NO175640B - - Google Patents

Abstract

1. Claims for the Contracting States : BE, CH, DE, FR, GB, IT, LI, LU, NL, SE A fusion protein of the general formula Met - Xn - D' - Y - Z in which n is zero or 1, X is a sequence of 1 to 12 genetically codable amino acids, D' is a sequence of about 70 amino acids in the region of the sequence of amino acids 23 - 93 of the D-peptide in the trp operon of E. coli, Y denotes a sequence of one or more genetically codable amino acids which permits the following amino acid sequence Z to be cleaved off, and Z is a sequence of genetically codable amino acids. 1. Claims for the Contracting State : AT A process for the preparation of a fusion protein of the general formula (1) Met - Xn - D' - Y - Z in which n is zero or 1, X is a sequence of 1 to 12 genetically codable amino acids, D' is a sequence of about 70 amino acids in the region of the sequence of amino acids 23 - 93 of the D-peptide in the trp operon of E. coli, Y denotes a sequence of one or more genetically codable amino acids which permits the following amino acid sequence Z to be cleaved off, and Z is a sequence of genetically codable amino acids, characterized by expressing a gene structure which codes for these fusion proteins in a host cell, and separating the fusion protein.

Description

The present invention relates to fusion proteins and the DNA that codes for them.

In the genetic engineering of small eukaryotic proteins with a molecular weight of up to 15,000 daltons, only a low yield is achieved in bacteria. The proteins formed are probably quickly degraded by host-specific proteases. Such proteins are therefore conveniently produced as fusion proteins, especially with a host-specific protein proportion that then split off.

It has now been found that one of only approx. The 70 amino acid segment of the D protein from the trp operon of E. coli is particularly suitable for the formation of fusion proteins, and especially in the region of amino acid sequence 23 to 93 (C. Yanofsky et al., Nucleic Acids Res. 9 (1981) 6647 ), hereinafter also referred to as "D'-peptide". Between the carboxy terminus of this peptide and the amino acid sequence for the desired eukaryotic protein is a sequence of one or more genetically codeable amino acids that allow a chemical or enzymatic cleavage of the desired protein. In preferred embodiments of the present invention, a short amino acid sequence consisting of Lys-Ala is joined to the amino terminal, possibly followed by a sequence of 1 to 10, in particular

1 to 3 further genetically codeable amino acids, preferably of 2 amino acids, especially of Lys-Gly. The present invention provides a fusion protein, characterized by the general formula

in which

n is zero or 1,

X is a sequence of 1 to 12 genetically coded amino acids,

D' is a sequence of approx. 70 amino acids in the region of the amino acid sequence 23-93 of the D-peptide in the trp operon for E. coli,

Y is a sequence of one or more genetically coded amino acids, which enables a cleavage of the following amino acid sequence Z, and

Z is a sequence of genetically coded amino acids.

Furthermore, the present invention provides DNA, characterized in that it codes for the fusion protein as discussed above, where the following sequence codes for D':

for X the DNA sequence (coding strand) codes:

5' AAA GCA AAG GGC 3'

in the nucleotide sequence of Y that codes for the amino acids, Y C-terminally contains Met, Cys, Trp, Arg or Lys or one of the groups

(Asp)m - Pro, respectively Glu - (Asp)m - Pro or Ile - Glu - Gly - Arg

wherein m means 1, 2 or 3, or consists of these amino acids or one of these groups.

It is of course advantageous when the unwanted (significant) proportion of the fusion protein is as small as possible, since in this case the cell only produces "ballast" and the yield of the desired protein thereby becomes high. Furthermore, the splitting off of the unwanted part results in less by-products, which makes processing easier. In the opposite direction, it seems that the (supposed) "protection function" for the unwanted proportion can only be expected from a certain size. Surprisingly, it has been shown that the segment chosen according to the invention from the D-protein fulfills this task, even though it only exhibits approx. 70 amino acids.

In many cases, above all in the preferred embodiment, in which X stands for Lys-Ala and the N-terminus contains this sequence, an insoluble fusion protein is formed. This can be easily separated from the soluble proteins, which greatly facilitates processing and increases the yield. The formation of insoluble fusion protein is surprising, as on the one hand the bacterial proportion is relatively low with only approx. 70 amino acids and, on the other hand, is part of a protein present in the host cell in solution.

"Approximately 70 amino acids in the area of the amino acid sequence 23 to 93 of the D-peptide" means that variations are possible in a manner known per se, i.e. that certain amino acids can be omitted, replaced or replaced, without thereby affecting the properties of the fusion protein according to the invention changes its significance.

The desired eukaryotic protein is preferably a biologically active protein such as a hirudin or is the precursor of such a protein as human proinsulin.

The fusion protein is recovered by expression in a suitable system and is isolated, in the particularly preferred embodiment, after digesting the host cells from the precipitate, in which it is enriched due to its low solubility. Thus, a simple separation from the soluble constituents in the cell is possible.

As host cells, all cells that are known for expression systems come into consideration, i.e. mammalian cells and micro-organisms, preferably bacteria, especially E. coli, the bacterial part of the fusion protein being a host-specific protein from E. coli.

The DNA sequence which codes for the fusion protein according to the invention is incorporated in a known manner into a vector which ensures good expression in the chosen expression system.

In bacterial hosts, the promoter and operator are appropriately selected from the group lac, tac, Pl or Pj) for the phage \, hsp, omp or a synthetic promoter, as for example described in the German specification no. 3 430 683 (European patent publication 0 173 149) .

Particularly suitable is a vector which from the trp operon from E. coli contains the following elements: The promoter, the operator and the ribosome binding position for the L-peptide. In the coding region, the first three amino acids of this L-peptide join particularly expediently, after which, over a short amino acid sequence, amino acids 23 to 93 of the D protein in the trp operon follow.

The intermediate sequence Y, which enables cleavage of the desired polypeptide, depends on the composition of this desired peptide: When this, for example, does not contain methionine, Y can stand for met, whereupon a chemical cleavage follows with cyanogen bromide. If in the linker Y at the carboxy terminal stands for cysteine or if Y stands for Cys, then an enzymatic cysteine-specific cleavage or a chemical cleavage, for example after S-cyanylation, can take place. If in the bridge link Y at the carboxy terminal stands for tryptophan or Y stands for trp, a chemical cleavage can take place with N-bromosuccinimide. If Y stands for Asp-Pro, then a proteolytic cleavage can take place in a manner known per se (D. Piszkiewicz et al., Biochemical and Biophysical Research Communications 40 (1970) 1173-1178). As it was found, the Asp-Pro bond can also be designed acid-labile, when Y is

whereby m is 1, 2 or 3. This results in cleavage products that begin the N-terminus with Pro or terminates the C-terminus with Asp.

Examples of enzymatic cleavages are also known, whereby modified enzymes with improved specificity can also be used (C.S. Craik et al., Science 228 (1985) 291-297). If the desired eukaryotic peptide is human proinsulin, then one can choose as sequence Y a peptide sequence in which a trypsin-cleavable amino acid (Arg, Lys) is bound to the N-terminal amino acid (phe) of the proinsulin, for example Ala-Ser-Met-Thr -Arg, as in this case the arginine-specific cleavage can take place with the protease trypsin. If the desired protein does not contain the amino acid sequence Ile-Glu-Gly-Arg, cleavage with factor Xa is also possible (EP-A 0 161 937).

When designing the sequence Y, one also has the opportunity to take into account the synthetic possibilities and incorporate suitable cleavage positions for restriction enzymes. The DNA sequence corresponding to the amino acid sequence Y can therefore also take over the function of a connecting link or an adapter.

The fusion protein according to the invention is advantageously expressed under the control of the trp operon from E. coli. A DNA segment containing the promoter and operator for the trp operon is commercially available. The expression of proteins under the control of the trp operon is often described, for example in EP-A2 0 036 776. The induction of the trp operon can be achieved in the absence of L-tryptophan and/or the presence of indolyl-3-acrylic acid in the medium.

Under the control of the trp operator, the transcription of the L-peptide, which consists of 14 amino acids, first takes place. This L-peptide contains in positions 10 and 11 L-tryptophan. The rate of protein synthesis of the L-peptide determines whether the subsequent structural gene is also translated, or whether protein synthesis is interrupted. In the case of a lack of L-tryptophan, a slow synthesis of the L-peptide takes place due to a low concentration of tRNA for L-tryptophan and a synthesis of the following proteins takes place. At high concentrations of L-tryptophan, on the other hand, the corresponding section of mRNA is read quickly and there is an interruption of protein biosynthesis as the mRNA assumes a terminator-like structure (C. Yanofsky et al., see above).

The frequency of translation for an mRNA is strongly influenced by the type of nucleotide surrounding the start codon. When expressing a fusion protein by means of the trp operon, it therefore seems advantageous to use the nucleotides for the first amino acids of the L-peptide for the beginning of the structural gene for the fusion protein. In the preferred embodiment of the invention in the trp system, the nucleotides for the first 3 amino acids of the L-peptide (hereinafter L'-peptide) were chosen as codons for the N-terminal amino acids of the fusion protein.

The invention therefore also relates to vectors, preferably plasmids for the expression of fusion proteins, whereby the DNA of the vectors seen from the 5' end has the following characteristics (in a suitable arrangement and in phase): A promoter, an operator, a ribosome binding site and the structural gene for the fusion protein, whereby the latter contains the amino acid sequence I (pendant) before the sequence of the desired protein. Before the structural gene, respectively as the first triplet for the structural gene, the start codon (ATG) and possibly additional codons for other amino acids are located between the start codon and the D' sequence or between the D' sequence and the gene for the desired protein. Depending on the amino acid composition of the desired protein, the selection of the DNA sequence is made before the structural gene, so that cleavage of the desired protein from the fusion protein is made possible.

In the expression of the fusion protein according to the invention, it may prove appropriate to change certain triplets for the first amino acids after the ATG start codon in order to prevent any base pair formation in the area of mRNA. Such changes, as well as changes, omissions or additions of certain amino acids in the D' protein, are known to the person skilled in the art.

As smaller plasmids exhibit several advantages, a preferred embodiment of the invention consists in eliminating a DNA section with the structural gene for tetracycline resistance from plasmids derived from pBR 322. Preferably, segments are removed from the HindiII cut site at position 29 to the PvulI cut site at position 2066. With these plasmids, it is particularly advantageous to remove a somewhat larger DNA section, using the PvuII cut site at the beginning (in the reading direction) of the trp the operon (which lies in a non-essential part). In this way, one can directly ligate the large fragment that occurs with both PvuII cut sites. It achieved, with approx. The 2 kbp reduced plasmid causes an increase in expression, which can possibly be attributed to an increased copy number in the host cell.

In the following examples, the invention is explained in more detail.

Example 1

a) Chromosomal E. coli DNA is cut using Hinf I and the 492 bp fragment is isolated, which from the trp operon contains

the promoter, the operator, the structural gene for the L-peptide, the attenuator and the codons for the first six amino acids of the trp-E structural gene. This fragment is completed using Klenow polymerase with deoxynucleotide triphosphates, joined at both ends with an oligonucleotide containing a recognition site for HindIII and it is cut with HindIII. The resulting HindIII fragment is ligated into the HindIII cut site of pBR 322. This results in the plasmid ptrpE2-1 (J.

C. Edmann et al., Nature 291 (1981) 503-506). This is transferred as described to the plasmid ptrpLl. Using the synthetic oligonucleotides (NI) and (N2)

which is complementary to the double-stranded oligonucleotide

(N3)

is built into the ClaI site of the plasmid ptrpLl the DNA sequence for the first three amino acids of the L-peptide, at the same time a restriction site (Stul) is formed for the introduction of further DNA (Fig. 1). The plasmid ptrpLl is reacted with the enzyme ClaI according to the manufacturer's instructions, the mixture is extracted with phenol and the DNA is precipitated with ethanol. The opened plasmid is reacted to remove the phosphate groups at the 5' ends with alkaline phosphatase from E. coli. The synthetic nucleotide is phosphorylated at the 5' ends and introduced with T4 DNA ligase into the opened, phosphatase-treated plasmid. After completion of the ligase reaction, transformation takes place in E. coli 294 and selection of the transformants by Amp resistance and the presence of a Stul restriction site.

About. 80% of the clones obtained display the expected restriction site; the one in fig. 1 the reproduced nucleotide sequence was confirmed by sequence analysis. The plasmid pH120/14 is obtained, which after the ribosome binding site for the L-peptide exhibits the nucleotide triplet for the first three amino acids of the L-peptide (L'-peptide), followed by a Stul site, which allows further introduction of DNA and thereby enables the formation of fusion proteins with the first three amino acids from the L-peptide.

b) In the following, based on the oligonucleotide (NI) used above, the chemical synthesis of

such oligonucleotides:

With the method according to M.J. Gait et al., Nucleic Acids Research 8 (1980) 1081-1096, the nucleoside standing at the 3' end, in this case guanosine, is covalently bound to a glass sphere carrier (CPG = controlled pore glass) LCAA (= long chain alkyl) amine) (from Pierce) via the 3'-hydroxy function. In this way, the guanosine is reacted as N-2-isobutyroyl-3'-0-succinoyl-5'-dimethoxy-trityl ether in the presence of N,N'-dicyclohexylcarbodiimide and 4-dimethylaminopyridine with the modified carrier, whereby the free carboxyl group of succinoyl- the remainder acylates the amino residue of the long-chain amine on the support.

In the following synthesis steps, the base components are used such as 5 '-O-dimethoxytrity 1-nucleoside-3'-phosphoric acid monomethyl ester dialkylamide or chloride, whereby the adenine is present as the N6-benzoyl compound, the cytosine as the 4N-benzoyl-for- bond, the guanine as the N2-isobutyryl compound and the thymine containing no amino group are present without a protecting group. 40 mg of the carrier containing 1 jjmol of bound guanosine is treated stepwise with the following reagents:

a) methylene chloride,

b) 10% trichloroacetic acid in methylene chloride,

c) methanol,

d) tetrahydrofuran,

e) acetonitrile,

f) 15 pmol of the corresponding nucleoside phosphite and 50 pmol tetrazole in 0.3 ml anhydrous acetonitrile (5 minutes), g) 20% acetic anhydride in tetrahydrofuran with 40% lutidine and 10% dimethylaminopyridine (2 minutes),

h) tetrahydrofuran,

i) tetrahydrofuran with 20% water and 40% lutidine,

j) 3% iodine in collidine/water/tetrahydrofuran in volume ratio

5:4:1,

k) tetrahydrofuran and

1) methanol.

By "phosphite" is meant here the deoxyribose-3'-monophosphoric acid monomethyl ester, whereby the third valence is desaturated with chlorine or a tertiary amino group, for example a diisopropylamino group. The yields in the individual synthesis steps can in any case after the detritylation reaction b) be determined spectrophotometrically by measuring absorption for the dimethoxytrityl cation at a wavelength of 496 nm.

After completion of the synthesis of the oligonucleotide, the methyl phosphate protection group for the oligomer is removed with the help of p-thiocresol and triethylamine.

The oligonucleotide is then separated from the solid support by treatment with ammonia for 3 hours. A 2- to 3-day treatment of the oligomer with concentrated ammonia cleaves the amino protecting groups quantitatively from the bases. The crude product obtained is purified by high-pressure liquid chromatography (HPLC) or by polyacrylamide gel electrophoresis.

The other oligonucleotides are synthesized in a similar way.

c) The plasmid ptrpE5-1 (R.A. Hallewell et al., Gene 9 (1980) 27-47) is reacted with the restriction enzymes HindIII and Sali

according to the manufacturer's specifications and the approx. The 602 bpDNA fragment is removed. The synthesized oligonucleotides (N4) and

(N5)

are phosphorylated, incubated together at 37°C and added by means of DNA ligase to the butt-ended DNA for proinsulin (W. Wetekam et al., Gene 19 (1982) 179-183). After reaction with HindIII and SalI, the now extended proinsulin DNA is incorporated covalently into the opened plasmid with the enzyme T4 DNA ligase (Fig.

2), whereby the plasmid pH106/4 arises.

The plasmid pH106/4 is first reacted once again with SalI, the overlapping ends are completed with Klenow polymerase to blunt ends and then incubated with the enzyme Mstl. A DNA fragment of approx. 500 bp, which contains the entire coding part of coinsulin as well as an approx. 210 bp fragment of the D protein from the trp operon of E. coli. The DNA fragment is butt-ended and introduced into the Stul site of the plasmid PH120/14, whereby the plasmid pH154/25 arises (Fig. 3). This is suitable for expression of a fusion protein under the control of the trp operon, in which the amino acid sequence Ala-Ser-Met-Thr-Arg is arranged after the L' and D' peptides, to which the amino acid sequence of proinsulin is connected.

Example 2

The plasmid pH154/25 (Fig. 3) is reacted with the restriction enzymes BamHI and XmalII. The excess ends are filled using Klenow polymerase and joined with T4 DNA ligase. The plasmid pH-254 (Fig. 4) is obtained which is suitable for expression of a fusion protein with the amino acid sequence L', D'-proinsulin under the control of the trp promoter. The plasmid is somewhat smaller than pH154/25, which may be advantageous.

Example 3

When incubating the plasmid pH254 (example 2; Fig. 4) with the restriction enzymes Mlul and Sali, a DNA section of 280 bp is set free, this is separated. The remaining plasmid is converted with Klenow polymerase to the butt-ended form and ring-closed with DNA ligase again covalently. This results in the plasmid pH255 (Fig. 4), which is suitable for the introduction of a structural gene at one of the restriction sites MluI, SalI and EcoRI. Under induction conditions, the formation of a fusion protein with the L',D' protein takes place. Naturally, further restriction sites can be introduced into the plasmid pH255 by means of suitable connecting links.

Example 4

The plasmid pH154/25 (Fig. 3) is incubated with the enzymes MluI and EcoRI and the released DNA fragment (approx. 300p) is removed. The remaining plasmid is ligated with Klenow polymerase. Ring closing takes place by the action of DNA ligase. The resulting plasmid pH256 (Fig. 5) can be used to introduce structural genes into the EcoRI site.

Example 5

By removing a 600 bp fragment from the plasmid pH256 (example 4; fig. 5) with the restriction enzymes BamHI and Nrul, the plasmid pH257 is obtained (fig. 5). For this purpose, pH256 is first incubated with BamHI and blunt ends are obtained with Klenow polymerase. After incubation with Nrul and separation of the 600 bp fragment, the formation of pH257 takes place after incubation with DNA ligase.

Example 6

Upon introduction of the lac repressor (P.J. Farabaugh, Nature 274

(1978) 765-769) in the plasmid pKK 177-3 (Amann et al., Gene 25

(1983) 167) the plasmid pJF118 is obtained. This is reacted with EcoEI and Sali and the residual plasmid is isolated.

From the plasmid pH106/4 (fig. 2) is obtained by exposure to Sali and incubation with Mstl et approx. 495 bp large fragment.

The synthetically produced oligonucleotides (N6) and (N7)

is phosphorylated and added with DNA ligase to the blunt-ended DNA fragment. By reaction with EcoRI and SalI, the overlapping ends are set free, which allows a ligation into the opened plasmid pJF118.

After transformation of the thus produced hybrid plasmid in E. Coli 294, the correct clones are selected on the basis of the size of the restriction fragments. This plasmid is designated as pJ120 (Fig. 6).

The expression of the fusion protein is carried out in a shake flask as follows: From an overnight culture of E. coli 294 transformants containing the plasmid pJ120, in LB medium (J.H. Miller, "Experiments in Molecular Genetics", Gold Spring Harbor Laboratory, 1972 ) with 50 µg/ml ampicillin, a fresh culture is started in a ratio of approx. 1:100 and the growth is monitored using OD measurement. At OD = 0.5, the culture is mixed with so much isopropyl-ε-D-galactopyranoside (IPTG) that its concentration is 1 mM, and the bacteria are separated after 150 to 180 minutes. The bacteria are boiled for 5 minutes in a buffer mixture (7M urea, 0.1% SDS, 0.1M sodium phosphate, pH 7.0) and the samples are applied to an SDS gel electrophoresis plate. After electrophoresis, a protein band corresponding to the size of the expected fusion protein and reacting with antibodies against insulin is obtained from bacteria containing the plasmid pJ120. After isolation of the fusion protein, the expected proinsulin derivative can be released by cyanogen bromide cleavage. After digestion of the bacteria (French Press; "Dyno-Muhle") and centrifugation, the L',D'-proinsulin fusion protein is found in the precipitate, so that considerable amounts of other proteins can already be separated with the supernatant.

The indicated induction conditions apply to shaking cultures; in the case of larger fermentations, correspondingly changed OD values and possibly slightly changed IPTG concentrations may be appropriate.

Example 7

From E. coli 294 transformants, which contain the plasmid pH154/25 (fig. 3), a culture is prepared overnight in LB medium with 50 µg/ml ampicillin and, as the next measure, diluted in a ratio of approx. 1:100 in M9 medium (J.H. Miller, see above) with 2000 pg/ml casamino acid and 1 pg/ml thiamine. At OD = 0.5, indolyl-3-acrylic acid is added so that the final concentration amounts to 15 µg/ml. After 2 to 3 hours of incubation, the bacteria are centrifuged off. An SDS gel electrophoresis shows at the site expected for the fusion protein a very clear protein band which reacts with antibodies against insulin. After digestion of the bacteria and centrifugation, the L',D'-proinsulin fusion protein is found in the precipitate, so that in this case also significant quantities of the other proteins are removed with the supernatant.

Also in the present case, the indicated induction conditions apply to shaking cultures. Fermentations in larger volumes require changed concentrations of casamino acids, respectively an addition of L-tryptophan.

Example 8

The plasmid pH154/25 (fig.3) is opened with EcoEI and the above DNA single chains are completed with Klenow polymerase. The DNA obtained in this way is incubated with the enzyme MluI and the DNA coding for insulin is excised from the plasmid. By gel electrophoretic separation, this fragment is separated from the residual plasmid and the residual plasmid is isolated.

The plasmid according to fig. 3 in the German interpretation document no.

3 429 430 (European patent publication EP-A1 0 171 024) is reacted with the restriction enzymes Accl and Sali and the DNA fragment containing the hirudin sequence is separated. After filling in the above ends of the SalI cut site with Klenow polymerase, the DNA segment is ligated with the synthetic DNA of formula (N8)

The ligation product is incubated with MluI. After inactivating the enzyme at 65 °C, the DNA mixture is treated with alkaline bovine phosphatase for 1 hour at 37 °C. The phosphatase and the restriction enzyme are then removed by means of phenol extraction from the mixture and the DNA is purified by ethanol precipitation. The DNA treated in this way is introduced with T4- ligase in the opened residual plasmid pH154/25, whereby the plasmid pK150 is obtained, which is characterized by restriction analysis and DNA sequencing according to Maxam and Gilbert (fig. 7).

Example 9

E. coli 294 bacteria containing the plasmid pK150 (Fig. 7) are allowed to stand in LB medium with 30 to 50 µg/ml ampicillin overnight at 37°C. The culture is diluted with M9 medium containing 2000 µg/ml casamino acids and 1 µg/ml thiamine, in a ratio of 1:100 and incubated at 37°C with constant stirring. At an OD^qq = 0.5 or 1, indolyl-3-acrylic acid is added to a final concentration of 15 pg/ml and incubated for 2 to 3 hours, respectively 16 hours. The bacteria are then centrifuged off and suspended in 0.1 M sodium phosphate buffer (pH 6.5) under pressure. The poorly soluble proteins are centrifuged off and analyzed using SDS-polyacrylamide electrophoresis. It turns out that cells whose trp operon is induced in the range below 20,000 daltons but above 14,000 daltons contain a new protein that is not found in non-induced cells. After isolation of the fusion protein and reaction with cyanogen bromide, hirudin is set free.

Example 10

In the following, constructions are described that allow DNA sequences to be introduced in front of the 5' end of the trp-D sequence that contain the most numerous recognition sites for different restriction enzymes, in order to be able to incorporate the trp-D sequence as universally as possible into the various pro - the karyotic expression systems.

Plasmids pUC12 and pUC13 (Pharmacia P-L Biochemicals, 5401 St. Goar: "The Molecular Biology Catalogue" 1983, Appendix, page 89) contain a poly linker sequence whereby in plasmid pUC13 between the restriction sites of Xmal and SacI the MstI-HindIII-trp fragment from the plasmid pl06/4 (Fig. 2), fused with the HindIII-hirudin-SacI fragment from plasmid pK150 (Fig. 7) should be introduced.

For this purpose, the DNA in the plasmid pUC13 is first treated with the restriction enzyme XmaI. The ends of the linearized plasmid are filled using Klenow polymerase reaction. After ethanol precipitation, the DNA is treated with the enzyme SacI and re-precipitated with ethanol from the reaction mixture. The DNA is then reacted in an aqueous ligation mixture with the MstI-HindIII-trp-D fragment, isolated from plasmid pH106/4 and the HindIII-SacI hirudin fragment, isolated from plasmid pK150, and T4 DNA ligase.

The plasmid pK160 produced in this way contains, immediately before the beginning of the trp-D sequence, a multiple restriction enzyme recognition sequence, which comprises the cutting sites for the enzymes XmaI, SmaI, BamHI, XbaI, HincII, Sali, AccI, Pstl and HindIII. With this construction, an EcoRI cut site is also obtained after the 3' end of the hirudin sequence (Fig. 8).

Example 11

The plasmid pH131/5 is prepared (according to the German patent publication no. P 35 14 113.1, example 1, fig. 1) in the following way:

The plasmid ptrpLl (J.C. Edman et al., Nature 291 (1981) 503-506) is opened with ClaI and ligated with the synthetically prepared, self-complementary oligonucleotide (N9)

The thus obtained plasmid pH131/5 (Fig. 9) is opened at the cut site for the restriction enzyme NcoI introduced in this way and the resulting single chain ends are filled in by means of the Klenow polymerase reaction. The linearised, smooth-ended DNA is now cut with the enzyme EcoRI and the largest of the DNA sequences formed is separated by means of ethanol precipitation from the smaller sequence. The remaining plasmid DNA for the plasmid pH131/5 formed in this way is now ligated with a fragment coding for trp-D'-hirudin from pK160 by T4 ligase reaction. This fragment was cleaved from the plasmid pK160 by opening the plasmid with HincII and EcoRI. The fragment is gel electrophoretically separated from the residual plasmid and then separated from the gel material. The ligation product is transformed by E. coli K12. The plasmid DNA containing clones are isolated and characterized by restriction analysis and DNA sequence analysis. The plasmid pK170 produced in this way contains, fused to the trp operator, a DNA sequence which codes for Met-Asp-Ser-Arg-Gly-Ser-Pro-Gly-trp-D'-(hirudin) (Fig. 9).

Example 12

The plasmid pJF118 (Example 6) is opened with EcoRI and the overhanging DNA ends are transferred by Klenow polymerase reaction to blunt ends. The DNA treated in this way is then cut with the enzyme SalI and gel electrophoretically separated from the short EcoRI SalI fragment.

The plasmid pK170 (Example 11) is cleaved with NcoI and the overhanging ends are transferred by means of Klenow polymerase to blunt ends. Plasmid DNA is separated from the reaction mixture by ethanol precipitation and treated with the enzymes HindIII and BamHI. From the fragments formed, two are isolated, namely the NcoI (with filled end )-trpD'-HindIII fragment and the HindIII-hirudin-BamHI fragment (German publication no. 3 429 430). Both fragments are isolated after gel electrophoretic separation.

Furthermore, the BamHI-SalI-hirudinII fragment is isolated according to fig. 2r in the German specification no. 3 429 430. In a ligation reaction, the four fragments, namely the pJF118 residual plasmid, the NcoI-trpD'-HindIII fragment, the HindIII-hirudin-BamHI fragment and the hirudinIII fragment are now reacted with each other and the obtained the plasmid pK180 (Fig. 10) is transformed after E. coli K12-W 3110. Correct plasmids are found when an EcoRI-trpD'-hirudin-SalI fragment can be detected in plasmid DNA. The trpD' hirudin sequence is now linked to the tac promoter. The fusion protein is expressed according to example 6.

Example 13

In the case of the plasmids derived from pBR322, such as pH120/14 (Example 1, Fig. 1), pH154/25 (Example 1, Fig. 3), pH256 (Example 4, Fig. 5), pK150 (Example 8, Fig. 7) and pK170 (example 11, fig. 9) there is - clockwise in the figures - between the start codon for the fusion protein and the closest HindIII site (corresponding to HindIII at position 29 in pBR322) a further PvuII site in the area of the fragment containing the trp promoter and operator, but admittedly outside the promoter region.

It has now been found that by removing the DNA section restricted by the mentioned PvuII site and the PvuII site corresponding to position 2066 in pBR322, the yield of a cloned protein (or fusion protein) can be significantly increased.

In the following, for example, the shortening of the plasmid pH154/25 to pH154/25<*> is described, which for the other proteins can take place in a similar way (whereby the shortened plasmids are in any case marked with an asterisk): pH154/25 is reacted (according to manufacturer's specifications) with PvuII whereby three fragments arise: Fragment 1: from the PvuII restriction site for proinsulin until the PvuII restriction site corresponding to position 2066 in pBR322,

fragment 2: from the PvuII restriction site near the trp promoter

to the PvuII site for proinsulin and

fragment 3: from the PvuII site near the trp promoter fragment to the PvuII site corresponding to position 2066 in pBR322.

The fragments can be separated by electrophoresis on agarose and then isolated (Maniatis "Molecular Cloning", Cold Spring Harbor 1982).

Fragments 1 and 2 are joined under "blunt end" conditions with the enzyme T4 DNA ligase. After transformation in E. coli 294, one examines in which colonies a plasmid with the complete proinsulin sequence is present and thus the fragments are present in the desired arrangement. The plasmid pH154/25<*> is reproduced in fig. 11.

During the expression, which takes place as described in the above examples, a significant increase in the proportion of the fusion protein is observed.

Example 14

The plasmid pH154/25<*> (Example 13, Fig. 11) is digested with HindIII and SalI and the small fragment (with the proinsulin sequence) is separated gel electrophoretically. The large fragment is isolated and ligated with the synthetic DNA (N10)

The plasmid plntl3 (fig.12) arises.

DNA (N10) codes for an amino acid sequence that contains several cleavage sites for a chemical cleavage:

a) Met for cyanogen bromide,

b) Trp for N-bromosuccinimide (NBS or BSI ),

c) Asp-Pro for proteolytic cleavage, whereby the preceding Glu further weakens the Asp-Pro bond to

the impact of acids.

The introduction of this HindIII-Sall link (N10) into the reading frame of an encoded polypeptide thus allows the stated possibilities for a chemical cleavage, depending on the amino acid sequence of the desired protein, respectively its sensitivity to the cleavage reagents.

The figures are not reproduced to the correct scale.

Amino acid sequence I

DNA sequence I

Claims (3)

1. Fusion protein, characterized by the general formula Met-Xn-D<*->Y-Z in which n is zero or 1, X is a sequence of 1 to 12 genetically codeable amino acids, D' is a sequence of approx. 70 amino acids in the region of the amino acid sequence 23-93 of the D peptide in the trp operon for E. coli, Y is a sequence of one or more genetically coded amino acids, which enables a cleavage of the following amino acid sequence Z, and Z is a sequence of genetically coded amino acids. 2. Fusion protein according to claim 1, characterized in that n is 1 and X consists of up to 5 amino acids, whereby preferably Lys-Ala is N-terminally in X and/or that Y C-terminally contains Met, Cys, Trp, Arg or Lys or one of the groups (Asp)m-Pro or Glu-(Asp )m-Pro or Ile-Glu-Gly-Arg wherein m is 1, 2 or 3, or consists of one of these amino acids, or one of these groups and/or that Z stands for the amino acid sequence from human proinsulin or a hirudin. 3. DNA, characterized in that it codes for the fusion protein according to claim 1 or 2, where the following sequence codes for D': for X codes the DNA sequence (coding strand): in the nucleotide sequence of Y that codes for the amino acids, Y C-terminally contains Met, Cys, Trp, Arg or Lys or one of the groups wherein m means 1, 2 or 3, or consists of these amino acids or one of these groups.