The present invention relates to a method of preparing protein and polypeptide products with high purity through recombinant DNA technology, and more particularly to the utilization of such technology to prepare novel gene products comprising desired proteins or polypeptides, which novel gene products are easily refinable, and optionally converting such gene products into the desired proteins or polypeptides. The invention also relates to such novel gene products.

The relatively new recombinant DNA technology or so-called genetic engineering, whereby novel recombinant DNA structures may be constructed from DNA segments derived from different biological sources and introduced into a prokaryotic or eukaryotic host cell to produce the corresponding protein or polypeptide, has made it possible to produce a great number of proteins which can otherwise only be obtained from natural sources and at considerable costs. Well-known examples are insulin and the growth hormone somatostatin. The proteins produced by the host cells are either trapped within the cells or secreted into the surrounding growth medium. In the former case the cells must be ruptured to permit the desired protein to be isolated, whereas in the latter case it can be separated from the growth medium. Even in case of secreted proteins, however, the preparation from which the protein is to be isolated is relatively complex containing a variety of other substances, and despite efficient separation techniques both the purity and yield of the desired protein may be low.

The present invention provides a solution to the above mentioned problem through a method based upon recombinant DNA technology which permits desired proteins and polypeptides to be produced with extreme purity. According to the invention this is achieved by utilizing the unique binding properties of protein A from Staphyiococci in combination with gene fusion technology as will be explained below.

Protein A is known as a cell wall component of the bacterium Staphylo- coccus aureus, hereinafter called S. aureus, and it is characterized by a specific serological reaction with mammal immunoglobulins. In contrast to the normal antigen-antibody reactions, however, protein A binds to the Fc-portion of all subclasses of human immunoglobulins type G, or IgG, except IgG _j , leaving the Fab-portioη thereof free, for antigen and hapten coupling. This property has given protein A a widespread use in both quantitative and qualitative immuno- chemical techniques; Covaiently bound to a carrier protein A is thus an excellent immunosorbeπt for the isolation of IgG. The exact structure of protein A may

vary depending on its origin. It has a molecular weight of about *2,000 and a markedly extended shape. The N-terminal part of the molecule comprises four or five highly homologous IgG-binding units, while the C-terminal part lacks Fc- binding ability. As used in the following description and claims, the term "protein A" is, however, not restricted to the above defined staphyloccal protein, but means any macromolecule having analogous immunologies! and biological activities to the protein A produced by staphyiococci, such as the natural strains of S. aureus, including any mutants thereof. Similarly "active fragments of protein A ^H or "active derivatives of protein A" are meant to comprise any polypeptide fragments or derivatives, respectively, of protein A as well as oligomeric forms of immunoreactive macromolecules or active fragments thereof or other macromolecules that are capable of binding to the constant regions of at least one immunogiobulin.

In our International patent application PCT/SE83/00297 (Swedish patent application No. 820*810-9), the disclosures of which are incorporated herein by reference, the isolation and characterization of the gene coding for staphylo- coccai protein A as well as the expression thereof in Escherichia coli, hereinafter called E. coli, are described. An E. coli strain transformed with a piasmid containing this staphylococcal protein A gene has been deposited on July 12, 1982 * ⁿ * ^e Deutsche Sammlung von Mikroorganismen (DSM), Gδttingen, Federal Republic of Germany, under No. DSM 2^.3'f. By using, e.g., this protein A gene containing piasmid, named pSPAl, protein A genetic material for the purposes of the present invention may be obtained.

It is assumed that the specific terms relating to gene technology, which will be used in the following description and claims, are well-known and accepted in the art. Definitions of a selection thereof may, however, be found in, for example, the above mentioned International patent application PCT/SE83/00297.

Gene fusion is a procedure wherein the coding sequence of two or more genes are spliced together to form a combined gene which on expression in a suitable host organism will produce a fusion product wherein the separate proteins or polypeptides coded for by the respective genes are fused together into a single molecule. The gene fusion technique is of growing importance and has so far been used to study various biological problems, such as protein transport mechanisms, piasmid replication and gene expression. Extensive use in this respect has been made of especially the E. coli lac Z gene coding for the enzyme β-galactosidase.

In accordance with the present invention gene fusion is used to combine a first DNA-sequence coding for protein A or an active polypeptide fragment thereof with a second DNA-sequence coding for a desired protein or polypeptide

into a functional gene capable of expressing the fusion product of said desired protein or polypeptide and the protein A component. Due to the IgG-binding ability of the protein A part, the produced protein or polypeptide can easily be isolated with high efficiency by conventional affinity chromatography utilizing immunoglobulin of type IgG immobilized to a suitable carrier. The carrier- bound fusion product may be used as such, e.g. if the desired protein is an enzyme, or it may be released from the carrier, either as a whole including the protein A part, or only the desired protein or polypeptide part thereof through cleavage with a suitable agent as will be further described below. A basic aspect of the present invention is thus the provision of a recombinant DNA cloning vehicle or vector comprising a DNA sequence coding foF a desired protein or polypeptide operatively linked to a DNA sequence coding for protein A or an active polypeptide fragment thereof, such that said DNA sequences together code for an IgG-binding fusion product of said desired protein or polypeptide and said protein A or active polypeptide fragment thereof. In order to be capable of transforming (which is also meant to include the case that the vector is a bacteriophage) a host organism to produce said fusion product, the vector in conventional manner further comprises a replicon and a promoter for the combined fusion product coding DNA sequence. For purposes which will be further elucidated below said combined DNA sequence may comprise a sequence coding for an appropriate cleavage site between the DNA sequences coding for the desired protein and protein A, respectively, such that the protein A part of the fusion molecule may be cleaved off as mentioned above. A more detailed description of the recombinant vector according to this aspect of the invention as well as the construction thereof will be described in more detail further on.

By transforming a compatible host organism with said vector to permit expression of the above combined DNA sequence and culturing the host in a nutrient medium the corresponding IgG-binding fused protein or polypeptide will be produced. Although bacterial hosts, such as strains of, for example, Escherichia, Bacillus and Staphylococcus, are preferred for the purposes of the invention, it is, of course, also within the scope thereof to use other hosts, such as yeasts and other fungi, plant cells in culture, etc. The transformation of the hosts may be effected with well-known methods.

Due to IgG-binding ability of the protein A moiety of the fusion molecule produced by the cultured host-organism the fusion molecule can be very effi¬ ciently isolated from the cell culture by means of IgG immobilized to a suitable carrier. If the fusion product is secreted into the surrounding medium the binding to the carrier may be performed directly from the medium. If, on the other hand,

the fusion product remains within the cells the latter have to be ruptured before such immunosorbance can be effected. Rupture of the cell walls may be effected in conventional manner by, e.g., high pressure, ultrasonication, homogenization, shaking with glass-beads etc. In cases where the product is trapped within the periplasmic space between two cell membranes, as in gram-negative bacteria, an osmotic shock procedure may be used to release the product into the suspension medium. Any other treatment of the cultured ceils or the growth medium prior to the IgG-aided isolation of the fusion product is, of course, also within the scope of the invention. In conventional manner the immobilization process may be performed batch-wise with the IgG-coupled carrier slurried in a suitable medium, or on a column of the activated carrier. IgG-coupled carriers for chromatographic use, e.g. IgG-Sepharose^ (Pharmacia AB, Sweden) are commercially available and may advantageously be used for the purposes of the invention. However, any conventional carrier material to which IgG can be sufficiently coupled for the present purposes may be used. The methods for coupling or immobilizing IgG to such carrier materials is well-known and need not be described in any detail herein.

Release or desorption of the fused protein or polypeptide which is bound to the IgG-carrier may be effected by conventional methods, such as lowering the pH, e.g. by means of glycine buffer (pH 3.0), treatment with high salt concentrations or chaotrophic ions, or by competitive elution using excess soluble protein A or IgG or fragments thereof to displace the fusion protein or polypeptide from the IgG-carrier adsorbent. The choice of desorption method should, of course, be made with regard to the particular desired protein or polypeptide, such that a desired activity thereof is not lost or considerably reduced thereby. From the resulting eluate the fusion protein or polypeptide may readily be isolated and, if desired, subjected to further purification steps, such as gel filtration, ion exchange etc. The purified fusion protein or polypeptide obtained may in itself be a valuable product as will be described below, and another aspect of the present invention is therefore the provision of a method of producing a highly purified fused protein or polypeptide product comprising the steps of transforming a compatible host with the above vector, culturing said host in a nutrient medium, isolating said fused protein or polypeptide from said host culture by selective binding thereof to an IgG-supporting carrier, and optionally releasing the fused protein or polypeptide from the carrier, as well as such an isolated fused product obtained thereby.

One valuable use of such a fusion product is when the protein fused to

the protein A part is an enzyme. In such cases the IgG-binding activity of the fusion product is utilized for immobilizing the enzyme to a carrier material having IgG coupled thereto. Such an enzyme system offers several advantages. Since the enzyme is bound to the carrier via the protein A-IgG coupling, all the enzyme molecules will be bound to the carrier in exactly the same way and maximum activity thereof will thus be obtained. When the enzyme activity has decreased to an unacceptably low level, such a system can easily be regenerated by conventionally desorbing the enzyme from the carrier through a pH change, e.g. glycine buffer treatment, and then binding fusion product containing active enzyme thereto. The binding or adsorption of the fused enzyme in question to the IgG-coupled carrier may be effected either directly from the.appropriately pre- treated cells or cell medium, or in purified state after adsorption and desorption ■ from another IgG-coupled adsorbent.

Immobilization of enzymes as above may be applied to such enzyme systems which are already used industrially as well as to enzyme systems not yet commercialized, as long as the DNA fragment coding for the selected enzyme is available. As examples of such enzyme systems may be mentioned amino-acid acylase, giucose-isomerase, penicili_n-amidase,.aspartase, fumarase, β-galactosi¬ dase, alkaline phosphatase, etc. Another case when the IgG-binding ability of the fused protein or polypeptide is desirable is, for example, for providing certain protein A conjugates which may be used in the well-known variant of immuno-chemical analysis named ELISA (enzyme linked immunosorbent assay). Two examples of such conjugates, which are frequently used and also are commercially available, are β-galactosidase and alkaline phosphatase. When these conjugates are prepared in conventional manner, i.e. by chemically binding protein A to the respective enzyme, only part of the two components will be correctly bonded to each other, the resulting conjugate mixture thus containing a relatively high proportion of inactive or poorly active conjugates. In contrast thereto, the corresponding conjugates prepared according to the present invention in the form of a fused gene product will always have the correct coupling relationship between protein A and the enzyme and consequently always maximum and de inable activity.

Still another case when the fused protein or polypeptide obtained may be used is when the protein A residue thereof does not inactivate or otherwise prevent the intended use of the desired protein or polypeptide fused to the protein A part. In such a case the usion product may well be used instead of the respective pure protein or polypeptide, and it will thus not be necessary to cleave off the protein A part therefrom as will be described below as a further aspect of

the Invention.

The protein part of the fused protein or polypeptide may under certain conditions be cleaved off, the pure desired protein or polypeptide thereby being obtained. In another aspect the present invention therefore provides a method of producing a desired protein or polypeptide of high purity comprising the steps of transforming a compatible host with the above mentioned vector, culturing said host in a nutrient medium, isolating said fused protein or polypeptide from the cell culture by selective binding to an IgG-supporting carrier, and cleaving off the desired protein or polypeptide from the protein A part of said fused protein or polypeptide, either directly from the carrier bound fusion product or after desorption thereof from the carrier.

A necessary condition to permit such cleavage of the fused protein or poly¬ peptide is, of course, that it contains a unique cleavage site which may be recog¬ nized and cleaved by suitable means. Such a cleavage site may be a unique amino- acid sequence recognizable by chemical or enzymatic means and located between - the desired protein or polypeptide and protein A sections, respectively, of the fused product to be produced. Such a specific a ino acid sequence must not occur within the desired protein or polypeptide and preferably not in the protein A part of the fusion product. Examples of enzymatic agents include proteases, such as collagenase, which in some cases recognizes the amino acid sequence NH ₂ -Pro-X-Gly-Pro— COOH, wherein X is an arbitrary amino acid residue, e.g. . leucine; chymosin (rennin), which cleaves the Met-Phe bond; kallikrein B, which cleaves on the carboxyl side of Arg in X-Phe-Arg-Y; enterokinase, which recognizes the sequence X-(Asp) -Lys— Y, wherein n=2— "., and cleaves it on the carboxyl side of Lys; thrombin which cleaves at specific arginyl bonds. Examples of chemical agents include cyanogen bromide (CNBr), which cleaves after Met; hydroxylamine, which cleaves the Asn— Z bond, wherein Z may be Gly, Leu or Ala; formic acid, which in high concentration ( 70%) specifically cleaves Asp-Pro. Thus, if the desired protein or polypeptide does not contain any methionine sequences, which is the case for, e.g., the hormone somatostain, the cleavage site may be a methinonine group which can be selectively cleaved by cyanogen bromide. Often it may be preferred to use chemical cleaving agents because protease recognition sequences may be stericaliy hindered in the produced fused protein. The techniques or introducing the corresponding DNA sequences coding for such cleavage susceptible peptide units or residues into the DNA sequence coding for the fused protein or polypeptide are well-known per se in the art and need not be discussed in any detail herein. In case a specific cleavage sequence which does . not occur in the desired protein, occurs in the protein A molecule,

this amino acid sequence may, without changing the activity of the protein A part, be converted into sequences which are not recognized and cleaved by the particular cleavage means by methods known per se in the art.

As mentioned above the cleavage may be effected either with the fusion product bound to the IgG-coupled carrier or after desorption therefrom. A batch-wise procedure may be carried out as follows. The carrier, e.g. IgG- Sepharos^ (Pharmacia AB, Sweden) having the fusion protein or polypeptide bound thereto is washed with a suitable medium and then incubated with the cleaving agent, such as protease or cyanogen bromide. After separation ^Ω of the carrier material having the protein A residue bound thereto, a solution containing the desired protein or polypeptide and the cleavage agent is obtained, from which the former may be isolated and optionally further purified by conventional techniques such as gel filtration, ion-exchange etc.

In case of the carrier being in column form and the fusion protein or polypeptide comprising a protease recognition site, the cleavage procedure may be performed in the following way. The column of carrier having the fusion protein or polypeptide bound thereto is washed with a suitable medium, and then eluted with an appropriate agent which is gentle to the desired protein or polypeptide to be produced as mentioned hereinbefore. Such an agent may, ⁰ depending on the particular protein or polypeptide, be a pH-lowering agent, such as, e.g. a glycine buffer, or a protein A solution (competitive elution). The eluate, containing the pure fusion protein or polypeptide together with the cleavage agent, is then passed through a second column comprising the immobilized protease, e.g. collagenase when the cleavage site is a collagenase ^ susceptible sequence. When passing therethrough the fusion protein or poly¬ peptide is cleaved into the desired protein or polypeptide and a protein A residue. The resulting solution is then passed through the same or another IgG-coupled column than used above, whereby the protein A component of the' solution is adsorbed and the resulting through-flow is a pure solution of the desired protein 0 or polypeptide. When the desorption agent is a protein A solution, the protein A adsorbed in the last step may be eluted and recycled to cleavage solution, the system then being a regenerative system with respect to protein A.

By means of the above aspect of the present invention a desired protein or polypeptide may readily be obtained with extreme purity and high yields using 5 only a small number of process steps. Such highly purified forms of desired proteins and polypeptides are, for example, excellently suited for the production of anti-bodies through immunisation of an animal, such as rabbit. Another possible application is the combination thereof with the so-called hybridoma technique for the production of monoclonal antibodies.

Desired proteins and polypeptides which can be produced with extreme purity in accordance with the present invention aspect are, or example, enzymes, such as various oxido-reductases, transferases, hydrolases, lyases, isomerases and ligasesj hormones, such as parathyroid hormones, growth hormone, gonadotropins (FSH, iuteinizing hormone, chorionogonadotropin and glycoproteins), insulin, ACTH, somatostatin, prolactin, placenta! lactogen, melanocyte stimulating hormone, thyrotropin, parathyroid hormone, calcitonin, enkephalin and angiotensin; and other proteins, such as serum proteins, fibrinogen, fibronectin, prothrombin, thromboplastin, globulin, e.g. gamma- globulins or anti-bodies, heparin, coagulation factors, complement factors, plasma proteins, oxytocin, albumins, actin, myosin, hemoglobin, ferritin, cyto- chrome, myogiobin, lactoglobulin, histones, avidin, thyroglobulin, interferon, transcorticai kinins, etc., and as suggested above peptide antigens for use in making vaccines. As appears from the above a crucial part of the present invention is the provision of the recombinant DNA structure or vector comprising the combined gene coding for the present fusion protein or polypeptide and capable of transforming a host cell to permit expression thereof and production of the fusion product. The present invention is meant to encompass any such vector irrespective of how it has been obtained using, for example, various restriction enzyme cutting, ligating, transforming and screening techniques well-known in the art as well as any appropriate vector materials and host-organisms. Thus, the DNA sequence coding for the desired protein or polypeptide may be inserted into a suitable vector and the protein A coding DNA sequence inserted subsequently, or vice versa; or the two DNA sequences may be introduced simultaneously into the vector. It is also possible to insert the respective DNA sequences in parts thereof into the vector. Further the two DNA sequences may be arranged with either the protein A coding sequence or the sequence coding for the desired protein or polypeptide at the .F-end or start of the combined gene. The special techniques for accomplishing such insertions and combinations with maintained correct reading frames, including the provision of suitable restriction sites therefore, are well-known per se in the art.

The source of the DNA sequence coding for protein A or an active polypeptide thereof may be any structure from which the corresponding gene or DNA-segment may be obtained such as a protein A gene containing piasmid. A suitable source is, for example, one of the protein A gene containing piasmids pSPAl, pSPA3 and pSPA5 as constructed and described in our International patent appli _¬ cation PCT/5E83/00297 (Swedish patent application 820f810-9). From such a vector the

whole or an appropriate part of the protein A gene may be cut out with restric¬ tion enzymes xre-ponding to suitably located restriction sites in or dose to the gene. The extent of the protein A gene to be included in the combined gene coding for the fusion product should be sufficient for imparting IgG-binding activity to the fusion product, which usually means at least a major part of the gene segment coding for the IgG-binding part of the protein A molecule. As will appear from the experimental part below, it may, however, at least in certain cases, be favourable to also include at least part of the gene segment coding for the non- IgG-binding part of the molecule. This part of the molecule will then serve as a spacer between the IgG-active part and the desired protein part of the fusion product.

The source of the DNA sequence coding for the desired protein or polypeptide, which may be of prokaryotic as well as eukaryotic origin, may likewise be any structure from which a corresponding gene or gene segment may be obtained. A suitable source is a piasmid containing such a gene or gene segment. From such a piasmid an appropriate part of the gene in question, i.e. that codes for a sufficient part of the corresponding protein or polypeptide to have the desired activity, may be cut out with suitable restriction enzymes corresponding to suitably located restriction sites in or close to the gene. The origin of the vector part of the recombinant vector of the invention is preferably a piasmid but it may also be f viral or phage origin. The particular choice of vector depends on the host-organism to be transformed. As mentioned above the latter may .be selected from bacteria, fungi, plants and algae. The preferred host is, however, bacteria, and bacteria susceptible to transformation comprise, e.g. members of Enterobacteriaceae, such as strains of E. coli and Salmonella, Bacillaceae, such as Bacillus subtilis, Pneumococcus, Streptococcus, Staphylococcus, Micrococcus and Hemophilus.

When constructing a recombinant DNA vector of the invention it is preferable to first construct an expression or fusion vector comprising a functional DNA sequence coding for protein A or an active polypeptide fragment thereof, and at least one unique restriction site at or near the end of the protein A coding gene. Such a fusion vector may be constructed by providing a suitable vector, e.g. a piasmid vector, containing the whole or a sufficient part of the protein A coding gene as discussed above. A unique restriction site, or preferably a multilinker containing several different restriction sites, is then inserted into the protein A gene after the IgG-binding region but before the stop codon. Such an insertion may, as will appear from the following experimental part, be effected in several steps and plasmids. The resulting fusion vector may then be used for insertion of any DNA sequence coding for a desired protein or

polypeptide. Such a fusion vector is also part of the present invention.

To insert a DNA sequence coding for a desired protein or polypeptide into the fusion vector, it is preferably provided as part of a piasmid. If suitable restriction sites giving complementary ends to those obtained when cutting the fusion vector in one of the unique restriction sites are not present in the gene, such sites may be inserted by conventional methods. They should be inserted as far upstream as possible, i.e. near the 5 ^, -end, in the gene or before the start codon provided that there is no stop codon in-between. By cleaving the fusion vector and the desired gene containing piasmid at the appropriate restriction sites and ligating the mixture, a recombinant vector containing the combined gene may be obtained. Although it may be preferred to cut out the gene coding for the desired protein and insert it into the piasmid, it is also possible to cut the piasmid only at the start of the gene and combine the two plasmids. The above is, however, only a rough example and many variations are possible. To provide a recombinant vector coding or a fusion protein or polypep¬ tide from which the protein A part may be cleaved off, a synthetic sequence coding for an oiigopeptide, which can be recognized by a protease or a peptide cleaving chemical agent, may be inserted between the .two fused genes or gene segments. Such insertions may be performed with conventional methods. A proviso is, of course, that the fusion protein or polypeptide, or at least not the desired protein, does not contain other peptide sequences that may be cleaved by. the protease or chemical agent.

For the case that a gram-positive bacterium, such as Bacillus subtilis or any staphylococcal species, is to be transformed by the recombinant vector, the control regions of the protein A gene (promoter and ribosome bindning sequence) may advantageously be used. For gram-negative bacteria, such as E. coli, it may be preferable to insert a control region of such origin, e.g. from the E. coli phage lambda.

Apart from the fact that the control regions of the protein A gene function well in gram-positive bacteria, these hosts having a single membrane are favourable from another point of view, viz. that the signal peptide, coded for by the signal sequence of the protein A gene, may serve to secrete the fusion protein or polypeptide into the surrounding medium. In a gram-negative bacterium, such as E. coli, the fusion product will be trapped between the two cell-membranes. Secretion of the product offers great advantages in that the cells need not be ruptured for recovery of the product but can readily be separated and the fusion protein or polypeptide be adsorbed directly from the medium.

As is well-known foreign proteins produced by recombinant DNA

techniques in E. coll may be subjected to proteolytic degradation of the product. Such degradation may be minimized by using a temperature-sensitive represser which is inactivated at higher temperatures. This permits the gene to be switched off when culturing the bacteria and to be switched on just before the cells are harvested. A DNA sequence coding for such a temperature-sensitive represser may be introduced into the recombinant vector of the invention by conventional methods.

The invention will now, by way of illustration only, be described in more detail in the following non-limiting examples, reference being made to the accompanying drawings. The disclosures of all the patent and literature references mentioned hereinafter are incorporated by reference herein. In the drawings:

Fig. 1 is a schematic illustration of a circular restriction map of a piasmid DNA (pSPAl) coding for protein A. The size of the map is given in kilobases starting at the Eco RI restriction site at 12 o'clock, which is a restriction site within the vector pBR322. The positions of the Eco RI, Eco RV, Hind III, Pst I and Bam HI restriction sites are indicated. The junctions between the vector and the inserted DNA are indicated with arrows.

Fig. 2A is a schematic illustration of the protein A coding gene indicating its different regions. Heavy line represents the DNA of the vector pBR322. S is a signal sequence, A-D are IgG-binding regions previously identified, E is a region nearly homologous to A-D, and X is the C-terminai part of protein A which lacks IgG-binding activity.

Fig. 2B is a detailed restriction map of the DNA-sequence corresponding to Fig. 2A, and showing the restriction sites for Tag I, Hind III, Eco RV, Pst I,

Bel I and Sau 3A. The size is given in kilobases starting at the same Eco RI restriction site as indicated in Fig. 1. The junction between the vector pBR322 and the inserted DNA fragment is indicated with an arrow. The restriction sites or Tag I (two) and Sau 3A (one) within the vector sequences have been omitted. Fig. 3A shows the base sequence around the Sau 3A restriction site at position 1,8 kb in Fig. 2B and Fig. 3B around the Pst restriction site at position 2,1 kb. The amino acid sequence as deduced from the DNA sequence is also shown (the

IUPAC amino acid abbreviations are used, 3. . Biol. Chem. 2_4, 527 and 2*91

(1966). Fig. is a schematic illustration of the constructions of plasmids containing the whole or parts of the protein A gene. A few restriction sites are shown. Boxes represent structural genes and the arrows indicate the orientation

(from start codon towards stop codon). The replication origin is also indicated by

Ori. AMP and TET are the genes coding for ampicillin and tetracycline

resistance, respectively. PROT A is the gene coding for protein A and lac Z* is the gene coding for the N-terminal part of β-galactosidase. (ROther et al, Nucl. Acids Res. 9, 4087-4098 (1981)).

Fig. 5A and 5B are schematic map illustrations of two fusion vector plasmids. The abbreviations are the same as in Fig. Ψ. An M13 multiiinker has been inserted in the gene coding for protein A at different positions in the two plasmids. The nucleotide sequence and the deduced amino acid sequence of these regions are also shown above the respective piasmid map. Fig. 5A illustrates piasmid pSPAll and Fig. 5B illustrates piasmid pSPA12. Fig. 6A and 6B are schematic illustrations of two plasmids containing the genes coding for protein A and β-galactosidase fused together. The abbreviations are the same as in Fig. 4. LAC Z represents the whole gene coding for β- galactosidase except a few nucleotides in the 5 ^, -end thereof. Fig. 6A illustrates piasmid pSPA13 and Fig. 6B illustrates piasmid pSPAl*. Fig. 7 (C and D) is a schematic illustration of the fused protein A and β- galactosidase genes, of the plasmids in Fig. 6A and 6B. In the Figure A and B are schematic drawings of the protein A gene corresponding to the restriction map of Fig. 2A and 2B, respectively, and in alignment therewith. The sizes are given in base pairs starting at the Tag I site. (The two Bel sites at nucleotides 355 and 1572 are also Sau 3 A sites).

Fig. 8A is a presentation of the nucleotide sequence around the fusion point of piasmid pSPA13 (Fig. 6 A). The restriction sites and the corresponding deduced amino acid sequence are indicated. The origins of the different parts of the sequence are also indicated, and Fig. 8B is a corresponding presentation of the nucleotide sequence around the fusion points of piasmid pSPAl* (Fig. 6B).

Fig. 9 shows the nucleotide sequence of a synthetized oligonucleotide and the corresponding phage mp9/lGF-l sequence at the point of mutagenesis. "*" indicates a non-complementary base pair. The decuded amino acid sequence after mutagenesis is also shown. Fig. 10 is a schematic illustration of the construction of shuttle vectors containing the IFG-1 and the protein A genes. A few restriction sites are shown. Boxes represent structural genes and the arrows indicate the orientation (from start codon towards stop codon). The replication origins are also indicated by ORI-E (E. coli) and ORI-S (S. aureus). AMP, TET and CML are the genes for a picillin, tetracycline and chloramphenicol resistance, respectively. PROT A is the gene coding for protein A and IGF-1 is the gene coding for the modified

human insulin-like growth factor, type 1.

Fig. 11 is a presentation of the nucleotide and the deduced amino acid sequence of the gene fusion between protein A and IGF-i in piasmid pUN201.

Only the DNA-sequence coding for the mature protein A (lacking the region coding for the signal peptide) is shown. Possible amino acid cleavage points with formic acid treatment (Asp-Pro) are underlined. Cleavage sites for Eco RI and

Hind III are shown, said cleavage sites representing the end points of a synthetic modified IFG-1 gene.

Fig. 12 is a presentation of the nucleotide sequence, and the correspon- ding deduced amino acid sequence, around the 3'-end of the protein A gene in piasmid pUN202. ^'

Fig. 13 A is a presentation of two DNA-strands representing the IGF-1 gene and flanking sequences, divided into oligomers. The sequence has been provided with a start codon (ATG in block A2), a stop codon (TAG in block A 17) and recognition sequences for Eco RI (block Al) and Hind III (block A17). Fig 13B is a ligation pattern for the IGF-1 gene. In the following Examples the starting materials, buf ers, cell media and routine method steps were as ollows.

STARTING MATERIALS Bacterial hosts. Four strains of E. coli were used in the Examples: HBlOl, described by Boyer et al, J. Mol. Bioi. 41, 4 9-472 (1969); XAC lac, (Miller et al,

3. Mol. BioL, 109, 275-301 (1977); RR ^' I del 15 (Langey et al, Proc. Natl. Acad.

Sci., USA, 72, 1254-1257 (1975); UM 83 (Viera and Messig, Gene 19, 259-268

(1982)). Also S. aureus SA1I3, described by Iordanescu et al _y J. Gen. Microbiol. 96_: 277-281 (1976), was used. (The strains are available at the Department of

Microbiology (N), Biomedical Centre, Uppsala, Sweden).

Cloning vehicles. The cloning vehicles used in the Examples were pBR322 as constructed and described by Bolivar et al, Gene 2, 95-113 (1977); pUR 222 as constructed and described by Rϋther et al, Nucl. Acids Res., 9, 4087- ψ 098 (1981);

PTR262 as constructed and described by Roberts, T.M. et al, Gene Ϊ2, 123-127

(1980); pUC8 as constructed and described by Viera and Messig, Gene JL9, 259-268 (1 82); pHV14 as constructed and described by Ehrlich, S.D., Proc. Natl. Acad. Sci. USA 70 _j 3240-3244 (1978); pS S104 and pS S106 as constructed and described by Casadaban, M.J.,

Martinec-Arias, A., Shapiro, F., and Chou, J., Methods in Enzymology, 100, p.

293-308 (1983); plasmids pSPAl, ρSPA3, pSPA5 and pSPA16 containing the gene coding for

staphylococcal protein A as constructed and described in the International patent application PCT/SES3/00297 (Swedish patent application 8204810-9). Cultures of an E. coli 259 strain containing piasmid pSPAl and of an S. xylosus KL117 strain containing piasmid pSPA16 have been deposited with the Deutsche Sammlung von Mikroorganismen (DSM), Gδttingen, Federal Republic of Germany, under No. DSM 2434 on July 12, 1982 and No. DSM 2706 on August 15, 1983, respectively; Phage .vectors M13 mp8 and mp9 RFI DNA, supplied by New England Biolabs, Beverly, MA, USA (catalogue No. 408 and 409)..

BUFFERS AND MEDIA

Tris-EDTA buffer 0.001 M EDTA and 0.01 M Tris (pH 7.S) CTE"):

Coating buffer 1.59 g Na ₂ CO ₃ , 2.93 g NaHCO ₃ and 0.2 g NaN-, (carbonate- bicarbonate - pH 9.6) made up to 1 liter with distilled H-O

PBS TWEEN: 8.0 g NaCl, 0.2 g KH _j PO^, 2.9 g Na ₂ HPO^x 12 H ₂ O,

(Phosphate bufferedsaline plus 0.05% TWEENY 0.2 g KCi, 0.2 ml TWEEf^ 20 and 0.2 g NaN ₃ , made up to 1 liter with distilled H ₂ O; pH 7.4

Diethanolamine 97 ml diethanolamine, 800 ml distilled H ₂ O, 0.2 g buffer 10% ^* NaN ₃ , and 100 mg MgCl ₂ x 6H ₂ O; pH adjusted to 9.8 with 1 M HC1; made up to 1 liter with distilled H ₂ O

Luria-broth ("LB"): 10 g Difco tryptone, 5 g Difco yeast extract, 0.5 g NaCl, 2 ml 1M NaOH; adjusted to pH 7.0 with 1 M NaOH; 10 ml 20% glucose added after autoclaving.

LA-medium: Luria broth supplemented with 1% Difco agar TEB buffer 0,09 M TRIS-borate, 0,09 M boric acid and 0,002 M EDTA. ONPG buffer 2 mM o-nitrophenyl-β-D-galactoside (ONPG, Sigma product No. N-1127) in 0.1 M potassium phospate buffer, pH 7.3, containing 0,1 M 2-mercaptoethanol and 1 mM MgCl ₂ «

Xgal-medium LA-medium supplemented with 40 mg/1 of 5-bromo-4- chlbro-3-indolyi-β-D-galactoside. (Xgal).

AXI LA-medium supplemented with 50 mg/1 of ampicillin,

40 mg/1 of 5-bromo-4-chloro-3-indoyl-β-D-galactosi- de (Xgal) and 0.1 mM isopropyl-β-D-thiogalactosi- de (IPTG).

ROUTINE METHODS Certain procedures were carried out repeatedly in the Examples. Unless otherwise specified, they were done exactly as follows each time that they were carried out. Transformations. Transformation of E. coli 12 with piasmid DNA, was performed exactly as described by Morrison, D.A., Methods in Enzymology, Academic Press 6S, 326-331 (1979). The transformed cells were selected in a conventional manner on plates by plating for single colonies on LA plates containing suitable antibiotics, i.e. 35 μg/mi of ampicillin or 25 μg/ l of chloroamphenicol.

Isolation of plasmids. Large scale piasmid preparation was performed exactly as described by Tanaka, T. and Weisburn, B., J. Bacteriol. 121, 354-362 (1975). For scoring a large number of clones for plasmids the "mini alkali method" was used exactly as described by Birnboim, H.C. and Doly, 3., Nucl. Acids Res. 7, 1513- 1523 (1979).

Gel elution. DNA fragments were eluted from either polyacrylamide or agarose gel pieces exactly as described by Maxam et al, P.N.A.S. 74, 560-564 (1977). DNA sequencing. DNA fragments were 5 ^l -end labelled, and their DNA sequences were determined exactly as. described by Maxam et al, supra. The 5 ^, -end of

32 endonuclease generated DNA fragments was labelled with (tf - P) ATP (New

England Nuclear, USA; 2700 Ci/mmol) using T4 polynucleotide kinase (Boeh- ringer, Mannheim, West Germany).

Restriction enzyme digestion of DNA. DNA was cleaved with conventional restriction enzymes purchased from New England. Biolabs, Beverly MA, USA. The restriction enzymes were added to DNA at conventional concentrations and temperatures and with buffers as recommended by New England Biolabs.

Ligating DNA fragments. All DNA fragments were ligated at 14 C over-night with T4 DNA ligase purchased from New England Biolabs, Beverly MA, USA, in a buffer recommended by the supplier. Agarose gel electrophoresis . 0.7% agarose gel electrophoresis for separating cut piasmid fragments, supercoiled plasmids, and DNA fragments 1000 to 10,000 nucleotides in length was performed exactly as described by Helling et al, 3. Vir.

.14, 1235-1244 (1974).

Polyacrylamide gel electrophoresis. 8% polyacrylamide gel electrophoresis for the separation of DNA fragments 100 to 4000 nucleotides in length was performed exactly as described by Maxam et al, P.N.A.S. 74, 560-564 (1977).

Preparation of cell lysate for detection of protein A. E. coli clones were grown overnight at 37°C in 50 ml Luria-broth (LB) with ampicillin . added at 35 Mg/ml.

After centrifugation the cells were resuspended in 5 ml Tris-EDTA (0.05 M. pH

8.5, 0.05 M) and centrifuged. The cells were resuspended in 5 ml of the same buffer and lysozyme was added to a final concentration of 2 mg/ml. After 1 hour at 37°C the lysate was centrifuged in a Sorvail SS-34 rotor at 15.000 rpm for 15 minutes. The supernatant was collected and assayed for protein A. Detection and quantification of protein A from E. coli clones. An ELISA-test (enzyme linked immunosorbent assay) was used for detection and quantification of produced protein A. The test makes use of a special microtiter plate (Titertek, Amstelstad, the Netherlands) having no net charge (neutral), the wells of which are coated with human IgG (Kabi, Sweden). Test samples are then added to allow protein A to bind to the Fc-part of the IgG-molecules. The amount of remaining free Fc-sites is then titrated by adding alkaline phosphatase linked to protein A. After washing of the wells, p-nitrophenyl-phosphate is added as a substrate for alkaline phosphatase. Assay: The wells of a microtiter plate were filled with 50 μl of a solution of human IgG (Kabi, Sweden) at 500 μg/ml in a coating buffer and the plate was incubated at room temperature for l ⁴ hour. The wells were then washed three times with PBS +0.05% Tween® 20, which was used in all washes in the assay, and 50 μl of the lysate to be tested was added. For quantitative determinations twofold serial dilutions of of PBS + 0.1% Tween*^ 20 was then added and incubation was allowed for 1 hour at room temperature. The wells were again washed three times, and 50 μl of protein A-aikaline phosphatase conjugate (prepared exactly as described in Immunochemistry, Pergamcn Press 1969, Vol. 6 pp. 43-52) was added. After 1 hour of incubation at room temperature the wells were again washed three times and 100 μl of alkaline phosphatase substrate (Sigma 104 = p-nitrophenyl- phosphate at 1 mg/ml) was added. The enzyme reaction was interrupted after 30 minutes by the addition of 10 μl of 3 M NaOH. The result was determined visually. A positive result, i.e. presence of protein A, is a colour-less reaction mixture, since no free Fc-sites of IgG are available to bind the conjugate. A negative result, i.e. no protein A, is observed as a yellow colour due to the activity of the alkaline phosphatase of the bound conjugate. Quantitative determinations of protein A were made by running serial twofold dilutions of a protein A standard solution of known concentration in parallell with the test samples. β-galactosidase assay

Recombinants containing a functional lac Z gene were scored by plating on Xgal medium. Cell free β-galactosidase activity was assayed by a colorimetric procedure using o-nitrophenyl-β-D-galactoside (ONPG, Sigma product No. N- 1127) as substrate as described by Miller, 3. H., Experiments in Molecular

Genetics, Cold Spring Harbor, New York; Cold Spring Harbor Laboratory, 1 7?, with the following modifications. The assay was performed at +8 C and the activity was measured at 405 nm. One unit of activity represents the change in absorbance at 405 nm per minute, β-galactosidase activities of the fused proteins coupled to l _g G-5epharose®were determined at 8°C, inverting the tube in order to prevent sedimentations.

Phage M13 cloning and seguencing.

All M13 cloning, purifications and sequencing was performed as described in the Instructions/Protocol obtained by the supplier (New England Biolabs, Beverly, MA., USA, catalogue No. 408 and 409).

EXAMPLE I

I. Analysis of the DNA sequence of the protein A gene In order to make fusions between genes or gene parts it is desirable to know the DNA sequence and its deduced amino acid sequence around the fusion point of the two genes or gene parts to be fused. With the knowledge thereof it will be possible to predict how the linkage should be designed in order to give the correct reading frame in both genes or gene parts and thus possibly the expression of a functional hybrid protein.

In our Swedish patent application No. 8204810-9 (the disclosure of • which is incorporated herein . by reference) the construction of three plasmids containing the whole structural gene coding for staphylococcal

protein A is described, viz. plasmids pSPAl, pSPA3 and pSPA5. However, only the DNA sequence of the 5*-end of the protein A gene (regions S, E, D and part of A in present Fig. 2A) is disclosed. A preliminary sequence of the whole protein A gene was therefore determined in order to obtain more detailed information of the DNA sequence in the 3'-end of the gene. This sequencing operation was effected as described above under Routine Methods, the DNA source being purified piasmid pSPA3, which is the smallest of the three protein A gene containing plasmids and therefore the least difficult one to sequence. In Fig. 3 the obtained DNA sequences around the Sau 3A restriction site at position 1,8 kb and around the Pst I restriction site at position 2,1 kb in the protein A gene restriction map of Fig. 2B are shown together with the corresponding deduced amino acid sequence . The particular interest in the above two restriction sites for the present purposes will be explained below.

Based upon the DNA sequence obtained it was decided to construct two different gene fusion vectors by inserting an M 13 multiiinker (an oligonucleotide containing restriction sites for several restriction enzymes) into the above mentioned Sau 3A site at nucleotide 1096 of Fig. 3A and the Pst I site at nucleotide 1541 of Fig. 3B. These sites are located before and after the repetetive part of region X of the protein A gene (Fig. 2A) which is thought to be involved in the binding of protein A to the peptidoglycan of the cell wall in S. aureus. The possible influence thereof on the fused proteins to be produced on expression of genes fused by means of such fusion vectors could then be determined. The construction of the two gene fusion vectors is described hereinafter. II. Construction of fusion vector piasmid pSPAll (Fig. 5A)

In the following steps A-E the construction of a piasmid containing the protein A gene without region X and having a unique Eco RI site inserted at the Sau 3A site at position 1098 (Fig. 7) is described. A. Subcloning of the 5'-end of the protein A gene from pSPAl into piasmid pTR262 to obtain piasmid pSPA2 (Fig. 4)

1 μg of piasmid pSPAl (see Fig. 1) and 1 μg of piasmid pTR262 were cut with restriction enzymes Hind III and Pst I, mixed, treated with T4-ligase and used to transform E. coli HBlOl. Cleavage, ligation and transformation were effected as described under Routine Methods. Piasmid pTR262 contains a lambda repressor gene which on expression inactivates the gene for tetracyciine resistance. The lambda repressor gene has a Hind in site and insertion of a DNA . sequence into the latter therefore inactivates the lambda repressor gene and activates the tetracyciine resistance gene. Piasmid pTR262 thus permits positive selection for tetracyciine resistant

recombinants.

Colonies containing recombinants were thus selected as being tetra¬ cyciine resistant. 1 colony out of 20 of these recombinants was discovered to be protein A positive using the ELISA method described under Routine Methods. Restriction analysis indicated that it contained vector piasmid pTR262 having a 2.1 kb protein A gene insert derived from the fragment corresponding to 0.0 to 2.1 kb of the pSPAl restriction map of Fig. 1 and 2B. This piasmid was designated pSPA2 and is shown schematically in Fig. 4. It has a unique Pst I restriction site at the 3'-end of the protein A gene fragment which will be utilized in the following step E.

B. Preparation of a DNA fragment containing the protein A gene

100 μg of piasmid pSPA5(plasmid vector pHV14 having a protein A gene insert; see Starting Materials above) were cut with restriction enzyme Eco R V for 1 hr at 37°C. This produced two DNA fragments, viz. the inserted DNA fragment containing the protein A gene (2.1 kb) between positions 0.2 kb and 2.3 kb in Fig. 2B and the vector pHV14 (7.2 kb). This digest was heat inactivated, precipitated with ethanol, dissolved in 100 μl of TE and sedimented through a 10-30% sucrose gradient in TE buffer. A Beckman SW40 rotor was used at 5°C, 35,000 rpm, for 20 hrs. The gradient was fractionated into 0.5 ml fractions, each of which was analyzed by agarose gel electrophoresis. The fractions containing the 2.1 kb fragment were pooled, precipitated with 2 volumes of ethanol and dissolved in TE buffer. As appears from Figs. 2A and B the fragment contains, in addition to the whole protein A gene, an E. coli sequence derived from piasmid pBR322 and a staphylococcal gene residue. C. Preparation of a DNA fragment containing » ^• part of the protein A gene

5 μg of the purified 2.1 kb fragment from step B were cut with restriction enzyme Sau 3A for 1 hr at 37 C. The digest was run on a preparative 8% polyacrylamide gel electrophoresis in TEB buffer. The gel was stained with ethidium bromide (1 μg/ml) and a DNA fragment of approximately 600 base pairs was cut out. This fragment corresponds to the part of the gene between positions 1.15 and 1.8 kb in Fig. 2B. The DNA was eluted overnight at 37°C in 5 ml of TE + 0.3 M NaCl. The eluate was passed over a column containing approximately 300 μl of sedimented DE-52 (Whatman, England) equilibrated with 5 ml of TE. After a 2 ml wash with TE + 0.3 M NaCl the DNA was eluted with two volumes of each 0.5 ml of TE + 0.6 M NaCl. The eluate containing the DNA fragment was diluted with one volume of TE, precipitated with ethanol and dissolved in TE buffer. The resulting purified protein A gene fragment has cohesive ends corresponding to a Sau 3A restriction site and an intermediate Hind HI site.

D. Preparation of vector piasmid pUR222

Piasmid pUR222 is a commercially avilable vector containing the gene coding for the enzyme β-galactosidase (lac Z). The gene comprises a multiiinker having several restriction sites, such as Pst I, Bam Hi and Eco RI. Since β- galactosidase is easily detectable by enzymatic assays, recombinants having a DNA fragment inserted in one of the restriction sites can easily be scored with the use of appropriate host strains. Often Xgal plates are used (Xgal is a chromogenic substrate, 5-bromo-4-chloro-3-indolyi-β-D-galactoside, which releases a blue indolyl derivative when cleaved by β-galactosidase) upon which β- galactosidase negative recombinants appear as white colonies in contrast to the blue-green colour of colonies containing plasmids without an insert.

To cleave piasmid pUR222 in the β-galactosidase coding gene to provide cohesive ends complementary to the cohesive ends of the protein A fragment of step C for insertion thereof into the piasmid, the Bam HI restriction site was used. 1 μg of pUR222, supplied by Boehringer-Mannheim, Germany, was digested with the restriction enzyme Bam HI for 1 hr at 37 C, whereupon the enzyme was inactivated at 65 C for 10 minutes. This cleavage preparation was used in the following step E for ligation with the protein A fragment.

E. Construction of a hybrid piasmid pSPAlO containing pSPA2 and pTR262 (Fig. 4)

200 ng of pUR222 digested with Bam HI. as described in step D, and 200 μg of eluted protein A fragment, as described in step B, were mixed and ligated in a total volume of 20 μl overnight at +1 C The enzyme was inactivated at 65 C for 10 minutes, precipitated with ethanol and dissolved in TE buffer. The whole DNA-mixture containing i.a. recombinant plasmids having the protein A insert in the β-galactosidase gene was cut with restriction enzymes Hind III and Pst I for 1 hr at 37°C in the recommended buffer for Hind in. This cleaves the recombinant piasmid in the β-galactosidase gene (Pst I) and in the protein A gene (Hind HI) producing two fragments, viz. a small fragment consisting of a minor β-galactosidase DNA sequence linked to the part of the protein A gene fragment from the Sau 3A site at position 1.15 kb to the Hind III site in Fig. 2B, and a large fragment consisting of the rest of the recombinant piasmid, which comprises the major part of the β-galactosidase gene linked to the protein A gene fragment from the Hind III site to the Sau 3A site at position 1.8 kb in Fig. 2B. As appears from Fig. 4 the β-galactosidase fragment has an Eco RI restriction site close to the point of fusion with the protein A fragment (the Bam HI site).

200 ng of piasmid pSPA2 from step A were cut with the restriction enzymes Hind m and Pst I in the same way as above to cleave the piasmid

V into (see Fig. 4) three fragments, viz. one fragment extending from the — H —in —d HI site located between the Tet-gene and the 5 ^, -end of the protein A gene to the Hind in site within the protein A gene, a protein A gene fragment extending from the latter Hind III site to the Pst I site at the S'-end of the protein A gene, and a larger fragment of pTR262 origin comprising the rest of the piasmid.

The two digests prepared above were inactivated at 65 C for 10 minutes, mixed and precipitated with ethanoL The DNA was dissolved in iigation buffer and treated with T4-ligase. The desired recombinant piasmid comprises the above mentioned large fragment, obtained on cleavage of the pUR222 recombinant, inserted in pSPA2 between the Hind HI site within the protein A gene and the Pst I site and com¬ prising the 5*-end of the protein A gene, one part thereof thus being derived from p5PA2 and the other originating from the pUR222 recombinant. Further, the piasmid is ampicillin and tetracyciine resistant and should give blue colour on Xgal plates as will be explained below. The ligated DNA-mixture was therefore used to transform E. coli RRI del M15. Cleavage, Iigation and transformation were effected as described above. Recombinants were plated out on Xgal plates containing ampicillin and tetracyciine. One of the clones appeared as light blue, and restriction analysis was performed on its piasmid. This revealed a piasmid, designated pSPAlO (Fig. 4), which consists of parts of piasmid pUR222, piasmid pTR262 and the protein A gene originating from piasmid pSPAl. In piasmid pSPAlO the protein A gene fragment is fused to the lac Z gene through its Sau 3A site at position 1096 as appears from Fig. 7.

Although piasmid pSPAlO does not contain the whole lac Z gene coding for β-galactosidase but only the gene coding for the α-fragment thereof (lac Z ), it is active in cleaving the Xgal substrate thereby producing blue colour under the above used conditions. This is due to a complementation between the o- fragment coded by the piasmid and a chromosomal gene product containing the carboxy terminal fragment of β-galactosidase resulting in an active enzyme. The E. coli RRI del M15 host strain used above has such chromosomal gene material and therefore complements the α-fragment produced by the pSPAlO piasmid to an active β- galactosidase molecule.

The above described steps A-E thus produced a piasmid vector pSPAlO containing a desired protein A fragment which has a unique Eco RI site adjacent to its downstream end. In order to construct a convenient fusion vector a DNA- linker containing multiple restriction sites was introduced at that site simul¬ taneously with removal of the non-desired fragment (containing the lac Z 'gene and the gene for ampicillin resistance) between the Eco RI and Pst I sites, as will be described in the following section HI.

F. Construction of fusion vector piasmid pSPAll containing a multi-linker from phage M 13 mp 8 (Fig. 5A)

1 μg of piasmid pSPAlO from step E and 2 μg of phage vector M13 mp 8 (supplied by New England Biolabs, Beverly, MA, USA) were cut separately with the restriction enzyme Eco RI, precipitated and dissolved in an appropriate buffer for the restriction enzyme Pst I. Digestion with the Pst I enzyme was performed and the two DNA digests were mixed, ligated and used to transform E. coli HBlOl. Cleavage, Iigation and transformation were effected as described above under Routine Methods. The desired recombinants were selected as being tetracyciine resistant and ampicillin sensitive. 52 tetracyciine resistant clones were picked onto plates containing ampicillin. 3 of these clones were found to be ampicillin sensitive and restriction analysis on one of them revealed the piasmid schematically shown in Fig. 5A. This piasmid, which contains an inserted M13 multi-linker at the end of region C of the protein A gene (position 1.8 kb, Fig. 2B), was designated pSPAll. The deduced amino acid sequence is also indicated providing a guide for obtaining the correct reading frame after gene fusions. Piasmid pSPAll is a vector well suited for gene fusions with the protein A fragment as will be demonstrated in the following step III. πi. Fusion of piasmid pSPAll to the E. coli lac Z ene 1 μg of piasmid pSKSI04 (containing a unique E. coli RI site) was digested with the restriction enzyme Eco RI. This piasmid, together with the piasmid pSKS106 used below, are examples of a set of plasmids constructed to aid gene fusions between the E. coli lac Z gene and other genes. The hybrid protein produced by such fusions contains, at its carboxyterminus, enzymatically active β-galactosidase (minus a few amino acids at the N-terminus) and can be assayed by the enzymatic activity thereof. 1 μg of piasmid pSPAll from step 2F (also having a unique Eco RI site) was digested separately with restriction enzyme Eco RI. Both DNA digests were heat inactivated, mixed, ligated and used to transform E. coli XAC lac (which strain lacks the β-galactosidase gene) as described above. Recombinants were scored on Xgal plates containing both tetracyciine and ampicillin. Approximately half of these clones were light blue (the cleaved pSKS104 can be inserted in correct or wrong direction with equal probability) and restriction analysis on one of these revealed the piasmid schematically represented in Fig. 6A. This piasmid, designated pSPA13, contains the lac Z gene fused to the protein A gene at the nucleotide of position 1.8 kb in Fig. 2B. This is schematically shown in Fig. 7 and the deduced amino acid sequence over the fusion point is shown in Fig* 8A. Cultures of this clone have been deposited on February 4, 1983 with the collection of the Deutsche Sammlung von Mikroorga¬ nismen (DSM), Gδttingen, Federal Republic of Germany, where it was assigned

No. DSM 2591.

IV Fusion of piasmid p5PA2 to the E. coli lac Z gene Piasmid p5PA2, as constructed in step IIA above and shown in Fig. 4, contains a unique Pst I site at position 1541 of the protein A gene (see Fig. 7). This piasmid was therefore used for gene fusion of the corresponding protein A gene fragment to the lac Z gene of piasmid pSKS106.

1 μg of piasmid pSPA9 from step 2C and 1 μg of piasmid pSKS106 were cut separately with restriction enzyme Pst I. The resulting DNA-fragments were mixed, ligated and used to transform E. coli XAC lac as described under Routine Methods. Recombinants were scored on Xgal plates containing tetracyciine and ampicillin. As in section ffl above approximately half of these clones were light blue and restriction analysis on one of these revealed a piasmid schematically represented in Fig. 6B. This piasmid, designated pSPA14, contains the lac Z ene fused to the protein A gene at position 2.1 kb in Fig. 2B. This is schematicaUy shown in Fig. 7, and the deduced amino acid sequence over the fusion point is shown in Fig. 8B. Cultures of this clone have been deposited on February 4, 1983 with the collection of the Deutsche Sammlung von' Mikroorganismen (DSM), Gδttingen, Federal

Republic of Germany, where it was assigned No. DSM 2592.

V. Construction of fusion vector piasmid pSPA12 (Fig. 5B) In order to make fusions of the protein A gene fragment of piasmid pSPA9 more generally possible a corresponding fusion vector as pSPAll above was constructed by cutting out the lac Z gene from piasmid p5PA1 and retaining the multi-linker sequence preceding the 5'-end of the gene.

This was achieved by cutting 1 μg of piasmid pSPA14 from section IV with restriction enzyme Eco RI, ligating and transforming E. coli XAC lac as described above under Routine Methods and scoring for tetracyciine resistance and lack of β-galactosidase activity. Clones were plated on Xgal plates containing tetracyciine. Approximately 80% of these colonies were white and restriction analysis on one of these revealed a piasmid schematically shown in Fig. 5B. This piasmid, designated pSPA12, contains the M13 multi-linker at position 2.1 kb in Fig. 2B. The reading frame at the fusion point is shown in Fig.

5B.

VI. Subcloning of the whole protein A coding gene into piasmid pBR322 for construction of piasmid pSPA8 (Fig. 4) The above constructed fusion vectors pSPAll and pSPA12 and the corresponding fused genes containing plasmids pSPA13 and pSPA14 all lack any E. coli residues upstream of the protein A gene originating from starting piasmid pSPAl, including the E. coli promoter. In order to construct, for comparative purposes, a piasmid which contains the whole structural gene of protein A,

including the protein A promoter sequence, but which lacks any E. coll promoter upstreams thereof, the 2.1 kb protein A fragment from step HB above was cloned into the piasmid vector pBR322 as follows below.

1 μg of the purified 2.1 kb protein A fragment from step πB was cut with restriction enzyme Tag I for 1 hr at 60°C to cleave it within the DNA of staphylococcal origin. The enzyme was inactivated by extraction with an equal volume of phenol, followed by repeated ether extraction, and finally the DNA was precipitated with ethanol and dissolved in TE buffer. 1 μg of piasmid pBR322 was cut with restriction enzymes Cla I and Eco RV (which cleave in the same way and thus provide complementary cohesive ends) for 1 hr at 37 C in Bam HI buffer and then heat inactivated for 10 minutes at 65 C. The DNA samples were mixed, ligated and used to transform E. coli HBlOl as described above under Routine Methods. Transformants were streaked out on ampicillin (35 μg/ml). Colonies were picked on plates containing 10 μg/ml of tetracyciine and 35 μg/ml of ampicillin, respectively. Transformants that grew on ampicillin but not on tetracyciine were considered as recombinants. 4 colonies out of 12 of these recombinants were discovered to be protein A positive using the ELISA method described under Routine Methods. Restriction analysis in which purified piasmid was cut with one, two or three restriction enzymes were performed on one of these clones. The resulting restriction map of this piasmid, designated pSPA8, is shown in Fig. 4. The thus constructed piasmid lacks any E. coli promoter upstream of the protein A gene. The protein A gene . fragment is preceded by its own staphylococcal promoter only.

VII. Detection and quantitation of protein A from E. coli clones To evaluate the protein A activity of the two plasmids pSPA13 and pSPA14 constructed above, comparisons were made with piasmid pSPA8 from section VI above, containing the whole structural protein A gene, and piasmid pSKS106, containing the β-galactosidase gene, with regard to total content of protein A and ability to bind to IgG-Sepharose^ columns as follows. ^" Cell suspensions of 300 ml carrying plasmids pSKS105, pSPAδ, pSPA13 and pSPAI4, respectively, were separately grown to OD-- _Q = 1.0 in LB medium containing 35 μg/ml of ampicillin without added glucose. Each cell culture was then centrifuged at 6000 rpm with a Sorvall GSA-rotor for 10 min. and the cell pellets were washed in 20 ml of TE (0.05 M tris, pH.8.5, 0.05 M EDTA) and again centrifuged as above. Finally the cell pellets were resuspended in 15 ml of a protease inhibitor buffer (0.02 M potassium phosphate, pH 7.5 0.1 M NaCl, 0.5% sodium deoxycholate, 1% Triton X-100, 0.1% sodiumdodecyl sulfate (SDS), and 1 mM phenylmethylsulfonyi fluoride (PMSF) ). The cells were then sonicated in an MSE Sonicator for 3 x 30 sec on an ice-bath and centrifuged at 15,000 rpm

(Sorvall SS-34 rotor) for 10 min.

@

3 ml of supernatant was passed over an IgG-Sepharose ^** *B column (Pharmacia AB, Uppsala, Sweden), as described by Hjelm et al, FEBS Lett. 28 (1972), that had been equilibrated with PBST buffer. The column was then washed with PBST and the adsorbed proteins were eluted with 3 ml giycine buffer (0.1 M giycine, 2% NaCl, pH 3.0). The eluate was dialyzed overnight against PBST and the concentration of protein A was determined by the ELISA- test as described in Routine Methods.

In order to determine whether any of the cultures contained a protein A - β-galactosidase fusion protein a modification of the test was performed. Before the addition of the protein A - alkaline phosphatase conjugate 100 μl of ONPG buffer was added to the wells of the microtiter plate and the colour change from colourless to yellow indicating β-galactosidase activity was determined visually. After three washes with PBST 50 μl of protein A - alkaline phosphatase conjugate was added and the test was continued exactly as described in Routine Methods. The results are shown in the following Table 1.

TABLE 1

In the Table "Total" gives the value for cell lysate and "Eluate" gives the corresponding value after binding and eiution from IgG-Sepharose . As to the β- galactosidase activity a negative result represents no detectable colour change after incubation for 30 minutes at room temperature, whereas a positive result represents a clearly visible colour change after 5 minutes under the same conditions.

The test shows that the β-galactosidase from the control (pSKS106) does not bind to IgG coated wells in detectable amounts. In contrast, fusion proteins from cultures containing plasmids pSPA13 and pSPA14 bind to the wells and have enzymatic activity. The β-gaiactosidase activity is, however, not recovered after eiution from the IgG-Sepharose column with glycine-buffer (see Table 1). This is due to inactivation of the enzyme in giycine buffers under pH 4.

The ELISA-test shows that the protein A concentrations of the three protein A containing clones (pSPA8, p5PA13 and pSPA14) vary although the same protein A gene sequence with the same promoter was used. The two clones having the lac Z gene fused to the protein A gene (pSPA13 and pSPA14) contain less protein A than the pSPA8 clone. However, the protein A of the orotein A containing clones (pSPA8, pSPA13 and pSPA1 ) binds to IgG-Sepharose^ and can be eluted with high efficiency with giycine buffer of pH 3.0, as appears from Table 1, although the β-galactosidase of pSPA13 and pSPA14 is irresversibly inactivated. In accordance with the above results desired enzymes may in this way be immobilized directly to an IgG affinity column from a cell crude lysate. Thus the specific affinity between IgG and protein A assures a one-step procedure giving a pure and immobilized enzyme. Viπ. Detection and quantitation of β-galactosidase activity from E. coli clones

Cells carrying plasmids pSKS106, pSPA8, pSPA13 and pSPA1 were grown and lysed exactly as described in section vπ. 10 ml of supernatants were mixed with 1 ml of sedimented IgG-Sepharose*^ 4B (Pharmacia AB, Uppsala, Sweden) that had been washed with PBST buffer. The mixtures were slowly inverted at 8 C f or 1 h and the supernatants were collected. After 4 washes with 12 ml of PBST the supernatants of the last wash were collected. The Sepharose^was resuspended in 10 ml of PBST and aliquots were transferred to smaller tubes.. The β-galactosidase activities were measured exactly as described in Routine Methods and the results are shown in the following Table 2.

TABLE 2

Values are calculated per ml of cell lysate 1 unit is defined as described under Routine Methods above. N.D. = not detected

The β-galactosidase from ceUs containing piasmid pSKSlOδ (control) does not bind to IgG-Sepharose (__) , in accordance with the inability to bind to IgG coated wells (see step VII). In contrast, β-galactosidase from the ceils containing plasmids pSPA13 and pSPA14 (protein A fusion proteins) binds efficiently, in fact more than 70% of the activity is immobilized.

From table 2 it appears that the fusion protein lacking the non-IgG- binding region of the protein A molecule (X in Fig. 2) as produced through piasmid pSPA13 gives 3-4 times more fusion product than the corresponding fusion product containing substantially the whole protein A gene, as obtained via piasmid pSPA14. It seems that in this particular case the "spacer" region X is less favourable.

IX. Eiution of bound β-galactosidase-protein A-fusion protein from IgG-

Sepharos

Aliquots of IgG-Sepharose*^suspension with bound fusion proteins from step Vin were transferred to columns. The fusion proteins were eluted from 50 μl of sedimented gel by adding 0,5 mi buffer containing purified protein A

(Pharmacia, Uppsala, Sweden) at various concentrations at room temperature.

(S) The β-galactosidase activity of the eluates and the IgG-Sepharose gels after eiution were determined as described under Routine Methods above. The results are shown in Table 3, the values being expressed as percent of Sepharose^bound β-galactosidase activity.

TABLE 3

From the above results it appears that at least half the β-galactosidase activity may be eluted by this procedure.

-^ifREΛ ^"

EXAMPLE H I. Construction of a shuttle piasmid containing a fusion between the genes encoding protein A and IGF-1

In the following steps A-D the construction of a piasmid pUN20i, containing the protein A gene, without region X, fused to a synthetic gene encoding a modified human IGF-1 (human insulin-like growth factor type 1), is de¬ scribed.

A. Synthesis and cloning of the gene encoding human IGF-1 The oligomers shown in Fig. 13A were synthesised on an automatic DNA- synthesising machine developed by KabiGen AB, Sweden (Chow et al, Nucleic Acids Res. 9, 2807-2817) with N-protected nucleoside chlorophosphites as reagents (Elmblad et al, Nucleic Acids Res. W, 3291-3301 (1982)). After purification and phosphorylation the oligomers were assembled into seven blocks as shown in Fig. 13B, which were then ligated as described in Fig. 13B. In the last step, block A and block B in Fig. 13B were ligated giving the complete IGF-1 gene. This segment was digested with EcoRI and Hind III restriction enzymes and after purification inserted into piasmid pUC8 and transformed into E. coli 3M83. The transformants were screened by colony hybridisation using A15 as probe and one of the positive clones, designated 3M83/pKG3, was sequenced, confirming that the sequence matched the IGF-1 gene.

The synthesis of the IGF-1 gene is also described in the Swedish patent application 8303626-9, the disclosure of which is incorporated by reference herein. The DNA-sequence of the IGF-1 gene and the corresponding amino acid sequence appears from Fig. 11, except that a giycine residue (Gly) has been changed into an aspartic acid residue (Asp) as will be described in the following step B.

B. In vitro mutagenesis of the synthetic gene to encode a modified human IGF-1. Oligonucleotide mediated in vitro mutagenesis was performed on the cloned synthetic IGF-1 gene, in order to change the part encoding the N-terminal amino acid residue of the mature protein. By changing this amino acid residue from a giycine to an aspartic acid residue the dipeptide aspartic acid-proline was formed. This allows for gene fusions encoding hybrid proteins that can be cleaved apart, before or after purification, by formic acid treatment which cleaves between aspartic acid and proline (Landon, Methods in Enzymology 47, 132-145, 1977). Thereby, mature IGF-.1, lacking the N-terminal giycine, can be produced.

10 μg of piasmid pKG3 was cleaved with Eco RI and Hind III and a 0.22 kb fragment thereof was cut out from a 5% polyacrylamide gel after electrophore¬ sis. The fragment was eluted and purified as described in the International patent application PCT/SE83/00297 (the disclosure of which is incorporated by reference herein) . '. 50 ng of purified fragment was mixed with

200 ng of phage 13 mp9 and cleaved with Eco RI and Hind in in a total volume of 20 μl. After treatment with T4-ligase the DNA was used to transform E. coli 3M83 and the cells were spread on AXI-plates. Cleavage, Iigation and transformation were performed as described under Routine Methods. Phage purification from one white plaque was performed as described under Routine Methods. Using the universal primer (Bio-Labs, New England, USA) the phage insert was confirmed to be the 220 bp synthetic IGF-1 gene. This phage, designated mp9/IGF-l, was used for the following mutagenesis.

Two oligonucleotides were synthesized and purified exactly as described above and in the Swedish patent application No. 8303626-9. One primer oligonucleotide consisting of 24 bases (5'-GTGAATTCTATGGACCCCGAAACT-

3') which was used for the mutagenesis and one probe (P) oligonucleotide consisting of 14 bases (5'-AATTCTATGGACCC-3 which was used to identify successfully, mutageπized phage clones. The mismatches between the synthetic IFG-1 gene and the primer are shown in Fig. 9.

16 p oles of mp9/IFG-l and 80 pmoles of primer were mixed in a total volume of 80 μl containing 100 mM NaCl, 20 M Cl ₂ and 40 mM TRIS-HC1, pH 7.5. The mixture was heated to 65°C for 3 minutes and allowed to cool to 23°c for 30 minutes. After transfer to an ice-bath, 190 μl of H ₂ O and 30 μl of a solution containing 100 mM MgCl ₂ , 50 M DTT and 200 M TRIS-HC1, pH 7.5 _t was added. 50 units of Klenow fragment (Boehringher-Mannheim, West-Germany) was added and after 10 minutes in an ice-bath the sample was brought to 23°C for 30 minutes. Another 50 units of Klenow fragment was added and after 60 minutes at 23°C the polymerase was heat inactivated at 65°C for 10 minutes. The sample was precipitated once with ethanol, followed by cleavage with Eco RI and Hind III according to Routine Methods above. The 0.22 kb fragment was cut out from a 5% polyacrylamide gel after electrophoresis and the fragment was eluted and purified as described in the above mentioned International patent application PCT/SE83/00297. 50 ng of purified fragment were mixed with 200 ng of phage 13 mp9 cleaved with Eco RI and Hind III in a total volume of 20 μl. After Iigation and transformation to E. coli 3M83, exactly as described above, white plaques were found in a background of blue plaques. 48 white plaques were

further analysed by hybridization witn two synthetic probes as described by

Winter et al (Nature 299, 21 October, 1982). The filters were hybridized at room temperature with 32 P-labelled oligonudeotides and washed at different tempera¬ tures. Using probe A2, 5 ^, -ATGGGTCCCGAAAC-3', (Swedish patent application No. 8303626-9.), all dones except four show strong hybridization after wash at 44°C indicating that these dones contain the original IGF-1 gene. Using probe P, 5*-AATTCTATGGACCC-3 ^l , all four of the previously negative dones showed significant hybridization. One of the four phages, designated mp9/IGF-l.M3, was further sequenced using the universal primer as described above. This confirmed a successful mutagenesis as shown in Fig. 9.

200 ng of phage mp9/IGF-l.M3 and 200 ng of piasmid pUC8 were separately deaved with Eco RI and Hind III. After T4-ligase treatment, in a total volume of 20 μl, the DNA was used to transform E. coli 3M83 and the cells were plated out on AXI-plates. Cleavage, Iigation and transformation were performed as described above under Routine Methods. Restriction analysis of a white colony revealed the expected piasmid, pUC8, containing a 0.22 kb Eco RI/Hind III insert. This piasmid was designated pKGll and was used for the following steps.

C. Construction of shuttle piasmid pUN200 containing pKGll .

1 μg of. pKG 11 from step B and 2 μg of pHV14, both digested with Hind • ni, were mixed and ligated in a total volume of 100 μl overnight at +14°C. After digestion with Eco RV, the DNA-mixture was transformed to E. coli HBlOl and plated on LA-plates containing 50 μg ampicillin per milliliter. 52 single colonies were picked to LA plates containing 10 μg/ml of chloramphenicol and 50 μg/ml of ampicillin. After two days at 28°C one done appeared and the piasmid in this done was further characterized by restriction analysis. This revealed piasmid pUN200 schematically shown in Fig. 10. This piasmid, which contains the IGF-1 gene, can replicate -both in E. coli and S. aureus.

D. Construction of shuttle plasmids pUN201 and pUN202

1 μg of pUN200 and 1 μg of piasmid pSPA16, both digested with Eco RI, were mixed and ligated in a total volume of 100 μl overnight at 14°C. The ligase was heat inactivated at 65 C for 10 minutes. After digestion with Eco RV to decrease the number of background dones containing pSPA16, the DNA mixture was transformed to E. coli HBlOl and plated out on LA-plates containing 50 μg of ampicUlin. Plasmids from 48 dones were analyzed by restriction mapping and 3 thereof were found to contain pUN200 with a 1,1 kb Eco RI insert from pSPA16, corresponding to the 5'-end of the protein A gene. The orientation of the insert in these three plasmids was further analyzed by deavage with Hind III and

two were found to contain a predicted fusion between the genes encoding protein A and IGF-1. This piasmid was designated pUN201 (Fig. 10). The nudeotide sequence and the deduced amino acid sequence of this gene fusion are shown in Fig. 11. The predicted molecular weight of the mature hybrid protein is 38,701. One of the three dones was found to contain a piasmid, designated pUN202, with opposite orientation of the protein A gene versus the IGF-1 gene (Fig. 10). This piasmid codes for a truncated protein A with a predicted molecular weight of 30,963 (Fig. 12). H. Transformation of shuttle plasmids pUN201 and pUN202 to S. aureus SA113

10 μg of plasmids pUN201 and pUN202 from step ID above were used to transform protoplasts of S. aureus SA113 as described by Gδtz, F. et al, 3. Bacterial. 145, 74-81 (1981) and in the International patent application PCT/SE83/00297 (Step niA). Chloramphenicol resistant dones were found after 3 days at 37 C and these transformants were restreaked on TSA-plates (Trypticase Soy Agar) with chloramphenicol (10 μg/ml). One transformant of the respective piasmid (pUN201 and pUN202) was chosen for further analysis. Restriction mapping of the purified plasmids revealed that the intact piasmid had been introduced into the S. aureus SA113 host. HI. Quantification and localization of the protein ^" A activity from dones carrying pUN201 and pUN202

E. coli cdls carrying pUN200, pUN201 and pUN202 respectivdy (from Step IC and D) above and S. aureus cdls carrying pUN201 or pUN202 were cultivated in 200 ml of liquid medium overnight. E. coli strains were grown in LB medium with ampiciHin (50 μg/ml) and S. aureus strains in TSB (Trypticase Soy Broth) with chloramphenicol (10 μg/ml). The cells were pelleted by centrifugation at 6000 rpm with a Sorwall GSA-rotor for 10 minutes and the supernatant, designated medium, was saved. The ceU pellet was washed in 10 ml of PBS + TWEEN and again centrifuged as above. This time the cell pellet was resuspended in 10 ml of a protease inhibitor buffer (θ.02 M potassium phosphate, pH 7.5, 0.1 M NaCl, 0.5% sodium deoxycholate, 1% Triton X-100, 0.1% sodiumdodecyl sulfate (SDS), and 1 mM phenylmethylsulfonyl fluoride (PMSF)J. The cdls were then sonicated in a MSE sonicator for 4 x 40 sec. on an ice-bath and centrifuged at 15,000 rpm (Sorvall SS-34 rotor) for 10 in. The supernatant, designated cdl extract, was collected and the ELISA-test described under Routine Methods was performed to determine the amount of protein A in the samples. The results are shown in the following Table 4.

Table 4

Amount of protein A per ml of sonicated ceU culture determined by the ELISA-test. Zero-values correspond to less than 0.1 μg/ml.

Host (Piasmid) CeU extract Medium (μg/ml) (μg/ml)

E. coli HBlOl (pUN200) 0 0

E. coli HBlOl (pUN201) 2 0

E. coli HBlOl (pUN202) 2 0

S. aureus SA113 (pUN20I) _. 0,2 5

S. aureus SA113 (pUN202) 0 5

E. coli HB 101 0 0

S. aureus SA113 0 0

Table 4 shows that in both E. coli and S. aureus i the amoum produced is not influenced by the orientation of the fragment containing the protein _. A gene (piasmid pUN201 versus pUN202). Thus, the protein A IGF-1 hybrid protein encoded by pUN201 is produced at approximatdy the same levd as the truncated protein A encoded by pUN202. Both proteins are, as expected, found in the cdl extract of E. coli and in the medium of S. aureus. IV. Purification of IGF-1 by IgG-affinity chromatography and formic acid treatment The media of S. aureus SA113 carrying ρUN201 and pUN202 respectivdy, from Step III, were each passed over an IgG-Sepharose^ 4B column (Pharmacia AB, Uppsala, Sweden) (Hjdm et al, FEBS Lett. 28, 73-76 (1972)} that had been eqialibrated with a sodium acetate buffer (0.1 M sodium acetate, 2% NaCl, pH 5.5). The column was then washed with the same buffer as above and the adsorbed protein A duted with a glydne buffer (0.1 M giycine, 2% NaCl, pH 3.0). The eluted fraction was dialyzed against distilled water and thereafter lyophUized in two diquots. The protein pdlet of one of the aliquots was analysed on a 13% SDS-polyacrylamide gd at 100 V for 12 hours. The gd was stained with amidoblack (0.1%, in 45% methanol, 10% acetic acid). This revealed a major protein with the molecular weight of 38,701 for pUN210 and 30,963 for pUN202. This is in accordance with the predicted sizes deduced from the DNA sequences (see Step ID above).

The second aliquot was resuspended in 0.5 ml of 70% formic acid and

further incubated at 37°C for 2 days. This process deaves proteins at the dipeptide sequence aspartic acid-proline. The predicted degradation products from the hybrid protein encoded by pUN201 were, apart from the IGF-1 moiety lacking the N-terminal glydne, five oligopeptides of the molecular weights of 6800, 6600, 6600, 6600 and 600. SDS-polyacrylamide dectrophoresis, as described above, also confirms that the major protein bands are shifted from approximately 38000 to several bands around 7000.

The formic acid treated proteins were lyophilized and resuspended in distiUed water. The sample from pUN201 was passed over an IgG-Sepharose*- 4B column as described above. The flow through and the duted material (with the glydne buffer) were saved for further analysis. V. An ysis of the protein products by radio receptor assay (RRA)

The radio receptor assay (RRA) was performed according to HaU et al, 3. Clin. Endocrinol. Metab. 4_8, 271-278 (1974) using a participate fraction of human placental membrane as matrix. The standard used in the assay consisted of pooled normd human serum having a potency of 1 unit (U) of IGF-1 per ml. The lower limit of the assay was 10 mU per ml. The peptide used for labdling was purified from Cohn fraction IV to a specific activity of 500 U/mg protein (according to RRA). Labdling of the peptide was performed according to Thordl et al. Biochem. Biophys. Acta 251, 363-369 (1971). Purification of the tracer was done on a carboxymethyl cell ulose column using an eiution gradient of 0.1 M NH^OAc from pH 4.0 to pH 6.8. The specific activity of the tracer was approximately 20 μCi/μg. The assay was performed as follows:

The standard or unknown sample (100 μl) was incubated together with 100 μl of placental membrane and 100 μl of labdled IGF-1 overnight at +4°C. After centrifugation the peUet was washed once and counted in a gamma counter. The sample potency was calculated using an "in house" computer program.

Samples before and after the formic acid treatment, from Step IV above, were andyzed by the RRA-test and the resdts are shown in Table 5 bdow.

• IT

Table 5

Radio receptor andysis (RRA) for IGF-1 activity in growth medium from S. aureus 5A113 (pUN202) and S. aureus SA113 (pUN201) after isolation and purification by IgG affinity chromatography. Zero corresponds to less than 1 U/l medium.

Piasmid Treatment Activity/! (U/l) pUN202 Before treatment with formic acid After treatment with formic add pUN201 Before treatment with 0 ormic acid

After treatment with 143 formic acid

Flow through, 106

Eluate, ^x 19

A formic acid treated sample was passed over an IgG-Sepharose^column and the activity was measured for bound (duate) and not bound (flow through) IGF-1.

From Table 5 it appears that the hybrid protein encoded by pUN201 has no detectable IGF-1 activity. Treatment with formic acid yields IGF-1 activity, and most of this activity does not bind to the IgG affinity column indicating a success ul deavage between the protein A and the IGF-1 moiety.

While embodiments of the invention have been presented above, the invention is, of course, not restricted thereto, but many variations and modi- fications are possible without departing from the scope thereof as de ined by the subsequent claims.

INTERNATIONAL FORM

r

Martin Un berg Department of Microbiology Blomedteal Center RECEIPT IN THE CASE OF AN ORIGINAL DEPOSIT Box 581 issued pursuant to Rule 7.1 by the INTERNATIONAL DEPOSITARY AUTHORITY

S-75123 Uppsala Identified at the bottom of this page Schweden

L _l

I. IDENTIFICATION OF THE MICROORGANISM

Identi ication reference given by the Accession number given by the

0EPOSΣT0R: INTERNATIONAL DEPOSITARY AUTHORITY.

E. con SPAH DSM 2-3"i

II. SCEENTIFIC DESCRIPTION AND/OR TAXONOMIC DESIGNATION

The microorganism identified under I above was accompanied by:

P I a scientific description

5(1 a proposed taxonocnic designation

(Mark with a cross where applicable)

IH. RECEIPT AND ACCEPTANCE

This International Depositary Authority accepts the nicroorgonism identified under I above, which wos received by it on ^' July 12, 1982 {Date of the original deposit) "

! IV. INTERNATIONAL DEPOSITARY AUTHORITY

Nome: DEUTSOE SAMMLUNG Signature(s) of person's) having the power

VON MIKROORGANISMEN to represent the International Depositary

Address: Grisβbachstrasse 8 Authority or of authorized official(s):

D-3-ιOO Gottingen

^ .Qλjuu-r

Date: J _u Jy J3 _f 1982

Where Rule έ.'.'d) applies, such date is the date an which the. status of international depositary authority wos acquired* where a deposit made outside the Budapest Treaty after the ocquisition of the status of international depositary authority is converted into a deposit under the Budapest Treaty, such date is the date on which the microorganism was received by the international depo¬ sitary authority.

~1 VIABILITY STATEMENT

Martin Lin db erg issued pursuant to rule 10.2 by the Department of Microbiology INTERNATIONAL DEPOSITARY AUTHORITY identified on the bottom of this pogβ Blomedlcal Center Box 581

S-75123 Uppsala Schweden

J

I. DEPOSITOR U. IDENTIFICATION OF THE MICROORGANISM

Name: Martin LIndberg Accession number given by the INTERNATIONAL. DEPOSITARY AUTHORITY

Department of Microbiology Address* Blomedlcal Center DSM 24 4 Box 581 Date of the deposit or of the transfer: S-75123 Uppsala, Schweden July 12, 1982 in. VIABILITY STATEMENT

The viability of the microorganism identified under H above was tested on On that date, the said microorganism was

0 viable

|~1 no longer viable

IV. CONDITIONS UNDER WHICH THE VIABILITY TEST HAS BEEN PERFORMED

V. INTERNATIONAL DEPOSITARY AUTHORITY

Name: DEUTSCHE SAJAJLUNG Signature(s) of person (s) having the power

VON MIKROORGANISMEN to represent the International Depositary Authority or of authorized official (3} :

Address: Grisebcshstr. 8 D-34O0 Gδttingen

Date: July 13, 1982

Indicate the date of original deposit or, where a new deposit or a transfer has been mode, the most recent relevant date (date of the new deposit or date of the transfer).

2

In the cases referred to in Rule 10.2(a)(li) and (ill), refer to the meet recent viability test.

3

Mark with a cross the applicable box.

. Fill In if the Information has been requested and if the results of the tests were negative.

3S

r

Martin Lindberg RECEIPT IN THE CASE OF AH ORIGINAL DEPOSIT Department of Microbiology issued pursuant to Rule 7.1 by the Biomedical Center INTERNATIONAL DEPOSITARY AUTHORITY Box 581 Identified at the botto* of this page

S-75123 Uppsala Sweden J

1

Where Rule 6.*ι(d) applies, such dote is the date on which the status of international depositary authority wos acquired} where a deposit made outside tha Budapest Treaty after the ocquisition of the status of international depositary authority is converted into a deposit under the Budopest Treaty, such date Is the date on which the microorganism was received by the international depo¬ sitary authority.

-+0

r VIABILITY STATEMENT Issued pursuant to rule 10.2 by the INTERNAΠONAL DEPOSITARY AUTHORITY identified on the bottoβ of this page

Martin Lindberg Department of Microbiology Biomedical Center Box 581

S-751 23 Uppsala L_ Sweden J

I. DEPOSITOR n. IDENΠFICATION OF THE MICROORGANISM

_N αm _e x Martin Lin berg,Dept.of Micro- Accession number given by the biology, Biomedical Center INTERNATIONAL DEPOSITARY AUTHORITY

Address: Box 581 DSM 2706

S-751 23 Uppsala Sweden Date of the deposit or of the transfer:

August 15,1983 in. VIABILITY STATEMENT

The viability of the nicroorgonism identified under II above was tested on Aug.22,1983 On that date, the sold microorganism wots

[ ³ • viable r~l no longer viable

IV. CONDITIONS UNDER WHICH THE VIABILITY TEST HAS BEEN PERFORMED

V. INTERNATIONAL DEPOSITARY AUTHORITY

None: . DEUTSCHE SAMMLUNG Signature (s) of person (s.) having the power

VON MIKROORGANISMEN to represent the International Depositary Authority or of authorized offidal(s) :

Address: Grisebochstr. 8 . D-3 00 GSttingen

^Datβ « August 23,1983

Indicate the date of original deposit or, -where a new deposit or a transfer has been node, the ^β o ^β t recent relevant date (date of the new deposit or date of the transfer).

In the cases referred to In Rule 10.2(a)(11) and (ill), refer to the βoβt recent viability test.

3

Mark with a cross the applicable box.

Ψ2

INTERNATIONAL FORM

Dr.Martin Lindberg RECEIPT IN THE CASE OF AN GRIdNAL DEPOSIT Department of Microbiology issued pursuant to Rule 7.1 by the Biomedical Center INTERNATIONAL DEPOSITARY AUTHORITY Box 581 identified at the bottoβ of this page

S-75123 Uppsala Sweden

L

I. IDENTIFICATION OF THE MICROORGANISM

Identification reference given by the Accession number given by the DEPOSITOR: INTERNATIONAL DEPOSITARY AUTHORITY:

SPA 23 DSM 2591

II. SCIENTIFIC DESCRIPTION AND/OR TAX0N0MIC DESIGNATION

The microorganism identified under I above was accompanied by: fX. a scientific description

f l a proposed taxonoaic designation

(Mark with a cross where applicable)

IH. RECEIPT AND ACCEPTANCE

This International Depositary Authority occepts the microorganism identified under I above, which was received by it on Februar 4, 1983 (Date of the original deposit) ¹

IV. INTERNATIONAL DEPOSITARY AUTHORITY

Name: DEUTSCHE SAMMLUNG Signature(s) of person(s) having the power

VON MIKROORGANISMEN to represent the International Depositary

Address: Grisebachstrasse 8 Authority or of authorized official(s): D-3<t00 Gottingen , .

Date: February 17, 1983

Where Rule &.<ι(d) applies, such date is the date on which the status of international depositary authority wos acquired} where a deposit made outside the Budapest Treaty after the acquisition of the status of international depositary authority is converted into a deposit under the Budapest Treaty, such date is the dote on which the microorganism wos received by the International depo¬ sitory authority.

-.3

1 VIABILITY STATEMENT issued pursuant to rule 10.2 by the INTERNATIONAL DEPOSITARY AUTHORITY identified on the bottoe. of this page

Dr. artin Lindberg Department of Microbiology Biomedical Center Box 581

S-75123 Uppsala Sweden J

I. DEPOSITOR II. IDENTIFICATION OF THE MICR00RGANIS

Name: Dr. .Lindberg

Department of Microbiology Accession number given by the INTERNATIONAL DEPOSITARY AUTHORITY

Biomedical Center

Address: DSM 2591

Box 581 S-75123 Uppsala Date of the deposit or of the transfer: February 4. 1983

IH. VIABILITY STATEMENT

The viability of the microorganism identified under II above was tested on F eb . 8 , 1983. On. that date, the said microorganism was .

Gj viable r~l no longer viable

IY. CONDITIONS UNDER WHICH THE VIABILITY TEST HAS BEEN PERFORMED

V. INTERNATIONAL DEPOSITARY AUTHORITY

Name: DEUTSCHE SAMMLUNG Signature(s) of person(s) having the power

VON MIKROORGANISMEN to represent the International Depositary Authority or of authorized official(s):

Address: Grisebochstr. 8 D-300 Gδttingen

Date: February 17 , 1983

Indicate the date of original deposit or, where a new deposit or a transfer has been node, the most recent relevant date (date of the new deposit or date of the transfer). In the cases referred to in Rule 10.2(a)(ii) and (ili), refer to the most recent viability test.

3

Mark with a cross the applicable box.

{,

Fill in if the inforn ^β tion hos been requested ond if the results of the tests were negative.

MICROORGANISMS

Optional Sheet In connection with tha microorganism rafarred to on page. 23 ., lin — __Ξ_1_L__ — of the description *

A. IDENTIFICATION OF DEPOSIT *

Further deposits are identified on an additional sheet | |»

Name of depositary institution *

Deutsche Sammlung von Mikroorganismen (DSM)

Address of depositary Institution (including postal cod* and country) *

Grisebachstrasse 8, D-3400 Gottingen, Federal Republic of Germany

Oat* αl deposit * Accession Number •

February 4, 1983 DSM 2592

B. ADDITIONAL INDICATIONS ' (leave blank if not applicable). This information is continued oα a separate attached sheet Q

C. DESIGNATED STATES FOR WHICH INOICΛTiONS ARE MADE * (if th* Inc . .stlons are not for ali designate SUtes)

D. SEPARATE FURNISHING OF INDICATIONS » (leave blank If not applicable)

The indications listed below will be submitted to tha International Bureau later • (Specify the general natur* of th* Indication* *.o, " Accession Number ol Deposit") . ^•

E. Q This sheet was received with th* International application when filed (to be checked by th* receiving Office)

INTERNATIONAL FORM

r i

Dr.Martin Lindberg RECEIPT IN THE CASE OF AN ORIGINAL DEPOSIT Department of Microbiology issued pursuant to Rule 7.1 by the Biomedical Center INTERNATIONAL DEPOSITARY AUTHORITY Box 581 identified at the bottom of this page

S-75123 Uppsala Sweden. L

I. IDENTIFICATION OF THE MICROORGANISM

Identification reference given by the Accession number, given by the DEPOSITOR: INTERNATIONAL DEPOSITARY AUTHORITY.

SPA 24 DSM 2592

II. SCIENTIFIC DESCRIPTION AND/OR TAXONOMIC DESIGNATION

The nicroorgonism identified under I above was accompanied by: Q a scientific description

rΑ a proposed taxonomic designation

(Mari with a cross where applicable)

in. RECEIPT AND ACCEPTANCE

This International Depositary Authority accepts the microorganism Identified under I above, which was received by it on February 4,1983 (Date of the original deposit)

IV. INTERNATIONAL DEPOSITARY AUTHORITY

Name: DEUTSCHE SAMMLUNG Signature(s) of person(s) having the power

VON MIKROORGANISMEN to represent the International Depositary

Address: Grisebachstrasse 8 Authority or of authorized official(s): D-3"*00 Gottingen

°ete February 17, 1983

Where Rule c.«iι(d) applies, such date is the date on which the status of International depositary authority was acquired] where a deposit made outside the Budapest Treaty after the acquisition o the statue of international depositary authority is converted into a deposit under the Budapest Treaty, such date is the date on which the microorganism was received by the international depo¬ sitary authority.

r VIABILITY STATEMENT issued pursuant to rule 10.2 by the INTERNATIONAL DEPOSITARY AUTHORITY identified on the bottom of this page

Dr. artin Lindberg Department of Microbiology Biomedical Center Box 581

S-75123 Uppsala Sweden J

I. DEPOSITOR II. IDENTIFICATION OF THE MICROORGANISM _ame: Dr . M. Lindberg Accession number given by the

Department o f Microbiolo gy INTERNATIONAL DEPOSITARY AUTHORITY

Biomedical Center Address: DSM 2592

Box 581

S-75123 Uppsala Date of the deposit ar of the transfer: February 4 _f 1983

III. VIABILITY STATEMENT

The viability of the microorganism identified under H above was tested on Feb.8.1983 _* On that date, the said microorganism was

^' C3 ³ vioble I I no longer vioble

IV. CONDITIONS UNDER WHICH THE VIABILITY TEST HAS BEEN PERFORVCD

V. INTERNATIONAL DEPOSITARY AUTHORITY

Name: DEUTSCHE SAMMLUNG Signature(s) of person(s) having the power

VON MIKROORGANISMEN to represent the International Depositary Authority or of authorized official(s):

Address: Grisebachstr, 8 D-3<t∞ Gottingen - ..CC& -cS

Date: February 17 , 1983

Indicate the date of original deposit or, where a new deposit or a transfer has been made, the most recent relevant date (date of the new deposit or date of the transfer).

2

In the cases referred to in Rule 10.2(a)(li) and (iii), refer to the most recent viability test.

3

Mark with a cross the applicable box.