WO1995007300A1 - Methods and means for producing a plasmaprotein-binding protein - Google Patents

Abstract

Method for producing a protein having the serum albumin binding activity as protein MAG and for a recombinant DNA molecule coding for said protein (or fragments thereof), and for microorganisms containing this recombinant DNA molecule.

Description

Method and Means for producing a Plasmaprotein-Bindincr Protein

The invention relates to the field of gene technology and is concerned with recombinant DNA molecules, which contain a nucleotide sequence coding for a protein or polypeptide having a broader serum albumin-binding specificity than earlier reported albumin binding properties of protein G from the group G streptococcal strain G148 or other reported streptococcal albumin-binding proteins. Moreover the invention comprises microorganisms containing the aforesaid molecules, and the use thereof in the production of the aforesaid protein or polypeptide.

The existence of bacteria that bind specifically to serum albumin of different species has been known for more than a decade (Kronvall et al 1979, Myhre and Kronvall, 1980) . Based on the binding of serum albumins from different species Wideback et al 1983 classified five different classes (a-e) of bacterial cell surface receptors present on different serological groups and species of streptococci. The best studied serum albumin binding protein is protein G (originating from a human group G strain called G148) a protein shown to have both IgG and serum albumin binding activity ( Bjδrck et al 1987; Akerstrδm et al 1987). The gene encoding protein G from G148 has been cloned and characterized (Guss et al 1986; Olsson et al 1987) . Sjδbring et al 1989 reported on the structure and distribution of IgG-binding and albumin-binding domains among protein G genes. Further analysis of the serum albumin binding domains of protein G have been reported by Nygren et al 1988; Falkenberg et al 1992; Nygren 1992. The biotechnological use of the serum albumin binding domains of protein G has been documented in several publications and patent applications (WO 9220805; WO 9101743; EP 333691; EP 327522) . For example Nygren et al (1988) showed that the albumin binding domains of protein G could, if they were immobilized, be used to affinity purify serum albumins of different species. The use of the albumin-binding regions as affinity handle to be used as a fusion partner to foreign polypeptides in order to facilitate purification was reported by Nygren et al 1988,

SUBSTITUTESHEET H-ammarberg et al 1989. Dual expression systems including the albumin- binding domains were used to generate, analyse and purify antibodies to a repeated sequence of a Plasmodium falciparum antigen (Stahl et al 1989) . Furthermore it was shown that the albumin binding domains of protein G could be used to stabilize recombinant proteins in vivo, a discovery important in vaccin development (Nygren et al 1991) .

The use of the serum albumin binding domains of protein G in biotechnology has some drawbacks since not all albumins from different species bind to this protein (Nygren et al 1988; Nygren et al 1990; Bjδrck and Akerstrδm 1990) . For instance serum albumins from cow, goat, rabbit, sheep, chicken and horse bind poorly to the albumin binding part of protein G (Nygren et al 1990) or the class a receptors as described by Wideback et al 1983.The animal species from which these albumins originates are from biotechnological, agricultural and commercial aspects very important. In the vaccin development concerning these animal species it should be of great value to have an albumin binding protein with a broader albumin binding spectrum than reported for protein G. Also in the biotechnological field a more versatile albumin binding protein would be of value. For instance in several techniques such as cell culturing and hybridom techniques albumins are present in high amounts added as specific additives or present in a more complex solution such as fetal calf serum (Lillehoj and Malik 1993) . The presence of the albumins could in some cases be a problem when for instance the cell culture media contain products of interest. The presence of albumins in such a solution could be unfavourable since it could have a negative influence on the purification procedure for the product of interest (Lillehoj and Malik 1989, Schneider, 1989). Therefore a protein which for instance binds specifically to albumins from e.g. cow should be of great interest. Such a protein could in analogy to what Nygren et al 1988 reported be immobilized and used to remove the undesired albumins in the sample. Furthermore a protein (or a fragment thereof) with a broader albumin-binding spectrum could be used to detect and measure the amount of albumin in a sample. Wideback et al 1983 showed that the albumin

SUBSTITUTESHEET receptors expressed by S. dysgalactiae (type c) could be of interest. Generally it might be difficult to obtain a homogenious and reproducible product if such a receptor were prepared from streptococcal cells directly. Moreover streptococci are pathogenic and need complex culture media which involves complications in large-scale cultures. There is thus a need for a new method for producing an albumin-binding protein (or fragments thereof) with a broad binding spectrum. The present invention relates to a recombinant DNA molecule comprising a nucleotide sequence which codes for a protein or polypeptide having a broader albumin-binding activity than reported for the class c receptors and protein G or the albumin- binding protein described by Sjδbring 1992 or Raeder et al 1991. The natural source of this nucleotide sequence is of course the S. dysgalactiae strain 8215 but with the knowledge of the nucleotide and deduced amino acid sequence presented here the gene or parts of the gene can be isolated or made synthetically. In particular the knowledge of the deduced amino acid sequence for the part of the protein responsible for the albumin-binding activity can be used to produce syntethic polypeptides which retain or inhibit the albumin binding. These polypeptides can be labelled with various componds suchs as enzymes, fluorescence, biotin (or derivatives of), radioactivity, etc and used for e.i. in diagnostic tests such as ELISA- or RIA-techniques. For production of a recombinant DNA molecule according to the invention a suitable cloning vehicle or vector, for example a plasmid or phage DNA, may be cleaved with the aid of a restriction enzyme whereupon the DNA sequence coding for the desired protein or polypeptide is inserted into the cleavage site to form the recombinant DNA molecule. This general procedure is known per se, and various techniques for cleaving and ligating DNA sequences have been described in the literature (see for instance US 4,237,224; Ausubel et al 1991; Sambrook et al 1989), but to our knowledge these techniques have not been used for the present purpose. If the S. dysgalactiae strain 8215 is used as the source of the desired nucleotide sequence it is possible to isolate said sequence and to introduce it into a suitable vector

SUBSTITUTESHEET in manner such as described in the experimental part below or, since the nucleotide sequence is presented here, use a polymerase chain reaction (PCR)-technique to obtain the complete or fragments of the mag gene.

Hosts that may be used are, microorganisms (which can be made to produce the protein or active fragments thereof) , which may comprise bacterial hosts such as strains of e. g. Escherichia coli , Bacillus subtilis, Streptococcus sp., Staphylococcus sp., Lactobacillus sp. and furthermore yeasts and other eucaryotic cells in culture. To obtain maximum expression, regulatory elements such as promoters and ribosome-binding sequences may be varied in a manner known per se. The protein or active peptide thereof can be produced intra- or extracellularly. To obtain good secretion in various bacterial systems different signal peptides could be used. To facilitate purification and/or detection the protein or fragment thereof could be fused to an affinity handle and /or enzyme. This can be done on both genetic and protein level. To modify the features of the protein or polypeptide thereof the gene or parts of the gene can be modified using e.g. in vitro mutagenesis; or by fusion of other nucleotide sequences that encode polypeptides resulting in a fusion protein with new features.

The invention thus comprises recombinant DNA molecules containing a nucleotide sequence which codes for a protein or polypeptide having a broad serum albumin-binding activity. Furthermore the invention comprises vectors such as e.g. plasmids and phages containing such a nucleotide sequence, and organisms, especially bacteria as e.g. strains of E. coli, B. subtilis and Staphylococcus sp. , into which such a vector has been introduced. Alternatively, such a nucleotide sequence may be integrated into the natural gene material of the microorganism.

The application furthermore relates to methods for production of a protein or polypeptide having the broad albumin-binding activity of protein MAG or active fragments thereof. According to this method, a microorganism as set forth above is cultured in a suitable medium whereupon the resultant product is isolated by affinity chromatographic purification with the aid of IgG or

SUBSTITUTE SHEET albumin bound to an insoluble carrier, or by means of some other separating method, for example ion exchange chromatography. Vectors, especially plasmids, which contain the protein MAG encoding nucleotide sequence or parts thereof may advantageously be provided with a readily cleavable restriction site by means of which a nucleotide sequence that codes for another product can be fused to the protein MAG encoding nucleotide sequence, in order to express a so called fusion protein. The fusion protein may be isolated by a procedure utilizing its capacity of binding to serum albumin and/ or IgG, whereupon the other component of the system may if desired be liberated from the fusion protein. This technique has been described at length in WO 84/03103 in respect of the protein A system and is applicable also in the present context in an analogous manner. The fusion strategy may also be used to modify, increase or change the albumin -binding activity of protein MAG (or albumin binding part thereof) by fusion of other albumin-binding molecules.

Starting materials

Bacterial strains and cloning vectors

Streptococcus dysgalactiae strain 8215 was obtained from The National Veterinary Institute, Uppsala, Sweden.

E. coli strain DH5α was used as bacterial host for the plasmids to be constructed. E. coli strain P2392 was used in cloning with the lambda vector EMBL3 (Frischauf et al 1983) and as a host for expression of the lambda SD1 encoded plasma protein binding protein, termed protein MAG.

The lambda EMBL3. vector, E. coli strain P2392 and in vitro packaging extract (Gigapack ®II gold) were otained from Stratagene, La Jolla, CA, USA. The plasmid vector pGEMllzf (+) was obtained from Promega, Madison,WI. USA.. A protein fusion and purification system obtained from NewEngland Biolabs,USA was used to construct clones expressing various parts of the mag gene fused to the vector pMALC2. The Protein Fusion and Purification System obtained from New England Biolabs (Beverly, MA, USA) was used to construct the expression clones pMAGl-4. The system was

SUBSTITUTE SHEET used essentially according to the manufacturers recommendations but the host strain was E. coli strain DH5α.

All strains and plasmid- or phage-constructs used in the examples are available at the Department of Microbiology at the Swedish University of Agricultural Sciences, Uppsala, Sweden.

Buffers and media

E. coli was grown on LB (Luria Bertani broth) agar plates or in LB broth (Sambrook et al 1989) at 37 °C. Ampicillin was in appropriate cases added to the E. coli growth media to a final cone, of 50 μg/ml. Streptococci were grown at 37 °C on bloodagar- plates (containing 5% final cone, bovine blood) or in Todd-Hewitt broth (obtained from Oxoid, Ltd Basingstoke, Hants., England) supplemented with Yeast Extract (Oxoid) to a final cone, of 5%. PBS: 0,05M sodium phosphate pH 7,1, 0,9 % NaCl. PBS-T: PBS supplemented with TWEEN 20 to a final cone, of 0,05%.

Preparation of DNA from streptococci

S. dysgalactiae strain 8215 was grown overnight in Todd-Hewitt Broth supplemented with 0.6% yeast extract and 10 mM glycine. The next morning glycine was added to a cone. of 0.67 M and the incubation was continued at 37 °C for additional 2 hours. After harvest the cells were washed three times in a buffer consisting of 50 mM Tris-HCl pH 7.0 + 50 mM EDTA and resuspended to 1/20 of the original culture volume in the same buffer including 25% sucrose. Lysozyme ( Boeringer,Germany) was added to a final cone, of 30 mg/ml and the suspension was incubated with gentle agitation for 2 hours at 37 °C. The cells now converted to protoplasts were then pelleted by centrifugation and resuspended in buffer consisting of 50 mM Tris-HCl. pH 7.0 + 50 mM EDTA including 1% SDS (sodium dodecyl sulphate) and incubated at 65°C for 15 minutes. Cell debris was removed by centrifugation and the viscous supernatant further treated as described for chromosomal DNA preparations (Sambrook et al 1989) .

SUBSTITUTE SHEET Proteins and other reagents

Protein G Sepharose 4FF; IgG Sepharose 6FF; CNBr-activated Sepharose 4B was obtained from Pharmacia LKB Biotechnolo-gy, Uppsala, Sweden. BSA Agarose (A3790); BSA biotin labelled (A8549); serum albumins from dog (A 9263), horse (A 9888), pig (A 4414), rabbit (A 0764), rat (A 4538), sheep (A 4289), goat (A- 4162) and chicken ovalbumin. Grade VII (A7641) were obtained from Sigma, St. Louis, USA. Human serum albumin was obtained from Serva (cat. no.11860). Bovine serum albumin (fraction V, ria grade) was obtained from USB (cat. no.10868)

DNA probes were labelled with α³²P-ATP by a random-priming method (Multiprime DNA labelling system; Amersham Inc, Amersham, England)

Nitrocellulose(nc)-filters(Schleicher&Schϋll,Dassel,Germany) were used to bind DNA in hybridization experiments or proteins i dot- blot or Western-blot techniques.

In order to analyze protein samples by native or SDS-PAGE the PHAST-system obtained from Pharmacia LKB Biotechnology, Uppsala, Sweden was used according to the suppliers recommmendations. The preparation and labelling of serum albumin from goat or human and bovine fα₂M with horse-radish peroxidase has earlier been reported by Rantamaki and Mϋller (1992)

Routine methods

Methods used routinely in molecular biology are not described such as restriction of DNA with endonucleases, ligation of DNA fragments, plasmid purification etc since these methods can be found in commonly used manuals (Sambrook et al 1989, Ausubel et al 1991) . For polymerase chain reaction amplification the Gene Amp™ kit, obtained from Perkin Elmer Cetus, was used with the exeption that the Taq polymerase was replaced with Pfu DNA Polymerase (Stratagene) .

SUBSTITUTESHEET Example 1 Isolation of a lambda clone which codes for polypeptides possessing α,M-, BSA-, and IqG-bindinq activity

A gene library of S. dysgalactiae strain 8215 was produced in a manner analogous to that described by Frischauf et al (1983) . Streptococcal DNA was partially digested with Sau3Al and ligated into .BamHI-cleaved lambda EMBL3 vector arms. The ligated DNA was packaged in vitro into phage particles, which were then allowed to infect E. coli P2392 cells. The resultant phage library was analysed for α₂M-, BSA- and IgG-binding activity. The resultant phage library was analysed for this purpose on agar plates, in a soft agar layer, which were incubated overnight at 37 °C. The next day plates having a plaque frequency of 10³-10⁴ were selected. The plaques from each plate were transferred by replica plating to nc-filters. Transfer was allowed to proceed at room temperature for about 15 minutes. The filters were subsequently removed and soaked, using gentle agitation, for 30 minutes in a PBS-T solution (250ml/10 filter with three changes of the PBS- T solution in order to remove loosely bound material such as cell debris and components from the growth media. The filters were then sorted into three groups where replicas originating from the same agarplate were represented in each group. The respective group of filters were transferred to a petri dish which either contained approximately 10⁷ cpm of ¹²⁵I-labelled rabbit-IgG- antibodies (specific activity 7 mCi/mg) in PBS-T, approximately 10⁷ cpm of ¹²⁵I-labelled BSA (specific activity 7 mCi/mg) or approximately 10⁷ cpm of ¹²⁵I-labelled fα₂M (125 MBq/ g) . After incubation for 2 hours at room temperature using gentle agitation the respective group of filters was washed separately for 3X 10 minutes in 250 ml PBS-T at room temperatur (this washing procedure is important to reduce background signals and has to be prolonged if necessary) . After washing the filters were dried and autoradiographed for several days. Analysis of the autoradiogra revealed several plaques reacting with the three labelled (cc₂M, IgG and BSA) ligands. By comparing autoradiograms corresponding to the different ligands clones binding all three ligands were selected on the original agar plate and replaqued,

SUBSTITUTESHEET and the binding activties verified in another two rounds of binding assay as described above. Finally, one clone called "lambda SD1" expressing all three activities was chosen for the subsequent procedures. The description of and biotechnologocal use of the a₂M -binding activity of protein MAG is described in a copending patent application entitled "Methods and means to produce serum proteinas inhitor binding proteins" filed Sept. 6, 1993, Application No. SE 9302855-3.

Example 2. DNA sequencing

Purfied phage DNA from the 1-ambda SD1 clone was analysed by restriction mapping and a preliminary restriction map was constructed. After HindiII digestion of the lambda SD1 DNA, fragments were cloned into the pGEMUZf (+) previously cleaved with Hindlll. After ligation and transformation into E. coli strain DH5α recombinant clones were screened for expression of α₂M-,BSA- and/or IgG-binding activity. This was done as follows: clones were grown over night on nc-filters on agar plates. On the next day the nc-filters were replica plated to a masterplate whereupon the filters were incubated for 10 minutes in chloroform vapour in order to release the proteins from the bacterial cells. After washing the filters in a large excess of PBS-T (in this step it is important to reduce the bacterial debris attached to the filter to avoid unspecific binding of the labelled ligands used in the next step, thereby reducing the background signals) the filters were transferred to petri dishes containing respective ¹⁵I-labelled <X₂m, BSA or IgG respectively as in Example 1 above. After incubation for 2 hours at room temperature under gentle agitation the filters were washed in PBS-T, dried and autoradiographed as mentioned in Example 1. After 2 days of exposure of the filters the autoradiograms were analysed and clones expressing α₂M-binding activity were isolated. One such clone called pSDlOO, harbouring an approximately 0,9 kb Hindlll insert, was chosen for further studies. Clones expressing BSA-or IgG-binding activity could not be identified. In order to identify clones encoding the BSA and IgG-binding activity, a ³²P

SUBSTITUTESHEET labelled DNA (random-priming) probe homologous to the domain encoding the IgG-binding regions of protein G (Guss et al 1986) was used to identify the presence of homologous sequences among the recombinant plasmid clones. Among several Hin lll clones hybridizing to this probe, one called pSDlOl containing a approximately 0,7 kb insert was chosen for further studies.

Example 3. Sequence analysis

The nucleotide (nt) sequences of the Hindlll inserts of pSDIOO and pSDlOl were determined and the nt and deduced amino acid (aa) sequences were compared with the corresponding sequences from earlier published sequences of type III Fc receptors (Fahnestock et al. 1986, Guss et al. 1986, Olsson et al. 1987). This analysis revealed only short stretches of homology within the 862 bp HindHI insert of pSDIOO to the other type III Fc receptor genes while the 693 bp HindHI insert of pSDlOl was highly homologous to the earlier studied receptor genes. The combined size of the directly linked inserts of pSDIOO and pSDlOl containing the whole gene, called mag, is 1555 nt (see sequence list, page 27) . There is a potential ATG start codon at nt position 288 preceeded by a nt sequence resembling a ribosome-binding site. Upstream this site there are several potential promoter sequences. Starting at the ATG codon there is an open reading frame of 1239 nt terminating in a TAA stop codon at nt 1527. Thus the gene encodes a 413 aa protein, termed protein MAG, with a calculated molecular mass of approximatively 44 kDa including a putative signal peptide. The N-terminal part of the protein, constituting the signal peptide, shows a high degree of homology to the corresponding domains of the other type III receptors. A possible signal peptidase cleavage site should be after the alanine residue at aa position 34 (sequence list, page 27) . Downstream the signal peptide there is a unique strech of 158 aa. No repeated motifs can be seen in this part of the protein from strain 8215. Further to the C-terminal end the deduced aa sequence starting with -ALK- in position 193-195 shows homology to parts of the albumin-binding domain of protein G, the type III Fc receptor from the group G streptococcal strain G148 (Bjδrck

SUBSTITUTESHEET et al. 1987, Nygren et al. 1988). In the 8215 receptor this part is only 50 aa long and is directly followed by a region, which is highly homologous to the IgG-binding domains found in other type III Fc receptors. Also in the C-terminal part of the protein, which is responsible for anchorage of the protein to the bacterial cell surface, there is a striking sequence homology to other cell wall associated streptococcal proteins. This part of the protein can be divided into two structural and functional different domains. The N-terminal part, called W, (sequence list, page 27) is extremely hydrophilic and consists of mainly charged residues and prolines. This part of the protein probably mediates the attachment to the cell wall. Following the W-region, there is a hexapeptide LPTTGE, matching the consensus sequence LPXTGX commonly found in wall achored surface proteins of Gram-positive cocci (sequence list, page 27) . Downstream this motif there is a region of hydrophobic residues, called M, spanning the cell membrane, followed by a stretch of positively charged residues in the C-terminal end of the protein.

Example 4 (a) . Localization of binding domains in protein MAG.

On the basis of the sequence analysis we used a fusion protein expression system (obtained from Biolabs) to produce the various domains of protein MAG (Fig. 1A) . The high homology between the 3' end of the gene and the genes of earlier described type III Fc receptors strongly suggested that this part of the sequence encoded the IgG-binding activity. In Western-blot experiments we could show that this was indeed the case (see legends to Fig.l A, B and C. The fusion protein encoded by the construct pMAG3 showed no reactivity with α₂M but a strong signal was obtained with labelled albumin and IgG (Figs. IB and C) . This construct encodes an IgG-binding domain and also the 50 aa sequence upstream that domain which is partially homologous to the albumin-binding domains of protein G (Bjόrck et al. 1987, Nygren et al. 1988). The subclone pMAG4, encoding the same 50 aa and 10 aa from the IgG-binding domain, was reactive with albumin but not with IgG or fα₂M (Figs. IB and C) . In addition, the subclone pMAG2, encoding the unique 158 aa long stretch in the N-terminal part of the

SUBSTITUTESHEET protein reacted only with α₂M. The product resulting from expression of the clone pMAGl, (representing the whole gene), bound all three ligands.

(b) .Schematic presentation of the construction of the expression clones pMAGl-4.

pMAGl: lambda SDl was cleaved with PvuII. The approximately 1.1 kb PvuII fragment representing almost the complete mag gene was purified using preparative agarose gel-electrophoresis and ligated into the vector pMALC2 (the vector had previously been cleaved with EcσRI and the sticky ends converted to blunt ends using T4 DNA polymerase. pMAG2: the 870 bp Hindlll fragment from pSDIOO was purified by preparative agarose gel-electophoresis. The purified Hindlll- fragment was cleaved by PvuII and a part of the cleaved material was subsequently ligated into the pMALC2 vector ( the vector had earlier been cleaved with EcoRI and the generated sticky ends converted to blunt ends using T4 DNA polymerase. After inactivation of the T4 DNA polymerase the vector was cleaved with Hindlll) . pMAG3: the 670 bp Hindlll fragment from pSDlOl was purified by preparative agarose gel-electrophoresis. The purified fragment was cleaved with PvuII and made blunt end with T4 DNA polymerase. The cleaved material was ligated into pMALC2 (the vector had previously been cleaved with -BcoRI and -Ba-π-HI and the sticky ends converted to blunt ends using T4 DNA polymerase. pMAG4: the 670 bp Hindlll fragment from pSDlOl was purfied using preparative gel-electrophoresis. The purified fragment was cleaved with Hgal and the sticky ends converted to blunt ends using T4 DNA polymerase and ligated into the pMALC2 vector (the vector had earlier been cleaved with EcoRI and SamHI and converted to blunt ends using T4 DNA polymerase. After ligation and transformation into E. coli DH5α the generated clones were screened for binding activities as described above. Clones expressing various binding activities were identified and called pMAGl-4. The presence of the expected inserts in pMAGl-4 were verified by nt sequencing (including nt sequencing over the

SUBSTITUTESHEET junctions between vector and insert) .

Example 5. Purification of protein MAG using the lambda SDl clone

The lambda SDl phages was allowed to adsorb to cells of 3 individual tubes containing 3 ml E. coli P2392 (from an over night culture grown at 37°C in LB-medium supplemented with maltose final cone. 0,2% and MgCl₂ final cone. 10 mM) at a m.o.i. of -0,2 for 15 minutes at 37°C. After adsorption, the phage/bacteria solutions were transferred to three E-flasks, each containg 500 ml LB-medium prewarmed to 37°C. The E-flasks were shaken vigorously at 37°C until lysis occured. Two ml of chloroform was added to the respective flask and the flasks shaken for additional 10 minutes. The lysed cultures were subsequently centrifugated at ~15000 g for 40 minutes and the respective supernatants removed pooled and sterile filtered (using 0,45 urn nc-filters otained from Schleicher & Schύll) . The filtered media were passed over a column containing IgG-Sepharose 6FF (3 ml sedimentated gel which had previously been washed with PBS, preeluated with 1 M HAc pH 2,8, washed and equilibrated with PBS. The column was sequentially washed with 30 ml PBS, 60 ml PBS-T and 30 ml dH₂0. The bound protein material was eluted with 12 ml 1 M HAc pH 2,8. The eluted fraction was lyophilized and the dried material was dissolved in a TE-buffer (10ml Tris/HCl pH 7,5; ImM EDTA). The use of SDS-PAGE (under reducing conditions) using a 8-25 % gradient-gel in the PHAST-system (staining the gel with Coomassie brilliant blue) showed that the affinity purified material was of high yield and homogenious in size with a band having the relative migration corresponding to a protein of --45 kDa (using the low molecular weight marker kit obtained from Bio- Rad) . Thus the above described system is well suited to obtain protein MAG. Samples of. the affinity purified material were further analysed in several ways as described in Examples 6,7,8,9,10.

SUBSTITUTESHEET Example 6. Binding of protein MAG to serum albumins of different species.

From the binding of serum albumins of from different species to whole bacterial cells Wideback et al (1983) were able to classify streptococci into five classes regarding their binding patterns. According to this classification S. dysgalactiae belongs to class c, a class showing a broad binding spectrum. For example this class should bind efficiently to serum albumins from mouse, rat, cow and goat but poorly to serum albumins from humans and horse. In order to determine the serum albumin binding spectrum of protein MAG the following experiments were performed: albumins of different species were dissolved in PBS and twofold stepwise diluted in PBS in a microtiterplate. The sample in the first well contained ~10μg of the respective kind of albumin while the last well (no.8) contained -0,08. After dilution the samples (50 ul) were transferred to a nc-filter (previosly soaked in PBS) using a dot-blot apparatus (BioRad, CA, USA) . After applying the samples the dot-blot wells were washed twice with 100 μl PBS and the filter removed and placed in a PBS-T solution (150 ml) . After 30 minutes at room temperature, under gentle agitation, and changing of the PBS-T solution twice the filter was removed and placed in a PBS-T solution (25 ml) containing -50 μg of protein MAG. After incubation at room temperature for 1 hour with gentle agitation, the filter was transferred to a 150ml PBS-T solution to wash away any unbound protein MAG. After 30 minutes at room temperature using gentle agitation and three changes of PBS-T the filter was transferred to a 25 ml PBS-T solution containing goat- anti-rabbit IgG peroxidase conjugate (BioRad) diluted 1/1000. After incubation at room temperature for 1 hour with gentle agitation the filter was transferred to a PBS-T solution (150 ml) to wash away any unbound IgG. After 30 minutes at room temperature using gentle agitation and three changes of PBS-T solution the filter was finally washed twice for 5 minutes in PBS (150 ml) . To visualize the bound IgG conjugate the filter was transferred to a solution containing a substrate for peroxidase (containing 25 ml PBS+ 6 ml 4-chloro-l-naphtol (otained from Sigma, St. Louis, USA) 3mg/ml in methanol + 20 ul H₂0₂(35%). After

SUBSTITUTESHEET -30 minutes the degree of colour was measured by eye and the result is presented in Table 1.

As seen in Table 1 protein MAG binds efficiently to albumins of different species. Surprisingly protein MAG binds efficiently to albumins from both human and horses which is not in agreement with the class c receptor type reported by Wideback et al (1983) . This implies that protein MAG is not to be considered as a class c receptor type but instead must be classified into a new type which to our knowledge has not earlier been reported. Furthermore the use of the albumin binding properties of protein MAG has several biotechnological advantages compared to other earlier reported streptococcal albumin binding proteins. This since protein MAG binds efficiently to albumins from different species, which are important from agricultural, veterinary, biotechnological and commercial point of view.

To verify that the binding of human albumin to protein MAG is not due to contamination of IgG in the commercially obtained albumin sample, the following experiment was performed. A human serum albumin solution (250 μg in 150 μl PBS) was passed over a column containing protein G-Sepharose 4FF (300 μl sedimented gel, previously washed and equilibrated in PBS) The flow through was collected and recirculated over the column three times. After the last recirculation the flow through was collected and the column washed by the addition of 350 μl PBS. The washing solution, 350 μl PBS, was collected and pooled with the earlier collected 150 μl albumin solution. The amount of albumin in this solution was measured using the same dot-blot procedure as described above. The result of this test showed that the binding of protein MAG to human serum albumin solution was not due to any IgG contamination of the human serum albumin solution.

Example 7. The use of protein MAG in Western-blot techniques

Serum albumins from pig, rat, horse, cow, human, goat, mouse and chicken (-0,5 μg in PBS of respective albumin per well) were run under non-reducing (native) conditions using 8-25% gradient gel in the Phast-system (Pharmacia, Uppsala, Sweden) . After electrophoresis was completed a nc-filter previously soaked in

SUBSTITUTESHEET PBS was put on gel and the temperature raised to 45°C. After -45 minutes the nc-filter was wetted with 1 ml PBS and the filter was gently removed and placed in 15ml PBS-T solution for 30 minutes

(with two changes of the PBS-T solution) at room temperature under gentle agitation. The gradient-gel was after transfer removed and stained with Coomassie-blue using the PHAST-system. The nc-filter was subsequently removed and placed in a 15 ml PBS- T solution containing -10 μg purified protein MAG. After 1 hour at room temperature using gentle agitation the nc-filter was removed and washed in a 15ml PBS-T at room temperature using gentle agitation for 30 minutes (with two changes of PBS-T solution) . The filter was removed and placed in 15 ml PBS-T solution containing goat-anti-rabbit IgG (Bio-Rad Laboratories, Richmond, CA, USA. cat no.170-6515) dilution 1/1000 for 1 hour at room temperature using gentle agitation. The filter was subsequently removed and washed in 15 ml PBS-T solution for 30 minutes at room temperature using gentle agitation (with two changes of PBS-T solution) . Finally the filter was removed and washed twice in 15 ml PBS. To visulize the bound IgG conjugate the filter was transferred to a solution containing a substrate for peroxidase (containing 25 ml PBS + 6 ml 4-chloro-l-naphtol

(otained from Sigma, St. Louis , USA) 3mg/ml in methanol + 20 μl H₂0₂(35%) . After -30 minutes the degree of colour was measured by eye. The bands appearing on the nc-filter and the bands appearing on the corresponding Coomassie-blue stained PAGE was compared. The obtained result clearly showed that; (i) the relative band position on the nc-filter corresponded to the bands appearing on the PAGE with one exception (the exception was albumin from chicken which was seen on the PAGE but, as aspected, not on the nc-filter), (ii) protein MAG efficiently bound to all tested albumins of different species except albumin from chicken, (iii) it was the albumin from the respective albumin solution that was binding to protein MAG (with the exception of albumin from chicken) .

Using the same procedure as described above another type of experiment was performed in order to determine if protein MAG can bind reduced forms of albumins. The exception was that the samples were boiled 2 minutes in sample buffer (containing 2,5%

SUBSTITUTESHEET SDS and 5 % beta-mercaptoethanol) prior loading them on the PAGE. The samples were subsequently run under reducing conditions using SDS-buffer strips. After the run was completed the albumins were transferred to a nc-filter and treated according to the procedure described above while the SDS-PAGE after transfer was stained with Coomassie-blue. Finally the results showed that protein MAG bound poorly to the reduced forms of respective albumins except to albumin from chicken which showed no reactivity.

Example 8. Binding of biotinylated bovine serum albumin to immobilized protein MAG

Purified protein MAG was dissolved in PBS (lmg/ml) and serially diluted using PBS. Samples from the respective dilution step were transferred to a nc-filter using a dot-blot apparatus (BioRad) The nc-filter was washed with PBS-T and incubated in a solution of PBS-T containing biotinylated bovine serum albumin dilution 1/500. After 1 hour at room temperature using gentle agitation the filter was removed and washed in PBS-T. After washing the filter was incubated 1 hour at room temperature in PBS-T solution containing a streptavidin-horse radish peroxidase conjugate (Amersham,cat.no. RPN.1231) dilution 1/400. The filter was removed and washed in PBS-T and finally twice in PBS. The bound bovine serum albumin was visualized by transfering the filter to a solution containing a substrate for horse radish peroxidase (containing 25 ml PBS+ 6 ml 4-chloro-l-naphtol (obtained from Sigma, St. Louis , USA) 3mg/ml in methanol + 20 μl H₂0₂. After -30 minutes the degree of colour was measured by eye. The result clearly showed that biotinylated bovine serum albumin bound to immobilized protein MAG.

Example 9. The use of protein MAG to affinity purify bovine serum albumin from a complex solution

Purified protein MAG was coupled to CNBr-activated Sepharose 4B (Pharmacia LKB Biotechnology, Uppsala Sweden following the suppliers general recommendation described in the handbook called "Affinity Chromatography" principles & methods Pharmacia Fine

SUBSTITUTESHEET Chemicals) . Protein MAG immobilized on Sepharose 4B was subsequently used to affinity purify bovine serum albumin from a complex sample. The sample used was bovine blood (citrate) obtained from The National Veterinary Institute, Uppsala, Sweden. Prior to applying the blood on a column containing protein MAG- Seharose 4B the sample was treated as follows: 10 ml of blood was centrifugated at 12 OOOg for 2 minutes, the supernatant removed and centrifugated at 12 OOOg for 3 minutes, the supernatant taken care of, and 1 ml of this supernatant was diluted ten times using PBS. The diluted sample was passed over a protein MAG-Sepharose 4B column (-0,5 ml sedimentated gel) in order to affinity purify bovine serum albumin. After the sample had passed the column, the column was washed sequentially with 5 ml PBS, followed by 5 ml PBS-T, followed by 5 ml dH20. The bound material on the protein MAG-column was eluted by lowering the pH using 2,4 ml 0,5 M HAc pH 2,8. The eluted sample was lyophilized and the dried sample resuspended in dH₂0. The sample was analysed using reducing conditions on a SDS-PAGE using a 8-25% gradient gel. After the electrophoresis was completed the gel was stained with Coomasssie-blue. The result showed that the protein MAG efficiently bound bovine serum albumin (additional bands were seen corresponding to the heavy and light chain of IgG which is in agreement with the function of protein MAG) .

As a control experiment purfied bovine and chicken serum albumin (40 μg of each) were mixed in 1 ml PBS. The mixed albumin solution was passed over a protein MAG-Sepharose 4B column (-0,5 ml sedimentated gel) . The flow through was passed over the column four times to allow maximal binding. After the last passage the flow through was collected. Additional 0,5 ml PBS was added to the column to wash away any unbound material, and the obtained flow through was pooled to the earlier collected 1 ml flow through. The flow through material was desalted and lyophilized. The protein MAG-column was sequentially washed and eluted as described above. The eluted sample was lyophilized. The eluted sample and the sample corresponding to the flow through were resuspended in dH₂0 and analysed using reducing conditions on a 8-25% gradient-gel using PHAST-system (Pharmacia) . After elecrophoresis the gel was stained with Coomassie-blue. The

SUBSTITUTE SHEET result clearly showed that; (i) bovine serum albumin was efficiently bound and eluted from the protein MAG column and no detectable bovine serum albumin was found in the flow through fraction, (ii) the albumin from chicken did not bind to the column since the the chicken albumin was found in the flow through fraction.

Example 10. Binding of protein MAG to bovine serum albumin Agarose

Purified protein MAG (50 μg) was dissolved in 1 ml PBS. The protein MAG solution was passed over a column containing 0,5 ml bovine serum albumin Agarose (Sigma) previously washed and equilibrated with PBS. The flow through was collected and passed over the column two additional times. The column was sequentially washed with 5 ml PBS, 5 ml PBS-T, 5 ml dH₂0. The bound material was eluted with 2 ml 0,5 M HAc pH 2,8. The eluted sample was lyophilized and the dried material resuspended in dH₂0 and analyzed using SDS-PAGE as described above. The result showed that protein MAG was efficiently bound and eluted from the bovine serum albumin column.

Example 11. The use of polymerase chain reaction (PCR) to construct clones expressing various number of the albumin-binding domain of protein MAG.

The albumin-binding part of protein MAG was cloned using PCR. The Alb-region (sequence list, page 27) was amplified using two synthetic oligonucleotide primers. The upstream primer 5 -CGTAGACGCTTTAAAAGCTGCAGCAC-3 'was used. The underlined nucleotides correspond to nucleotides 864-882 in. the sequence list. The downstream primer 5 '-CGTCTACTGAAGCTAAAATTTCATCTTTAAG- 3 'was used. The underlined nucleotides correspond to nucleotides 1013-990 in the complementary strand in the sequence list. Plasmid pSDlOl was used as template. After PCR the amplified Alb- region was analysed and purified using agarose-gel electrophoresis. The amplified region was ligated into the plasmid vector pUC19 which had previosly been cleaved with

SUBSTITUTE SHEET Hindi . After ligation and transformation into E. coli DH5α several clones were isolated which contained the amplified fra-gment. One of these called pMAGALB(PCR) #1 was cleaved with AccI and the -160 bp Alb-fragment was isolated using agarose-gel electrophoresis. This fragment has the ability during ligation to be joined "head to tail" so that after ligation one, two, three etc Alb-regions are generated in a way that the correct reading frame is maintained. The -160 bp AccI fragment was ligated into AccI cleaved vector pEZZ18 (Pharmacia) . Previous to ligation the AccI cleaved vector was purified using agarose-gel electrophoresis in order to omit the DNA fragments coding for the Z domains. After ligation and transformation into E. coli DH5α cells transformants were screened for production of serum albumin binding activity. This was done as follows; clones were grown over night on nc-filters on agar plates. Next day the nc-filters were replica plated to a master plate whereupon the filters were incubated for 10 minutes in chloroform vapour in order to release the proteins from the bacterial cells. After washing the filters in large excess of PBS-T the filters were transferred to petri dishes containing horse-radish peroxidase labelled human serum albumin. After incubation for one hour at room temperature using gentle agitation the filters were washed and the bound conjugate was visualized as described in Example 6. In this way several positive clones were obtained (to verify that these clones did not contain any remaining Z fragments the clones were tested for the lack of IgG-binding activity) . Plasmid DNA from a number of positive clones (clones expressing albumin-binding) were prepared and double cleaved with NotI and Sphl. Analysis of samples using agarose-gel electrophoresis showed that the insert of the clones could be grouped mainly into three size classes corresponding to one, two or three copies of the PCR amplified Alb-region. One representant of each class was chosen for further studies. These clones was called pMAGALB(PCR)#3,#11 and #21 respectively. When plasmids from these clones where cleaved with AccI as aspected only one fragment of ~160bp was seen on an agarose gel. DNA sequence analysis also confirmed the presence of the PCR amplified Alb-fragment in these clones. The PCR amplified Alb-fragments are inserted in the modified

SUBSTITUTE SHEET plasmid pEZZlδ in a way that they are preceeded by a sequence encoding a signal sequence which allow secretion of the expressed Alb regions of pMAGALB(PCR)#3,#11 and #21 respectively. The clones pMAGALB(PCR)#3,#11 and #21 were separately grown in LB broth supplemented with ampicillin (50μg/ml, final cone.) over night. To affinity purify the expressed and secreted albumin binding polypeptides cell free growth media were passed over an column of human serum albumin(HSA) -Sepharose. The preparation, washing and elution of the column was as decribed for (BSA- Agarose)in Example 10. The eluted samples were lyophilized and the dried material resuspended in dH₂0 and analyzed using SDS-PAGE as described above. The result showed that pMAGALB(PCR)#3, #11 and #21 expresses extracellular polypeptides corresponding to one, two and three Alb-binding domains, respectively(Fig.2A) . To test if the affinity purified Alb-binding domains retain their albumin-binding activity after SDS-PAGE under reduced conditions the samples were transfered by Western-blot as described in Example 7 to a nc-filter. The filter was incubated with horse¬ radish peroxidase conjugated human serum albumin for one hour (room temperature under gentle agitation) . The filter was subse-quently washed with PBS-T and PBS and finally the bound conjugate was visulized according to Example 6. The result is shown in Fig. 2B. As seen the main bands binding the conjugate in the different lanes correspond to one, two and three domains of the Alb-region.

Table 1

Comparison of the binding of protein MAG to albumins of different species.

serum albumin from dilution μg

Pig 1/64 -0,16

Rabbit

Rat >1/128 <0,08

Sheep 1/32 ' -0,31

Dog >1/128 <0,08

SUBSTITUTESHEET (Table 1. continued)

Horse 1/64 -0,16

Cow 1/32 -0,31

Chicken

Human >1/128 <0,08

Goat 1/128 -0,08

Mouse >1/128 <0,08

Key to the table: The binding of protein MAG to different albumins expressed as the highest dilution step of detectable binding. The corresponding amount of albumin for the respective dilution step is written out. (-) no detectable binding.

After the submission of the priority application (Method and means for producing a plasma protein-binding protein, SE application no.9302856-1) , the plasma protein binding properties of protein MAG has been further characterized in a serie of publications which are a part of a doctoral thesis by one of the applicants, Hans Jonsson. These publications are; ¹¹ "Cell suface proteins of animal group C streptococci mediating binding of plasma proteins" by Hans Jonsson, Dissertation (1994) , Swedish University of Agricultural Sciences, Department of Microbiology, Uppsala Genetic Center, ISSN 0348-4041. ²¹ Jonsson, H. , Frykberg, L., Rantamaki, L. and Guss, B. MAG, a novel plasma protein receptor from -Streptococcus dysgalactiae. GENE 143(1994)85-89.

³⁾ Jonsson, H. , Burtsoff-Asp, C. -and Guss, B. Streptococcal protein MAG- a new type of albumin-binding protein with broad specificity, manuscript IV in ¹⁾ .

References

Ausubel,F.A. , Brent, R. , Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl. K. (eds) (1991) Current

SUBSTITUTESHEET Protocols in Molecular Biology, Greene Publishing and Wiley-

Intersciences, New York. Bjδrck, L.,Kastern, W. , Lindahl, G and Wideback, K. (1987)

Streptococcal proteinG, expressed by streptococci or by

Escherichia coli has separate binding sites for human albumin and IgG. Mol. Immunol. 24, 1113-1122. Bjδrck, L. and Akerstrδm,B. (1990) Streptococcal protein G. In

Bacterial Immunoglobulin-binding proteins. Boyle, M.P.D. (ed)

Academic Press Inc.CA,USA. 1, 113-126. Fahnestock, S.R., Alexander, P., Nagle, J. and Filipula,D. (1986)

Gene for an immunoglobulin-binding protein from a group G streptococcus. J. Bacteriol. 167, 870-880. Falkenberg,C. , Bjδrck, L. and Akerstrόm, B. (1992) Localization 6 the binding sites for streptococcal protein G on human serum albumin.Identification of a 5.5-kilodalton protein G binding albumin fragment. Biochem. 31,1451-1457. Guss, B. , Eliasson, M. , 01sson,A., Uhle'n, M. , Frej , A.-K.,

Jδrnvall,H. ,Flock, J.-I. and Lindberg, M. (1986) Structure of the IgG-binding regions of streptococcal protein G. EMBO

J.5,1567-1575. Frischauf, A.-M., Lehrach, H., Poustka; A. and Murray,N. (1983)

Lambda replacement vectors carrying polylinker sequences. J

Mol. Biol.170, 827-842. Hammarberg, B., Nygren, P.-A., Holmgren, E. , Elmblad, A., Tally,

M., Hellman, U. , Moks, T. and Uhle'n, M. (1989) Dual affinity fusion approach and its use to express recombinant human insulin-like growth factor II. Proc. Natl. Acad. Sci.

USA. 86,4367-4371. Laemmli, U.K. (1970) Cleavage of structural proteins during the assemly of the head of bacteriophage T4. Nature 227,680-685. Kronvall, G. , Simmons, A., Myhre, E.B. and Jonsson, S. (1979)

Specific adsorption of human serum albumin, immunoglobulin

A and immunoglobulin G with selected strains of group A and

G streptococci. Infect. Immun. 25, 1-10. Lillehoj , E.P. and Malik, V.S. (1989) Adv.

Biochem.Eng. /Biotechnol. 40,19-71. Lillehoj, E.P. and Malik, V.S. (1993) The new antibody technologies, in Advances in Applied Microbiology.38,149-

SUBSTITUTESHEET 209. Myhre,E.B. and Kronvall, G. (1980) Demonstration of specific binding sites for human serum albumin in group C and G streptococci. Infect. Immun. 27, 6-14. Meth.Enzymol.185 Nygren, P.-A. (1992) Characterization and use of the serum albumin binding region of streptococcal protein G. Thesis.

Department of Biochemistry and Biotechnology, Royal

Institute of Technology, Stockholm, Sweden. ISBN 91-7170-

087-0. Nygren, P.-A., Eliasson, M. Palmcrantz, E., Abrahmse'n, L. and

Uhle'n, M. (1988) Analysis and use of the serum albumin binding domains of streptococcal protein G. J. Mol.

Recognit. 1,60-74. Nygren, P.-A. ,Flodby, , P. Andersson, R., Wigzell,H. and Uhle'n,

M. (1991) in vivo stabilization of a human recobinant CD4 derivative by fusion to a serum albumin-binding receptor.

Vaccines ) 1, Modern approaches to vaccine development.

Chanock, R. M. et al (eds) Cold Spring Harbor Laboratory

Press, New York. USA 363-368. Nygren, P.-A., Ljungquist,C. , Trδmborg, H. , Nustad, K. and

Uhle'n, M. (1990) Species-dependent binding of serum albumins to the streptococcal receptor protein G. Eur. J. Biochem.

193, 143-148. Raeder, R., Otten, R.A. and Boyle, M.D.P. (1991) Comparison of albumin receptors expressed on bovine and human group G streptococci. Infect. Immun.59, 609-616. Rantamaki, L. K. and Miiller, H.-P. (1992). Isolation and characterization of a2-macroglobulin from mastitis milk.

J.Diary Res. 59, 273-285. Sambrook, J. , Fritsh, E.F.and Maniatis,T. (1989) Molecular cloning, A laboratory manual, second ed. Cold Spring Harbour

Laboratory Press, New York. Schneider, Y.-J. (1989) J. Immunol.Methods 166, 65-77. Sjδbring, U. (1992) Isolation and molecular characterization of a novel albumin-binding protein from group G streptococci.

Infect. Immun.60, 3601-3608. Sjδbring, U. , Bjδrck, L. and Kastern, W. (1989) Protein G genes:

SUBSTITUTESHEET structure and distribution of IgG-binding and albumin- binding domains. Mol. Immunol. 3,319-327.

Sjδbring, U., Bjόrck, L. and Kastern, W(1991) Streptococcal protein G: gene structure and protein binding properties. J Biol. Chem. 266, 399-405.

Sjδbring, U., Falkenberg, C. Nielsen, E., Akerstrόm, B. and Bjδrck,L. (1988) Isolation and characterization of a 14-kDa albumin-binding fragment of streptococcal protein G. J. Immunol. 140, 1595-1599.

Stάhl, S. ,Sjδlander, A., Nygren, P.-A., Berzins, K. , Perlmann, P and Uhle'n, M. (1989) A dual expression system for the generation, analysis and purification of antibodies to a repeated sequence of the Plasmodium falciparum antigen Pfl55/RESA. J. Immunol. Meth. 124, 43-52.

Wideback, K. (1987) Binding of albumin fragments to surface receptor in A, C and G streptococci. Acta path, microbiol. immmunol. scand. Sect B 95, 303-307.

Wideback, K. , Havlicek, J. and Kronvall, G. (1983) Demonstration of a receptor for mouse and human serum albumin in Streptococcus pyogenes. Acta path, microbiol. immunol. scand. Sect B. 91, 373-382.

Wideback, K. and Kronvall, G. (1987) Isolation of a specific albumin receptor from a group G streptococcal strain. Acta path, microbiol. immunol. scand. Sect B. 95,203-210.

Akerstrόm, B. , Nielsen, E. and Bjδrck, L. (1987) Definition of IgG- and albumin binding regions of streptococcal protein G J. Biol. Chem. 262, 13388-13391.

Patents or patent applications cited;

WO 84/03103

WO 9220805

WO 9101743

EP 333691

EP 327522

US 4,237,224

SUBSTITUTE SHEET Legends to the figures

Figure 1. (A) Schematic presentation of protein products coded by the expression clones pMAG 1-4 derived from S. dysgalactiae strain 8215. Restriction sites used in the construction work are indicated. The two upper lines represent inserts of streptococcal DNA in pSDIOO and pSDlOl. The upper bar represents protein MAG. (The Mai E portion is not drawn to scale) .

(B) SDS-PAGE of total cell lysates from clones harbouring pMAGl-4. Lanes: (M) molecular size markers, lanes 1-4 correspond to lysates from E. coli pMAGl-4,respectively. Methods: After IPTG induction the cells were harvested, lysed and protein samples were subjected to SDS-PAGE by the method of Laemmli (1970) using a 4% spacer- and 12% separation-gel. After electrophoresis the gel was stained with Coomassie Brilliant Blue, destained and photographed.

(C) Western-blot analysis of the proteins encoded by pMAGl-4. Three parallel gels were run using the same conditions as described in Figure 1(B) . After the electrophoresis the proteins were electrophoretically transfered to a nitrocellulose membranes. The membranes were washed and blocked in a PBS-T solution and separately incubated with horse-radish peroxidase labelled bovine fα₂M, goat albumin and polyclonal goat IgG, respectively. After incubation for one hour at room temperature the membranes were washed with PBS-T and the bound labelled serum proteins were visualized by the addition of a 4-chloro-l-naphtol solution.

Figure 2. (A) SDS-PAGE of HSA-Sepharose affinity purified albumin binding domains expressed by various E. coli clones. Lanes 1-3 corresponds to samples purified from pMAGALB(PCR). #3, #11 and #21, respectively. (B) Western blot analysis of HSA-Sepharose affinity purified albumin binding domains. Lanes 1-3 corresponds to samples purified from pMAGALB(PCR) #3, #11 and #21, respectively.

SUBSTITUTESHEET - α,-κ

Sequence list. Nucleotide sequence of the mag gene from S. dysgalactiae strain 8215 and the deduced aa sequence. Underlined, the putative transcription initiation signals. The ribosome binding site is double underlined. The start of the signal sequence (S) , α₂M-, Albumin- (Alb) and IgG-binding domains, respectively, are indicated as well as the cell wall binding (W) and membrane spanning (M) regions. In the C-terminal the LPXTGX motif is underlined. Methods: The nucleotide sequence was determined for both DNA strands by the dideoxy chain-termination method of Sanger et al. (1977) . The juxtaposition in lambda SDl of the inserts in pSDIOO and pSDlOl was verified using oligonucleotide primers (hybridizating on both sides of the Hindlll site) allowing sequencing over the Hindlll site in position 862. Sequence reactions were performed using "Sequenase, version 2.0" kit (United States Biochemical Corporation, Cleveland, Ohio, USA) . The obtained nucleotide sequences were analyzed using the PC/GENE computer software package, Intelligenetics Inc. CA, USA.

SUBSTITUTESHEET

Sequence list. Nucleotide sequence of the mag gene from S. dysgalactiae strain 8215 and the deduced aa sequence. Underlined, the putative transcription initiation signals. The ribosome binding site is double underlined. The start of the signal sequence (S) , α₂M-, Albumin- (Alb) and IgG-binding domains, respectively, are indicated as well as the cell wall binding (W) and membrane spanning (M) regions. In the C-terminal the LPXTGX motif is underlined. Methods: The nucleotide sequence was determined for both DNA strands by the dideoxy chain-termination method of Sanger et al. (1977) . The juxtaposition in lambda SDl of the inserts in pSDIOO and pSDlOl was verified using oligonucleotide primers (hybridizating on both sides of the Hindi11 site) allowing sequencing over the Hindlll site in position 862. Sequence reactions were performed using "Sequenase, version 2.0" kit (United States Biochemical Corporation, Cleveland, Ohio, USA) . The obtained nucleotide sequences were analyzed using the PC/GENE computer software package, Intelligenetics Inc. CA, USA.

SUBSTITUTESHEET AAGCTTTATTTTATTATAAAAGAAAGTAATTTTTGAAAAATTATAGAAAATCACTTTTAT 60 GCTAATAAAATAGCCATAAATATAAATTGATGAGTCTATGATAGGAGATTTATTTGCCAG 120

GATTTCCTAATTTTATTAATTTAACGAAAATTGATAGAAAAATTAAATGAAATCCTTGAT 180

TTAATTTTGTTAAGTTGTATAATAAAAGGTGAAATTATTAGATTGTAGTTTCAAATTTTT 240

→ S

TGGTTTTTTAATATGTGCTGGCGTATTAAATAAAAAAGGAGAAAGTAATGGAAAAAGAAA 300

M E K E 4

AAAAAGTAAAATACTTTTTACGTAAATCAGCTTTTGGATTAGCGTCCGTATCAGCTGCGT 360

K K V K Y F L R K S A F G L A S V S A A 24

→ a₂-M

TTTTAGTTGGGACTGCGGTAGTAAATGCGGAAGAGTCAACTGTTTCGCCTGTGACAGTTG 420

F L V G T A V V N A E E S T V S P V T V 44

CTACAGATGCAGTTACTACTTCTAAGGAAGCGCTTGCGATAATTAACAAGCTAAGTGAAG 480 A T D A V T T S K E A L A I I N K L S E 64

ATAATTTAAATAATCTTGACATCCAGGAAGTATTGGCCAAAGCGGGGAGGGACATTTTAG 540 D N L N N L D I Q E V L A K A G R D I L 84

CCTCTGACTCAGCAGATACTATCAAAGCACTTCTTGCTGAAGTTACCGCTGAAGTTACTC 600 A S D S A D T I K A L L A E V T A E V T 104

GTTTGAATGAGGAAAAGATGGCACGTGATGCAGTAGACAAAGCTATTGCAGCAGATGCAG 660 R L N E E K M A R D A V D K A I A A D A 124

CCGCTTTTTCTGAATTAAAAGATGCTCAACTGAAAGCATATGAAGATCTTGCGAAACTCG 720 A A F S E L K D A Q L K A Y E D L A K L 144

CAGCAGATACAGACTTAGATTTAGATGTTGCTAAAATTATAAATGACTACACTACAAAAG 780 A A D T D L D L D V A K I I N D Y T T K 164

TTGAAAATGCAAAAACAGCAGAAGATGTTAAAAAAATTTTTGAAGAATCTCAAAATGAAG 840 V E N A K T A E D V K K I F E E S Q N E 184

→ Alb

TGACACGTATTAAAACAGAAAAAGCTTTAAAAGCTGCAGCACTAGCTAAAGCAAAAGCAG 900

V T R I K T E K A L K A A A L A K A A 204 ATGCTATTGAAATTCTGAAGAAATACGGAATTGGCGATTACTATATTAAATTAATTAATA 960 D A I E I L K K Y G I G D Y Y I K L I N 224

→ IgG

ATGGTAAAACTGCAGAAGGTGTGACTGCTCTTAAAGATGAAATTTTAGCTTCAAAACCAG 1020

N G K T A E G V T A L K D E I L A S K P 244

CAGTGATTGACGCACCTGAATTAACACCAGCTTTGACAACCTACAAACTTGTTATCAATG 1080 A V I D A P E L T P A L T T Y K L V I N 264

GTAAAACATTGAAAGGCGAAACAACTACTAAAGCAGTAGACGCAGAAACTGCAGAAAAAG 1140 G K T L K G E T T T K A V D A E T A E K 284

CCTTCAAACAATACGCTAACGAAAACGGTGTTGATGGTGTTTGGACTTACGATGATGCGA 1200 A F K Q Y A N E N G V D G V W T Y D D A 304

→W

CTAAGACCTTTACTGTAACTGAAATGGTTACTGAAGTTCCTGGTGATGCACCAACTGAAC 1260

T K T F T V T E M V T E V P G D A P T E 324

CAAAAAAACCAGAAGCAAGTATCCCTCTTGTTCCGTTAACTCCTGCAACTCCAATTGCTA 1320 P K K P E A S I P L V P L T P A T P I A 344

AAGATGACGCTAAGAAAGACGATACTAAGAAAGACGATACTAAGAAAGAAGATGCTAAAA 1380 K D D A K K D D T K K D D T K K E D A K 364

AACCAGAAGCTAAGAAAGAAGAAGCTAAGAAAGCTGCAACTCTTCCTACAACTGGTGAAG 1440 P E A K K E E A K K A A T L P T T G E 384

→M

GAAGCAACCCATTCTTCACAGCTGCTGCGCTTGCAGTAATGGCTGGTGCGGGTGCTTTGG 1500

G S N P F F T A A A L A V M A G A G A L 404

CAGTCGCTTCAAAACGTAAAGAAGACTAATTGTCATTGCTTTTGACAAAAAGCTT 1555 A V A S K R K E D * 413

SUBSTITUTESHEET

Claims

Claims

1. Recombinant DNA molecule containing a nucleotide sequence coding for a protein or polypeptide having serum albumin binding specificity as protein MAG.

2. Plasmid or phage containing a nucleotide sequence coding for a protein or polypeptide having serum albumin binding specificity as protein MAG.

3. Microorganism containing at least one recombinant DNA molecule according to claim 1.

4. Microorganism containing at least one plasmid or phage according to claim 2.

5. Method for producing protein MAG or a polypeptide thereof, characterized in that

- at least one recombinant DNA molecule according to claim 1 is introduced in a microorganism,

- said microorganism is cultured in a suitable medium,

- the protein thus formed is isolated by affinity chromatographic purification.

6. Recombinant DNA molecule according to claim 1, characterized by containing one or more of the following nucleotide sequences:

AGCTTTAAAAGCTGCAGCACTAGCTAAAGCAAAAGCAGATGCTAT GAAATTCTGAAGAA

ATACGGAATTGGCGATTACTATATTAAATTAATTAATAATGGTAAAACTGCAGAAGGTGT

GACTGCTCTTAAAGATGAAATTTTAGCTTC-AAAACCAGCAGTGATTGACGCACCTGAATT

7. Recombinant DNA molecule according to claim 1, characterized by encoding one or more of the following sequences:

ALKAAALAKAKADAIEILKKYGIGDYYIKLINNG TAEGVTALKDEILAS

8. Plasmid or phage containing one or more nucleotide sequences according to claim 6.

9. Microorganism containing at least one plasmid or phage according to claim 7.

10. The use of immobilized protein MAG or fragments thereof to isolate plasma components from complex solutions such as blood.

11. The use of immobilized protein MAG or fragments thereof to isolate serum albumins from complex solutions such as blood.

12. The use of immobilized serum albumins to remove protein MAG or fragments or derivatives thereof from a solution.

13. The use of protein MAG or fragments thereof to detect and/or quantify blood components.

14. The use of protein MAG or fragments thereof to detect and/or quantify serum albumins.