, mEGF1-47, mEGF1-48 or mEGF1-51.

3. A DNA sequence according to claim 1 or 2, wherein the EGF or EGF analogue is immediately followed by a stop codon.

4. A DNA sequence according to claim 3, which incorporates the following DNA sequence: (Lys) (mEGF1) G ATC CTT AAG AAT TCT TAT CCG GGT TGT CCG TCT TCT (1û) (20) TAC GAC GGT TAC TGC CTG AAC GGT GGT GTT TGC (30) ATG CAC ATC GAA TCT CTG GAC TCT TAC ACT TGC (40) AAC TGC GTT ATC GGT AC TC GGT GAC CGT TGC CAG (SO) (STOP) ACT CGT GAC CTG CGT TGG TGG GAA CTG CGT TAA GG

5. A DNA sequence according to claim 1 or 2, wherein the fusion protein comprises two or more runs each of a said EGF or EGF analogue.

6. A DNA sequence according to claim 5, comprising repeats of a DNA sequence of a said EGF or EGF analogue linked in tandem.

7. A DNA sequence according to any one of the preceding claims, wherein the carrier protein is part of the TrpE protein.

8. A vector which incorporates a DNA sequence as claimed in any one of the preceding claims and which is capable, in a transformed host, of expressing the said fusion protein.

9. A vector according to claim 8, which is a plasmid.

10. A vector according to claim 8 or 9, which incorporates a part of the Trp operon such that expression of the fusion protein is under the control of the trp promoter-attenuator and the carrier protein is part of the TrpE gene.

11. A vector according to claim 8 or 9, wherein expression of the fusion protein is under the control of the tac promoter-operator.

12. A host transformed with a vector as claimed in any one of claims 8 to 11.

13. A transformed host according to claim 12, wherein the host is a strain of E. coli.

14. A process for the preparation of an EGF or EGF analogue which does not contain an internal Lys residue, which process comprises culturing a transformed host as claimed in claim 12 or 13 such that the said fusion protein is expressed and treating the fusion protein with a Lys-specific protease to release the EGF or EGF analogue.

15. A process according to claim 14, wherein the Lys-specific protease is endoproteinase LysC.

16. A method of depilating an animal, which method comprises administering to the animal an EGF or EGF analogue which has been obtained by a process as claimed in claim 14 or 15.

17. A method according to claim 16, wherein the EGF or EGF analogue is used for defleecing sheep.

18. A method according to claim 17, wherein sufficient of the EGF or EGF analogue is administered to the sheep to reduce the mean staple plucking force to from 2 to 6 N/ktex.

19. A vector which is capable, in a transformed host, of expressing a fusion protein comprising a carrier protein linked through a Lys residue to an EGF or EGF analogue, which EGF or EGF analogue does not contain an internal Lys residue and from which fusion protein the EGF or EGF anlogue is cleavable by a Lys-specific protease, said vector being substantially as hereinbefore described in any one of Examples 2, 5 and 6.

20. A transformed host substantially as hereinbefore described in any one of Examples 2, 5 and 6.

21. A process for the preparation of an EGF or EGF analogue which does not contain an internal Lys residue, said process being substantially as hereinbefore described in Examples 3 and 4 together or in Example 5 or 6.