DE3731874A1 - METHOD FOR THE PRODUCTION OF PEPTIDES BY SPECIFIC CUTTING OF GENETICALLY OBTAINED FUSION PROTEINS WITH COLLAGENASES - Google Patents

Abstract

Fusion proteins are disclosed having the formula (I) H2N-Z1-X-(Pro-Y-Gly)n-Pro-Z2-COOH, in which n >/= 2, X and Y are each of the 20 amino-acids determined by the genetic code, Z1 is a bacterial amino-acid sequence and Z2 is the target peptide composed of any desired amino-acids of the genetic code. Also disclosed are their production process and their enzymatic cleavage to form desired eukaryotic proteins.

Description

Genetic engineering has new ways of making peptides and proteins opened, which of the conventional chemical Synthesis, especially with longer amino acid sequences, far are superior. As a result, otherwise difficult to access are available for the first time Substances like growth hormone, interferons and numerous peptide hormones available in sufficient quantities (F.A.O. Marston, Biochem. J., 240, 1-12 [1986]).

The genetic engineering processes for the production of polypeptides are based on the transfer of a suitable deoxyribonucleic acid (DNA), which codons for the amino acid sequence to be produced contains, on suitable host cells in a form which  the replication and expression of the newly introduced genetic Information enables. As host cells are small to medium-sized peptides (less than approx. 100 amino acids) preferably bacteria such as Escherichia coli are used.

It has now been shown that polypeptides of the size range mentioned very easily from intracellular bacterial proteases be attacked because they are not characterized by a pronounced tertiary structure are protected. This can lead to a severe reduction in yield or also for the complete loss of the primary biosynthesis product lead if the expression is directed so that the polypeptide to be produced directly in the final form or at most with an additional N-terminal methionine residue is formed (so-called direct expression). Let against it the desired amino acid sequences are usually in Get good yield if they are initially considered to be much larger Precursor molecules, so-called fusion proteins, synthesized will.

A method variant is therefore preferred for the production of smaller and medium-sized peptides, in which a hybrid gene is formed on a plasmid vector by in vitro recombination from a bacterial gene which can be expressed well and, if possible, also regulated, and the DNA sequence for the peptide to be produced. The hybrid gene contains the original bacterial regulatory elements (promoter, ribosomal binding site and terminator) and as large a part of the codons for the N-terminal amino acids of the bacterial gene product as possible, followed by the codons for the new peptide and a stop codon. After conversion of the plasmid vector into a suitable bacterial host cell by transformation, the hybrid gene controls the biosynthesis of a fusion protein, the N-terminal part of which is derived from the corresponding bacterial protein, while the C-terminal part consists of the amino acid sequence of the peptide to be produced. This fusion protein can easily be produced and isolated by fermentation of the transformed cells in large quantities. Various bacterial genes have been used for the construction of hybrid genes, but preferably those for the enzymes β- galactoidase, β- lactamase, anthranilate synthetase, alkaline phosphatase and chloramphenicol acetyltransferase.

The last step in the production of a peptide after The procedure outlined above is the exact cleavage of the fusion protein. To make this possible, the hybrid gene must be constructed in this way be that in the later fusion protein between the bacterial part and the new peptide sequence to be produced easily and selectively cleavable peptide bond is present. The Development of a specific and as broadly applicable as possible Process for the cleavage of fusion proteins has proven to be the largest and not yet completely satisfactorily resolved Problem in the genetic engineering of small and medium-sized Peptides proved.

The present invention relates to a specific one and therefore generally applicable method for cleaving fusion proteins,  which is the exact amino acid sequence of the desired Peptide delivers without undesirable at the N-terminal end to leave additional amino acids.

The first genetically engineered peptides, such as B. Somatostatin (K. Itakura et al., Science, 198, 1056-1063 [1977]) or the A and B chains of human insulin (DV Goeddel et al., Proc. Natl. Acad. Sci., USA , 76, 106-110 [1979]), had been released by cleavage with cyanogen bromide from the corresponding fusion proteins which were derived from β- galactosidase. Bromocyanine selectively cleaves the bond between methionine and the nearest carboxylic amino acid (Met-X, where X can be any of the 20 amino acids specified by the genetic code). Naturally, this method can only be used for peptides which themselves do not contain methionine; this applies to the three peptides mentioned above.

Similar restrictions apply to all splitting processes of fusion proteins, which are based on the recognition of a single Amino acid based. This is how fusion proteins are made which are the amino acids lysine, arginine or glutamic acid as a link between the N-terminal and the C-terminal Part included. The enzymes trypsin (J. Shine et al., Nature, 285, 456-461 [1980]) and Clostripain (A.D. Bennett et. al., EPA 01 31 33), both of which are the bond between Open Lys-X and Arg-X as well as endoproteinase Glu-C (A.D. Bennett et al., EPA 01 31 363), which bind the Glu-X splits, used. Endoproteinase Lys-C (G. Allen et al., Biol.  Chem. Hoppe-Seyler, 367, Suppl. Aug. 1986, p. 162, Abstr. 19/01/03), which specifically cleaves for lysine, is also been used.

Processes are much more specific and therefore more widely applicable that have a specific bond within a well defined one Split sequence from several amino acids. So could P. R. Szoka et al. (DNA, 5, 11-20 [1986]) a fusion protein, in which is a partial sequence from anthranilate synthetase and bovine Growth hormone linked together via the Asp-Pro sequence are due to the action of acid in the two components disassemble because the Asp-Pro bond is very acid-labile. In addition to the fact that when exposed to acid also with the cleavage of further peptide bonds as a side reaction The main disadvantage of this method is that the proline from the coupling sequence at the N-terminal end of the Peptide remains, which in many cases is intolerable. The same limitation applies to the implementation with Renin which has been used to make up a fusion protein selectively within the sequence Pro-Phe-His-Leu-Leu-Val-Tyr Cleave bond between the two leucine residues (J. S. Boger et al., EPA 01 63 573). In this case there are three additional ones and often unwanted N-terminal amino acids.

Factor Xa, a plasma protease whose normal function is to convert prothrombin to thrombin, has been used by K. Nagai and HC Thøgersen (Nature, 309, 810-812 [1984]) and to release human β- globin from a fusion protein, whose N-terminal part consisted of 31 amino acids of the λ cII protein. Similarly, T. Imai et al. (J. Biochem., 100, 425-432 [1986]) used the enzyme to produce human prorenin via a fusion protein with 9 additional N-terminal amino acids. In both cases the sequence Ile-Glu-Gly-Arg-X was present as a transition between the N- and C-terminal region, which is specifically recognized by factor Xa and cleaved at the bond between Arg and X (where X is the first N- terminal amino acid of human protein). Factor Xa therefore seems to meet the requirements that must be placed on an enzyme which is to be used for the cleavage of fusion proteins. However, two cases have also been reported in which a cleavage of fusion proteins containing the above recognition sequence was not possible with factor Xa (G. Allen et al., Biol. Chem. Hoppe-Seyler, 367, Suppl. Aug. 1987 , P. 162. Abstr. 19.01.03; H. Mayer et al., Biol. Chem. Hoppe-Seyler, 367, Suppl. Aug. 1986, p. 285, abstr. 05.01.50.). A general application of this method is therefore not possible.

Similar restrictions have to be made for the enzyme enterokinase, which in the sequence (Asp) ₄-Lys-X the bond between Lys and X split. R.M. Belagaje et al. (DNA, 3, 120 [1984]) with this enzyme human growth hormone from a fusion protein released in which this sequence is not located far from the N-terminal end, similar to that in  Trypsinogen, the natural substrate of enterokinase.

Attempts to find fusion proteins that contain alkaline Phosphatase and adrenocorticotropic hormone via the above Sequence were linked to cleave with enterokinase however not successful.

EP 20 290 describes a method which the high specificity of collagenases for the cleavage of fusion proteins appropriate structure is used.

Collagenases recognize the sequence Pro-Y-Gly-Pro, where Y is any of the 20 amino acids of the genetic Codes can be and split between Y and Gly. From a fusion protein, which contains this sequence becomes a peptide released, which at the N-terminal end still from the Contains collagenase sequence-derived amino acids Gly and Pro. These can be used as a dipeptide with another enzyme, the postproline Dipeptidyl aminopeptidase (PPDA) are removed.

Attempts to cleave peptides and proteins which the Sequence Pro-Y-Gly-Pro included, with clostridiopeptidase A, one Collagenase from Clostridium histolyticum (EC 3.4.24.3) now shown that the application of the above Sequence of reactions for the production of peptides via fusion proteins is only possible to a very limited extent. M. Töpert could et al. (Communication at the 14th FEBS meeting [1981]) show that a relatively small, semi-synthetic model peptide, and the connection though

Cbo-Gly-Pro-Leu-Gly-Pro Insulin A Chain (Bovine)

according to the scheme shown above with clostridiopeptidase A and PPDA can be implemented, with the expected reaction products were obtained in good yield. Similar results received E. Wünsch et al. (Hoppe-Seylers Z. Physiol. Chem., 362, 1285-1287 [1981]) in the implementation of another model peptide. In contrast, four of the natural proteins with the sequence Pro-Y-Gly-Pro only two (Neurophysin II and Bovine catalase) specifically clostridiopeptidase A, and this only after converting the cysteine side chains into the S-sulfonates. Despite a relatively large excess of enzyme however, no quantitative implementation could be achieved (M. Töpert et al., Communication at the 14th FEBS meeting [1981]).

In the German patent application (file number P 37 31 875.6) a process for the production of adrenocorticotropic hormone described about fusion proteins that differ from the alkaline Derive phosphatase from E. colie. In this case too, it was found that fusion proteins with a tetrapeptide sequence Pro-Y-Gly-Pro between bacterial protein sequence and target peptide, where Y means each of the 20 amino acids of the genetic code, specifically with the predetermined position Clostridiopeptidase A cleaved, but the reaction rate was very low.

J. Germino and D. Bastia (Proc. Natl. Acad. Sci., USA, 81, 4692-4696 [1984]) have produced a fusion protein which, as a link between the N- and C-terminal region, has a sequence of approx. Contained 60 amino acids from Pro α -2 collagen. This protein could be cleaved with collagenases just as well as the corresponding collagen itself. However, it is not known at which locations within the sequence of approximately 60 amino acids the fusion protein is cleaved exactly. It was also found that a more or less large residue of this sequence is still present at the N-terminal end of the C-terminal cleavage product. This process variant is therefore unsuitable for the production of polypeptides with a precisely specified N-terminal end.

The present invention is a method of manufacture of target peptides, which is characterized in that from a Fusion protein of formula I.

H₂N-Z₁-X- (Pro-Y-Gly) _n -Pro-Z₂-COOH (I)

in the
n is 2,
X and Y represent each of the 20 amino acids defined by the genetic code,
Z₁ is a bacterial amino acid sequence and
Z₂ is the target peptide from any amino acids of the genetic code,

the C-terminal to the amino acid sequence -X- (Pro-Y-Gly) _n -Pro-, in which X and Y are each genetically encodable amino acid and n 2, enzymatically cleaves the following protein sequence with collagenase and postproline dipeptidylaminopeptidase (PPDA).

The invention also relates to fusion proteins of the formula I. and processes for their preparation according to claims 1-6 as well as gene structures which are necessary for fusion proteins of the formula I encode and plasmids or comparable vectors (e.g. Phage vectors), the codons for repetitive collagenase Contain interfaces and thus for the production of fusion proteins of formula I are suitable.

As bacterial amino acid sequence Z₁ alkaline phosphatase, anthranilate synthetase, β- lactamase, β- galactosidase, chloramphenicol acetyltransferase etc. come into question.

The reaction scheme below explains the proteolytic Cleavage of fusion proteins with collagenase and Postproline dipeptidyl aminopeptidase (PPDA):  

X and Y can be any of the 20 amino acids specified in the genetic code, n preferably means 2 to 10.

The advantage of the new process for cleaving fusion proteins compared to the process from EP 20 290 consists in that the rate of implementation z. B. with clostridiopeptidase A can be increased by orders of magnitude. Fusion proteins that contain the recognition sequence for collagenase in repetitive form, can be split much better than that corresponding fusion proteins with a simple cleavage site for collagen noses.  

To illustrate the method according to the invention, in Examples 1-3 the preparation and enzymatic cleavage of 3 model proteins described, which are different, by collagenases contain cleavable sequences. The amino acid sequences these proteins are summarized in Tab. 1. You guide yourself on the prephosphatase from E. coli K 12, an intermediate the biosynthesis of alkaline phosphatase, which consists of 471 amino acids consists. By cleaving an N-terminal signal sequence Prephosphatase is converted into active from 22 amino acids Enzyme transferred. In contrast, the modified phosphatases the signal sequence of 22 amino acids was not significant cleaved.

The amino acid sequence of alkaline phosphatase (RA Bradshaw et al., Proc. Natl. Acad. Sci., USA, 81, 4692-4696 [1981]) and the nucleotide sequence of the corresponding gene from E. coli K 12 (phoA) are known ( M. Simonis, PhD, Free University Berlin [1983]); CN Chang et al., Gene, 44, 121-125 [1986]); Vectors which can be used for the construction of fusion proteins based on alkaline phosphatase have also been described (W. Boidol et al., Mol. Gen. Genet., 185, 510-512 [1982]; W. Boidol, Dissertation, Technical University Berlin [1984]). Starting from one of these plasmids, which bears the designation pSB94, known methods of in vitro recombination and DNA cloning (T. Maniatis et al., Molecular Cloning, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA) were used [1982]) produced modified vectors which, after being converted into suitable E. coli host cells, bring about the biosynthesis of the proteins shown in Table 1. For this purpose, oligonucleotides were inserted in two positions of the phoA gene on pSB94, namely in the NcoI and SphI sequences, which form codons in the modified genes for the amino acid sequences recognizable by collagenases. Restriction maps of pSB94 and the plasmids pSA302, pSA506 and pHS4133 derived therefrom are shown in FIGS. 1-4.

The results of the implementation of the 3 model proteins with clostridiopeptidase A are shown in Figs. 5 and 6 and in Table 2. Fig. 5 shows a gel electrophoretic separation, which shows the formation of cleavage products of the expected molecular weight and thus the specific cleavage at the predetermined location. Fig. 6 shows the dependence of product formation on the incubation time and the amount of enzyme used. The product formation rates determined from the curves, which are summarized in Tab. 2, make clear to what extent a double interface (HS 4133/1) and especially a five-fold interface (SA 506/1) the reactivity towards collagenases compared to a simple one Increase interface (SA 302/1).

In the German patent application (file number P 37 31 875.6), the subject of which is a process for the production of adrenocorticotropic hormone and peptides derived therefrom, fusion proteins are described in which the hormone sequence via simple or repetitive collagenase interfaces from partial sequences of alkaline phosphatase E. coli are coupled. The experiments described there for the implementation of these fusion proteins with clostridiopeptidase A fully confirm the results described in the present invention. The product formation rates of the most important ACTH fusion proteins are also shown in Table 2. While the protein with a simple collagenase interface (SA 186/1) could only be released incompletely and under conditions that are unsuitable for practical preparative use, the corresponding protein with a double interface (SA 343/1) already split better by a factor of at least 10³. The reaction rate was further increased by using a five-fold or eight-fold interface (SA 341/1 or SA 360/1). The sequence of the ACTH fusion protein SA 441/1 is shown in Table 1. Fig. 5 shows in lane 3 the products of the conversion of SA 341/1 with clostridiopeptidase A.

The implementation with PPDA according to the scheme already given above can occur when using repetitive collagenase interfaces only then to the authentic peptide with the exact amino acid sequence lead at the N-terminal end when within the repetitive Interface also cleaves the Y-Gly bond that is closest to the N-terminal end of the peptide. Analytical Investigations by gel electrophoresis and HPLC left no Heterogeneity of the cleavage products from fusion proteins with repetitive Recognize collagenase interfaces. This indicates that indicates that when reacting with clostridiopeptidase A, all Y-Gly Bonds of the repetitive sequence are cut. Like in the German patent application (file number P 37 31 875.6) this was described using the example  of the corresponding fission product from SA 341/1 by the determination of an N-terminal partial sequence confirmed: Edman degradation gave the expected for the first seven amino acids of Gly-Pro-ACTH Sequence. Apparently by splitting a first bond the further cleavage sites within a repetitive sequence activated by exposure and better accessibility, that they react much faster than the first fission point, so that intermediate products of incomplete cleavage cannot be detected are.

By using repetitive collagenase interfaces the specificity of the process is significantly increased, and even Peptides that happen to contain the Pro-X-Gly-Pro sequence because of the low reactivity of this sequence under the conditions used here. So that is the process of making virtually any peptide suitable, and also in the choice of host / vector systems as well the use of other collagenases and aminopeptidyl dipeptidases more similar There are no restrictions on specificity.

The method according to the invention can of course also be carried out in one step, i. H. with simultaneous implementation with collagenase and PPDA.

On July 16, 1979, the American Type Culture Collection (ATCC) in Rockeville, Md., USA, deposited (accession number): plasmid pBR 322 (ATCC 40015).
On July 16, 1979 the following was deposited with the DSM, Göttingen: Escherichia coli SB 44 (DSM 1606).
Since November 26th In 1973 the DSM deposited a deposit for: Escherichia coli K 12 wild type (DSM 498).

Examples 1. Production of recombined plasmids with genes for the Phosphatase proteins SA 302/1, SA 506/1 and HS 4133/1

The biosynthesis of the three proteins is transformed accordingly Bacterial cells through the recombined plasmids pSA302, pSA506 and pHS4133 determined. These plasmids are out the vector plasmid pSB94 was created. pSB94 was extracted from the plasmid pSB53 (ATCC 40020) according to a described method (W. Boidol, Dissertation, Technische Universität Berlin [1984], see also U.S. Patent No. 4,375,514).

Fig. 1 shows a restriction map of pSB94. The plasmid carries the gene for β- lactamase (ampicillin resistance) from pBR322 and the gene for alkaline phosphatase (phoA) from E. coli K 12. For the production of pSA302 and pSA506, corresponding oligonucleotides were included in the recognition sequences for the restriction endonucleases NcoI and SpI, which are located within the phoA gene, inserted by DNA cloning. pHS4133 was then prepared from pSA506 by removing a 27 bp PstI fragment. Fig. 2-4 show restriction maps of the three newly combined plasmids.

Known from the literature were used to prepare the recombined plasmids Methods of DNA cloning applied. These are by T. Maniatis et al. (Molecular cloning, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., USA [1982]) in detail  and have been described in summary. The serves as the host cell E. coli strain SB 44 (W. Boidol, dissertation, Technical University Berlin [1984]); G. Siewert et al., U.S. Patent No. 4,375,514 [1983]). This is a phosphatase negative mutant (phoA) of the widely used cloning strain E. coli HB 101.

Oligonucleotides were made using an automated DNA synthesizer, Model 381A from "Applied Biosystems, 850 Lincoln Center Drive, Foster City, Calif. 94 404, USA " synthesized according to the instructions of the device manufacturer and by gel electrophoresis cleaned.

DNA was sequenced by the dideoxy method (F. Sanger et al., Proc, Natl. Acad. Sci., USA, 74, 5463-5467 [1977]) using the single-strand phage vectors M13mp18 and M13mp19 (C. Yanisch-Perron et al., Gene, 33, 103-119 [1985]). Deoxyadenosine 5- [ α -³⁵S] thiotriphosphate was used for the radioactive labeling (MD Biggin et al., Proc. Natl. Acad. Sci., USA, 80, 3963-3965 [1983]).

The restriction endonucleases EcoRI, PpstI, SphI and HaeIII as well DNA ligase (T4) were obtained from Boehringer, Mannheim. NcoI was from New England Biolabs GmbH, 6231 Schwalbach, and Polynucleotide kinase (T4) from Pharmacia / P-L, 7800 Freiburg i. Br.  

1.1. Manufacture of pSA302

The plasmid pSB94 was with the restriction endonuclease NcoI linearized. For this purpose, a reaction mixture containing 14 µg Plasmid DNA and 17 units of NcoI in 50 µl reaction buffer (50 mM Tris HCl, 10 mM MgCl₂, 100 mM NaCl₂, 1 mM dithiothreitol, pH 7.5), incubated at 37 ° C. for 7 h. After extracting the Reaction mixture with phenol, the DNA was precipitated with ethanol, dried and in TE buffer (45 mM Tris HCl, 1 mM EDTA, pH 7.5) solved.

1 µg each of the two 18mer synthetic oligonucleotides
5′-CAT GGC CCT GCA GGA CCC-3 ′
and
5′-C ATG GGG TCC TGC AGG GC-3 ′,
which were not phosphorylated on the 5'-OH group were mixed in aqueous solution and preincubated at 37 ° C. for 15 minutes. They were then treated with 3.5 µg linearized pSB94 and 4 units of DNA ligase (T4) in 50 µl ligase reaction buffer (20 mM Tris HCl, 10 mM MgCl₂, 10 mM dithioerythritol, 0.6 mM ATP, pH 7.6) Implemented 2.5 h at 20 ° C. The DNA was processed by phenol extraction and ethanol precipitation as described above and used to transform the strain E. coli SB 44. The competent cells were produced and transformed using a method by M. Dagert and SD Ehrlich (Gene, 6, 23-28 [1979]).

The transformed cells were plated into single colonies. Colonies whose plasmid DNA contained the above-mentioned oligonucleotide sequences were identified by colony hybridization. Colony hybridization was carried out according to the method described by T. Maniatis et al. (Molecular Cloning, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA [1982]) given basic protocols in a JP Gergen et al. (Nucleic Acids Research, 7, 2115-2136 [1979]). The above-mentioned oligonucleotides, which had previously been labeled with 5′-P³²-phosphate by reaction with γ P³²-ATP and polynucleotide kinase (T4), served as the radioactive hybridization reagent.

In the newly combined plasmids of the positive clones, the Orientation of the synthetic 18 bp NcoI fragment by Restriction analysis determined. This was done according to standard procedures Plasmid DNA isolated and with the restriction endonuclease HaeIII implemented. The resulting fragment mixtures were by gel electrophoresis analyzed (molecular weight determined by Comparison with HaeIII-cut pBR322 in an 8% polyacrylamide gel).

The NcoI interface used for the integration of the synthetic fragment of 18 bp is located in pSB94 on a HaeIII fragment of 385 bp. This fragment must be missing in the recombined plasmids and replaced by two new fragments, since the 18 bp fragment itself carries a HaeIII sequence. The size of the two new fragments depends on the orientation of the 18 bp fragment: in the desired orientation shown in FIG. 2, fragments of 342 bp and 61 bp are expected, while the opposite orientation results in fragments of 330 bp and 73 bp . Several plasmids of both orientations were found. The pSA302 shown in Fig. 2 is one of the plasmids with the desired orientation.

The 349 bp EcoRI fragment from pSA302, which contains the synthetic DNA inserted by cloning, was cloned into the sequencing vector M13mp19 and sequenced as indicated above. The partial sequence of pSA302 shown in Fig. 2 between the two EcoRI interfaces could be fully confirmed.

1.2. Manufacture of pSA506

The preparation was similar to that of pSA302. pSB94 was linearized with the restriction endonuclease SphI; Reaction conditions and workup were the same as for pSA302.

The two 45mer oligonucleotides
5′-GTCCTGCAGGCCCGGCGGGTCCAGCAGGCCCTGCAGGTCCGCATG-3 ′
and
5′-CGGACCTGCAGGGCCTGCTGGACCCGCCGGGCCTGCAGGACCATG-3 ′
were reacted with SphI-linearized pSB94 and DNA ligase under the same conditions as described for pSA302, and worked up. After transformation of E. coli SB 44, positive clones were identified by colony hybridization with the P³²-end-labeled 45mer oligonucleotides.

The recombined plasmids obtained in this way have an intact SPhI sequence only on one side of the DNA section formed from the two 45mer oligonucleotides. In the plasmids sought, this is at a distance of 257 bp from the closest EcoRI sequence (see FIG. 3), while in the plasmids of the opposite orientation, as in pSB94, this distance is only 212 bp.

The size of the DNA fragments obtained by reacting the newly combined plasmids with the restriction sequences EcoRI and SphI by gel electrophoresis showed that pSA506 is one of the plasmids with the desired orientation of the synthetic DNA fragment. To confirm the sequence shown in FIG. 3, the EcoRI / SphI fragment of 257 bp was cloned into the sequencing vector M13mp18 and sequenced.

1.3. Preparation of pHS4133

A reaction mixture of 3 μg DNA of the plasmid pSA506 and 5 units of the restriction enzyme PstI in 50 μl reaction buffer (as in 1.1.) Was incubated for 2 hours at 37 ° C. and as in Example 1.1. worked up. 3 DNA fragments (27 bp, 2420 bp and 4007 bp) were obtained, which were separated from one another by electrophoresis in a 1% agarose gel. The two larger fragments were isolated from the gel by electroelution, combined and with DNA ligase (T4) under the reaction conditions as in Example 1.1. implemented, refurbished and used to transform E. coli SB 44. Plasmid DNA was isolated from some ampicillin-resistant clones. One of these is pHS4133, the restriction map of which is shown in Fig. 4. pHS4133 is derived from pSA506 by a deletion which leads to a reduction in the size of the EcoRI / SphI fragment from 257 to 230 bp. This fragment was as in Example 1.2. cloned and sequenced in M13mp18. The nucleotide sequence shown in Fig. 4 was confirmed.

2. Production of the proteins SA302 / 1, SA506 / 1 and HS4133 / 1

The three proteins were obtained by fermentation of the strains described in part 1 of the examples
E. coli SB 44 (pSA302),
E. coli SB 44 (pSA506),
and
E. coli SB 44 (pHS4133)
receive. Their biosynthesis is regulated in the same way as that of alkaline phosphatase: it is repressed by inorganic phosphate and, conversely, is induced in low-phosphate media (A. Torriani, Biochem. Biophys. Acta, 38, 460-469 [1960]).

To identify the phosphatase derivatives, u. a. the immunoblot H. Towbin et al. (Proc. Natl. Acad. Sci., USA, 76, 4350-4354 [1979]). The antiserum needed for this was more alkaline by immunizing rabbits with purified alkaline E. coli K 12 phosphatase under standard conditions  (see D. M. Weir, Handbook of Experimental Immunology, Vol. III, Blackwell Scientific Publ., Oxford, 1978). The Phosphatase used for immunization was known according to Methods (A. Torriani, in: "Methods in Enzymology", Vol. XIIB, 212-218, L. Grossman, K. Moldave [Eds.], Academic Press, New York [1968]) isolated. A goat antiserum served as the second antibody against rabbit IgG, to which alkaline phosphatase as Marker enzyme was coupled (Cooper Biomedical Inc., Malvern, USA). 5-bromo-4-chloro-3-indolyl phosphate was used as a color reagent.

The three proteins as well as the three above Tribes of which they produced are so similar to each other that for everyone the same fermentation and isolation process applied can be.

2.1. fermentation

20 ml of L medium (ES Lennox, Virology, 1, 190-206 [1955]), which additionally contained 100 μg / ml ampicillin, were inoculated with a single colony, shaken at 37 ° C. for 24 h and then in 800 ml of the same Transferred medium, which in turn were shaken at 37 ° C for 16 h. This preliminary cultivation was transferred to a 20 l glass fermenter containing 9.2 l of low phosphate medium (LP medium according to K. Kreuzer et al., Genetics, 81, 459-468 [1975]). The fermenter was stirred at 37 ° C. and aeration of 10 l / min at 220 min ^{-1 for} 6 h. The cells were then harvested in a flow-through centrifuge, about 30 g of moist cell mass being obtained.

2.2. Isolation of enriched protein extracts

30 g moist cells from 2.1. were in 400 ml digestion buffer (50 mM Tris HCl, 160 mM NaCl, pH 8.0) suspended. After encore of phenylmethylsulfonyl fluoride (final concentration 10 mM) and DNAse (DNAse I from bovine pancreas, Boehringer / Mannheim, Purity grade I; 10 µg / g wet cells) the suspension in a cell homogenizer (Manton-Gaulin) at a pressure of Opened 650-700 kg / cm² (10,000 psi) in a cold room (4 ° C). The exclusion was repeated twice, after each The suspension was cooled in ice. The insoluble Cell components were separated by centrifugation (20 min at 10,000 × g and 4 ° C), washed twice with digestion buffer and dissolved in 50 mM Tris HCl, pH 8.0 / 8 M urea (60 ml). The Solution was obtained by centrifuging insoluble matter freed and until further purification by gradient chromatography kept in ice.

2.3. Identification of the phosphatase derivatives

Main protein component (approx. 50-70%) of the 2.2. received Protein extracts in urea buffer is the respective phosphatase Derivative. After separation of the protein mixtures by electrophoresis in SDS polyacrylamide gels under standard conditions (U. K. Laemmli, Nature, 227, 680-685 [1970]) were the corresponding Protein bands identified using the following criteria:  

• a) The strongest band of the protein mixture has a molecular weight, which of the amino acid sequence given in Table 1 corresponds.
• b) This protein is repressible by phosphate, i.e. i.e., the Band is missing if the cells as in 2.1., But with the 10 times the phosphate concentration were fermented.
• c) The band is plasmid-specific, it is missing in cells of the plasmid-free strain E. coli Sb 44, which is among the same Conditions as in 2.1. was produced.
• d) The band reacts specifically in the immunoblot with antiserum against purified alkaline phosphatase from E. coli K 12.

From gel electrophoretic molecular weight determinations with Using a standard protein mixture (low molecular weight protein from Pharmacia) it can be seen that the phosphatase Derivatives in the so-called Pre-form, d. H. still the signal sequence of alkaline phosphatase. The proteins are relatively difficult to dissolve in dilute buffer. Therefore in the supernatants obtained after cell disruption and centrifugation detect only small amounts.

2.4. Purification of the phosphatase derivatives

The in 2.2. Protein extracts obtained were used for further purification the modified phosphatases by gradient chromatography using a medium pressure chromatography device (FPLC) from Pharmacia, Uppsala / Sweden. An anion exchange column from the same manufacturer (type Mono Q-HR 5/5)  was equilibrated with buffer A (50 mM Tris HCl, 8 M urea, pH 8.0), with the protein solution from 2.1. loaded and finally with one of buffer A and buffer B (50 mM Tris HCl, 1 M NaCl, 8 M urea, pH 8.0) formed gradients of 0-200 mM NaCl eluted at a flow rate of 1.0 ml / min. The entire gradient volume was 20 ml, fractions of 1 ml each collected.

The absorbance of the eluate at 280 nm was recorded with a flow plotometer. The protein composition of the individual fractions was analyzed by SDS-polyacrylamide gel electrophoresis. The phosphatase derivatives eluted at 50-70 mM NaCl. The corresponding fractions were combined and dialyzed against the 400-fold volume buffer (50 mM Tris HCl, 150 mM NaCl, pH 8.0). The proteins remained dissolved under these conditions and were checked in this form for the subsequent cleavage with collagenase SDS-polyacrylamide gel electrophoresis (see Fig. 5, lanes 4, 6 and 8 ), it was greater than 90%. The protein content was determined according to W. Schaffner et al. (Anal. Biochem., 56, 502-514 [1973]) with amido black 10B, the concentrations were 0.6-0.7 mg / ml. Approx. 160 mg protein was obtained from a 10 l fermentation.

3. Specific cleavage of the modified phosphatase proteins with collagenase 3.1. Purification of the collagenase

Clostridiopeptidase A from Clostridium histolyticum (EC 3.4.24.3) was from Sigma Chemie GmbH, D-8024 Deisenhofen, Grünwalder Route 30, related (Type VII, Order No. C0773). The enzyme was determined by gradient chromatography using the FPLC Device and the same anion exchange column as in example 2.4. further cleaned. 2 mg of collagenase were used and with a buffer A (10 Tris HCl, 5 mM CaCl₂, pH 8.0) and Buffer B (10 mM Tris HCl, 5 mM CaCl₂, 1 mM NaCl, pH 8.0) was formed Gradients of 0-200 mM NaCl with a flow rate of 0.5 ml / min. The total volume of the gradient was 20 ml. The collagenase activity was determined using one of E. Wünsch et al. (Hoppe-Seylers Z. Physiol. Chem., 333, 149-151 [1963]) test described. The most active faction had a protein content of 0.8 mg / ml and a specific activity of 2500 units / mg. The ones described below Cleavage tests were carried out with this fraction, which at -20 ° C.  

3.2. Cleavage of HS 4133/1

The single-chain protein has a length of 477 amino acids (molecular weight 49,919). It contains those that are cleavable by collagenases sequence

-Gly- (Pro-Y-Gly) ₂-Pro

where Y stands for His and once for Ala (see Table 1). Cleavage of both Y-Gly bonds creates two proteins with lengths of 447 amino acids (MW. 46 880) and 27 amino acids (MG. 2832) and the tripeptide Gly-Pro-Ala. About responsiveness of the different phosphatase derivatives compared to collagenase Education became able to compare with each other of the largest fission product depending on the reaction time and the amount of enzyme measured.

A reaction mixture containing 53 µg HS 4133/1 and 0.1 unit Clostridiopeptidase A in 100 µl reaction buffer (90 mM Tris HCl, 5 mM CaCl₂, 120 mM NaCl, pH 8.0) was 2.5 hours at 37 ° C. incubated. At intervals of 15 to 30 minutes, samples of Take 10 µl each and add 3 µl sample buffer (10% Sucrose, 10% sodium dodecyl sulfate, 1mM dithiothreitol, 2mM EDTA) and stopped immediately by heating to 95 ° C for five minutes.

The samples were separated by electrophoresis in an SDS-polyacrylamide gradient gel (gradient from 12.5% to 25% acrylamide with a constant 0.33% bisacrylamide). After staining with Coomassie Brilliant Blue G, the gel was cut into strips in accordance with the protein webs and quantitatively evaluated at 280 nm using a gel scanner (Isco, optical unit type 6, absorption monitor UA-5). With the help of a corresponding calibration curve, which was obtained with a purified phosphatase protein, the protein contents for the bands of the largest cleavage product were determined from the peak heights. The results are summarized in Fig. 6. The data gives a product formation rate of 21.0 pmol / min / enzyme unit for HS 4133/1.

Fig. 5 shows in lane 7 a gel electrophoretic separation of the largest cleavage product from a sample of HS 4133/1 after complete reaction with clostridiopeptidase A.

3.3. Splitting of SA 506/1

The single-chain protein has a length of 486 amino acids (MG. 50 596) and contains the sequence cleavable by collagenase

where Y stands for His and four times Ala (see Table 1). The same proteins are formed when all Y-Gly bonds are cleaved as from HS 4133/1 (see 3.2.) and 4 moles of Gly-Pro-Ala each Mole of substrate. Determining the reactivity of SA 506/1 towards clostridiopeptidase A was done in the same way as with HS 4133/1.

The reaction mixture contained 53 µg SA 506/1 in 100 µl and had the same enzyme and buffer concentrations as the mixture for HS 4133/1. It was incubated at 37 ° C for 80 min. 10 µl samples were taken at intervals of 10 to 15 minutes and, as in 3.2. described, stopped and analyzed after gel electrophoresis with a gel scanner. The results are shown in Fig. 6. The product formation rate was 60 pmol / min / enzyme unit.

Fig. 5 shows in lane 5 the largest reaction product from SA 506/1 after quantitative conversion with clostridiopeptidase A.

3.4. Splitting of SA 302/1

The single chain protein is 477 amino acids in length contains the sequence that can be cleaved by collagenase

(see Tab. 1). Two fission products with lengths of 301 amino acids (MW. 31 269) and 176 amino acids (MG. 18 668). Because of the low reactivity of SA 302/1 only one approximate value for the product formation rate determined.

A reaction mixture which contained 5 μg SA 302/1 and 1.0 unit of clostridiopeptidase A in 15 μl reaction buffer (as in 3.2.) Was incubated at 37 ° C. for 15 h and then, as in 3.2. described, stopped by adding sample buffer and heating and analyzed by gel electrophoresis and evaluation with a gel scanner. About 0.02 pmol / min / enzyme unit of the larger reaction product had been formed. Fig. 5 shows the corresponding gel electrophoretic separation in lane 9 .

Table 1

Amino acid sequences of the model and fusion proteins derived from alkaline phosphatase (AP)

Table 2

Product formation rates for the implementation of the modified pre-phosphatases (Tab. 1) and the pre-phosphatase / ATCH fusion proteins (see file number P 37 31 875.6) with clostridiopeptidase A

Claims (12)

1. Fusion protein of the general formula I H₂N-Z₁-X- (Pro-Y-Gly) _n -Pro-Z₂-COOH (I) in the
n is 2,
X and Y represent each of the 20 amino acids defined by the genetic code,
Z₁ is a bacterial amino acid sequence and
Z₂ mean the target peptide from any amino acids of the genetic code. 2. Fusion protein according to claim 1, characterized in that n is 2 to 10 and X represents Gly. 3. Fusion protein according to claim 1 or 2, characterized in that that Z₁ is a sequence of 444 amino acids in the range of the amino acid sequence 1-444 of the prephosphatase from E. coli K 12.   4. Fusion protein according to claims 1-3, characterized in that Z₂ is an amino acid sequence of human ACTH. 5. Process for the preparation of fusion proteins of the formula I, characterized in that one for proteins of formula I coding gene structure expressed in a bacterial host cell and obtain the fusion protein via separation processes. 6. The method according to claim 5, characterized in that one uses E. coli as the bacterial host cell. 7. Gene structures coding for fusion proteins of the formula I. 8. Plasmids or comparable vectors, the codons for repetitive Contain collagenase interfaces and thus for production fusion proteins according to claim 1 are suitable. 9. Plasmid pSA506. 10. Plasmid pHS4133. 11. A process for the production of a eukaryotic protein, characterized in that the C-terminal enzymatically from a fusion protein of the formula I to the amino acid sequence -X- (Pro-Y-Gly) _n -Pro-, in which X and Y are each genetically encodable amino acid and n 2 mean the following protein sequence is split off. 12. A method for producing a eukaryotic protein according to claim 11, characterized in that the Y-Gly bonds in the amino acid sequence -X- (Pro-Y-Gly) _n -Pro are selectively cleaved with a collagenase and then the N-terminal Gly-Pro residue cleaved with a postproline dipeptidylaminopeptidase.