PL203645B1 - MAGE family tumor-associated antigen derivative and use, nucleic acid sequence and use thereof, vector, host cell, vaccine, method of purifying MAGE protein or a derivative thereof, and method of making a vaccine - Google Patents

Description

Description of the invention The invention relates to a derivative of a tumor-associated antigen from the MAGE family and its use, a nucleic acid sequence and its use, a vector, a host cell, a vaccine, a method for purifying the MAGE protein or its derivative, and a method for producing a vaccine. Generally, the invention relates to protein derivatives including tumor-associated antigens that find use in anticancer vaccine therapy. Derivatives of the invention include fusion proteins comprising an antigen encoded by the MAGE gene family (e.g., MAGE-3, MAGE-1), fused to an immunological fusion component that provides epitopes to T helper cells, such as, for example, lipidated form of protein D from Haemophilus influenzae B; chemically modified MAGE proteins in which the antigen's disulfide bonds are reduced to form blocked thiols, and genetically modified MAGE proteins equipped with an affinity tag and/or genetically modified to prevent the formation of disulfide bridges. Also described are methods for purifying the MAGE protein and formulating vaccines for the treatment of numerous cancers, including, but not limited to, melanoma, breast cancer, bladder cancer, lung cancer, NSCLC, squamous cell carcinoma of the head and neck, colon cancer and gulp. The antigens encoded by the MAGE gene family are mainly expressed on melanoma cells (including malignant melanoma) and some other cancers, including NSCLC (non-small cell lung cancer), head and neck cancers, transitional to bladder cancer and esophageal cancer, but are not detected in normal tissues except the testes and gland beds (Gaugler, 1994; Weynants, 1994; Patard, 1995). MAGE-3 is expressed in 69% of melanomas (Gaugler, 1994) and can be found in 44% of NSCLC (Yoshimatsu, 1988), 48% of head and neck squamous cell carcinomas, 34% of transitional cell carcinomas of the bladder, 57% esophageal cancers, 32% of colon cancers and 24% of breast cancers (Van Pel, 1995); Inoue, 1995; Fujie, 1997; Nishimura, 1997). Tumors expressing MAGE are known as MAGE-related tumors. The immunogenicity of human melanoma cells has been demonstrated in experiments using mixed cultures of melanoma cells and autologous lymphocytes. These cultures often produce cytotoxic T lymphocytes (CTLs) that are capable of lysing only autologous melanoma cells, but not autologous fibroblasts or EBV-transformed B cells (Knuth, 1984; Anichini, 1987). Currently, some of the antigens recognized by CTL clones on autologous melanoma cells have been identified, including antigens belonging to the MAGE family. The first antigen that could be identified by being recognized by specific CTLs on autologous melanoma cells was termed MZ2-E (Van den Eynde, 1989) and is encoded by the MAGE-1 gene ( Van der Bruggen, 1991). CTLs directed against MZ2-E recognize and degrade autologous MZ2-E-positive melanoma cells as well as those from other patients, assuming that these cells have HLA.A1 alleles. The MAGE-1 gene belongs to a family of 12 closely related genes, MAGE-1, MAGE-2, MAGE-3, MAGE-4, MAGE-5, MAGE-6, MAGE-7, MAGE-8, MAGE-9, MAGE -10, MAGE-11 and MAGE-12, located on the X chromosome and sharing 64 to 85% coding sequence homology (De Plaen, 1994). These are sometimes referred to as MAGE A1, MAGE A2, MAGE A3, MAGE A4, MAGE A5, MAGE A6, MAGE A7, MAGE A8, MAGE A9, MAGE A10, MAGE A11 and MAGE A12 (MAGE A family). Two other groups of proteins are also part of the MAGE family, although more distantly related. These are the MAGE B and MAGE C groups. The MAGE B family includes MAGE B1 (also known as MAGE Xp1 and DAM 10), MAGE B2 (also known as MAGE Xp2 and DAM 6), MAGE B3 and MAGE B4 family MAGE C currently includes MAGE C1 and MAGE C2. In general terms, a MAGE protein can be defined as containing a core sequence signature located towards the C-terminus of the protein (for example, in the case of MAGE A1, a protein of 309 amino acids, the core signature corresponds to amino acids 195-279). . The core signature agreement pattern is described below, where x represents any amino acid, lowercase a residues are conserved (conservative variants are allowed), and uppercase a residues are completely conserved. Core sequence signature: LixvL(2x)I(3x)g(2x)apEExiWexl(2x)m(3-4x)Gxe(3- 4x)gxp(2x)11t(3x)VqexYLxYxqVPxsxP(2x)yeFLWGprA(2x)Et( 3x)kvPL 203 645 B1 3 Conservative substitutions are well known and provided as default evaluation matrices in sequence alignment computer programs. These programs include PAM250 (Dayhoft et al., (1978), "A model of evolutionary changes in proteins" in "Atlas of Protein sequence and structure", Dayhof (ed.) 345-352), National Biomedical Research Foundation, Washington and Blosum 62 (Steven Henikof and Jorja Henikof (1992), "Amino acid substitution matrices from protein blocks") Proc. Natl. Acad. Sci. USA 89 (Biochemistry): 10915-10919. Generally, substitution in the following groups is a conservative substitution, but substitutions between groups are not considered conservative. The groups are a i) aspartate/asparagine/glutamate/glutamine ii) serine/threonine iii) lysine/arginine iv) phenylalanine/tyrosine/tryptophan v) leucine/isoleucine/valine/methionine vi) glycine/alanine. Generally and in the context of the invention, the MAGE protein should be at least 50% identical in the core region to amino acids 195 to 279 of MAGE A1. Some CTL epitopes have been identified in the MAGE protein -3. One of these epitopes, MAGE- -3. A1 is a nonapeptide sequence located between amino acids 168 and 176 of the MAGE-3 protein, which constitutes a CTL-specific epitope when presented in association with a molecule MHC class I HLA.A1. Recently, two additional epitopes were identified in the MAGE-3 peptide sequence due to their ability to induce CTL responses in a mixed culture of melanoma cells and autologous lymphocytes. These two epitopes display a specific binding motif for the HLA.A2 (Van der Bruggen, 1994) and HLA.B44 (Herman, 1996) alleles, respectively. The invention provides a derivative of a tumor associated antigen from the MAGE family which is selected from the group consisting of: - a MAGE antigen comprising derivatized thiol groups; - a recombinantly expressed MAGE fusion protein, wherein the fusion component is selected from the group consisting of protein D from Haemophilus influenzae B or a fragment thereof, protein NS1 from influenza virus or a fragment thereof, or C-LYTA from Streptococcus pneumoniae or thereof fragment. Such derivatives are useful for therapeutic applications in vaccine formulations that are used in the treatment of many types of cancer. Therefore, the invention also relates to the use of the protein according to the invention for the production of a vaccine for the immunotherapeutic treatment of a patient suffering from melanomas or other MAGE-related cancers. In one embodiment of the invention, the derivative is a fusion protein comprising an antigen from the MAGE protein family combined with a heterologous component. The proteins can be chemically combined, but are preferably expressed as recombinant fusion proteins, which allows for their higher production in the expression system compared to the unfused protein. Thus, the fusion component may provide a T cell epitope (immune fusion component), preferably a T helper cell epitope recognized in humans, or facilitate the expression of a protein (expression enhancer) at a high level, as compared to native recombinant protein. The fusion component may be both an immunological fusion component and an expression enhancing component. In one embodiment of the invention, the immunological fusion component is derived from protein D, a surface protein of the Gram-negative bacterium, Haemophilus influenzae B (WO91/18926). The protein D derivative may, for example, comprise approximately the first 1/3 portion of the protein, in particular about the first 100-110 amino acids of the N terminus. Preferably, the protein D or fragment thereof is lipidated. For example, the first 109 residues of the lipidoprotein D fusion component are fused to the N terminus to provide a potential vaccine antigen, with an additional exogenous T-cell epitope, and increased levels of expression in E. coli (so they function also as an expressive enhancer). The lipid tail ensures optimal presentation of the antigen to antigen-presenting cells. Other fusion components include the influenza virus NS1 nonstructural protein (hemagglutinin). Typically, the N-terminal 81 amino acids are used, although different fragments may be used, assuming they contain T helper cell epitopes. The immunological component of the fusion may also be a protein called LYTA, wherein a fragment of the C-terminus of the protein is used. Lyta comes from Streptococcus pneumoniae, which synthesizes N-acetyl-L-alanine amidase and LYTA amidase (encoded by the lytA gene, (Gene 43 (1986) pp. 265-272)), PL 203 645 B1 4 autolysin e, which specifically degrades certain bonds in the peptidoglycan skeleton. The C-terminal domain of the LYTA protein is responsible for the affinity for choline or some choline analogues, such as DEAE. This property was used to develop E. coli - LYTA expression plasmids useful for the expression of fusion proteins. The purification of hybrid proteins containing a C-LYTA fragment at the amino terminus has been described (Biotechnology: 10 (1992) pp. 795-798). It is possible to use a part containing the C-terminal repeats of the LYTA molecule, starting from residue 178, in particular the form comprising residues 188-305. The immunological components of the fusion described above are also beneficial in promoting expression. In particular, such fusions are expressed with greater efficiency than recombinant native MAGE proteins. Such constructs have been shown to be capable of treating melanoma in clinical applications. In one case, in a patient with stage IV melanoma, after two doses of non-adjuvanted lipo D 1/3 MAGE 3 His fusion protein, the metastases were removed. Thus, the present invention in one embodiment provides fusion protein drugs comprising a tumor-associated antigen from the MAGE family in combination with an immunological fusion component. The immunological component of the fusion is protein D or a fragment thereof, most preferably lipoprotein e D. The lipoprotein e D portion, for example, may comprise the first 1/3 of lipoprotein D. The MAGE proteins in the antigen derivative according to the invention are preferably MAGE A1, MAGE A2, MAGE A3, MAGE A4, MAGE A5, MAGE A6, MAGE A7, MAGE A8, MAGE A9, MAGE A10, MAGE A11, MAGE A12, MAGE B1, MAGE B2, MAGE B3, MAGE B4, MAGE C1 and MAGE C2. The proteins of the invention can be expressed in E. coli and can be expressed with a histidine "tail" comprising from 5 to 9 and preferably 6 histidine residues. This is particularly useful for purification. The invention also provides nucleic acid encoding the fusion proteins of the invention. Such sequences can be introduced into appropriate expression vectors and used for DNA/RNA vaccination or for expression in a suitable host. Appropriate vaccines can be used microbial vectors expressing nucleic acid. Such vectors include, for example, poxvirus, adenovirus, alphavirus, listeria and monarphage. A DNA sequence encoding the fusion proteins of the invention can be synthesized using standard DNA synthesis techniques, such as enzymatic ligation as described by Roberts et al., Biochemistry 1985 24:5090-5098, chemical synthesis, in vitro enzymatic polymerization, or PCR technology using, for example, a thermostable polymerase, or by a combination e of these techniques. Enzymatic DNA polymerization can be performed in vitro using DNA polymerase, such as DNA polymerase I (Klenow fragment) in an appropriate buffer containing nucleoside triphosphates dATP, dCTP, dGTP and dTTP, at a temperature of 10-37°C , generally in a volume of 50 µl or less. Enzymatic ligation of DNA fragments can be performed using a DNA ligase such as T4 DNA ligase in a suitable buffer such as 0.05 M Tris (pH 7.4), 0.01 M MgCl2, 0.01 M dithiothreitol, 1 mM spermidine, 1 mM ATP and 0.1 mg/ml bovine serum albumin, at a temperature of 4°C to ambient temperature, generally in a volume of 50 µl or less. Chemical synthesis of polymer or DNA fragments can be accomplished by conventional phosphothioester, phosphite, or phosphite amide reactions using solid-state techniques such as those described in Chemical and Enzymatic Synthesis of Gene Fragments - A Laboratory Manual (ed. . Gassen and Lang), Verlag Chemie, Weinheim (1982) or in other scientific publications, for example, M.J. Gait, H.W.D. Matthes, M. Singh, B.S. Sproat, and R.C. Titmas, Nucl. Acid Res. 1982 10:6243; B.S. Sproat and W. Bannwarth, Tetrahedron Lett. 1983 24:5771; M.D. Matteuci and M.H. Caruthers, Tetrahedron Lett. 1980 21:719; M.D. Matteuci and M.H. Caruthers, J. Am. Chem. Soc. 1981 103:3185; S.P. Adams et al., J. Am. Chem. Soc. 1983 105:661; N/A Sinha, J. Biernat, J. McMannus, and H. Koester, Nucl. Acid Res. 1984 12:4539, and H.W.D. Matthes et al., EMBO J. 1984 3:801. The invention also relates to the use of the nucleic acid according to the invention for the production of a vaccine for immunotherapeutic treatment of a patient suffering from melanomas or other MAGE-related cancers. The antigen derivative of the invention can be obtained by conventional recombinant techniques as described in Maniatis et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor 1982-1989.PL 203 645 B1 5 In particular, the method for producing the white the fusion drug may include the steps of: i) producing a replicable or integration-competent expression vector capable in a host cell of expressing a DNA polymer comprising a nucleotide sequence that encodes the protein or an immunogenic derivative thereof ii) transforming the host cell with such a vector; iii) culturing the transformed host cell under conditions that allow expression of the DNA polymer to produce the protein; and iv) protein acquisition. The term "transforming" as used herein means introducing foreign DNA into a host cell. This may be achieved, for example, by transformation, transfection or infection with an appropriate plasmid or viral vector, using e.g. conventional techniques such as described in Genetic Engineering; eds. S. M. Kingsman and A. J. Kingsman; Blackwell Scientific Publications; Oxford, England, 1988. The term "transformed" or "transformant" will refer to the resulting host cell containing and expressing occupying a foreign gene of interest. The expression vectors comprising the nucleic acid of the invention are novel and also form part of the invention. Replication-competent expression vectors can be produced according to the invention by cutting a vector compatible with the host cell to obtain a linear DNA segment with a complete replicon and linking such linear segment, under ligation conditions, with one or more DNA molecules, such as a DNA polymer encoding a protein according to the invention or thereof derivative, which, together with the linear segment, encode the given product. Thus, the DNA polymer can be obtained in advance or during vector construction, as desired. The choice of vector will be determined in part by the host cell, which may be prokaryotic or eukaryotic, but is preferably an E. coli or CHO cell. Suitable vectors include plasmids, bacteriophages, cosmids and recombinant viruses. The production of a replicable expression vector can be performed conventionally using appropriate enzymes for DNA restriction, polymerization and ligation, procedures described, for example, in Maniatis et al., cited above. The invention also relates to a recombinant host cell transformed with a nucleic acid or vector according to the invention. A recombinant host cell is produced by transforming the host cell with a replication-competent vector according to the invention under transforming conditions. Suitable transformation conditions are conventional and are described, for example, in Maniatis et al., cited above, or in DNA Cloning, Vol. II, D.M. Glover (ed.), IRL Press Ltd., 1985. The choice of transforming conditions is determined by the host cell. Thus, a bacterial host such as E. coli may be treated with a solution of CaCl2 (Cohen et al., Proc. Natl. Acad. Sci. USA 1973 69:2110) or a solution comprising a mixture of RbCl2, MnCl2, potassium acetate and glycerol, and then with 3-[N-morpholino]-propane-sulfonic acid, RbCl and glycerol. Mammalian cells in culture can be transformed by calcium precipitation of the vector DNA into the cells. Culturing the transformed host cell under conditions permitting expression of the DNA polymer may be carried out in a conventional manner as described, for example, in Maniatis et al., supra, or in "DNA Cloning", cited above. Thus, media is provided and cultured at a temperature below 50°C. The product is obtained by conventional methods, depending on the host cell and the location of the expressed product (intracellular or product secreted into the culture medium or cell periplasm). .Thus, when the host cell is a bacterium such as E. coli, it can be lysed physically, chemically, or enzymatically, and the protein product isolated from the resulting lysate. When the host cell is a mammalian cell, the product can generally be isolated from culture medium or cell-free extracts. Conventional protein isolation techniques include selective precipitation, adsorption chromatography, and affinity chromatography, including monoclonal antibody affinity columns. The proteins of the invention may be provided either in soluble liquid form or in lyophilized form. It is generally expected that a human dose will contain from 1 to 1000 µg of protein, preferably 30- 300 µg. The subject of the invention is also a protein-containing vaccine according to the invention. The vaccine composition comprises the protein of the invention in a pharmaceutically acceptable excipient. The vaccine composition of the invention may include at least a lipoprotein D - MAGE-3 fusion. Such a vaccine preferably additionally contains one or more antigens. These antigens may be associated with cancer, for example, they may be other members of the MAGE and GAGE families. Other suitable tumor-associated antigens include MAGE-1, GAGE-1 or tyrosinase proteins. Vaccine production is described generally in Vaccine Design - The Subunit and Adjuvant Approach (Powell and Newman eds.), Pharmaceutical Biotechnology vol. 6, Plenum Press, 1995. Encapsulation in liposomes is described by Fullerton, US Pat. No. 4,235,877. Preferably, the vaccine according to the invention additionally contains an adjuvant and/or an immunostimulating cytokine or chemokine. Moreover, the vaccine according to the invention preferably contains the protein in a carrier consisting of an oil-in-water emulsion. Also preferably the adjuvant comprises 3D-MPL, QS21 or a CpG oligonucleotide. Suitable adjuvants include aluminum salts such as aluminum hydroxide gel (alum) or aluminum phosphate, but may also be a calcium, iron or zinc salt, or may be an insoluble suspension of acylated tyrosine or acylated sugars , cationic or anionic derivatives of polysaccharides, or polyphosphazenes. Other known adjuvants include CpG-containing oligonucleotides. Such nucleotides are characterized by the fact that the CpG dinucleotide is unmethylated. Such oligonucleotides are well known and described, for example, in WO 96/02555. The adjuvant composition should selectively induce TH1-type reactions. Suitable adjuvant systems include, for example, a combination of monophosphoryl lipid A, particularly 3-de-O-acylated monophosphoryl lipid A (3D-MPL) together with aluminum salts. CpG oligonucleotides so that they preferentially induce a TH1 response. The enhanced system includes a combination of monophosphoryl lipid A and a saponin derivative, in particular a combination of QS21 and 3D-MPL as disclosed in WO 94/00153, or a less reactogenic composition in which QS21 is quenched with cholesterol as disclosed in WO 96/33739. A particularly potent adjuvant formulation comprising QS21, 3D-MPL and tocopherol in an oil-in-water emulsion is described in WO 95/17210, and is a preferred formulation. According to one embodiment of the invention, vaccines are provided comprising a protein according to the invention, more preferably a fusion of lipoprotein D (or its derivative) - MAGE-3, supported by mono-phosphoryl lipid A or its derivative. Preferably, the vaccine additionally contains saponin E, more preferably QS21. The formulation may additionally include an oil-in-water emulsion and tocopherol. Also described herein is a method of formulating a vaccine comprising mixing a protein of the invention with a pharmaceutically acceptable excipient, such as 3D-MPL. The invention further provides a vaccine for use as a medicament for the treatment of MAGE-related diseases. Also described herein is a method for purifying recombinantly produced MAGE protein. The method includes solubilizing the protein, for example, in a strongly chaotropic agent (such as, for example, urea, guanidine hydrochloride) or a zwitterionic detergent, e.g. (Empigen BB - n-dodecyl-N,N-dimethylglycine), reducing intra- and inter-capillary disulfide bonds, blocking thiols to prevent them from recombining as a result of oxidation and subjecting the protein to one or more chromatography steps. The blocking agent may be an alkylating agent. Such blocking agents include, but are not limited to, alpha-halogen acids and alpha-haloamides. For example, iodoacetic acid and iodoacetamide cause carboxymethylation or carboxamidation (carbamidomethylation) of proteins. Other blocking agents that may be used have been described in the literature (see, for example, The Proteins, vol. II, eds. H. Neurath, R.L. Hill and C.-L. Boeder, Academic Press 1976, or Chemical Reagents for Protein Modification, Volume I, eds. R.L. Lundblad and CM. Noyes, CRC Press, 1985). Typical other blocking agents include N-ethylmaleimide, chloroacetyl phosphate, O-methylisourea and acrylonitrile. The use of blocking agents is advantageous because it prevents aggregation of the product and ensures stability in further stages of purification.PL 203 645 B1 7 Blocking agents can be selected in such a way that they induce the formation of a stable and irreversible derivative ( e.g. alpha-halogen acid or alpha-halogen amide). However, other blocking agents may be selected such that after purification of the protein, the blocking agent can be removed to release the non-derivatized protein. The subject of the invention is therefore a method for purifying the MAGE protein or its derivative, comprising reducing the disulfide bond, blocking the resulting thiol groups and subjecting the resulting derivative to one or more chromatographic purification steps. The subject of the invention is also a method for producing a vaccine, including the step of purifying the MAGE protein or its derivative using the method according to the invention and formulating the resulting protein as a vaccine. The thiol-derivatized MAGE antigen is new and constitutes one aspect of the invention. Preferably, the derivatized free thiols are carboxamidated or carboxymethylated. Moreover, the proteins according to the invention are preferably provided with an affinity tag. Such a tag may be CLYTA or a polyhistidine tail. In such cases, the protein is preferably subjected to affinity chromatography after the blocking step. For these proteins with a polyhistidine tail, immobilized metal ion affinity chromatography (IMAC) can be performed. The metal ion may be any suitable ion, for example zinc, nickel, iron, magnesium or copper, preferably zinc or nickel. A preferred IMAC buffer contains a zwitterionic detergent such as Empigen BB (hereinafter referred to as Empigen) because this results in a low level of endotoxin in the final product. If the protein is produced from part of Clyta, it is possible to use affinity with choline or choline analogues such as DEAE to purify the product. In an embodiment of the invention, proteins with a polyhistidine tail and a Clyta part are provided. They can be purified in a simple two-step affinity purification scheme. The invention is further described with reference to the following examples. EXAMPLE I Production of a recombinant E. coli strain expressing the D-MAGE-3-His lipoprotein fusion protein (LPD 1/3-MAGE-3-His or LpD MAGE-3-His) 1. E. coli expression system To produce lipoprotein D, DNA encoding protein D was cloned into the pMG81 expression vector. This plasmid uses signals from the lambda phage DNA to conduct transcription and translation of the introduced foreign gene. The vector contains a PL lambda PL promoter, an OL operator, and two utilization sites (NutL and NutR) to overcome transcriptional polarity when N protein is supplied (Gross et al., 1985, Mol. Cell. Biol. 5:1015) . Vectors containing the PL promoter are introduced into the lysogenic host E. coli to stabilize the plasmid DNA. Lysogenic host strains contain replication-incompetent lambda phage DNA integrated into the genome (Shatzman et al., 1983; Experimental Manipulation of Gene Expression, Inouya (ed.) pp. 1-14. Academic Press Press NY). Lambda phage DNA directs the synthesis of the Cl repressor protein, which binds to the OL repressor on the vector and prevents the binding of RNA polymerase to the PL promoter, thereby preventing transcription of the introduced gene. The cl gene of the AR58 expression strain contains a temperature-sensitive mutation, so that PL-directed transcription can be regulated by changes in temperature, i.e. an increase in temperature in the culture inactivates the repressor and initiates the synthesis of the foreign gene. This expression system enables the controlled synthesis of foreign drug proteins, especially those that may be toxic to the cell (Shimataka and Rosenberg, 1981, Nature 292:128). 2. E. coli AR58 strain The lysogenic E. coli AR58 strain used to produce the LPD-MAGE-3-His protein is a derivative of the standard E. coli strain NIH K12 N99 (F-sugalK2, lacZ-, thr- ). It contains a defective lysogenic lambda phage (galE::TN10, 1 Kil- ci857 DH1). The Kil-phenotype prevents the shutdown of host macromolecule synthesis. The cI857 mutation provides temperature-sensitive changes in the cl repressor. Deletion of DH1 removes the right operon of phage lambda and the host bio, uvr3 and chlA loci. The AR58 strain was produced by transducing N99 with lambda P phage, previously cultured on the SA500 derivative (galE::TN10, 1 Kil-, cI857 DH1). The introduction of a defective lysogen into N99 was selected for tetracycline due to the presence of the TN10 transposon encoding the tetracycline resistance gene in the neighboring galE gene. N99 and SA500 are E. coli K12 strains coming from the laboratory of Dr. Martin Rosenberg from NIH.PL 203 645 B1 8 3. Construction of a vector intended for expressing the recombinant LPD-MAGE-3-His protein The message was the expression MAGE-3 as a fusion protein using the N-terminal third of lipidated protein D as the fusion component, fused to the N-terminus of MAGE-3 and a sequence of several histidine residues (His tail) at the C-terminus. Bia D is a lipoprotein A (42 kDa immunoglobulin D-binding protein, exposed on the surface of the Gram-negative bacterium Haemophilus influenzae). The protein is synthesized as a precursor with an 18-amino acid signal sequence containing a sequence compatible with bacterial lipoproteins (WO 91/18926). When the lipoprotein signal sequence is processed during secretion, Cys (at position 19 of the precursor protein) becomes an amino-terminal residue and is then modified by covalent attachment of fatty acids, both via an ester and an amide bond. Fatty acids connected to amino-terminal residues act as an anchor in the membrane. The plasmid expressing the fusion protein was designed in such a way that it expresses a precursor protein containing an 18-amino-acid signal sequence and the first 109 residues of the processed protein D, two unrelated amino acids (Met and Asp), amino acid residues 2-314 of MAGE-3, two Gly residues acting as a hinge region to expose a further seven His residues. The recombinant strain produces a processed, lipidated fusion protein of 432 amino acids in length with a His tail, (see Fig. 1), with the amino acid sequence shown in ID. Sequ. No. 1, and the coding sequence shown in ID. Sequ. No. 2. 4. Cloning strategy used to produce the LPD-MAGE-3-His fusion protein (vector pRIT14477). A cDNA plasmid (from Dr. Thierry Boon from the Ludwig Institute) containing the coding sequence of the MAGE-3 gene (Gaugler B. et al., 1994) and the pRIT 14586 vector, containing part of the sequence coding the N-terminus of Lipo, were used -D-1/3 (prepared as described in Fig. 2). The cloning strategy included the following steps (Fig. 3). a) PCR amplification of the sequences presented in the MAGE-3 cDNA plasmid, using the sense oligonucleotide 5'gc gcc atg gat ctg gaa cag cgt agt cag cac tgc aag cct, and the antisense oligonucleotide 5' gcg tct aga tta atg gtg atg gtg atg gtg atg acc gcc etc ttc ccc etc tct caa); this amplification led to the following modifications at the N-terminus: changing the first five codons in codon usage by E. coli, replacing the Pro codon with an Asp codon at position 1, inserting an NcoI site at the 5' end and adding two codons Gly residues and 7 His codons at the 5' end, and then XbaI sites at the C end. b) Cloning the above amplified fragment into the TA cloning vector (Invitrogen) and generating the intermediate vector pRIT14647. c) Cutting out the Ncol XbaI fragment from the pRIT14647 plasmid and cloning it into the pRIT14596 vector. d) Transformation of host strain AR58. e) Selection and characterization of E. coli strain transformants containing the pRIT 14477 plasmid, expressing the LPD-MAGE-3-His fusion protein. Example II Production of the LPD1/3-MAGE-3-His antigen 1. Growth and induction of the bacterial strain - expression of LPD1/3-MAGE-3-His AR58 cells transformed with the pRIT14477 plasmid were cultured in 2-liter bottles containing 400 ml each. LY12 medium with the addition of yeast extract (6.4 g/l) and kanamycin sulfate (50 mg/l). After incubation on a shaking table at 30°C for 8±1 hours, a small sample was taken from each bottle for microscopic examination. The contents of the two bottles were pooled to supply the inoculation to the 20-liter fermenter. The inoculation (approximately 800 ml) was added to a previously sterilized 20-liter (total volume) fermenter, containing 7 liters of medium, with the addition of 50 mg/l kanamycin sulfate. The pH was adjusted and maintained at 6.8 by periodic addition of NH 4 OH (25% vol.), and the temperature was adjusted and maintained at 30°C. The aeration rate was set and maintained at 12 liters of air per minute, and the dissolved oxygen pressure was maintained at 50% saturation by feedback controlling the mixing rate. The overpressure in the fermenter was maintained at 500 g/cm 2 (0.5 bar). Batch cultivation was carried out by the controlled addition of a feed solution constituting a carbon source. The feed solution was initially added at a rate of 0.04 ml/min. and increased exponentially during the first 42 hours to maintain a growth rate of 0.1 hr -1. PL 203 645 B1 9 After 42 hours, the fermenter temperature was rapidly increased to 39°C and the feed rate was kept constant at 0.005 ml/g DCW/min. during the induction phase for an additional 22-23 hours, when intracellular expression of LPD-MAGE-3-His reached its maximum level. Portions (15 ml) of broth were collected at regular intervals during the growth/induction phase and at the end of fermentation in order to monitor the kinetics of microbial growth and intracellular expression of the product and, additionally, to provide samples for microbial identification tests and clean sc. At the end of fermentation, the optical density of the culture was from 80 to 120 (corresponding to a cell density of 48 to 72 g DCW/l), and the total volume of the culture was approximately 12 liters. The cultures were rapidly cooled to 6-10°C and the ECK32 cells were separated from the culture broth by centrifugation at 5000 x g at 4°C for 30 minutes. Concentrated ECK32 cells were quickly stored in plastic bags and immediately frozen at -80°C. 2. Protein extraction Frozen, concentrated ECK2 cells were thawed at 4°C and then resuspended in cell destruction buffer to a final density of 60 (equivalent to a log cell density of approximately 36 g DCW/L). . The cells were destroyed by passing through a high-pressure homogenizer (1000 bar) twice. The suspension of destroyed cells was centrifuged (x 10,000 g at 4°C for 30 minutes), and the sediment fractions were washed twice with Triton X100 (1% w/v) + EDTA (1 mM), and then with saline solution, phosphate-buffered (PBS) + Tween 20 (0.1% w/v) and finally PBS. Between washing steps, the suspensions were centrifuged x10,000 g for 30 minutes at 4°C, the supernatant was removed and the sediment fraction was recovered. EXAMPLE III Characterization of Lipo D protein - MAGE-3 1. Purification LPD-MAGE-3-His was purified from cell homogenate using the steps described below: a) dissolving the cleaved fraction of the precipitate after cell destruction; b) chemical reduction of intra- and inter-protein disulfide bonds, followed by blocking of thiol groups to prevent their oxidative re-combination; c) microfiltering the reaction mixture to remove particles and reduce endotoxins; d) capture and initial purification of LPD-MAGE-3-His by taking advantage of the affinity interaction between the polyhistidine tail and zinc-chelating Sepharose; e) removal of contaminating protein drugs by anion exchange chromatography. Purified LPD-MAGE-3-His was subjected to numerous purification steps: f) buffer exchange/removal of urea by size-dependent exclusion chromatography using Superdex 75; g) in-process filtration h) buffer exchange/desalting by size exclusion chromatography using Sephadex G25. Each stage is described in detail below. 1.1) Dissolution of the cell homogenate pellet The pellet fraction from the final washing step (as described above) was re-dissolved overnight in 800 ml of a solution of guanidine hydrochloride (6M) and sodium phosphate (0.1 M, pH 7.0) in 4 °C. 1.2) Reduction and carboxymethylation The dissolved material (pale yellow, cloudy suspension) was purged with argon to remove residual oxygen and a stock solution of 2-mercaptoethanol (14 M) was added to a final concentration of 4.3 M (corresponding to lo 0.44 ml of 2-mercaptoethanol per 1 ml of solution). The resulting solution was divided and transferred to two bottles, which were heated to 95°C in a water bath. After 15 minutes at 95°C, the bottles were removed from the bath and allowed to cool, then the contents were pooled in a beaker covered with foil and (5 l) was placed on ice and solid iodoacetamide was added with vigorous stirring to obtain a st. final yield 6M (corresponding to 1.11 g of iodoacetamide per ml of solution). The mixture was kept on ice in the dark for an hour to ensure complete dissolution of the iodoacetamide and then neutralized (with vigorous stirring and constant pH monitoring) by the addition of approximately 1 liter of sodium hydroxide (5 M) to give final pH 7.5-7.8. PL 203 645 B1 10 The resulting mixture was kept on ice in the dark for a further 30 minutes, after which the pH was again adjusted to 7.5-7.9. 1.3) Microfiltration The mixtures were microfiltered in an Amicon Proflux M12 unit equipped with a Minikross cartridge with hollow capillaries (ref. no. M22M-600-01; surface area 5600 cm 2 , 0.2 µm). The permeate was saved for further purification steps. 1.4) Metal (Zn 2+ ) chelation chromatography (IMAC) Metal chelation chromatography was performed using chelating Sepharose FF (Pharmacia Biotechnology, cat. no. 17-0575-01) packed with a BPG 100/500 column (Pharmacia Bio - technology, part number 18-1103-01). Bed dimensions: diameter 10 cm, cross-sectional area 79 cm 2, bed height 19 cm; volume of the lozenge per 1500 ml. The empty column was cleaned with sodium hydroxide (0.5 M) and then rinsed with water. The medium (supplied in 20 vol.% ethanol) was rinsed with purified water (8 liters) in a Buchner funnel (under vacuum a) and saturated with zinc by passing at least 15 liters of ZnCl 2 solution (0.1 M). Excess zinc was removed by rinsing the substrate with 10 liters of purified water until the pH of the flowing solution reached the pH of the ZnCl 2 solution (pH 5.0). The medium was then equilibrated with 4 liters of a solution containing guanidine hydrochloride (6 M) and sodium phosphate (0.1 M, pH 7.0). The microfiltration permeate containing LPD-MAGE-3-His was mixed with the medium (batch binding) before loading and packing the BPG column with a solution of guanidine hydrochloride (6 M) and sodium phosphate (0.1 M, pH 7, 0). The next steps of metal chelation chromatography were performed at an eluent flow rate of 60 ml/min. The columns were washed first with a solution containing guanidine hydrochloride (6 M) and sodium phosphate (0.1 M, pH 7.0), then with a solution containing urea (6 M) and sodium phosphate (0.1 M, pH 7, 0) until the eluent from the column shows zero absorbance OD 280 nm (t lo). The partially purified LPD-MAGE-3-His protein fraction was eluted with 2 column volumes of a solution containing urea (6 M), sodium phosphate (0.1 M, pH 7.0) and imidazole (0. 5M). The conductivity of this fraction is approximately 16 mS/cm. 1.5) Anion exchange chromatography Before anion exchange chromatography was performed, the conductivity of the partially purified LPD-MAGE-3-His protein fraction was reduced to approximately 4 mS/cm by dilution with a solution containing urea (6 M) and Tris-HCl ( 20 mM, pH 8.0). Anion exchange chromatography was performed using Q-Sepharose FF (Pharmacia Biotechnology, cat. no. 17-0510-01) packing a BPG 200/500 column (Pharmacia Biotechnology, cat. no. 18-1103-11). Dimensions with bed: diameter 10 cm, cross-sectional area 314 cm 2, height with bed 9 cm, volume with bed 2900 ml. The column was filled (in 20% vol. ethanol) and washed with 9 liters of purified water at an eluent flow rate of 70 ml/min. The packed column was cleaned with 3 liters of sodium hydroxide (0.5 M), rinsed with 30 liters of purified water, and then equilibrated with 6 liters of a solution containing urea (6 M) and Tris-HCl (20 mM, pH 8.0). Diluted, semi-purified LPD-MAGE-3-His was loaded onto the column and washed with 9 liters of a solution containing urea (6 M), Tris-HCl (20 mM, pH 8.0), EDTA (1 mM) and Tween (0.1%) until the absorbance (280 nm) of the eluent reached zero. A further rinsing step was performed using 6 liters of a solution containing urea (6 M) and Tris-HCl (20 mM pH 8.0). Purified LPD-MAGE-3-His was eluted from the column with a solution containing urea (6 M), Tris-HCl (20 mM, pH 8.0) and NaCl (0.25 M). 1.6) Size exclusion chromatography Removal of urea from purified LPD-MAGE-3-His and buffer exchange were performed using size exclusion chromatography. It was carried out using Superdex 75 (Pharmacia Biotechnology, cat. no. 17-1044-01) filling the XK 50/100 column (Pharmacia Biotechnology, cat. no. 18-8753-01). Bed dimensions: diameter 5 cm, cross-sectional area 19.6 cm 2, bed height 90 cm, bed volume 1800 ml. The columns were filled with ethanol (20%) and washed with 5 liters of purified water at an eluent flow rate of 20 ml/min. The columns were cleaned with 2 liters of sodium hydroxide (0.5 M), rinsed with 5 liters of purified water, and then equilibrated with 5 liters of phosphate-buffered saline with the addition of Tween 80 (0.1% by volume). PL 203 645 B1 11 The purified LPD-MAGE-3-His fraction (maximum 500 ml/desalting cycle) was introduced to the column at an eluent flow rate of 20 ml/min. Desalted LPD-MAGE-3-His was eluted from the column with 3 liters of PBS containing Tween 80 (0.1% v/v). The fraction containing LPD-MAGE-3-His was eluted at the free volume of the column. 1.7) In-process filtration LPD-MAGE-3-His batches from size exclusion chromatography were filtered through a 0.22 µm membrane in a laminar flow chamber (class 10000). The filtered batch was frozen at -80°C and stored until the desalination step. 1.8) Desalting chromatography Since the osmolarity of the final batch should not exceed 400 mOsM, another buffer replacement step was necessary to reduce the salt concentration. This was performed by a desalting chromatography step using Sephadex G25 (Pharmacia Biotechnology, cat. no. 17-0033-02) packed with a BPG 100/950 column (Pharmacia Biotechnology, cat. no. 18-1103-03). . Bed dimensions: diameter 10 cm, cross-sectional area 78.6 cm 2, bed height 85 cm, bed volume 6500 ml. Sephadex G25 was hydrated with 7 liters of purified water and allowed to swell overnight at 4°C. The gel was packed into a column in clean water at an eluent flow rate of 100 ml/min. The column was cleaned with 6 liters of sodium hydroxide (0.5 M), then equilibrated with 10 liters of a solution containing sodium phosphate (10 mM, pH 6.8), NaCl (20 mM) and Tween 80 (0.1% vol. ). The purified LPD-MAGE-3-His fraction (maximum 1500 ml/desalting cycle) was introduced to the column at an eluent flow rate of 100 ml/min. The desalted LPD-MAGE-3-His fraction was eluted at the free volume of the column, sterilized by filtering on a 0.22 µm membrane and stored at -80°C. The final batch of protein was thawed at 4°C, then portioned into vials and freeze-dried in lactose (3.2%) as an excipient. 2. Analysis on Coomassie-stained SDS-polyacrylamide gels. The purified LPD-MAGE-3-His antigen was analyzed by SDS-PAGE on a 12.5% acrylamide gel under reducing conditions. The protein charge for dyeing with Coomassie is 150 µg, and for staining with silver nitrate 5 µg. Clinical series 96K19 and informational series 96J22 were analyzed. A single major band was obtained, corresponding to a molecular weight of approximately 60 kDa. Two additional smaller bands of approximately 45 kDa and 35 kDa were also observed. 3. Western blot analysis Peptides detected during SDS-PAGE analysis of the LPD-MAGE-3-His protein were identified by Western blot analysis using mouse monoclonal antibodies. These antibodies were developed for personal use using purified preparations of MAGE-3His protein (this protein does not contain LPD). Two monoclonal antibody preparations (Mab 22 and Mab 54) were selected based on their suitability for Western blot analysis and were used in the batch-specific identity confirmation test. Figure 4 shows the band pattern obtained for batches 96K19 and 96J22 after labeling with Mabs 32 and 54. 600 ng of protein was separated on 12.5% SDS-PAGE, transferred to a nylon membrane, reacted with Mabs 32 and 54 (60 µg /ml) and imaged with α-peroxidase-conjugated anti-mouse antibodies. The 60 kDa and 30 kDa peptide from SDS-PAGE are detected by both MAbs. Example IV 1. Preparation of a vaccine using the LPD-MAGE-3-His protein. The vaccine used in the experiment was prepared from recombinant DNA encoding lipoprotein e D 1/3-MAGE-3-His, expressed in E. coli strain AR58, adjuvanted or not. As an adjuvant, mixtures of e 3-de-O-acylated monophosphoryl lipid A (3D-MPL) and QS21 in a water-oil emulsion were used. The SBAS2 adjuvant system has previously been described in WO 95/17210. 3D-MPL: is an immunostimulator derived from lipopolysaccharide (LPS) of the Gram-negative bacterium Salmonella minnesota. MPL has been deacylated and lacks a phosphate group on the lipid A group. This chemical treatment dramatically reduces toxicity while retaining its immunostimulatory properties (Ribi, 1986). Ribi Immunochemistry produces and sells MPL for SB-Biologicals. Experiments conducted by SmithKline Beecham Biologicals have shown that 3D-MPL, in combination with various carriers, strongly enhances humoral and cellular immunity of the TH1.PL 203 645 B1 12 QS21 type: is a natural saponin molecule extracted from the bark of the Native American tree Quillaja samonaria Molina. The purification technique developed for the separation of individual saponins from crude bark extracts enabled the identification of saponin, QS21, which is a triterpene glycoside showing stronger adjuvant activity and lower toxicity compared to the compound output. QS21 has been shown to activate CTLs in the context of MHC class I against several Ag subunits, as well as to stimulate the proliferation of Ag-specific lymphocytes (Kensil, 1992). Aquila (officially, Cambridge Biotech Corporation) manufactures and supplies QS21 to SB-Biologicals. Experiments conducted by SmithKline Beecham Biologicals showed a distinct synergistic effect of the combination of MPL and QS21 in inducing immune responses of the humoral type and TH1 cell type. The water/oil emulsion includes an organic phase made of 2 oils (tocopherol and squalene) and a water phase with PBS containing Tween 80 as an emulsifier. The emulsion consisted of 5% squalene, 5% tocopherol, 0.4% Tween 80, with an average particle size of 180 nm and is known as SB62 (see WO 95/17210). Experiments conducted at SmithKline Beecham Biologicals have shown that the addition of this water-oil emulsion to 3D-MPL/QS21 (SBAS2) further enhances the immunostimulatory properties against various subunit antigens. 2. Preparation of SB62 emulsion (2x concentrated) Tween 80 was dissolved in phosphate buffered saline (PBS) to obtain a 2% solution in PBS. To deliver 100 ml of twice-concentrated concentrate, 5 g of DL-alpha-tocopherol and 5 ml of squalene were centrifuged to mix thoroughly. 90 ml of the PBS/Tween solution was added and mixed thoroughly. The resulting emulsion was then passed through a syringe and then microfluidized using an M110S microfluidization device. The resulting oil drops were approximately 180 nm in size. 3. Preparation of the lipoprotein D1/3 - MAGE-3 - His QS21/3D MPL formulation in an oil-in-water emulsion (SBAS2) The adjuvant was formulated by combining MPL and QS21 in a water/oil emulsion. This preparation was provided in 0.7 ml vials for admixture with the lyophilized antigen (vials containing 30 to 300 μg of antigen). The composition of the adjuvant diluent for the lyophilized vaccine has the following composition: Ingredient Amount of sc (per dose) Adjuvant Emulsion SB62 squalene DL-a-tocopherol Tween 80 Monophosphoryl-lipid A QS21 Preservative Thimerosal Buffer Water for injection dibasic sodium phosphate monobasic phosphate potassium chloride potassium chloride sodium 250 µl 10.7 mg 11.9 mg 4.8 mg 100 µg 100 µg 25 µg q.s. to 0.5 ml 575 µg 100 µg 100 µg 4.0 mg The final vaccine was obtained by reconstituting the lyophilized LPD-MAGE-3-His preparation with the adjuvant or PBS alone. Adjuvant controls without antigen were prepared by replacing the protein with PBS. 4. Vaccine antigen: lipoprotein D1/3 fusion protein - MAGE-3 - His Lipoprotein D is a lipoprotein exposed on the surface of the Gram-negative bacterium Haemophilus influenzae.PL 203 645 B1 13 Combining the first 109 residues of processed protein D as a fusion component was carried out to deliver a vaccine antigen with T-cell epitopes. In addition to the LPD domain, the protein contained two unrelated amino acids (Met and Asp), amino acid residues 2 to 314 of MAGE-3, two Gly residues acting as a hinge region to expose the terminal seven His residues. Example V 1. Immunogenicity of LPD-MAGE-3-His in mice and monkeys In order to test the antigenicity and immunogenicity of the human MAGE-3 protein, a potential vaccine was injected into two different strains of mice (C57BL/6 and Balb /C), differing in genetic background and MHC alleles. For both strains, potential MHC class I and MHC class II peptide motifs were theoretically predicted for the MAGE portion of the LPD-MAGE-3-His fusion protein. a) Immunization protocol: 5 mice of each strain were injected twice with an interval of 2 weeks in the paw pad with 5 µg of LPD-MAGE-3-His formulated with or in the absence of SBAS2 at a concentration of 1/10 concentration - tion used in humans. b) Proliferation assay: Lymphocytes were isolated by crushing the spleens or inguinal lymph nodes of mice 2 weeks after the second injection. 2x105 cells were plated in triplicate in 96-well plates and restimulated in vitro for 72 hours with various concentrations (1 - 0.1 µg/ml) of His-MAGE-3 as such or coated on latex microbeads. Increased MAGE-3 specific lymphoproliferative activity was observed in both spleen cells (see Figs. 5 and 7) and lymph nodes (see Figs. 6 and 8) of C57BL/6 or Balb/C mice , which were injected with the LPD-MAGE-3-His protein, compared with the lymphoproliferative reaction in mice that received the SBAS2 or PBS formulation alone. Furthermore, a significantly higher proliferative response was obtained using lymphocytes from mice immunized with LPD-MAGE-3-His in SBAS2 adjuvant (see Figs. 6 and 8). c) Conclusions: LPD-MAGE-3-His is immunogenic in mice, and the immunogenicity of SC can be increased by using the SBAS2 adjuvant formulation. 2. Antibody reaction a) Immunization protocol: Balb/c or C57BL/6 mice were immunized by two injections into the paw pad at 2-week intervals with PBS or SBAS2 or 5 µg LPD-MAGE-3-His or 5 µg LPD -MAGE-3- His from SBAS2. In the control and test groups, 3 and 5 animals were used, respectively. b) Intermediate ELISA: Two weeks after the second injection, individual sera were collected and subjected to an intermediate ELISA. 2 µg/ml purified His MAGE-3 was used as antigen to coat the wells. After saturation for 1 hour at 37°C with PBS + 1% serum from newborn calves, the sera were successively diluted (starting at 1/1000) in saturation buffer and incubated overnight at 4°C or 90 minutes at 37°C. °C. After washing in PBS/Tween 20, 0.1%, anti-mouse biotinylated goat total IgG (1/1000) or anti-mouse goat IgG1, IgG2a, IgG2b sera (1/5000) were added as the second antibody. After 90 minutes of incubation at 37°C, streptavidin conjugated with peroxidase was added, and TMB (tetra-methyl-benzidine peroxide) was used as the substrate. After 10 minutes, the reaction was blocked by adding H 2 SO 4 , 0.5 M and O.D. was determined. c) Results: Figure 9 shows the comparison between different groups of mice (N=5/group e) of serum titers at the mean midpoint, which is the average dilution necessary to reach the midpoint on the curve. These results show that both tested strains of mice develop a weak anti-MAGE-3 antibody response after two injections of LPD-MAGE-3-His alone, but much stronger anti-MAGE-3 antibody responses are generated after the injection. elution of LPD-MAGE-3-His in the presence of SBAS2. Thus, only 2 injections of LPD-MAGE-3-His + SBAS2 at 2-week intervals are sufficient to produce the observed strong Ab.PL 203 645 B1 14 responses, and a better antibody response was observed in Balb/c mice in compared to the reaction obtained in C57BL/6 mice, which can be explained by differences in haplotypes or in the genetic background between these strains, although the antibody titer obtained in C57BL/6 mice is also higher after injection of LPD-MAGE-3-His + SBAS2 than after administration of LPD-MAGE-3-His alone. MAGE-3 specific Ig subclass responses after vaccination in different groups of mice are shown in Figures 10 and 11, which provide a comparison of the mean midpoints of dilutions of n sera. Neither IgA nor IgM was detected in any sera sample, even in samples from mice vaccinated with LPD-MAGE-3-His in SBAS2 adjuvant. In contrast, total IgG levels were slightly higher in the sera of mice vaccinated with LPD-MAGE-3-His alone and increased significantly in the sera of animals treated with LPD-MAGE-3-His in SBAS2. Analysis of the concentrations of the different IgG subclasses shows that a mixed Ab reaction was induced in the mice because the levels of all IgG subclasses tested (IgG1, IgG2a, IgG2b) were higher in mice vaccinated with the adjuvanted antigen than with the adjuvant alone. However, the nature of the mixed Ab reaction after LipoD-MAGE-3 vaccination in the presence of SBAS2 depends on the mouse strain, since IgG1 and IG2b occurred mainly in the sera of Balb/c and C57B1/6 mice, respectively. 3. Immunogenogenicity of sc lipoprotein D 1/3 MAGE-3-His + SBAS2 adjuvant in rhesus monkeys. Three groups of five rhesus monkeys (Macaca Mulatta) were selected. RTS,S and gp120 were used as positive controls. Groups: Group 1 right leg: RTS, S/SBAS2 left leg: gp120/SBAS2 Group 2 right leg: RTS, S/SB26T left leg: gp120/SB26T Group 3 right leg: lipoD1/3 MAGE 3 His/SBAS2 The animal will receive e vaccine on day 0, followed by booster doses on days 28 and 84, after which they were bled to determine antibody responses against MAGE 3 and protein D. Vaccines were administered intramuscularly as a bolus injection (0, 5 ml) in the back of the right leg. Small blood samples were collected every 14 days. Blood samples of 3 ml each, without the addition of heparin, collected from the femoral vein were left to clot for at least an hour and then centrifuged at room temperature for 10 minutes at 2500 rpm. The serum was collected, frozen at -20°C and sent for determination of the antibody level in a specific ELISA test. 96-well plates (maxisorb Nunc) were coated with 5 μg of His MAGE-3 or protein D overnight at 4°C. After an hour of saturation at 37°C with PBS with NCS 1%, subsequent dilutions of rabbit serum were added for 1.5 hours at 37°C (starting from a 1/10 dilution), after 3 washes in PBS Tween, rabbit serum was added biotinylated sera (Amersham RPN 1004 lot 88) (1/5000). The plates were washed and streptavidin a peroxidase complex (1/5000) was added for 30 minutes at 37°C. After washing, 50 µl of TMB (BioRad) was added for 7 minutes and the reaction was stopped by adding 0.2 M H 2 SO 4 and the OD was read at 450 nm. Midpoint dilutions were calculated using SoftmaxPro software. Antibody reaction Small blood samples were collected every 14 days to monitor the kinetics of MAGE 3 antibody reaction by ELISA. The results indicate that after one injection of LPD1/3 MAGE 3 His + SBAS2, the total MAGE-specific Ig titer 3 would be low. A clear enhancement was observed in 3 out of 5 animals after the second and third injection of LipoD1/3 MAGE 3 + adjuvant in the same monkeys. Poorly responding individuals remained negative even after 3 injections. 28 days after II and III, antibody titers returned to basic levels. The subclasses of these antibodies have been designated primarily IgG rather than IgM. The switch to IgG suggests that a T helper cell response was induced. The specific antibody reaction against protein D, although weak, is consistent with the antibody reaction to MAGE 3.PL 203 645 B1 15 EXAMPLE VI 1. LPD-MAGE 1 His LPD-MAGE 1- was prepared in a similar manner. His. The amino acid and DNA sequences are shown in ID. Sequ. no. 3 and 4. The resulting white protein was purified in a manner analogous to LPD-MAGE-3-His. Briefly, the cell culture was homogenized and treated with 4M guanidine hydrochloride and 0.5M beta-mercaptoethanol in the presence of 0.5% Empigen detergent. The product was filtered and the pulp was treated with 0.6 M iodoacetamide. The carboxamidated fractions were subjected to IMAC chromatography (zinc chelate - Sepharose FF). The column was first equilibrated and washed with a solution containing 4M guanidine, HCl and sodium phosphate (20 mM, pH 7.5), and then washed with a solution containing 4M urea in sodium phosphate (20 mM, pH 7.5) and 0.5% Empigen. The protein was eluted in the same buffer but with increasing concentration of imidazole (20 mM, 400 mM and 500 mM). The eluate was diluted with 4M urea. The Sepharose Q columns were equilibrated and quenched with 4M urea in 20 mM sodium phosphate (pH 7.5) and 0.5% Empigen. The second wash was performed in the same buffer but without detergent. The protein was eluted in the same buffer but with increasing concentrations of imidazole (150 mM, 400 mM, 1M). The eluate was ultrafiltered. EXAMPLE VII Construction of the expression plasmid pRIT14426 and transformation of host strain AR58 to produce NS1-MAGE 3-His Protein design The design of the NS1-MAGE-3-His fusion protein expressed in E. coli is shown in Fig. 12. Primary structure The sequence of the resulting protein has the sequence shown in Id. Sequ. No. 5. The coding sequence (SEQ. ID. No. 6) corresponding to the above protein was placed under the control of the α promoter. pL in an E. coli expression plasmid. Cloning strategy to produce the NS1-MAGE-3-His fusion protein. The starting material was a cDNA plasmid from Dr. Thierry Boon from the Ludwig Institute, containing the coding sequence of the MAGE-3 gene and the PMG81 vector, containing the 81 amino acid coding region of NS1 (non-structural protein) of the influenza virus. The cloning strategy shown in Fig. 3 included the following steps: a) PCR amplification of the sequences presented in the MAGE-3 cDNA plasmid, using the sense oligonucleotide 5'gc gcc atg gat ctg gaa cag cgt agt cag cac tgc aag cct , and the 5' antisense oligonucleotide gcg tct aga tta atg gtg atg gtg atg gtg atg acc gcc ctc ttc ccc ctc tct caa; This amplification led to the following modifications at the N end: change of the first five codons in codon usage by E. coli, replacement of the Pro codon by an Asp codon in position 1, insertion of an Ncol site at the 5' end and adding two Gly residue codons and 7 His codons at the 5' end, and then inserting an XbaI site at the C-terminus. b) Cloning the above amplified fragment into the TA cloning vector (Invitrogen) and generating the intermediate vector pRIT14647. c) Cutting out the NcoI XbaI fragment from the pRIT14647 plasmid and cloning it into the pRIT PMG81 vector. d) Transformation of host strain AR58. e) Selection and characterization of E. coli strain transformants containing the pRIT 14426 plasmid (Fig. 14), expressing the NS1-MAGE-3-His fusion protein. Characterization of recombinant NS1-MAGE-3-His (pRIT14426) Bacteria were cultured in LB medium supplemented with 50 µg/ml kanamycin at 30°C. When the culture reached OD = 0.3 (620 nm), heat induction was performed by increasing the temperature to 42°C. After 4 hours of induction, cells were harvested, resuspended in PBS, and lysed (by disruption) by passing through a French press three times. After centrifugation (60 minutes at 100,000 g), the sediment, supernatant and total extract were examined by SDS-PAGE. The proteins were visualized in gels stained with Coomassie B1, and the fusion protein constituted approximately 1% of the total E. coli protein. The recombinant protein was visible in the form of a single band with a MW of 44.9 kDa. The fusion protein was identified by Western blot analysis using monoclonal antibodies against NS-1. Example VIII Purification of NS1-MAGE-3-His (E. coli) for immunizing rabbits/mice Purification scheme PL 203 645 B1 16 In order to purify the antigen, the following scheme was used: cell lysis + centrifugation - antigen dissolution + centrifugation? agarose Ni 2+ +-NTA ? so ? preparative electrophoresis ("Prep cell") - TCA precipitation and dissolution in PBS a. Lysis Bacterial cells (23 g) were lysed in 203 ml of 50 mM PO 4 pH 7 buffer in a Rannie homogenizer, and then the lysate was centrifuged in a JA 20 rotor at 15,000 rpm for 30 minutes. The supernatant was removed. b. Antigen dissolution. 1/3 of the pellet was dissolved overnight at 4°C in 34 ml of 100 mM PO 4 - 6M GuHCl, pH 7. After centrifugation in a JA 20 rotor at 15,000 rpm for 30 minutes, the precipitate was removed and the supernatant was further purified by IMAC. c. Affinity chromatography: Ni 2+ -NTA agarose (Qiagen) Column volume: 15 ml (16 mm x 7.5 cm) Packing buffer : 0.1 M PO 4 - 6 M GuHCl pH 7 Sample buffer: as above Wash buffer: 0.1 M PO 4 - 6 M GuHCl pH 7 0.1 M PO 4 - 6 M urea pH 7 Elution: imidazole gradient (0 - 250 mM) in 0.1 M PO 4 pH 7 with 6 M urea. Flow rate: 2 ml/min d. Concentration: The fraction of the IMAC eluate containing antigen (160 ml) was pooled and concentrated up to 5 ml in a chamber with an Amicon mixer on a Filtron membrane (Omega type, cut-off weight 10,000). The purity of the sc at this stage is approximately 70% as assessed by SDS-PAGE. b. Preparative electrophoresis (Prep Cell BioRad) 2.4 ml of the concentrated sample was boiled in 0.8 ml of sample reducing buffer and loaded onto a 10% polyacrylamide gel. The antigen was eluted in Tris-Glycine pH 8.3 buffer with 4% SDS and fractions containing NS1-MAGE 3-His were pooled. e. TCA precipitation The antigen was precipitated using TCA and then centrifuged in a JA 20 rotor at 15,000 rpm for 20 minutes. However, the messages were removed. The pellet was redissolved in PBS buffer pH 7.4. The protein was soluble in PBS, after freezing/thawing it showed no signs of degradation when stored for 3 hours at 37°C, and its molecular weight was approximately 50,000 Da, as determined by SDS (12.5% PAGE). EXAMPLES IX Production of an E. coli strain expressing the CLYTA-MAGE-1-His fusion protein 1. Construction of the expression plasmid pRIT14613 and transformation of the host strain AR58: Protein design: Design of the Clyta-MAGE-1-His fusion protein, expressed ne in E. coli is shown in Fig. 15. The primary structure of the resulting protein has the sequence shown in ID. Sequ. No. 7. The coding sequence (SEQ. ID. No. 8) corresponding to the above protein was placed under the control of the α promoter. pL in the expression plasmid E. coli.PL 203 645 B1 17 Cloning: The starting material was the PCUZ1 plasmid, which contains 117 codons of the C-terminus of the LytA coding region from Streptococcus pneumoniae and the pRIT14518 vector in which the cDNA of the MAGE gene was previously subcloned. -1 from a plasmid obtained from Dr. Thierry Boon of the Ludwig Institute. The cloning strategy to express the CLYTA-MAGE-1-His protein (shown in Fig. 16) included the following steps: a) The first step was PCR amplification to surround the CLYTA sequence with NdeI-AfIII restriction sites. Amplification was carried out using the PCUZ1 plasmid as a template and the primers sense oligonucleotide 5' taa aac cac acc tta agg agg ata taa cat atg aaa ggg gga att gta cat tca gac and antisense: 5' GCC AGA CAT GTC CAA TTC TGG CCT GTC TGC CAG. This leads to the amplification of a CLYTA sequence of 378 nucleotides in length. b) In the second step, the CLYTA sequence was combined with the MAGE-1-His sequences to create a sequence encoding the fusion protein. This step involved cutting out the NdeI-AflII Clyta fragment and inserting into the pRIT14518 vector previously opened by the restriction enzymes AfllI and NcoI (compatible with NcoI and AfIIII) to give the plasmid pRIT14613. c) Transformation of the AR58 strain d) Selection and characterization of E. coli transformants (KAN-resistant) containing the pRIT 14613 plasmid (see Fig. 16) 1. Characterization of recombinant CLYTA-MAGE-1-His (pRIT14613) Bacteria were cultured on the medium LB supplemented with 50 µg/ml kanamycin at 30°C. When the culture reached OD = 0.3 (620 nm), induction was performed by raising the temperature to 42°C. After 4 hours of induction, cells were harvested, resuspended in PBS, and lysed (by disruption) in one go. After centrifugation, the pellet, supernatant and total extract were analyzed by SDS-PAGE. Proteins were detected in Coomassie B1-stained gels, with the fusion protein accounting for approximately 1% of total E. coli protein. The recombinant protein was visible as a single band with an apparent MW of 49 kDa. The fusion protein was identified by Western blot analysis using anti-MAGE-1 polyclonal antibodies. Reconstruction of an expression unit including a long promoter? pL (suitable for induction with nalidixic acid) and pRIT14614 encoding the CLYTA-MAGE-3 sequence. The EcoRI-NcoI restriction fragment containing the long pL promoter and part of the CLYTA sequence was prepared from the pRITDVA6 plasmid and inserted between the EcoRI-NcoI sites of the plasmid pRIT14613. Recombinant plasmid pRIT14614 was obtained. The recombinant plasmid pRIT14614 (Fig. 17) encoding the CLYTA-MAGE-1-His fusion protein was used to transform E. coli AR120. A potential KAN-resistant strain was selected and characterized. Characterization of the recombinant protein Bacteria were grown on LB medium supplemented with 50 µg/ml kanamycin at 30°C. When the agn ela culture OD = 400 (600 nm), nalidixic acid was added at a concentration of 60 mg/ml. After 4 hours of induction, cells were harvested, resuspended in PBS, and lysed by disruption (one-shot CLS degradation). After centrifugation, the pellet, supernatant, and total extract were analyzed by SDS-PAGE. Proteins were detected in the gels. stained with Coomassie blue, the fusion protein constituting approximately 1% of total E. coli protein. The fusion protein was identified by Western blot analysis using rabbit polyclonal anti-MAGE-1 antibodies. The recombinant protein was visible as a single band with an apparent MW of approximately 49 kD. EXAMPLE -lyt A leads to the expression of soluble protein, acts as an affinity tag and provides epitopes to T lymphocytes. Generation of an E. coli strain expressing the CLYTA-MAGE-3-His fusion protein. Construction of the expression plasmid pRIT14646 and transformation of the host strain AR120: Design proteins: The design of the CLYTA-MAGE-3-His fusion protein for expression in E. coli is described in Fig. 18. The primary structure of the resulting protein is shown in ID. Sequ. No. 9, here is the coding sequence in Id. Sequ. No. 10.PL 203 645 B1 18 The sequence encoding the above protein structure was placed under the control of the promoter pL in an E. coli expression plasmid. Cloning: The starting material used was the PCUZ1 plasmid, which contains 117 codons of the C-terminus of the LytA coding region from Streptococcus pneumoniae (Gene 43, 1986) pp. 265-272) and the pRIT 14426 vector, in which the cDNA of the gene was previously subcloned MAGE-3 from a plasmid obtained from Dr. Thierry Boon from the Ludwig Institute. The cloning strategy for the expression of CLYTA-MAGE-3-His (shown in Fig. 19) included the following steps: 1. Production of the CLYTA-MAGE-3-His 1.1 coding sequence module. The first step was PCR amplification to surround the CLYTA sequence with AflIII and AflIII restriction sites. PCR amplification was performed using the PCUZ1 plasmid as template and the primers: sense oligonucleotide 5' taa aac cac acc tta agg agg ata taa cat atg aaa ggg gga att gta cat tca gac and antisense primers: 5' ccc aca tgt cca gac tgc tgg cca att ctg gcc tgt ctg cca gtg. This leads to the amplification of a CLYTA sequence of 427 nucleotides in length. The above-mentioned amplified fragment was cloned into the TA vector (Invitrogen) to obtain ac pRH 14661. 1.2. The second step was to combine the CLYTA sequence with the MAGE-3-His sequences to generate a sequence encoding the fusion protein. This step included cutting out the Clyta Afllll-Afllll fragment and introducing the pRIT14426 vector previously opened with Aflll and Ncol restriction enzymes (Ncol and AflIII compatible) into the vector pRIT14662 to obtain the plasmid pRIT14662. 2. Reconstruction of an expression unit containing a long promoter - pL (suitable for induction with nalidixic acid) and the CLYTA-MAGE-3 coding sequence: A BglII-XbaI restriction fragment containing a short pL promoter and the CLYTA-MAGE-3-His coding sequence was generated from the pRIT14662 plasmid and introduced between the BglII and XbaI sites of the TCM67 plasmid (a derivative of pBR322 with the ampicillin e resistance gene and a long promoter - pL, PCT/EP92/01827). The recombinant plasmid pRIT14607 was obtained. The recombinant plasmid pRIT14607 encoding the CLYTA-MAGE-3-His fusion protein was used to transform E. coli AR120 (Mott et al., PNAS, 82:88). A potential ampicillin e-resistant strain was selected and characterized. 3. Production of plasmid pRIT14646 Finally, a plasmid similar to pRIT14607, but with a kanamycin resistance gene (pRIT 14646), was constructed. Characterization of the recombinant protein Bacteria were grown on LB medium supplemented with 50 µg/ml kanamycin at 30°C. When the agn axis culture was OD = 400 (at 600 nm), nalidixic acid was added at a concentration of 60 g/ml. After 4 hours of induction, cells were harvested, resuspended in PBS, and lysed by disruption (one-shot CLS degradation). After centrifugation, the pellet, supernatant, and total extract were analyzed by SDS-PAGE. Proteins were detected in the gels. stained with Coomassie blue, the fusion protein constituting approximately 1% of the total E. coli protein. The fusion protein was identified by Western blot analysis using rabbit polyclonal anti-MAGE-3 antibodies. The recombinant protein was visible as a single band with an apparent MW of approximately 58 kDa. Example The recombinant protein was induced by the addition of nalidixic acid at a final concentration of 60 μg/ml. Cells were harvested at the end of fermentation and lysed at 60 OD/600 by passing through a French press (20,000 psi) twice. Lysed cells were centrifuged at 15,000 x g at 4°C. The supernatant containing the recombinant protein was introduced into a DEAE Sepharose CL6B anion exchange column (Pharmacia) pre-equilibrated with 0.3 M NaCl, 20 mM Tris HCl, pH 7.6, Buffer A. After washing the column with buffer A, the fusion protein 2% choline was eluted in buffer A. Fractions containing antigen were pooled, as visualized by Western blot analysis using anti-MAGE-3 antibody. The antigen eluted from DEAE was introduced into 0.5% Empigen BB (zwitterionic detergent) and 0.5 M NaCl, and then charged to a metal ion affinity chromatography column, previously equilibrated with 0.5% Empigen BB, 0.5 M NaCl, 50 mM phosphate buffer pH 7.6 (Buffer B).PL 203 645 B1 19 The IMAC p columns were washed with buffer B until the basic absorbance value was reached at OD 280. The antigen was eluted with an imidazole concentration gradient of 0-250 mM in buffer C. Imidazole fractions 0.090-0.250 M were pooled, concentrated on a Filtron Omega 10 kDa membrane, and then dialyzed against PBS buffer. Conclusions The LPD-MAGE-3-His fusion protein has been shown to be immunogenic in mice, and this immunogenicity (proliferative response and antibody response) can be further enhanced by the use of the adjuvant described above. Purification can be improved by derivatizing thiols that form disulfide bonds. It has also been shown that a better antibody response is obtained by vaccination with LPD-MAGE-3-His in the presence of an adjuvant. The main isotype found in the serum of mice of the C57B1/6 strain is IgG2b, which indicates that a Th1 type reaction has been induced. In a clinical setting, a patient treated with LPD-MAGE-3-His in the unadjuvanted form experienced removal of melanoma. LITERATURE - Anichini A., Fossati G., Parmiani G. Immunol. Today, 8: 385 (1987). - De Plaen E., Arden K., Traversari C., et al. Immunogenetics, 40: 360 (1994). - Gaugler B., Van den Eynde B., van der Bruggen P., et al. J. Exp. Med., 179: 921 (1994). - Herman J., van der Bruggen P., Immanuel F., et al. Immunogenetics, 43: 377 (1996). - Inoue H., Mori M., Li J., et al. Int. J. Cancer, 63: 523 (1995). - Kensil C.R., Soltysik S., Patel U., et al. in: Channock R.M., Ginsburg H.S., Brown F., et al., (eds.), Vaccines 92, (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.), 36-40: (1992). - Knuth A., Danowski B., Oettgen H.F., et al. Proc. Natl. Acad. Sci. USA, 81: 3511 (1984). - Patard J.J., Brasseur F., Gil-Diez S., et al. Int. J. Cancer, 64: 60 (1995). - Ribi E., et al. in: Levine L., Bonventre P.F., Morello J., et al. (eds)., American Society for Microbiology, Washington DC, Microbiology 1986, 9-13; (1986). - Van den Eynde B., Hainaut P., Hérin M. et al. Int. J. Cancer, 44: 634 (1989). - Van der Bruggen P., Traversari C., Chomez P., et al. Science, 254: 1643 (1991). - Van der Bruggen P., Bastin J., Gajewski T., et al. Euro. J. Immunol., 24: 3038 (1994). - Van Pel A., van der Bruggen P., Coulie P.G. , et al., Immunol. Rev., 145: 229 (1995). - Weynants P., Lethé B., Brasseur F., et al. Int. J. Cancer, 56: 826 (1994). - Nishimura S, Fujita M, Terata N, Tani T, Kodama M, Itoh K, Nihon Rinsho Meneki Gakkai Kaishi 1997, Apr, 20 (2): 95-101. - Fuijie T et al, Ann Oncol 1997 Apr, 8 (4): 369-72.PL203 645 B1 20PL 203 645 B1 21PL 203 645 B1 22PL 203 645 B1 23PL 203 645 B1 24PL 203 645 B1 25PL 203 64 5 B1 26PL 203 645 B1 27PL 203 645 B1 28PL 203 645 B1 29 PL PL PL

Claims (18)

1. Patent claims 1. A derivative of a tumor associated antigen from the MAGE family, characterized in that it is selected from the group consisting of: - a MAGE antigen comprising derivatized thiol groups; - a recombinantly expressed MAGE fusion protein, wherein the fusion component is selected from the group consisting of protein D from Haemophilus influenzae B or a fragment thereof, protein NS1 from influenza virus or a fragment thereof, or C-LYTA from Streptococcus pneumoniae or thereof fragment. 2. Antigen according to claim. The process according to claim 1, characterized in that the derivatized free thiols are carboxyamidated or carboxymethylated. 3. Antigen according to claim. 1 or 2, characterized in that it comprises an affinity tag. 4. Antigen according to claim. 1-3, characterized in that protein D or a fragment thereof is lipidated. 5. Antigen according to claim. 1-4, characterized in that the MAGE protein is selected from the group consisting of MAGE A1, MAGE A2, MAGE A3, MAGE A4, MAGE A5, MAGE A6, MAGE A7, MAGE A8, MAGE A9, MAGE A10, MAGE A11, MAGE A12, MAGE B1, MAGE B2, MAGE B3, MAGE B4, MAGE C1 and MAGE C2. 6. A nucleic acid sequence encoding the fusion protein of claim 1. 1-5. 7. A vector characterized in that it comprises a nucleic acid according to claim 1. 6. 8. A host cell characterized in that it is transformed with a nucleic acid as defined in claim. 6 or the vector defined in claim 6. 7. 9. A vaccine characterized in that it contains the protein specified in claim 1. 1-5. 10. Vaccine according to claim 9, characterized in that it additionally contains an adjuvant and/or immunostimulating cytokine e or chemokine e. 11. Vaccine according to claim 9 or 10, characterized in that the protein is contained in a carrier consisting of an emulsion of oil in water. 12. Vaccine according to claim 10 or 11, wherein the adjuvant comprises 3D-MPL, QS21 or a CpG oligonucleotide. 13. Vaccine according to claim 9 - 12, characterized in that it additionally contains one or more antigens. 14. The specific vaccine in claim 9 - 13 for use as a drug for the treatment of MAGE-related cancer diseases. 15. The use of a protein according to claim 1. 1 for the preparation of a vaccine for the immunotherapeutic treatment of a patient suffering from melanomas or other MAGE-related cancers. 16. The use of the nucleic acid according to claim 1. 6 for the preparation of a vaccine for the immunotherapeutic treatment of a patient suffering from melanomas or other MAGE-related cancers. 17. A method for purifying a MAGE protein or a derivative thereof, characterized in that it includes reducing the disulfide bond, blocking the resulting thiol groups with a blocking group, and subjecting the resulting derivative to one or more chromatographic purification steps. 18. A method for producing a vaccine, characterized in that it includes the step of purifying the MAGE protein or its derivative by the method specified in claim 1. 17 and formulating the resulting protein as a vaccine.PL 203 645 B1 30 DrawingsPL 203 645 B1 31PL 203 645 B1 32PL 203 645 B1 33PL 203 645 B1 34PL 203 645 B1 35PL 203 645 B1 36PL 203 645 B1 37PL 203 645 B1 38PL 203 645 B1 39PL 203 645 B1 40PL 203 645 B1 41PL 203 645 B1 42PL 203 645 B1 43PL 203 645 B1 44PL 203 645 B1 45PL 203 645 B1 46PL 203 645 B1 47PL 2 03 645 B1 48 Publications Department, UP RP Price PLN 6.00 per year. PL PL PL