AU4270193A

AU4270193A - Protein l and process for its preparation by recombinant dna technology

Info

Publication number: AU4270193A
Application number: AU42701/93A
Authority: AU
Inventors: Antony Atkinson; Clive James Duggleby; Jonathan Paul Murphy; Angus Robert Trowern
Original assignee: Health Protection Agency; Public Health Laboratory Service Board
Current assignee: Actinova Ltd
Priority date: 1992-05-07
Filing date: 1993-05-07
Publication date: 1993-11-29
Anticipated expiration: 2013-05-07
Also published as: EP0640134A1; WO1993022438A1; AU674320B2; GB9209804D0; JPH07507925A; CA2135207A1

Description

PROTEIN L AND PROCESS FOR ITS PREPARATION BY RECOMBINANT DNA TECHNOLOGY

This invention relates to novel immunoglobulin binding proteins, processes for their production, recombinant DNA molecules coding therefor and recombinant DNA molecules useful as probes therefor.

More specifically the present invention relates to the protein designated Protein L in substantially pure and/or intact and/or homogeneous form and to recombinant DNA molecules coding for Protein L.

A multitude of Gram-positive bacteria species have been isolated that express surface proteins with affinities for mammalian i munoglobulins through interaction with their heavy chains. The best known of these immunoglobulin binding proteins are type 1 Staphylococcus Protein A and type 2 Streptococcus Protein G which have been shown to interact principally through the C2-C3 interface on the Fc region of human immunoglobulins. In addition, both have also been shown to interact weakly to the Fab region, but again through the immunoglobulin heavy chain.

Recently, a novel protein from Peptococcus magrrus , Protein L, has been reported that was found to bind to human, rabbit, porcine, mouse and rat immunoglobulins uniquely through interaction with their light chains. In humans this interaction has been shown to occur exclusively to the kappa chains. Since both kappa and lambda light chains are shared between different classes. Protein L binds strongly to all human classes, in particular to the multi-subunited IgM, and similarly is expected to bind^■ to all classes in species that show Protein L light chain binding.

Both peptococcus and peptostreptococcus have been reported to produce Protein L, which binds to the Kappa light chain of human immunoglobulins. It has been proposed that Protein L is a virulence factor; non-virulent peptococci and peptostreptococci appear to neither express Protein L nor have the structural gene for it (Kastern et al 1990). Protein L is of particular interest since it has been reported to bind to the Kappa light chain which is present in all classes and sub classes of immunoglobulins. As such it should prove to be a useful diagnostic reagent for use in ELISA and RIA techniques. Existing methods for producing Protein L typically rely on its extraction and purificarion from bacteria expressing Protein L on their cell surface. Such methods are by their nature inefficient, yielding a low yield of impure protein and taking a long time to perform. Protein L obtained in this way may in addition be incomplete.

EP-A-0255 ^ 7 describes the purification and attempted characterisation of Protein L by standard protein purification techniques. Subsequently, the authors of EP-A-0255 ^97 have published a number of scientific papers describing further investigations into the nature and structure of Protein L, but to date, attempts fully to characterize the protein have failed. Thus recently, in a paper entitled "Protein L a Bacterial Immunoglobulin-Binding Protein and Possible Virulence Determinant" by . Kastern et al (Infection and Immunity, May 1990, pp. 1217-1222) there are described unsuccessful attempts to isolate the gene coding for Protein L by determining N-terminal amino acid sequences of tryptic fragments of Protein L and using the derived sequence information to construct probes for isolating the gene.

However up until now, the problem of Isolating and characterising the gene for Protein L has defied solution, thereby preventing significant improvement in Protein L production.

Furthermore, in the absence of sequence information for Protein L it has not been possible to identify sequences associated with complex formation with immunoglobulin Kappa light chains.

This invention is based on a cDNA sequence comprising a cDNA insert coding for Protein L in its entirety which has now been isolated, thus enabling the above problems to be solved. This cDNA sequence, and the amino acid sequence corresponding to the longest open reading frame thereof, are depicted in Figure 1. The longest open reading frame of the sequence depicted in Figure 1 extends from TTG (103) to AAA(3l83) and the depicted DNA comprises a coding region extending from nucleotide 208 to nucleotide 3183 which codes for immature Protein L.

According to one aspect, the present invention thus provides a polypeptide designated Protein L and being capable of forming a complex with immunoglobulin Kappa light chains, said polypeptide being characterised by being in substantially homogeneous and/or intact and/or full length form.

The polypeptides of the invention preferably have at least two and more preferably three of the characterising features of (i) being in substantially homogeneous (ii) being intact and (iii) being in substantially full length form.

The full length polypeptides are preferably at least 900 amino acids in length, more preferably are at least 950 amino acids in length and most preferably are at least 975 amino acids in length.

The full length polypeptides provided according to the invention include both the mature and immature sequences. The triplet ATG (208) is believed to be the start of the signal sequence which extends for 30 - 35 amino acids. Thus the polypeptides of the invention include both the immature polypeptide extending from Met (208) to Lys (3183) as well as the mature polypeptides which omit 30 to 35 amino acids from the N-terminus.

By way of summary, the full length of immature protein L (or pre-Protein L) is believed to be 992 amino acids and the immature polypeptides of the invention optimally comprise polypeptides which are at least 990 amino acids in length.

The immature polypeptides according to the invention preferably have N-terminal sequences corresponding to at least f , more preferably at least 10 and most preferably at least 15 of the amino acids of the N-terminal sequence Met Lys lie Asn Lys Lys Leu Leu Met Ala Ala Leu Ala Gly Ala He Val Val Gly Gly These N-terminal sequences are:

Met Lys He Asn Lys Lys Leu

Met Lys He Asn Lys Lys Leu Leu Met Ala

Met Lys He Asn Lys Lys Leu Leu Met Ala Ala Leu Ala Gly Ala

The invention further includes polypeptides as defined above, but omitting a signal sequence. The signal sequence is between 20 and 35 amino acids in length, more specifically, between 23 and 27 amino acids in length. Thus the full length mature polypeptides of the invention preferably commence with one of the following N-terminal sequences: Gly Ala Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn .^"..

*Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn

Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Glu Glu Asp Asn Thr Asp Asn Asn ...

Glu Asp Asn Thr Asp Asn Asn ...

Asp Asn Thr Asp Asn Asn ...

Asn Thr Asp Asn Asn —

Thr Asp Asn Asn — and continuing with the sequence depicted in Figure 1 from Leu • 313) ■ The actual mature sequence of Protein L is believed to commence with the above sequence marked "*", i.e. the start of the mature sequence is marked "M" in Figure 1, with the signal sequence marked "SS".

The invention also includes variants of the polypeptides defined above, all variants being capable of forming a complex with immunoglobulin Kappa light chains.

Desirably, the variant polypeptides of the invention have at least 7 sequence homology, preferably at least 0 sequence homology with the amino acid sequence depicted in Figure 1. Most preferably they have at least 95% sequence homology, preferably at least 98% sequence homology with the amino acid sequence depicted in Figure 1.

The full length polypeptide depicted in Figure 1 has the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala Gly Ala Tyr Val Ser Leu Lys Lys Arg Lys.

Polypeptides according to the invention preferably have C-terminal sequences corresponding to at least 7. more preferably at least 10 and most preferably at least 15 of the amino acids of the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala Gly Ala Tyr Val Ser Leu Lys Lys Arg Lys.

These C-terminal sequences are:

Leu Ala Ala Ala Ala Leu Ser

Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala

Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala Gly Ala Tyr Val Ser

The invention further provides, in a second aspect, recombinant DNA molecules having an insert coding for polypeptides as defined above. In an embodiment of the invention, such recombinant DNA molecules which code for the immature polypeptides of the invention are selected from (a) the DNA coding sequence depicted in Figure 1 commencing with ATG in location 208 and extending to AAA in location 3183. and (b) degenerate DNA sequence coding for the same amino acid sequence.

In a further embodiment, the recombinant DNA molecules which code for the mature polypeptides of the invention are selected from (c) the DNA coding sequence depicted in Figure 1 commencing with a codon in the region between locations 211 and 313 and extending to AAA In location 3183, and (d) degenerate DNA sequence coding for the same amino acid sequence.

In a third aspect, the present invention provides a process for producing an immunoglobulin binding protein which comprises culturing a transformed host which has been transformed with an expression vector containing a DNA coding sequence as defined above. The transformed host may be eukaryotic or prokaryotic, e.g. a bacteria or yeast or an animal cell, a cultured mammalian cell or an Insect cell.

The invention further provides, in a fourth aspect, recombinant DNA molecules comprising a contiguous sequence of at least 12, preferably 15, and most preferably 17 bases depicted In the DNA sequence of Figure l and being useful as probes for isolating other DNA sequences coding for immunoglobulin binding protein. Thus, in embodiments of the fourth aspect, the DNA sequence depicted in Figure 1 is used to construct probes useful in probing gene banks for DNA sequences coding for polypeptides of related sequence to Protein L.

The present invention further provides, in a fifth aspect, a process for producing an immunoglobulin binding protein which comprises culturing an expression vector containing a DNA coding sequence isolated by probing- a gene bank with a probe as defined above.

The specific binding properies of Protein L, including its ability to bind immunoglobulin Kappa light chains, is believed to be attributable to the presence of sequences which have a recognisably repeated character within the amino acid sequence of the molecule. 3y the term "recognisably repeated character" as used herein is meant that the amino acid sequence comprises at least two sequences, each of from 20 to 45 amino acids in length, which have an at least 75%. preferably at least 90% and most preferably at least 95% homology with one another.

The polypeptide sequence depicted in Figure 1 includes ten sets of repeated sequences at least two of which are considered to be responsible for immunoglobulin Kappa light chain binding.

These ten sets of repeated sequences are labelled at their N-terminal ends as follows:

(1) Al, A2 and A3;

(2) Bl and B2;

(3) Cl, C2. C3, and C4;

(4) Dl, D2, D3 and D4;

(5) El, E2 and E3;

(6) FI, F2, F3 and F4;

(7) Gl and G2;

(9) HI and H2; and (10) Rl, R2, R3, R4, R5. R6, R7 and R8 - Each of these repeated sequences has a length of between 2 and 45 amino acids.

The ability to bind Kappa light chains is considered to be associated with one or more of the repeated sequences A, B, C and D (sequences (1) - (4) above). —

It is thus a further feature of the invention to provide synthetic immunoglobulin binding molecules comprising a plurality of recognisably repeated binding domains selected from the sequences which are labelled at their N-terminal ends in Figure 1 as Al, A2 and A3; Bl, and B2; Cl, C2, C3. and C4; and Dl, D2, D3 and D4. The synthetic immunoglobulin binding molecules preferably comprise from 2 to 1 of said domains. The selected domain or domains may be identical to the sequences which are labelled at their N-terminal ends in Figure 1 as Al, A2 and A3; Bl, and B2; Cl, C2, C3, and C4; Dl, D2, D3 and D4, or they may vary from said sequences, provided that they have an at least 75%^". preferably at least 90% and most preferably at least 95% homology therewith. The sequences labelled at their N-terminal ends as El. E2 and E3; and FI. F2, F3 and F4 are believed to be responsible for albumin binding and the synthetic binding molecules provided according to the Invention may include sequences selected from sequences El, E2 and E3; and FI, F2, F3 and F4 or related squences which vary from said sequences, provided that they have an at least 75%. preferably at least 90% and most preferably at least 95% homology therewith.

The characterisation and isolation of the gene coding for Peptococcus Protein L and a comparison with the isolation of the protein by standard protein purification techniques will be described In the following Examples.

EXAMPLE 1 - Attempted Isolation and Characterisation of

Protein L By Standard Protein Purification Techniques

A collection of 56 different clinical isolates of Gram-positive anaerobic cocci were obtained from the Luton Public Health Laboratory Anaerobe Reference Unit, Bath Public Health Laboratory and Salisbury Public Health Laboratory. Type strains of peptococci and peptostreptococci were from the National Collection of Type Cultures, Central Public Health Laboratory, Colindale, London.

1.1 Preparation of cell extracts

The above strains were routinely cultured anaerobically in Todd-Hewitt broth supplemented with 0.05% Tween 80. Cells were harvested by centrifugation at 8,000 g, washed with 50mM potassium phosphate buffer, pH 6.8, and then resuspended in the same buffer to a cell density of 10% (wet weight/volume) . Lysis was by the addition of mutanolysin (Sigma; 50 units per ml of bacterial suspension) . After incubation for two to three hours at 37°C, the cell extracts were centrifuged at 12,000 g for 15 min and either stored as aliquots at -20°C or boiled in SDS PAGE loading buffer for 10 min (Harlow & Lane, 1988). 1.2 SDS PAGE and Western blotting of proteins

Proteins from the cell extracts were separated by electrophoresis on 7-5% polyacrylamide SDS gels, and electrophoretically transferred to nitrocellulose membranes (Hybond C; Amersham) by Western blotting (Harlow & Lane, 1988) . After overnight transfer, the membranes were washed several times in phosphate-buffered saline containing 0.02% Tween 20 (PBS-T) and then the protein binding sites were blocked by incubation in 1% gelatine (Sigma) in PBS-T for one hour. Immunoglobulin-binding proteins were then detected by incubation with either alkaline phosphatase-coupled human IgG or phosphatase-coupled light chain (prepared from human IgG by reduction, alkylation and separation by FPLC - Harlow and Lane, 1988) . Immunoglobulin-binding proteins were detected using nitro-blue tetrazolium and X-phos (Harlow & Lane, 1988) .

1.3 Affinity purification of Protein L

A 4-litre culture of strain 1018 was grown up, harvested and lysed as described above. The cell extract was heat-treated (80°C, 10 min) prior to the centrifugation step. The supernatant was made to 250 mM NaCl and applied to a 5ml column of IgG-Sepharose (15 mm diameter). The column was washed with 50 M Hepes-NaOH, pH 8.0, and immunoglobulin-binding proteins were eluted with lOOmM glycine-HCl, pH 2.0. The protein-containing eluate was neutralised with Tris-HCl to pH 7-5 prior to analysis by Western blotting.

The sizes of the immunoglobulin-binding proteins in the P. magnus isolates corresponds with the reported size of Protein L. In contrast to the results of Kastern et al , 1990, we found that the clinical isolates that had immunoglobulin-binding proteins were from wound isolates or infected surgical wounds.

The immunoglobulin-binding proteins were detected using an alkaline phosphatase-labelled light chain preparation from human IgG. 1. Conclusions.

These results confirm that the immunoglobulin-binding protein(s) are significantly degraded during the process of affinity purification, to yield smaller products that retain the ability to bind to IgG light chain.

EXAMPLE 2 - Determination of the Complete Nucleotide Sequence of

Protein L, and Comparison of its Translated Amino Acid Sequence with other Se uenced Immunoglobulin-binding Proteins.

2.1 Materials

Radiochemicals were from Amersham International. X-Omat S X-ray film was from Kodak. Deoxy- and dideoxy-nucleoside triphosphates, DNA ligase, restriction endonucleases and other DNA-modifying enzymes were from Boehringer. Agarose, acrylamide, bisacrylamide and phenol were from Bethesda Research Laboratories. Chromatography media were from Pharmacla-LKB (Uppsala, Sweden) . Human Immunoglobulins and serum albumin were from Sigma. All other reagents were from Sigma or BDH. NitroceHulose was purchased from Anderman and Co. , Kingston-upon- Thames, Surrey, U.K.

2.2 Media and Culture conditions.

E.coli TGI was cultured in 2xYT broth (2% (w/v) tryρtone/1% (w/v) yeast extract/1% (w/v) NaCl) overnight at 37^°C. Media were solidified with 2% (w/v) Bacto-agar (Difco) . HT-agar for M13 overlays contained 1% (w/v) tryptone, 0.8% (w/v) NaCl and 0.8% (w/v) Bacto-agar (Difco) . Ampicillin at a concentration of 50μg/ml was used where necessary for the selection and growth of transformants. Functional β-galactosidase was detected by addition of 5-bromo-4-chlorindolyl-β- D-galactoside to a final concentration of 600 μg/ml and, where necessary, isopropyl-β-D-thiogalactopyranoside to a final concentration of 200 μg/ml. Plasmids and phage RF DNA were purified from E. co l i by Brij lysis and CsCl/ethidium bromide density-gradient centrifugation. Peptococcus chromosomal DNA was isolated as described elsewhere.

2.3 Genetic Manipulation Procedures

DNA-modifying enzymes were used in the buffer and under the conditions recommended by the supplier (Boehringer) . Transformation of E. co li was essentially as described previously. Electrophoresis of DNA fragments was performed on vertical 1% (w/v)-agarose slab gels in Tris-acetate buffer (40 mM-Tris/20 mM-sodium acetate/2 mM-EDTA, adjusted to pH 7-9 with acetic acid). DNA fragment sizes were estimated by comparison with fragments of lambda phage DNA previously digested with the restriction endonuclease Hind III. DNA fragments were purified by electroelution essentially as described previously.

2.4 Nucleotide Sequencing

Nucleotide sequences were determined by the chain- ermination procedure on M13 templates using a shotgunning protocol to generate random templates. Multiple overlapping sequences were compiled using the programmes supplied by DNASTAR Inc (Maidison, U.S.A.). These same programmes were used to analyse the sequenced gene and its translated protein. Oligonucleotide primers were synthesised by using the Applied biosysterns 380B DNA synthesiser.

2.5 Sonication of cells

A cell suspension was transferred to a MSE sonication tube and subjected to ultra sonication (3^χ30 sec bursts at 18MH with 30 sec intervals, at 4 C using an MSE Soniprep 150 Sonicator) .

2.6 Affinity Chro atography on IgG-Sepharose B

A sonication procedure was used to disrupt bacterial cells from small scale purification of PPL by affinity chromatography on IgG-sepharose FF. Cultures of 300ml were grown overnight then centrifuged (15000 g for 10 min at 4^°C) and resuspended in 3 ml of 100 mM Tris-HCl, pH 7-5. 250 mM NaCl. The suspension was sonicated, centrifuged (30000 g 10 min at 4°C) and the supernatant fluid passed through a 1ml column (1.6 cm x 0.90 cm i.d.) of IgG-Sepharose FF equilibrated and washed with 5ml of 100 mM tris-HCl, pH 7-5. 250 mM NaCl. The protein was eluted with 100 mM glycine-HCl, pH 2.0, and the pH raised to 7-5 using 1M tris, pH 8.0.

2.7 PAGE

Samples were solubilised under reducing conditions and electrophoresed on SDS-polyacrylamide slab gels. Acrylamide (7-5%. w/v) slab gels were run in an LKB vertical electrophoresis unit using the method of Laemmll. Proteins were stained with Commassie Brilliant Blue R-250, and protein bands were scanned with a Chomoscan-3 laser optical densitometer (Joyce-Loebl, Gateshead, Tyne and Wear, U.K.), to estimate the apparent M .

2.7 Western blotting

Proteins were applied to nitrocellulose membranes by electrophoretic transfer from SDS-polyacrylamide gels and probed with

125 I labelled protein.

2.8 N-Teπninal Sequencing

N-terminal amino acid sequences were determined on an Applied Biosystems 77A pulsed ^"liquid protein sequencer by automated Edman - phenylthiohydantoin degradation. PPL samples were dialysed against 0 mM-NaCl, and about 500 pmol was applied to the gas-phase sequencer. The equipment was operated essentially according to the manu acturer's instructions. Repetitive Edman degradations provided sequential removal of amino acids from the peptide, which were identified by using reversed-phase HPLC.

2.9 Screening and identification of 31

From a partial Sa 2A digest of 3316 chromosomal DNA, fragments between 6.0 to 8.0 Kb were isolated by electro-elution and cloned into dephosphoryiated BamHl digested pMTL23- 1152 recombinant clones were picked into 12 microtitre trays and incubated at 37°C overnight. Addition of glycerol to a final concentration of 10% (w/v) in each well allowed the gene bank to be stored at -70^°C. To rapidly identify possible immunoglobulin-binding proteins, the clones were pooled into 24 lots of 48 (2 microtitre tray) and incubated shaking at 37^°C overnight. Soluble cell protein released by sonication was subjected

125 to western blotting with I radiolabelled human immunoglobulin light chain. From identified pools, the original clone stock was re-pooled into groups of 8 and re-probed. Finally, clones in positive pools were re-screened individually resulting in the identification of

3 human immunoglobulin light chain binding proteins.

2.10 Characterisation

7-5% SDS polyacryiamide gels of human IgG affinity purified samples for all 3 clones showed three major bands following commassie blue staining of 104, 96, and 90 Kdal. This same banding pattern was

125 also seen after I probing of western blots using crude sonicated samples with human IgG, IgA, IgM, IgD, IgE, and kappa light chain, but not wih lambda light chain. In addition, human serum albumin was also found to bind.

Knowing of the extreme stability of Protein A and G to heat, we tested the stability of Protein L by heating the crude sonicate for 10 min. at 80 C. Protein precipitated from solution by this treatment was discarded and the soluble fractions analysed by SDS-PAGE.

Commassie staining showed that most of the E. co li proteins had been precipitated but protein L remained in solution. By comparison with IgG-affinity purified Protein L, heat treatment appears to be an alternative quick method to purifying the protein. By western blot analysis, the protein still retains its activity for kappa light chains and HSA. It is interesting to note that smaller proteins presumed to be proteolysed fragments of Protein L only bind HSA in contrast to that found for Protein G. 2.11 DNA SEQUENCING

By DNA restriction mapping of the three clones, ρPPL9 was found to contain the smallest insert of 6.2Kb, and through sub-cloning the ppl gene was narrowed down to a 4.2Kb Pstl fragment. This Pstl fragment was subsequently excised, and sonicated by shotgunning.

Sequencing the inserts enabled the full DNA sequence depicted in Figure 1 to be derived and the associated amino acid sequence.

2.12 CONCLUSIONS

The pure homogeneous Protein L obtainable according to the invention can form the basis of numerous systems where Kappa light chain binding is desired. Thus Protein L can be used as a reagent for immobilising antibodies, e.g. on columns, in diagnostic tests and in assays. Additional uses are as pharmaceuticals and as reagents for preparing pharmaceuticals.

Important characteristics of Protein L according to the invention include:

(i) the ability to bing light chains, i.e. F .-binding as opposed to F -binding (ii) heat stability, i.e. to at least 8θ°C for 10 minutes, (iii) ability to bind human serum albumin (iv) a pK of about 4.68

Synthetic immunoglobulin binding molecules according to the invention can also form the basis of numerous systems where Kappa light chain binding is desired, e.g. test kits, biochemical reagents, protocols. In addition, these synthetic molecules may omit sequences selected from sequences El, E2, E3, FI, F2 and F3 so as to be substantially free of albumin binding ability.

Claims

1. A polypeptide designated Protein L and being capable of forming a complex with immunoglobulin Kappa light chains, said polypeptide being characterised by being in substantially homogenous and/or intact and/or full length form.

2. A polypeptide according to Claim 1 having at least two of the characterising features of (i) being substantially homogeneous, (ii) being intact and (iii) being in substantially full length form.

3- A polypeptide according to Claim 1 having at least three of the characterising features of (i) being substantially homogeneous, (ii) being intact and (iii) being in substantially full length form.

4. A polypeptide according to any preceding claim having substantially the amino acid sequence depicted in Figure 1.

5. A polypeptide according to any preceding claim being at least 900 amino acids in length.

6. A polypeptide according to any preceding claim being at least 950 amino acids in length.

7. A polypeptide according to any preceding claim being^{^}at least 975 amino acids in length.

8. A polypeptide according to any preceding claim being at least 990 amino acids in length.

9. A polypeptide according to any preceding claim being 992 amino acids in length.

10. A polypeptide according to any preceding claim having an N-terminal sequence corresponding to at least 7 amino acids of the N-terminal sequence Met Lys He Asn Lys Lys Leu.

11. A polypeptide according to any preceding claim having an N-terminal sequence corresponding to at least 10 amino acids of the N-terminal sequence Met Lys He Asn Lys Lys Leu Leu Met Ala.

12. A polypeptide according to any preceding claim having an N-terminal sequence corresponding to at least 15 amino acids of the N-terminal sequence Met Lys He Asn Lys Lys Leu Leu Met Ala Ala Leu Ala Gly Ala.

13- A polypeptide according to any preceding claim having an N-terminal sequence corresponding to at least 20 amino acids of the N-terminal sequence Met Lys He Asn Lys Lys Leu Leu Met Ala Ala Leu Ala Gly Ala He Val Val Gly Gly.

14. A polypeptide according to any preceding claim having at least 75% sequence homology, preferably at least 90% sequence homology with the amino acid sequence depicted in Figure 1.

15. A polypeptide according to any preceding claim having at least 95% sequence homology, preferably at least 98% sequence homology with the amino acid sequence depicted in Figure 1.

16. A polypeptide according to any preceding claim having a C-terminal sequence corresponding to at least 7 amino acids of the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser.

17. A polypeptide^according to any preceding claim having a C-terminal sequence corresponding to at least 10 amino acids of the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala.

18. A polypeptide according to any preceding claim having a C-terminal sequence corresponding to at least 15 amino acids of the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala Gly Ala Tyr Val Ser.

19. A polypeptide according to any preceding claim having a C-terminal sequence corresponding to at least 20 amino acids of the C-terminal sequence Leu Ala Ala Ala Ala Leu Ser Thr Ala Ala Gly Ala Tyr Val Ser Leu Lys Lys Arg Lys.

20. A polypeptide according to any preceding claim but omitting a signal sequence, especially wherein said signal sequence is between 20 and 35 amino acids in length, more specifically, between 23 and 27 amino acids in length.

21. A polypeptide according to Claim 20 commencing with one of the following N-terminal sequences:

Gly Ala Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Asn Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn

*Tyr Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Ala Glu Glu Asp Asn Thr Asp Asn Asn ...

Glu Glu Asp Asn Thr Asp Asn Asn ...

Glu Asp Asn Thr Asp Asn Asn ...

Asp Asn Thr Asp Asn Asn ...

Asn Thr Asp Asn Asn ...

Thr Asp Asn Asn ... and continuing with the sequence depicted in Figure 1 from Leu (315) •

22. A polypeptide according to Claim 20 commencing with the following N-terminal sequence:

Asn Leu Ser Met Asp Glu He Ser Asp Ala Tyr... and continuing with the sequence depicted in Figure 1 from Phe (343) •

23- --• recombinant DNA molecule having an insert coding for a polypeptide as defined in any preceding claim.

24. A recombinant DNA molecule according to Claim 23 selected from (a) the DNA coding sequence depicted in Figure 1 commencing with ATG in location 208 and extending to AAA in location 3183. and (b) degenerate DNA sequence coding for the same amino acid sequence.

25- A recombinant DNA molecule comprising a contiguous sequence of at least 12 bases depicted in the DNA sequence of Figure 1 and being useful as a probe for Isolating other DNA sequences coding for immunoglobulin binding protein.

26. A recombinant DNA molecule comprising a contiguous sequence of at least 15 bases depicted in the DNA sequence of Figure 1 and being useful as a probe for isolating other DNA sequences coding for Immunoglobulin binding protein.

27. A recombinant DNA molecule comprising a contiguous sequence of at least 17 bases depicted in the DNA sequence of Figure 1 and being useful as a probe for isolating other DNA sequences coding for immunoglobulin binding protein.

28. A process for producing an immunoglobulin binding protein which comprises culturing a transformed host, said host being transformed by an expression vector containing (a) a DNA coding sequence as defined in Claim 23 or Claim 24, or (b) a DNA coding sequence Isolated by probing a gene bank with a probe as defined in any of Claims 25 to 27.

29. A synthetic immunoglobulin binding molecule comprising a plurality of recognisably repeated binding domains selected from the sequences which are labelled at their N-terminal ends in Figure 1 as Al, A2 and A3; Bl, and B2; Cl, C2, C3, and C4; and Dl, D2, D3 and D4.

30. A synthetic immunoglobulin binding molecule according to Claim 29 comprising from 2 to 15 of said domains.

31. A synthetic immunoglobulin binding molecule according to Claim 29 or Claim 30 wherein the selected domain or domains are identical to the sequences which are labelled at their N-terminal ends in Figure 1 as Al, A2 and A3; Bl, and B2; Cl, C2, C3, and C4; and Dl, D2, D3 and D4, or vary from said sequences, provided that they have an at least 75%. preferably at least 90% and most preferably at least 95% homology therewith.

32. A synthetic immunoglobulin binding molecule according to any of Claims 29 to 31 additionally including domains selected from sequences El, E2 and E3; and FI, F2, F3 and F4 or related squences which vary from said sequences, provided that they have an at least 75%. preferably at least 90% and most preferably at least 95% homology therewith.

33. A synthetic immunoglobulin binding molecule according to any of Claims 29 to 32 wherein said domains are from 20 to 4 amino acids in length.