FI93471B - Process for producing fusion proteins as well as products useful in realizing the process - Google Patents

Abstract

Fusion proteins in which the C- or N-terminal essentially corresponds to the first 100 units of interleukin 2 are new. Also new are gene structures coding for the above fusion proteins, vectors containing these gene structures, and hot cells containing these vectors. Hirudin derivs. with an amino acide sequence begining N-terminally with Pro are new and claimed. Human interleukin 2 derivs. contg. Asp. C-terminally are new and claimed.

Description

93471

Method for the preparation of fusloprotelinle and the products used in carrying out the method

The invention relates to a method for producing a fusion protein, to a gene construct encoding such a fusion protein, to a vector containing said gene construct, and to bacterial hosts containing the vector, and to plasmids useful in the production of a fusion protein.

The invention is based on the "open reading frame" of DNA encoding interleukin-2 and the use of this DNA as an expression aid for the production of peptides or proteins.

The production of eukaryotic proteins by gene-15 technology often results in low yields in bacteria, especially in the case of small proteins with a molecular weight of up to about 15,000 daltons and a structure containing disulfide bridges. It is hypothesized that the host's own proteases rapidly degrade the 20 proteins formed. Therefore, it is expedient to generate gene constructs encoding fusion proteins in which the undesired part of the fusion protein is the host's own protein, which is removed after isolation of the primary product by methods known per se.

It has now surprisingly been found that the N-terminal part of interleukin-2, which corresponds approximately to the first 100 amino acids, is particularly well suited for the preparation of fusion proteins. Thus, the primary product is a fusion protein which consists wholly or very largely of eukaryotic protein sequences. Surprisingly, the host organism in question apparently does not recognize this protein as a foreign protein and does not immediately degrade again. A further advantage is that the resulting fusion proteins are sparingly soluble or insoluble, and can thus be separated from soluble proteins simply by convenient centrifugation.

___ I

93471 2

Since the function of the interleukin-2 moiety as a "ballast portion" of the fusion protein is not dependent on the fact that this moiety is a biologically active molecule, it is also not dependent on the exact ra-5 claim of the interleukin-2 moiety. For this purpose, it is sufficient that approximately the first 100 N-terminal amino acids are present. Thus, for example, it is possible to make changes at the N-terminus that allow the fusion protein to be removed if the desired protein is N-terminally associated with it. Similarly, changes can be made to the C-terminus to allow or facilitate cleavage of the desired protein when it is, as usual, associated with the C-terminus of the fusion protein.

The natural DNA sequence encoding human interleukin-2, hereinafter "nIL-2", is described in EP-A-0 091 539. The literature therein deals with murine and rat IL-2. This mammalian DNA can be used in accordance with the invention to synthesize proteins. However, it is more expedient to start from synthetic DNA, particularly preferably DNA corresponding to human IL-2, as proposed in (unpublished) DE application 3 419 995 (corresponding to EP application 0 163 249). This synthetic DNA sequence is given in the appendix (DNA sequence 1). The advantage of this synthetic DNA sequence is not only that it is suitable for the most frequently used host of codon selection, E. coli, but also that it contains a number of restriction endonuclease cleavage sites that can be utilized in accordance with the invention. Table 1 below shows a selection of suitable 30 cleavage sites from the beginning of the cycle, i.e. in the region of 100 triplets. However, this does not preclude DNA alterations in the intervening region, in which case the other cleavage sites provided in the above-mentioned patent application can be exploited.

3 93471

table 1

Restriction Identification- Enzyme sequence of the recognition sequence First nucleotide position _ (coding strand) _ 5 '3' 10 Aha II, Ban I,

Hae II, Nar I, GGCGCC 8

Ban II, Sac I, Sst I GAGCTC 291 15

Hha I GCGC 9

Hinf I GACTC 35 20 Pvu I CGATCG 346

Taq I TCGA 387 If the nucleases BanII, SacI or SstI are used, an IL-2 subsequence encoding about 95 amino acids is obtained. This length is generally sufficient to obtain an insoluble fusion protein. If, for example, the low solubility of the desired hydrophilic eukaryotic protein is not yet sufficient and it is not desired to use cleavage sites closer to the C-terminus to provide the least possible "ballast", the DNA sequence can be extended with a suitable adapter at the N- and / or C-terminus. and "tailor" the "ballast" part in 35 ways.

The invention thus relates to a method for producing a fusion protein, characterized in that the bacterial host cell expresses a gene encoding a fusion protein having a C- or N-terminal portion substantially corresponding to at least one of interleukin 2 (IL-2). The first 96 amino acids but lack interleukin-2 activity.

4 93471

Preferably, the gene encodes a fusion protein of general formula

Met - X - Y - z or Met - Z - Y - X (Ia) (Ib) 5 wherein X is an amino acid sequence comprising at least the first 96 amino acids of human interleukin 2, Y is a direct bond or a bridging sequence of genetically encodable amino acids, which possibly lists the cleavage of the amino acid sequence Z, and Z is a sequence of genetically encodable amino acids.

As can be seen from formulas Ia and Ib - and as already mentioned above - it is possible to express the desired protein-15 before or after the IL-2 moiety. For the sake of simplicity, the first possibility, which corresponds to the general method for the preparation of fusion proteins, is mainly illustrated below. Thus, when this "classic" option is described below, another 20 options are thus not excluded.

The cleavage of the fusion protein can be performed chemically or enzymatically in a manner known per se. The choice of a suitable method depends above all on the amino acid sequence of the desired protein. For example, if this does not contain methionine, Y may be Met, which undergoes chemical cleavage with chlorine or bromo-cyano. If the carboxyl terminus of the linking moiety Y is cysteine or Y is Cys, enzymatic cysteine-specific cleavage or chemical cleavage, e.g. 30 characters after specific S-cyanylation. If the carboxyl terminus of the linking moiety Y is tryptophan or Y is Trp, chemical cleavage with N-bromosuccinimide can be performed.

Proteins whose amino acid sequence does not contain the pair 35 Asp-Pro 5 93471 and which are sufficiently acidic can be proteolytically cleaved in a manner known per se. This results in proteins containing an N-terminal proline group or a C-terminal aspartic acid group. Thus, modified proteins can also be synthesized in this way.

The Asp-Pro bond can be made even more sensitive to acids when this intermediate is (Asp) n-Pro or Glu- (Asp) n-Pro, where n is 1 to 3.

Examples of enzymatic cleavages are also known, in which modified enzymes with improved specificity can also be used [cf. C.S. Craik et al., Science 228 (1985) 291-297]. If the desired eukaryotic peptide is proinsulin, it is appropriate to select as sequence Y the peptide sequence in which the trypsin-cleavable amino acid (Arg, Lys) is linked to the N-terminal amino acid (Phe) of proinsulin, for example the sequence Ala-Ser-Met-Thr-Arg , because then arginine-specific cleavage can be performed with trypsin protease.

20 Unless the desired protein contains an amino acid sequence

Ile-Glu-Gly-Arg, the fusion protein can be cleaved by factor Xa (EP-A-0 025 190 and 0 161 973).

The fusion protein is formed in a manner known per se through expression in a suitable expression system. All known host vector systems are suitable for this purpose, i.e. for example mammalian cells and microorganisms, for example yeasts and preferably bacteria, in particular E. coli.

. The DNA sequence encoding the desired protein is inserted in a known manner into a vector that provides good expression in the selected expression system.

In the case of bacterial hosts, the promoter and operator are suitably selected from the group consisting of Lac, tae, 35 trp, λ-phage PL or PR, hsp, omp or a synthetic pro-_ Γ 6 93471 engine, which are proposed, for example, in DE-A-3 430 683 (EP-A 0 173 149). The preferred promoter-operator sequence is a guarantee which, moreover, is commercially available (e.g. the expression vector pKK223-3, Pharmacia, "Molecular Biologicals, Chemicals and Equipment for Molecular Biology", 1984, p. 64).

In the expression of the fusion protein of the invention, it may be appropriate to modify the individual triplets corresponding to the first amino acids 10 following the ATG start codon to prevent possible base pairing at the mRNA level. Such alterations, as well as alterations, deletions, or additions of individual amino acids in the IL-2 protein moiety, are within the skill of the art and are within the scope of the invention.

The invention is illustrated in more detail in the following examples and drawings. Thus, Figure 1 and its extension, Figure 1a, show the synthesis of plasmid pK360, which encodes a fusion op-20 protein containing a hirudin sequence; Figure 2 and its extension, Figure 2a, show the synthesis of plasmid pK410, which also encodes a fusion protein containing a hirudin amino acid sequence; Figure 3 and its extensions, Figures 3a to 3c, show the construction of plasmids pPH15, 16, 20 and 30, which encode fusion proteins containing the amino acid sequence of monkey proinsulin; Figure 5 and its extension, Figure 5a, shows plasmid pK370 encoding the hirudin amino acid sequence si-. . 30-containing fusion protein; and Figure 6 and its extension, Figure 6a, show the synthesis of plasmid pKH10I, which encodes a fusion protein containing the amino acid sequence of monkey proinsulin.

The scale of the figures is usually not correct, especially at the polylinker, the scale is "stretched".

Example 1 5 By locating a Lac repressor [P. J. Farabaugh,

Nature 274 (1978) 765-769] to plasmid pKK177-3 [Amann et al. Gene 25 (1983) 167] gives plasmid pJE118 (1) (Figure 1; cf. DE patent application P 35 26 995.2, Example 6, Figure 6). 6). This is opened at its single Aval restriction site 10 and truncated in a manner known per se by exonuclease treatment of about 1000 base pairs. After ligation, plasmid pEW100 (2) is obtained (Figure 1), which retains the entire Lac repressor gene but which, due to its smaller size, is clearly more abundant.

Instead of the plasmid pKK177-3, the above-mentioned commercially available plasmid pKK223-3 can also be used as a starting product, a Lac repressor can be placed in it and the product obtained can be shortened accordingly.

Plasmid pEW100 (2) is opened by restriction enzymes

EcoRI and Sali.

Plasmid (4) encoding hirudin, prepared according to Example 4 of DE-A-3 429 430 (EP-A-0 171 024) (Figure 3), is opened with restriction enzymes Acc1 and SalI, and a small fragment (5) containing for the most part the hirudin sequence.

Plasmid p159 / 6 (8), prepared according to Example 4 of DE-A-3 419 995 (EP-A-0 163 249) (Figure 5), is opened with restriction enzymes; 30 EcoRI and PvuI, and a small fragment containing most of the IL-2 sequence is isolated. This subsequence, and hereinafter also the other truncated IL-2 sequences, is referred to in the figures as AIL2 ".

8 93471

Sequences (3), (5), (7) and the synthetic DNA sequence s1 (8; Figure 1a) are then treated with T4 ligase. Plasmid pK360 (9) is obtained.

Competent E. coli cells are transformed with the ligase-5 thio product and inoculated onto NA plates containing 25 pg / ml ampicillin. Plasmid DNA of the clones is characterized by restriction and sequence analysis.

An overnight E. coli cell culture containing plasmid (9) is diluted in LB medium 10 (J. H. Miller, Experiments in Molecular Genetics, Cold

Spring Harbor Laboratory, 1972) containing 50 pg / ml ampicillin in a ratio of approximately 1: 100, and growth is monitored by measuring optical density (OD). With an OD of 0.5, the isopropyl-β-galactopyrra-15 noside concentration (IPTG concentration) of the shake culture is adjusted to 1 mM, and the bacteria are separated by centrifugation after 150 to 180 minutes. The bacteria are boiled for 5 min in a buffer mixture (7 mol / l urea, 0.1% SDS, 0.1 mol / l sodium phosphate, pH 7.0), and the samples are transferred to an SDS gel electrophoresis plate. After electrophoresis, a protein band corresponding to the expected size of the fusion protein is obtained from the bacteria containing plasmid (9). After bacterial disruption (French Press; Dyno-MUhle) and centrifugation, the fusion protein is in the precipitate, so that considerable amounts of other proteins can already be separated with the supernatant. After isolation of the fusion protein, the desired hirudin peptide is released by bromocyanine cleavage. It is characterized after isolation by protein sequence analysis.

The induction conditions given are valid for shake cultures; in the case of larger fermentations, correspondingly modified OD values and possibly slightly modified IPTG concentrations are appropriate.

Example 2

Plasmid (4) (Figure 1) is opened with Accl, and single-stranded ends are filled in with Klenow polymerase. Its

II

After 93471, digestion with SacI is performed, and fragment (10) containing most of the hirudin sequence is isolated.

The commercially available vector pUC13 is opened with the restriction enzymes SacI and SmaI, and a large fragment is isolated (11).

Fragments (10) and (11) are then ligated into plasmid pK400 (12) (Figure 2) using T4 ligase. Plasmid (12) is shown in Figure 2 as two drawings, the lower 10 of which show the amino acid sequence of the hirudin derivative thus obtained.

Plasmid (4) (Figure 1) is opened with the restriction enzymes KpnI and SalI, and a small fragment (13) containing the hirudin subsequence is isolated.

Plasmid (12) is reacted with the restriction enzymes Hindi and KpnI, and a small fragment (14) containing the hirudin subsequence is isolated.

Plasmid (9) (Figure 1a) is partially digested

With EcoRI, fill the free ends with a Kulow polymerase-assisted fill-in reaction, and digestion with SalI. A derivative of plasmid pK360 is obtained (15).

Ligation of fragments (3), (13), (14) and (15) gives plasmid pK410 (16), shown twice in Figure 2a, with the lower drawing showing the amino acid sequence of the fusion protein and thus the acid cleavage hirudin derivative. .

After expression and treatment according to Example 1, a new hirudin derivative having the amino acids proline and histidine at positions 1 and 2 is obtained. At hiru-. The dine derivative has similar activity to the natural product of DE, which has the amino acids threonine and tyrosine at these positions, but is more resistant to the action of aminopeptidases, which may be an advantage in vivo.

10 93471

Example 3

The commercially available vector pBR322 is opened with BamHI to give a linearized plasmid (17). The single-stranded ends are partially filled in using dATP, 5 dGTP, and dTTP, and the excess nucleotide G is removed with S1 nuclease to give the pBR322 derivative (18).

The HaellI fragment (19) of the monkey proinsulin gene [Wetekam et al., Gene 19 (1982) 181] is ligated into the modified plasmid (18) to give plasmid pPH1 (20). Because the insulin subsequence was inserted into the tetracycline gene, clones containing this plasmid are not tetracycline resistant and can thus be identified.

Plasmid (20) is opened with BamHI and DdeI, and a small fragment (21) is isolated.

In addition, the DdeI-PvuII subsequence is isolated from the monkey proinsulin sequence (22).

The vector pBR322 is opened with BamHI and PvuII, and the linearized plasmid is isolated (23).

Ligation of the insulin subsequences (21) and (22) to the opened plasmid (23) gives plasmid pPH5 (24). This is opened with BamHI and PvuII, and a small fragment is isolated (25).

To complete the insulin structure, a DNA sequence is synthesized ·· 25 (26).

The commercially available vector pUC8 is opened with BamHI and SalI, and the residual plasmid is isolated (27). This is ligated into DNA sequences (25) and (26) to plasmid pPH15 (28). This is opened with Sall, and filled with one -... 30 stranded ends. From the obtained plasmid derivative (29), the DNA sequence (30) is cleaved with BamHI.

The commercially available vector pUC9 is opened with BamHI and SmaI, and a large fragment is isolated (31). This is ligated to the DNA sequence (30) to give plasmid pH16 (32).

11 93471

Plasmid! (32) is opened with SalI, the plasmid-linked plasmid (33) is partially filled in with dCTP, dGTP and dTTP, and the remaining nucleotide is cleaved with T S1 nuclease. The plasmid derivative (34) thus obtained is treated with BamHI, and an additional single strand of S1 nuclease is removed from the product (35) to give the plasmid derivative (36).

The blunt ends of the plasmid derivative (35) are cyclized to plasmid pPH 2 O (37).

Competent E. coli Hb 101 cells are transformed with the ligation mixture and seeded on a selective medium. Clones containing the desired plasmid produce proinsulin, and 28 of the 70 clones examined contained radioimmunologically detectable proinsulin.

Plasmids are further characterized by DNA sequence analysis. They contain DNA with a codon corresponding to arginine above the codon corresponding to the first amino acid (Phe) of the B chain.

Plasmid (37) is digested with HindIII, filled into single-stranded ends and digested again with DdeI. Isolate a small fragment (38).

Plasmid (28) (Figure 3a) is digested with SalI and DdeI, and a small fragment (39) is separated.

Plasmid (9) (Figure 1a) is first digested with Accl, the free ends are filled in, and then digested with EcoRI. A fragment (40) containing a truncated IL-2 sequence is isolated.

The linearized plasmid (3) (Figure 1) and DNA segments (38), (39) and (40) are thus ligated into plasmid pPH30, 30 (41). This plasmid encodes a fusion protein comprising the following amino acid sequence after amino acids 1-114 of IL-2:

Asp-Phe-Met-Ile-Thr-Thr-Tyr-Ser-Leu-Ala-Ala-Gly-Arg.

12 93471 Arginine, the last amino acid in this intermediate Y, allows the insulin chain to be cleaved by trypsin.

Plasmid (9) (Figure 1a) can also be accessed to plasmid-5 (41) by the following route: (9) opening with Accl, filling single-stranded ends, digestion with SälI, and ligating the resulting plasmid derivative (42) to segments (3), (38) and (39).

Example 4 The commercially available vector pUC12 is opened with the restriction enzymes EcoRI and SacI. An IL-2 subsequence is cleaved into this linearized plasmid (46), which is cleaved from plasmid (6) (Figure 1) by the restriction enzymes EcoRI and SacI. This sequence (47) contains complete triplets corresponding to the first 94 amino acids of IL-2. Ligation (46) and (47) gives plasmid pK300 (48).

Plasmid (9) (Figure 1a) is opened with EcoRI, filled into single-stranded ends and then digested with HindIII. A small fragment (49) containing a portion of the pUC12 polylinker after the DNA segment encoding hirudin 20 is isolated.

Plasmid (48) is opened with the restriction enzymes SmaI and HinII, and a large fragment (50) is isolated. Ligation to (50) (49) gives plasmid pK301 (51).

“25 Competent E. coli 294 cells are transformed with the ligation mixture. Clones containing plasmid (51) are characterized by restriction analysis. They contain DNA with a 6 amino acid linker after the codons corresponding to the first 96 amino acids of IL-2. Codons corresponding to 30 sequences followed by codons corresponding to hirudin.

Plasmid (51) is treated with EcoRI and HindIII, and a fragment (52) containing the DNA sequence corresponding to said eukaryotic fusion protein is isolated. Plasmid (2) (Figure 1) is opened with EcoRI and HindIII.

tv K

13 93471 I: 11a. Obtained linearized Plasmid! (53) is ligated to the DNA sequence (52) to give Plasmid! pH370 (54).

When Plasmids! (54) is expressed in E. coli in a manner similar to Example 1, a fusion protein-5 having an intermediate after the first 96 amino acids of IL-2 is obtained.

Ala-Gln-Phe-Met-Ile-Thr followed by the amino acid sequence of hirudin.

Example 5 Plasmid (41) (Example 3; Figure 3c) is digested with restriction enzymes EcoRI and HindIII into a DNA segment encoding monkey proinsulin, and single-stranded ends are filled in. A DNA segment (55) is obtained.

Plasmid (48) (Example 5, Figure 5) is opened with 15 SmaI and treated with alkaline bovine phosphatase. The linearized plasmid (56) thus obtained is ligated to the DNA segment (55) to give plasmid pK302 (57). E. coli 294 cells are transformed with the ligation mixture, and clones containing the desired plasmid-20 are characterized by subjecting plasmid DNA first to restriction and then to sequence analysis.

A segment (58) encoding IL-2 and monkey proinsulin is cleaved from the plasmid (57) with EcoRI and HindIII.

Plasmid (2) (Example 1, Figure 1) is similarly digested with EcoRI and HindIII, and segment (58) is ligated to the linearized plasmid (3). Plasmid pKHIOI (59) is obtained.

Expression according to Example 1 results in a fusion process. 30 teins in which the first 96 amino acids of IL-2 are followed by a 14 amino acid fusion sequence [corresponding to sequence Y in the DNA segment (58)] followed by the amino acid sequence of monkey proinsulin.

14 93471

Annex I: DNA sequence of interleukin-2

Triplet No. 0 1 2 5 Amino acid Met Ala Pro

Nucleotide No. 1 10

Coding thread 5 'AA TTC ATG GCG CCG

Uncoded thread_3J_G TAC CGC GGC

10 5 4 5 6 7 8 9 10 11 12

Thr Ser Ser Ser Thr Lys Lys Thr Gin Leu 20 30 40

ACC TCT TCT TCT ACC AAA AAG ACT CAA CTG

TGG AGA AGA AGA TGG TTT TTC TGA GTT GAC

15 13 H 15 16 17 18 19 20 21 22

Gin Leu Glu His Leu Leu Leu Asp Leu Gin 50 60 70

CAA CTG GAA CAC CTG CTG CTG GAC CTG CAG

GTT GAC CTT GTG GAC GAC GAC CTG GAC GTC

2 0 23 24 25 26 27 28 29 30 31 32

Met Ile Leu Asn Gly Ile Asn Asn Tyr Lys 80 90 100

ATG ATC CTG AAC GGT ATC AAC AAC TAC AAA

2 5 TAC TAG GAC TTG CCA TAG TTG TTG ATG TTT

33 34 35 36 37 38 39 40 41 42

Asn Pro Lys Leu Thr Arg Met Leu Thr Phe 110 120 130

30 AAC CCG AAA CTG ACG CGT ATG CTG ACC TTC

: · TTG GGC TTT GAC TGC GCA TAC GAC TGG AAG

15 93471 43 44 45 46 47 48 49 50 51 52

Lys Phe Tyr Met Pro Lys Lys Ala Thr Glu 140 150 160

AAA TTC TAC ATG CCG AAA AAA GCT ACC GAA

5 TTT AAG ATG TAC GGC TTT TTT CGA TGG CTT

53 54 55 56 57 58 59 60 61 62

Leu Lys His Leu Gin Cys Leu Glu Glu Glu 170 180 190

10 CTG AAA CAC CTC CAG TGT CTA GAA GAA GAG

GAC TTT GTG GAG GTC ACA GAT CTT CTT CTC

63 64 65 66 67 68 69 70 71 72

Leu Lys Pro Leu Glu Glu Val Leu Asn Leu 15 200 210 220

CTG AAA CCG CTG GAG GAA GTT CTG AAC CTG

GAC TTT GGC GAC CTC CTT CAA GAC TTG GAC

73 74 75 76 77 78 79 80 81 82 20 Ala Gin Ser Lys Asn Phe His Leu Arg Pro 230 240 250

GCT CAG TCT AAA AAT TTC CAC CTG CGT CCG

CGA GTC AGA TTT TTA AAG GTG GAC GCA GGC

25 83 84 85 86 87 88 89 90 91 92

Arg Asp Leu Ile Ser Asn Ile Asn Val He 260 270 280

CGT GAC CTG ATC TCT AAC ATC AAC GTT ATC

GCA CTG GAC TAG AGA TTG TAG TTG CAA TAG

30 93 94 95 96 97 98 99 100 101 102 --- Val Leu Glu Leu Lys Gly Ser Glu Thr Thr 290 300 310

GTT CTG GAG CTC AAA GGT. TCT GAA ACC ACG

35 CAA GAC CTC GAG TTT CCA AGA CTT TGG TGC

ie 93471 103 104 105 106 107 108 109 110 111 112

Phe Met Cys Glu Tyr Ala Asp Glu Thr Ala 320 330 340

TTC ATG TGC GAA TAC GCG GAC GAA ACT GCG

5 AAG TAC ACG CTT ATG CGC CTG CTT TGA CGC

113 1H 115 116 117 118 119 120 121 122

Thr Ile Val Glu Phe Leu Asn Arg Trp He 350 360 370

10 ACG ATC GTT GAA TTT CTG AAC CGT TGG ATC

TGC TAG CAA CTT AAA GAC TTG GCA ACC TAG

123 124 125 126 127 128 129 130 131 132

Thr Phe Cys Gin Ser He He Ser Thr Leu 15 380 390 400

ACC TTC TGC CAG TCG ATC ATC TCT ACC CTG

TGG AAG ACG GTC AGC TAG TAG AGA TGG GAC

133 134 135 20 Thr 410 ACC TGA TAG 3 'TGG ACT ATC AGC T 5' • · * «·

Claims (12)

1. A method for producing a fusion protein, characterized in that a gene encoding a fusion protein with a C- or N-terminal moiety substantially corresponds to at least the first 96 amino acids of interleukin 2 (IL-2). ), but lacking interleukin-2 activity, for expression in a bacterial host cell. A method according to claim 1, characterized in that the gene encodes a fusion protein of the general form Met-XYZ or Met-ZYX, (Ia) (Ib) in which X is an amino acid sequence comprising essentially the at least 96 first the amino acids of human interleukin 2, Y are a straight link or a bridging portion consisting of genetically encodable amino acids, which enable the cleavage of amino acid sequence Z, and Z is a sequence consisting of genetically encodable amino acids. Method according to Claim 2, characterized in that Y next to Z contains a group "Met, Cys, Trp, Arg or Lys or consists of these amino acids. Method according to claim 2, characterized in that Y next to Z contains a sequence Asp-Pro or consists of this sequence. 5. A method according to any one of claims 1-4, characterized in that Z is the amino acid sequence of proinsulin or hirudin. 93471 ---- ί 6. A method according to any of claims 1-5, characterized in that fusion proteins are separated from soluble proteins by centrifugation. Process according to any of claims 1-6, characterized in that the host cell is E. coli. Use of a fusion protein obtainable by a method according to any one of claims 1 to 7 for the production of a protein substantially corresponding to the amino acid sequence Z, by chemical or enzymatic cleavage. 9. A gene structure, characterized in that it encodes a fusion protein presented in claim 1 or 2. 10. A vector, characterized in that it contains a gene structure according to claim 9. 11. The plasmid, characterized in that it is the pEW 1000 shown in Figure 1, in pK360 shown in figure 1a, in pK410, in figure 2a, in pPH30, in figure 3c, in pK370 indicated in figure 5a, or in pKH101 shown in figure 6a. Bacterial host cell, characterized in that it contains a vector according to claim 10. «