EP0973909A2 - Process for extracting and purifying recombinant, non-lipidised osp-protein - Google Patents

Abstract

A recombinant, non-lipidised OspC is disclosed, as well as a process for preparing Osp-proteins and vaccines which contain these recombinant proteins.

Description

 Process for the production and purification of recombinant, non-lipidated Osp protein

The present invention relates to a method for obtaining and purifying recombinant Osp proteins, in particular recombinant OspC, and a recombinant, non-lipidated OspC. The invention further relates to a vaccine containing recombinant, non-lipidated OspC and the use of the vaccine for the immunization of vertebrates.

Lyme disease is an infectious disease of vertebrates transmitted by ticks and caused by the spirochete Borrelia burgdorferi sensu lato. In temperate latitudes (North America, Central Europe), Lyme disease is the most common tick-borne infectious disease. The high infection rate (up to 60%) of ticks in certain areas has recently made Lyme disease an epidemiological problem.

The symptoms most frequently associated with Lyme disease include, in some cases, serious diseases of the skin, the nervous system, the heart and the joints. Lyme disease is usually treated with antibiotics, a treatment that works very efficiently in the early stages of infection. Because of the heterogeneity of the different Borrelia strains, however, early detection of a Borrelia infection is very difficult. If the disease has progressed to the chronic stage, antibiotic treatment is more difficult and not always successful. With the duration of the infection, the likelihood of developing severe long-term consequences also increases. An efficient, preventive vaccination against Lyme disease is therefore very desirable.

The term "Borrelia" as used here corresponds to Borrelia burgdorferi sensu lato and includes the causative agents of Lyme disease, for example B.burgdorferi sensu stricto, B.garinii sp. nov and B.afzelii.

For the surface lipoproteins OspA and OspC from Borrelia described that they protect against Borrelia infections in animal experiments. Therefore, OspA and OspC are considered to be preferred candidates as vaccine antigens against Lyme disease (EP 418 827; Fikrig et al. (1990), Science 250: 553-6; Preac-Mursic et al. (1992), Infection 20: 342 -9). OspA and OspC are both plasmid-encoded proteins (Barbour et al. (1987), Science 237: 409-11; Marconi et al. (1993), J. Bacteriol. 175: 926-32) at their N-terminus in one post-translational process a lipid portion is attached. This lipid portion serves as a "membrane anchor" that anchors the lipoprotein in the outer bacterial membrane. The part of an Osp protein not anchored in the membrane is exposed on the cell surface (Howe et al. (1985), Science 227: 645-6; Bergström et al. (1989), Mol. Microbiol. 3: 479-86 ). The genes coding for OspA and OspC occur in almost all Borrelia strains and both surface proteins occur in many serologically different forms (Wilske et al. (1988), Ann. New York Acad. Sci. 539: 126-43; Livey et al. (1995) Mol. Microbiol. 18: 257-69). "Serologically different forms" means that different variants of the individual surface proteins (eg OspA, OspB, OspC) occur in different Borrelia strains, the difference being a lack of cross-reactivity with antibodies and in different RFLP patterns (RFLP = Restriction fragment length polymorphism). Livey et al. (1995, Mol. Microbiol. 18: 257-69) have described 35 different RFLP types of OspC; these correspond to the serological variants of OspC mentioned in this text.

The presence of many serologically different forms complicates the development of a vaccine based on these lipoproteins. For example, Fikrig et al. (1992, J. Immunol. 148: 2256-60) showed that immunization with a serological variant of OspA does not protect against infection with a strain which carries another OspA variant. As a result, the majority of the opinion in the literature is that it is necessary to develop a vaccine that contains a mixture of serologically different forms of the antigen and thus provides vaccination protection against as many different strains of Borrelia as possible. This requires for efficient and reproducible production, that the extraction and cleaning process can be used uniformly for all variants.

The main interest in the development of vaccines against Lyme boron reliosis has so far been the lipoprotein OspA, of which seven different serotypes have been described. However, the OspC lipo protein is also immunogenic during natural infection and protects against homologous challenge in animal experiments (WO 94/25596, Probert et al. (1994), Infect. Immun. 62: 1920-26).

The terms "homologous" and "heterologous" are used here in connection with vaccination protection and "challenge" (infecting a mammal with a virulent pathogen). In this context, "homologous" means "originating from the same serological variant of a pathogen" and "heterologous" means "not originating from the same serological variant of a pathogen". Homologous protection means that the vaccine antigen protects against pathogens that carry an antigen that belongs to the same serological variant as the vaccine antigen. Heterologous protection means that the vaccine antigen also provides cross protection, which means that the protection also acts against pathogens which carry the antigens of other serological variants. So far, the following strategies for the production of Osp lipoproteins (OspA, OspB, OspC from Borrelia) have been pursued: on the one hand, the wild-type Osp lipoproteins were isolated directly from the bacterial cell wall of the Borrelia, on the other hand, genes coding for Osp lipoproteins were recombinantly isolated in E. .coli or the yeast Pichia pastoris expressed. Furthermore, various recombinant Osp fusion proteins were expressed in E. coli.

Wild-type Osp lipoproteins are integrated with their membrane anchors into the cell wall of the Borrelia. In order to isolate them, the membrane fraction must first be prepared, then the Osp lipoproteins are extracted from the membrane. This can be achieved in an efficient manner only with detergents or substances which act like detergents (for example EP 522 560, WO 94/25596, WO 93/08299, Belisle et al. (1994), J. Bacteriol. 176: 2151-57). Detergents have the property that they bind to lipoproteins and thus release them from the membrane surrounding them. However, they can only be separated from the lipoproteins with great difficulty and with great losses of product. However, the detergents must be completely separated off for human application. Furthermore, lipoproteins are very difficult to handle due to their poor solubility in the absence of detergents. In the absence of detergents, lipoproteins tend to form aggregates, which complicates, for example, sterile filtration and exact dosing. The use of detergents is therefore associated with many difficulties in the efficient production of vaccines.

Gondolf et al. (1990, J. Chromatography 521: 325-34) and Ma and Weis (1993, Infect. Immun. 61: 3843-53) avoided the use of detergents by extracting the lipoproteins OspA and OspB with n-butanol. Here, n-butanol fulfilled a detergent-like function, but without being a real detergent. The Borrelia were subjected to extraction with n-butanol after sonification. The following centrifugation led to the formation of three phases, a butanol phase, an intermediate phase and an aqueous phase. The lipoproteins OspA and OspB were partly in the aqueous phase, from which they were purified after dialysis using chromatographic methods.

E. coli is the host organism that is most commonly used in genetic engineering to produce recombinant proteins. E. coli is easy to cultivate and usually gives very high yields. E. coli can be used as a host organism in all cases in which no post-translational modifications or only those which are also made by E. coli are required in order to obtain a functional protein.

Borrelia Osp proteins are naturally lipidated, ie post-translationally a lipid portion is attached to the N-terminus of the protein. Many authors have already indicated that lipidation is essential for the immunogenicity of the Osp lipoproteins from Borrelia (WO 93/08299, Erdile et al. (1993), Infect. Immun. 61: 81-90; Belisle et al. (1994) J. Bacteriol. 176: 2151-57; Weis et al. (1994) Infect. Immun. 62: 4632-36). The finding that E. coli recombinant Osp- Lipidifying proteins correctly has led to the recombinant expression of Borrelia Osp proteins in E. coli in many cases replacing the direct isolation of Osp proteins from Borrelia (WO 93/08299, Weis et al. (1994) Infect. Immun. 62: 4632-36).

As in Borrelia, the lipidated Osp protein in E. coli is associated with the cell membrane. As in the case of Borrelia, the Osp protein must consequently be extracted from the cell membrane with detergents. The expression of recombinant Osp proteins in E. coli therefore does not prevent the problem of adding detergents described above. However, the use of E. coli for the expression of Osp proteins allows modifications to be introduced into the recombinant genes or proteins.

Variants of OspA shortened in the N-terminus that can no longer be lipidated are no longer anchored in the cell membrane. Thin et al. (1990, Prot. Expr. Purif. 1: 159-68) expressed a recombinant OspA lacking the first 17 wild-type amino acids in E. coli. The N-terminal truncated OspA was not lipidated and consequently soluble in the absence of detergents. This non-lipidated OspA construct was detectable with sera from Lyme disease patients. This means that at least some of the Borrelia epitopes which trigger an immune response in patients were present in the N-terminally shortened, non-lipidated OspA. Another recombinant, non-lipidated OspA, which lacked the first wild-type amino acid, was not shown to be immunogenic in animal experiments. (Erdile et al. (1993); Infect. Immun. 61: 81-90). The lack of immunogenicity is reported by Erdile et al. experimentally clearly attributed to the lack of lipid content. N-terminally truncated, non-lipidated OspA protein from E. coli therefore offers insufficient immunogenicity in order to be used in a vaccine.

According to the prior art, fusion proteins are a further way of expressing Osp proteins in E. coli. Fusions of OspA with NS-1 (Non-Structural Protein from Influenza; Gern et al. (1994), Immunol. Lett. 39: 249-58; Telford III et al. (1995), J. Infect. Dis. 171: 1368-70), by OspA, OspB and OspC with glutathione-S-transferase (Padula et al. (1993), Infect. Immun. 61: 5097-105; Telford III et al. (1993), J. Exp Med. 178: 755-58) and OspC with maltose binding protein (Probert et al. (1994), Infect. Immun. 62: 1920-26) have already been described. In these cases, the respective Osp protein is not lipidated due to the N-terminal blockade by the fused protein. The fusion partners mentioned in the prior art either have a stimulating effect on the immune system or they facilitate expression or purification of the fusion protein. The idea of fusion proteins is thus a further attempt to solve two major problems in the production of Osp proteins: the greatly reduced immunogenicity of the non-lipidated Osp proteins and the difficulty in cleaning Osp lipoproteins for a human usable vaccine. Since the fusion proteins are not lipidated, they are not associated with the cell membrane, so that no detergent is required to isolate the respective Osp fusion protein. Fusions with glutathione-S-transferase and maltose binding protein allow the fusion protein to be purified via the specific affinities of the respective fused protein.

For a fusion of OspA with the 81 amino acid long N-terminus of NS-1 it could be shown that both relatively high expression rates are achieved in E. coli and that the fusion protein reacts with homologous immune serum (WO 93/04175 ). The authors conclude from the reaction with homologous immune serum that this fusion protein has the ability to protect against infection with virulent B.burgdorferi. However, this assumption is not supported by experimental data in WO 93/04175.

For fusions that block the N-terminus of the Osp proteins, it is believed that the fusion product is not lipidated. The immunogenicity of these non-lipidated Osp fusion proteins is due on the one hand to a possible immunostimulating effect of the fused protein or protein portion, and on the other hand it is attributed in the literature to very high doses of adjuvants (WO 93/08299). In addition, such fusion proteins have the disadvantage that it is not only Osp protein-specific Antigens are present in the immunogen. High doses of adjuvants have to be used if the vaccine antigen has poor immunogenicity. Since high doses of adjuvants greatly increase the risk of side effects of the vaccine, vaccine antigens that require high doses of adjuvant are not favorable for use in humans (Gupta et al. (1995); Vaccine 13: 1263-76). WO 93/08299 also expressly points out that an adjuvant-free vaccine is desirable. Therefore, the inventors of WO 93/08299 prefer native, that is lipidated Osp protein, to the non-lipidated one for vaccine development.

Probert et al. (1994) describe another way, which is above all a solution to the cleaning problem. They express an OspC-MBP (maltose binding protein) fusion protein in E. coli and use the affinity of MBP for amylose to specifically enrich the fusion protein by binding it to an amylose-containing carrier. After the specific isolation of the fusion protein, the MBP portion is split off from the fusion protein, so that only OspC is present in the end product. In contrast to the native OspC, the end product thus prepared lacks the N-terminal cysteine. At the DNA level, an interface for factor Xa was inserted between MBP and OspC, so that the two proteins could be cleaved by factor Xa activity. A disadvantage of this method is that additional steps, namely the cleavage reaction with factor Xa and the subsequent separation of factor Xa and the MBP fraction which has been split off, are necessary. The risk that traces of both or one of the two substances is present in the vaccine speaks against the use of such an additional cleavage reaction. However, the use of a protease is also problematic because additional cleavage sites could be present in the Osp protein itself. In such a case, the protease reaction would destroy the end product.

WO 94/25596 describes, inter alia, the production of recombinant OspC in the yeast Pichia pastoris. The OspC obtained in this way proved to be immunogenic and protected against homologous challenge in animal experiments. The preparation of OspC from Pichia was carried out without the use of detergents. The Cells were mechanically broken up in an isotonic buffer solution, filtered and centrifuged. OspC present in the supernatant was then enriched over several chromatographic purification steps. The disadvantage of this method arises from the fact that the centrifugation has no specificity for OspC. As a result, some of the proteins originating from host cells and the medium are also present in the supernatant. These proteins are undesirable and must be separated from the product via several successive chromatographic steps. Every chromatographic step involves a high loss of product. In addition, it cannot be guaranteed beyond any doubt that traces of contaminating proteins or other substances such as lipids, nucleic acids and carbohydrates are present in the end product.

WO 95/21928 describes the expression of heterologous proteins in methylotrophic yeasts, including Pichia. OspA is described as an example of a heterologous protein. WO 95/21928 is mainly concerned with high yields. In the case of OspA, these are determined after the cells have been broken up using a "French press" using an ELISA. OspA is not specially cleaned and is also not specifically characterized.

Despite the methods already known in the prior art for producing a vaccine against Lyme disease, there is still a need for a high-purity vaccine which can be used unconditionally on humans and which offers broad vaccination protection against the various Strains of Lyme borrelia mediated.

It is an object of the present invention to provide an improved process for the production and purification of recombinant Osp protein, in particular of recombinant OspC protein, which allows simple and efficient purification of immunologically active, recombinant Osp .

The present invention provides a process for the recovery and purification of recombinant Osp derived from Borrelia Osp from eukaryotic host cells, preferably yeast cells. The method according to the invention is characterized in that the host cells are lysed and the lysate is mixed with an organic solvent, under conditions in which cellular proteins are substantially completely precipitated and recombinant Osp protein remains selectively in solution, and the Osp protein is purified from the solution and recovered. This step is very selective due to the extraordinary dissolution behavior of the recombinant Osp proteins, in particular the recombinant OspC, which is very different from that of other proteins.

The organic solvents used in the process according to the invention are known as protein-precipitating solvents: alcohols, short-chain ketones, sulfoxides and nitriles, preferably methanol, propanol isomers, butanol isomers, DMSO, acetonitrile, dioxane, acetone and particularly preferably ethanol. Protein-precipitating organic solvents are usually used to precipitate proteins from solutions and not to keep them in solution. In this case, the organic solvent can be used as an efficient precipitation reagent for contaminating substances. Most proteins precipitate under the conditions used in the present method; due to their association with proteins, 99.99% of the nucleic acids also precipitate.

The method according to the invention therefore offers a very high selectivity even in the first cleaning step. The high accumulation of the recombinant Osp proteins, in particular of recombinant OspC, and the strong depletion of contaminating substances in the solution reduce the number of subsequent cleaning steps required, so that the high yields of the expression in Pichia in the further course cleaning can be reduced only minimally and the effort for further cleaning can also be kept to a minimum.

The addition of the organic solvent leads to the precipitation of contaminating proteins and a large part of the nucleic acids. The separation of the precipitated substances from the solution containing, for example, recombinant OspC can be carried out according to the present the invention by sedimentation or filtration. The sedimentation is preferably accelerated by centrifugation, but it can also be carried out by leaving it to stand for a long time, the precipitated substances collecting at the bottom of the vessel containing the solution due to the gravity of the earth. After the sedimentation, the recombinant solution containing OspC can be removed from the precipitated sediment.

In the filtration, the recombinant solution containing OspC is separated from the precipitated substances using a filter. Sterile filters, for example, are used for the filtration. The filtration is preferably carried out in two steps. A relatively coarse filter with a relatively large pore size is used for a prefiltration, for example a 1 μm sterile filter. A finer filter with a smaller pore size is then used for the main filtration, for example a 0.2 μm sterile filter.

After the precipitation with the organic solvent and the sedimentation or filtration, the solution containing recombinant OspC also contains phospholipids, dyes and small traces of protein and nucleic acids from the host cell. Compared to the solution before the precipitation, approximately 10% proteins from the host cells and medium, approximately 0.001% nucleic acids and approximately 5% lipids remain in the solution after the precipitation. These can preferably be separated very efficiently from the recombinant Osp protein, in particular OspC, by a subsequent reverse phase chromatography.

According to a further aspect of the invention, the reverse phase chromatography can also be used without a preceding purification step, such as the precipitation of contaminating proteins or nucleic acids by organic solvents (as described above), for the purification of an Osp-containing starting material. The present invention therefore also includes a process for the purification of Osp proteins from an Osp-containing starting material, in which the Osp protein is selectively bound to a solid support suitable for reverse-phase chromatography, optionally washed, and eluted. Suitable Materials can be carriers made of silicate, preferably Spherisorb 5C18, or plastic polymers, preferably Source PR15 ™ (Pharmacia) or Amberchrome ^® CG 300 (Tosohaas).

The principle of reverse phase chromatography is based on the fact that, depending on their specific affinities, molecules are distributed between a solid, hydrophobic and a liquid, hydrophilic phase. The Osp proteins according to the invention, in particular OspC, have a high affinity for the solid, hydrophobic carrier, while contaminating substances - the remaining proteins from host cells and medium, nucleic acids, phospholipids and dyes - remain in the liquid, hydrophilic phase and are separated from the solid support to which Osp protein is bound. The solid support is usually immobilized on a column matrix, so that the liquid phase, which contains the contaminating substances, flows through the column. When the carrier is in bulk, the carrier particles are separated from the liquid phase and washed with an eluent with low elution power. Under these conditions, the Osp protein has a higher affinity for the carrier than for the solution. In the method according to the invention, this step is particularly efficient in the elimination of unwanted proteins from host cells and medium, nucleic acids, phospholipids and dyes, an elimination which does not take place completely when precipitated with the organic solvent. Reverse phase chromatography takes advantage of the special solution behavior of the recombinant Osp protein obtained in accordance with the invention, in particular OspC, in organic solvents. In the reverse phase chromatography, the organic solvent is used according to the invention to elute the Osp protein. First, the solution containing, for example, recombinant OspC is diluted with an aqueous buffer solution and brought into contact with a solid carrier. The conditions are set so that Osp protein selectively binds to the carrier. So that the Osp protein according to the invention binds to the carrier, the solution containing Osp protein is diluted with the organic solvent so that the solvent concentration has a low elution power. Under these conditions, the Osp protein produced according to the invention has a higher affinity for the carrier than for the solution. The wearer is ger can of silicate, preferably is Spherisorb 5C18, or made of plastic polymers, preferably source RP15 ™ (Phar¬ macia) or Amber Chrome ^® CG 300 (TosoHaas). The carrier can be packed in a column, but can also be used in bulk with the solution. Recombinant Osp protein present in the diluted solution binds selectively to the carrier, while contaminating phospholipids, dyes, proteins and nucleic acids do not bind. The selectivity here is based on the concentration of the concentration gradient used and the type of organic solvent.

In a preferred embodiment of the present invention, ethanol is used as the solvent, preferably in a concentration of around 45%. After washing the carrier material by rinsing with an eluent with low elution power, for example a physiological buffer solution with 35%, 40% or 45% ethanol, the Osp protein, in particular OspC, becomes selective with an organic solvent in high concentration or eluted from the column with high elution power. A solvent with high elution power causes the Osp protein, in particular OspC, bound to the support to detach from it and to dissolve in the solvent.

In a further embodiment of the present invention, ethanol is used in a concentration between 50% and 60%, preferably between 52% and 57%, particularly preferably 55%. The organic solvent used for elution can be the same solvent that was used for the precipitation in the first step, but it can also be a different organic solvent. It can be used in a selected concentration or as a gradient between two different concentrations. In any case, the concentrations are in the range from 10% to 90%, preferably from 20% to 70% and particularly preferably in the range from 40% to 60%. Depending on the desired concentration, the organic solvent is diluted in aqueous buffer solution. The aqueous buffer solution can be, for example, a phosphate-buffered sodium chloride solution, Tris buffer or citrate buffer. Preferred organic solvents in the context of the present invention are selected from the group of protein-precipitating substances, in particular monohydric or polyhydric alcohols, short-chain ketones, sulfoxides or nitriles, preferably methanol, n-propanol, 2-propanol and other isomers of Propanols, t-butanol, 2-butanol and other isomers of butanol, DMSO, acetonitrile, dioxane or acetone and particularly preferably ethanol.

The organic solvent can be a single organic solvent in aqueous, buffered solution or a mixture of various organic solvents. The organic solvent used in the different cleaning steps of the process according to the invention can be the same in all steps, but different organic solvents can also be used in different steps in a cleaning process.

As already mentioned above, a vaccine that provides effective protection against Lyme disease should protect against all serological subtypes of Lyme disease. Therefore, it makes sense to provide a polyvalent mixed vaccine with broad effectiveness. "Polyvalent mixed vaccine" means here that the vaccine contains various serological subtypes of the Osp proteins, in particular of OspC. For the large-scale production of such a polyvalent vaccine, it is desirable to have a uniform extraction and purification method for the various components. However, in order to separate the respective Osp protein from contaminating substances, the chromatographic methods previously available had to be adjusted very specifically to each Osp variant. Because even the smallest differences in the amino acid sequence, as they occur with the different serological variants of the Osp proteins, can change the behavior of the proteins in the highly specific, chromatographic processes so that the cleaning efficiency is impaired.

Many of the already known chromatographic methods are based on the fact that Osp, in particular OspC, does not or only binds very weakly to the carrier materials used for classic protein purification, as are used in ion exchangers and hydrophobic interaction gels. The cleaning process accordingly takes place as so-called "negative chromatography", which means that the substances to be eliminated are gradually removed from the solution containing Osp (OspC) by binding to various carriers, while Osp (OspC) always can be found in the flow. After a series of such steps, Osp (OspC) should then be in a purified form. This method is labor-intensive, and the risk of "taking" contaminating substances in the flow is very great. The "negative chromatography" must always be matched to the respective Osp variant, although Osp variants have also been found to which such a method cannot be applied.

Methods that have to be adjusted to the respective Osp variant are labor-intensive and expensive and therefore unsatisfactory in the large-scale area. Each process step must be checked and validated for its depleting effect on contaminating substances. Many successive process steps also lead to large losses of product, since certain losses must be expected with each step.

The process according to the invention for the production and purification of recombinant Osp, in particular OspC, is maximally selective with a minimum of process steps, as a result of which the product loss is reduced. The method according to the invention can be used for all serotypic variants of Osp, in particular OspC, and can therefore be implemented on an industrial scale. A minimum of procedural steps also requires a minimum of contamination options. The Osp (OspC) purified by the process described in WO 94/25596 contains up to 20% of fragments and degradation products. Due to the small number of necessary process steps, the process according to the invention only leads to a maximum of 5% of fragments and degradation products.

In a special embodiment of the inventive In this way, yeast cells, preferably Pichia pastoris, which express recombinant OspC according to the invention are suspended in a Tris buffer (pH 7.5) and broken up in a Manton-Gaulin homogenizer. The solid components of the cells are separated from the solution by centrifugation. Ethanol is added to the solution up to a final concentration of 50%. After incubation for one hour at room temperature, the precipitate formed is separated from the solution by centrifugation. The solution is diluted to 45% ethanol with Tris buffer (pH 7.5) and applied to a chromatography column filled with Source RP15 ™ (Pharmacia). With an ethanol gradient between 45% and 70%, OspC is eluted from the column, the UV absorption of the effluent being measured continuously. Under these conditions, recombinant OspC elutes at an ethanol concentration between 52% and 57%. The OspC-containing fractions are combined and re-buffered against a physiologically compatible buffer, for example phosphate buffer, by means of gel filtration, for example over Sephadex G25. This is followed by concentration by means of ultrafiltration through membranes with an exclusion limit of 5000 Da, for example Omega 5K (Filtron).

According to a further aspect of the present invention, there is provided a recombinant OspC which is characterized in that it is non-lipidated and consists only of the protein portion of the B. burgdorferi OspC lipoprotein. This non-lipidated OspC protein accordingly represents an OspC apoprotein, ie a lipoprotein which completely lacks the lipid portion. The OspC according to this aspect of the invention is therefore a complete, unfused OspC. Surprisingly, it was found that the recombinant OspC from Pichia pastoris according to the invention is non-lipidated and highly immunogenic. This was particularly surprising since it was found for non-lipidated OspA that its immunogenicity is greatly reduced compared to lipidated OspA (Weis et al., 1994; Infect. Immun. 62: 4632-36). The OspC according to the invention is further characterized in that it is soluble in organic solvents with a protein-precipitating effect, such as, for example, in 50% ethanol. It was shown in the context of the present invention that the expression of a gene coding for non-truncated OspC in yeast, preferably in Pichia pastoris, surprisingly led to an OspC protein which lacked the lipid portion. "Not shortened" means that the recombinant OspC at the amino acid level corresponds to the processed, native OspC and the N-terminal cysteine is present. The reason for the lack of the lipid portion is that Pichia pastoris lacks an enzyme system necessary for this post-translational modification of the OspC, as it occurs in bacteria (Borrelia, E. coli). As a consequence of the absence of the lipid portion, the OspC protein according to the invention is not associated with the membrane of the host cell, but is present in the cytosol of the host cell.

Solvents, as are used according to the present invention, usually serve to precipitate proteins from solutions. It was therefore all the more surprising that the recombinant OspC according to the invention did not coexist with the other proteins, e.g. cellular proteins or protein components of the medium, fails if organic solvents are added, but is soluble in organic solvents. This special property has never been associated with OspC. This particular form of solubility is very important for the development of a new cleaning process for OspC. It is also important for vaccine production that this property applies to all serological OspC variants tested so far.

The recombinant OspC purified according to the invention provides highly efficient vaccination protection against homologous challenge and also surprisingly high vaccination protection against heterologous challenge. The immunogenicity of the recombinant, non-lipidated OspC purified according to the invention is all the more surprising since it is described in the available literature that the Osp proteins from Borrelia are only immunogenic if they carry a lipid portion. The recombinant OspC purified according to the invention is furthermore highly pure and free of all contaminations which are dangerous for humans, such as proteins of the host cell, proteins from the nutrient medium, nucleic acids, pyrogenic and toxic see substances and lipids. Because of the method according to the invention for its extraction and purification, the recombinant OspC according to the invention is 100% free of detergents. The OspC purified according to the invention proves to be completely pyrogen-free according to USP (United States Pharmacopoia) in animal experiments. Its tolerance has also been demonstrated in abnormal toxicity tests according to USP and in tests for dermoreactivity according to USP. It can be safely contained in human vaccines. Because of its high immunogenicity, the OspC according to the invention requires only small additions to adjuvants.

According to a further aspect, the present invention relates to an OspC preparation which can be obtained by the cleaning method according to the invention.

To produce recombinant OspC according to the present invention, a polynucleotide molecule which codes for OspC was introduced into the host cell in a corresponding vector. A polynucleotide molecule which codes for OspC is hereinafter referred to as "OspC gene".

The OspC gene can be prepared from Borrelia by methods known to those skilled in the art and inserted into a vector molecule (PCR, “screening” of gene banks, cloning; for example: Maniatis et al., MOLECULAR CLONING: A LABORATORY MANUAL; Cold Spring Harbor Laboratories (1982)). The OspC gene can also be produced synthetically using a so-called gene assembler. The OspC gene is inserted 3 'from expression-controlling sequences (promoter; enhancer, ribosome binding site) into the vector molecule. Expression-controlling sequences bring about the transcription and translation of a corresponding gene; they can also have an enhancing effect on transcription or translation. The expression-controlling sequences are chosen according to criteria known to the person skilled in the art so that they deliver high production rates in the selected host cell. The production rates of the process according to the invention are in the range from 0.1 mg / g wet cell mass (FZM) to 100 mg / g FZM, preferably in the range from 1 mg / g FZM. The host cell used in the method according to the invention is a eukaryotic cell, preferably a yeast cell. The yeast cell is preferably selected from the following group: Hansenula sp., Saccharomyces cerevisiae, Schizosaccharomyces pombe, Pichia pastoris is particularly preferred. According to the invention, the expression-controlling sequences are chosen such that they are suitable for the expression of recombinant Osp proteins, in particular OspC, in yeast cells. Both constitutive and inducible promoters can be selected for expression. Inducible promoters only become active when an inductor molecule is added; they can be "switched on" specifically for production. Inducible promoters are, for example, the promoters of ADH2, AOX, isocytochrome C, azide phosphatase, enzymes which are associated with nitrogen metabolism, metalothionein, glyceraldehyde-3-phosphatase dehydrogenase and enzymes which are involved in galactose and maltose metabolism are active. Constitutive promoters are, for example, PGKl and ADHl promoters and the promoters of 3-phosphoglycerate kinase and other glycolytic enzymes such as, for example, enolase, hexokinase, pyruvate decarboxylase, phosphofructokinase, glucose-6-phosphate isomerase, pyruvate kinase and glucokinase.

The vector molecule for the expression of Osp (OspC) can be present in the yeast cell in the cytoplasm, for example as YEp, YRp, YCp or YAC, or it can be integrated into the genome of the yeast cell, for example as YIp. Furthermore, so-called "shuttle vectors" are favorable, which can be propagated both in yeast cells and in bacteria, for example E. coli. An example of a "shuttle vector" is pHIL-Al (Phillips Petroleum) which can be propagated in Pichia pastoris and in E. coli (WO 94/25596).

The OspC gene codes for an OspC molecule which is identical to or derived from a Borreia wild-type variant. Derived here means that the gene codes for a recombinant OspC protein which is so far identical to the original variant that antibodies directed against the original variant recognize the recombinant OspC. The gene can be genetically engineered modified, it can contain genetically introduced mutations, and it can contain nucleotide sequences in the 5 'and 3' region which influence its expression in a desired manner. If the host cell is Pichia pastoris, the gene can be placed, for example, under the control of a Pichia-specific promoter, for example the A0X-1 promoter which is inducible with methanol. The gene is contained in a vector for transformation and if necessary also for expression in the host cell. The type of vector here depends primarily on the host cell, and the selection of the correct vector is to be made according to criteria known to those skilled in the art. According to the invention, the vector used should allow efficient expression over a long period of time. If the host cell is Pichia pa storis, the vector is preferably pHIL-Al (Phillips Petroleum) and the OspC gene is under the control of the methanol-inducible AOX-1 promoter. In a special embodiment, the pHIL-Al constructs are constructed as in WO 94/25596, for example pPC-PP4. As in WO 94/25596, the inserted OspC gene comes from the B.burgdorferi strains B31, PKO, ZS7, KL10 or E61, but it can also be obtained from anyone else in Livey et al. (1995, Mol. Microbiol. 18: 257-69) described RFLP type, or be derived.

The present invention provides a new method for obtaining and purifying recombinant OspC protein derived from Borrelia OspC. Derived means that the recombinant OspC according to the invention has so much similarity to a native OspC from Borrelia that it still triggers the production of protective antibodies against Lyme borreliosis in mammals, or that it is still from mammals against OspC from homo ¬ Logenous pathogens of Lyme disease-produced antibodies reacted. The OspC according to the invention can be derived from all serological subtypes of Borrelia OspC. According to the invention, Borrelia is the causative agent of Lyme borreliosis, preferably B.burgdorferi sensu stricto, B.garinii sp. nov and B.afzelii.

The present invention provides a new, recombinant protein derived from Borrelia-OspC, which thereby is characterized that it is not lipidated. "Non-lipidated" means that no lipid portion is covalently bound to the N-terminal cysteine of the OspC. The recombinant OspC according to the invention can represent any serological variant of Borrelia OspC, but also OspC derived from serological variants.

The recombinant OspC according to the invention is furthermore characterized in that it is soluble in an organic solvent under conditions under which contaminating proteins and nucleic acids associated with proteins precipitate. According to the invention, the organic solvent is selected from the group of protein-inducing substances, in particular from the group of mono- or polyhydric alcohols, short-chain ketones, sulfoxides or nitriles. The organic solvent is preferably methanol, n-propanol, 2-propanol or another isomer of propanol, t-butanol, 2-butanol or another isomer of butanol, DMSO, acetonitrile, dioxane or acetone and particularly preferably ethanol . The organic solvent is in an aqueous buffer solution. The organic solvent can be a single organic solvent in aqueous buffered solution, or it can be a mixture of different organic solvents. The aqueous buffer solution can be, for example, a phosphate-buffered sodium chloride solution, Tris buffer or citrate buffer. According to the invention, the organic solvent is present in a concentration of 10% to 90%, preferably 20% to 70% and particularly preferably 40% to 60% in aqueous buffer solution. The aqueous buffer solution can be, for example, a phosphate-buffered sodium chloride solution, Tris buffer or a citrate buffer.

In a preferred embodiment, the organic solvent is a mono- or polyhydric alcohol, particularly preferably ethanol. Ethanol is present in an aqueous buffer solution in a concentration between 40% and 60%, preferably between 45% and 55%.

The recombinant OspC according to the invention is preferably produced using the method for obtaining and purifying prepared by Osp. It is free of cellular proteins, pyrogenic and toxic substances, nucleic acids and lipids. Since the recombinant OspC according to the invention was prepared without the addition of detergents, it is also 100% free of detergents.

The recombinant OspC produced according to the invention is highly pure. It has a purity of at least 90%, preferably 95%, and particularly preferably 98%. The purity was demonstrated using common biochemical methods (HPLC, electrophoresis and mass spectroscopy) (see Example 3).

According to the invention, the recombinant OspC can be a single serological variant, but it can also be an OspC mixture composed of several different variants. The various serological variants are usually prepared separately using the method according to the invention. Since the procedure can be used uniformly for all serological variants of OspC, it is also possible to clean different variants at the same time. The advantage of the method according to the invention is not only that the method does not have to be adapted to the respective serotypes, but also that the different variants can be mixed without problems before or after the preparation. This is particularly useful for the efficient and uniform production of a polyvalent vaccine, which is necessary for the prevention of Lyme disease.

Functional studies have shown that the recombinant OspC according to the invention is highly immunogenic. The recombinant OspC according to the invention induces the formation of antibodies in mice. Furthermore, the non-lipidated OspC mice according to the invention protect one hundred percent against homologous challenge. However, it also provides surprisingly high protection against a heterologous challenge, ie it provides protection against Borrelia strains whose OspC variants were not present in the vaccine. If a single OspC variant was used as the vaccine antigen, up to 70 percent protection could be observed (see Example 5). When a mixture of 3 different OspC variants was administered as vaccine antigen, the heterologous protection was 80 percent (see example 6). These results indicate that a vaccine that efficiently protects against Lyme disease should preferably contain several different OspC variants. Because of its broad applicability, the process according to the invention for the purification of OspC is particularly favorable for the production of such a polyvalent mixed vaccine. The high immunogenicity of the non-lipidated OspC according to the invention shown in Examples 5 and 6 also stands in contrast to the widespread assumption in the prior art that the Osp proteins from Borrelia only show efficient immunogenicity if they have one Wear lipid portion (WO 93/08299, Weis et al. (1994), Infect. Immun. 62: 4632-36; Erdile et al. (1993), Infect. Immun. 61: 81-90).

The recombinant OspC according to the invention was able to be produced with the process according to the invention in particularly high purity and free of all contaminations dangerous to humans, such as proteins of the host cell, proteins from the nutrient medium, nucleic acids, pyrogenic and toxic substances, lipids and detergents become. In animal experiments (pyrogenicity test, anomalous toxicity test, dermoreactivity) proof of tolerance was also provided. The recombinant OspC according to the invention can be safely contained in vaccine preparations to be used on humans. It is therefore very suitable for pharmaceutical use. Pharmaceutical use here means that it can be administered to mammals, in particular humans, in a corresponding physiological vehicle and with any substances added, and that it has a prophylactic or therapeutic effect. The prophylactic or therapeutic effect of OspC is that it is a vaccine antigen due to its immunogenicity. A vaccine antigen is an antigen that elicits a specific immune response in a mammal. This immune response leads to humoral and / or cellular immunity. Immunity means that the mammal is protected against a Borrelia infection or that the Borrelia infection is easier than in non-immune mammals. The prophylactic effect comes into effect when a healthy mammal is immunized. OspC is contained in a vaccine that provides immunological protection against Lyme disease in healthy mammals. The therapeutic effect Kung can be used if the mammal is already suffering from a Borrelia infection and the OspC present in the vaccine strengthens the immune response against Borrelia and thus supports the healing process. Because of its high purity, the OspC according to the invention is also very suitable as a vaccine antigen for children.

The recombinant OspC according to the invention can be used in a suitable manner as an immunogen for the production of antisera or antibodies. For this purpose, test subjects can be immunized with an immunologically effective dose of vaccine antigen and the immunoglobulins isolated and purified by methods known in the art. In addition, by detecting the formation of antibodies against recombinant OspC protein, a statement can be made about the success of the vaccination.

Physiological buffer solutions, such as phosphate buffer or citrate buffer, are usually used as the physiological carrier. The substances that may be added include adjuvants, for example aluminum compounds such as aluminum hydroxide and aluminum phosphate, Freund's adjuvant or other mineral oil or vegetable oil emulsions, lipopolysaccharides, lipid A, saponin, Quil A, QS21 and ISCOMS. The preferred adjuvant is aluminum hydroxide. In addition to adjuvants, carriers can also be added to the vaccine. Typical carriers are liposomes, latex particles or bentonite. In order to increase the immunogenicity of the vaccine, further immunostimulating substances, such as interleukins, for example IL-1 or IL-2, can also be added.

Furthermore, the substances that may be added also include other antigens, for example other Borrelia antigens or viral antigens. Other Borrelia antigens can be, for example, OspA, OspB, OspE or OspF (Nguyen et al. (1994); Infect. Immun. 62: 2079-84) or derivatives thereof. Other viral antigens can be, for example, antigens of the TBE virus. Since both the TBE virus and Borrelia are transmitted by ticks, such a combination is suitable for a vaccine for people living in tick-contaminated areas. Furthermore, so-called "carrier substances" can be added to the vaccine. "Weary gerSubstanzen "are, for example, tuberculin PPD, serum albumin, ovalbumin or keyhole limpet hemocyanin.

All substances present in the vaccine in addition to the vaccine antigens are preferably non-toxic and non-allergenic.

The formulation of vaccines is known to the person skilled in the art. The vaccine antigen of the present invention can be lyophilized and can only be dissolved in a physiologically acceptable buffer solution for use. Physiological salt solutions or other physiological solutions are suitable as a buffer solution. The vaccine antigen is mixed with the physiological buffer solution in order to obtain a desired concentration of the vaccine antigen. The desired concentration is immunologically effective; it is in the range from 1 μg to 100 μg of the respective vaccine antigen per dose, preferably in the range from 10 μg to 50 μg. If necessary, the vaccine antigen dose can be reduced in a children's vaccine, for example to a dose in the range between 5 and 30 μg.

Finally, the present invention provides a vaccine which contains recombinant, non-lipidated OspC according to the invention. The vaccine is suitable for administration to mammals, preferably humans. Depending on the region and the desired width of the vaccination protection, the vaccine contains one or more serologically different variants of the OspC according to the invention. The serological variants correspond to the 35 by Livey et al. (1995, Mol. Microbiol. 18: 257-69) described RFLP types. Vaccine compositions can, however, also correspond to the region-specific compositions described in WO 94/25596, containing various OspC variants.

In one embodiment, the vaccine contains three to eight different serological variants of the OspC according to the invention and provides broad heterologous protection against the Borrelia serotypes occurring in a specific area. In a particular embodiment, the vaccine contains three different serological variants of OspC, for example a composition of the RFLP variants Orth, PKO and E61 or the RFLP variants ZS7, PKO and E61. However, it is any composition of two or more RFLP variants according to Livey et al. (1995, Mol. Microbiol. 18: 257-69) possible.

The present invention is further illustrated by the following examples and the drawing figures, but without being restricted thereto.

Show it:

1 to 3 precipitation optimizations;

Fig. 4 an elution optimization and

5 to 8 purity studies of conventionally or according to the invention obtained OspC proteins

EXAMPLES:

Example 1: Expression of recombinant OspC (rOspC) in Pichia pastoris.

Pichia pastoris strain GS115 NRRL-Y 11430 (Cregg et al. (1985); Mol. Cell. Biol. 5: 3376-85) was prepared according to the method of Dohmen et al. (1991; Yeast 7: 691-92) with the vector pPC-PP4 (described in WO 94/25596), which contains the OspC gene of the Borrelia burgdorferi strain Orth, the strain ACA1 or the strain ZS7. Transformants were grown at 30 ° C in MD medium to an optical density (OD 600) of 2-10. Induction and analysis of the OspC expression were carried out as described in WO 94/25596.

Example 2: Purification of rOspC from Pichia pastoris. (According to the applicant, example 2 is currently the best form for carrying out the invention)

500 g of wet cell mass of the transformed rOspC (strain Orth or ZS7) expressing Pichia yeast cells was in 500 ml Tris-Puf- fer (20mM Tris-HCl, pH 9.0) suspended and broken up in a Manton-Gaulin homogenizer. Solid components of the yeast cells were removed from the suspension by centrifugation (15 min at 10,000 g).

95% ethanol was added with stirring until a final concentration of 50% ethanol was reached. After an hour's incubation at 4 ° C., the precipitate formed was removed by centrifugation (15 min at 20,000 g). The clear, cell-free supernatant was diluted with Tris buffer (20 mM Tris-HCl, pH 9.0) to an ethanol concentration of 45% and then applied to a chromatography column filled with Source RP15 ™ (Pharmacia). The column was washed with 45% ethanol in phosphate buffered sodium chloride solution.

ROspC was eluted from the column with an ethanol gradient of 45% to 70% in phosphate-buffered sodium chloride solution. The UV absorption of the effluent was measured continuously at a wavelength of 214 nm. rOspC eluted at an ethanol concentration of 55%. OspC positive fractions were pooled. The running buffer was replaced by gel filtration over Sephadex G25 with a Tris-HCl + 0.9% NaCl buffer (pH 7.4). The solution was then concentrated by ultrafiltration over an Omega 5K membrane (Filtron).

Example 3: Analysis of rOspC purified according to Example 2.

The yield and purity of the rOspC purified according to Example 2 were determined by means of the BCA method, HPLC, SDS-PAGE, Limulus amebocyte lysate method, gas chromatography and phenol / H2S04 method. The results are summarized in Tables 1 (strain Orth) and 2 (strain ZS7). The example shows that rOspC proteins of very different subtypes (Orth, ZS7: homology = 69.2%) can be efficiently cleaned with the method according to the invention. The values described in the tables demonstrate the high purity of the respective rOspC. Table 1

Analysis of rOspC of the strain Orth purified according to Example 2,

Property method applied

Protein BCA method 1.5 mg / ml

HPLC purity 98%

Purity SDS-PAGE> 95%

Endotoxin content LAL <1.5 EU / ml

Lipid gas chromatography <0.01 mg / ml

Carbohydrates phenol / H ₂ SO ₄ <0.02 mg / ml

Table 2:

Analysis of rOspC of strain ZS7 purified according to Example 2.

Property method applied

Protein BCA method 0.9 mg / ml

HPLC purity 98%

Purity SDS-PAGE> 95%

Endotoxin content LAL <1.5 EU / ml

Lipid gas chromatography <0.01 mg / ml

Carbohydrates phenol / H ,, SO ₄ <0.02 mg / ml

Example 4: Comparison of different organic solvents for cleaning rOspC from Pichia pastoris.

1 ml each of a suspension of yeast cells broken open according to Example 2, which expressed rOspC of the Orth strain, were mixed with the organic solvents in Table 3 using various organic solvents specified concentrations brought. After an hour's incubation at 4 ° C, the precipitate formed is centrifuged off (15 min, 20,000 g). An aliquot of the supernatant is subjected to protein analysis using SDS-PAGE. The yield of rOspC is determined by densitometric evaluation of the SDS-PAGE gel stained with Coomassie blue. The yield is based on the value of the precipitation with ethanol. The example shows that in addition to ethanol, other organic solvents, for example t-butanol, 2-butanol, acetonitrile and dioxane, could also be used as precipitation reagents for contaminating substances, rOspC being obtained predominantly as a soluble protein in the precipitation supernatant.

Table 3:

SOLVENT YIELD (in%) PURITY (in%)

(Final conc. In%) (after the 1st precipitation step) (after the 1st precipitation step)

Ethanol (50%) 100 80

2-propanol (50%) 70 25 t-butanol (50%) 80 65

Acetonitrile (40%) 80 60

Dioxane (50%) 70 60

Example 5: Immunization of mice with rOspC.

C3H / HeJ laboratory mice were immunized twice with a dose of 3 μg rOspC in a formulation with Al (OH) ₃ (0.2%). The antibody titer against the immunization antigen was determined by ELISA using Borrelia OspC. Two weeks after the second immunization, the animals were challenged with a dose of 10 ⁴ infectious Borrelia (10 id 50) of the homologous or a heterologous strain. After a further week, blood and several organs (heart, bladder, kidney, spleen) were taken from the mice for diagnosis of infection. The results are summarized in Table 4; they show the excellent protective effect which is achieved with rOspC purified according to Example 2 against a homologous challenge. All three tested rOspC antigens (from the strains Orth, ZS7 and ACA1) provided 100% protection against homologous challenge. Table 4 also shows that a partial protective effect of up to 70% could be achieved in the heterologous challenge.

Table 4:

ANTIGEN CHALLENGE TITER INFECTED ANIMALS (strain) (10 ⁴ Germs) before challenge (%) no orth <1: 1000 100

Orth Orth 1: 50000 0

ZS7 ZS7 1: 50000 0

ACA 1 ACA 1 1: 50000 0

Orth ZS7 1: 50000 80

ZS7 Orth 1: 50000 30

Example 6: Immunization of mice with various mixtures of recombinant OspCs.

As described in Example 5, C3H mice were immunized, but instead of 3 μg of a single rOspC variant, mixtures of different rOspC variants were administered in this experiment. The composition of the various mixtures is summarized in Table 5. The total dose of antigen was 9 μg in all cases. The challenge was carried out with different Borrelia strains (one homologous and one heterologous strain). The results of this experiment are summarized in Table 5. The results show that the mixed vaccine, which was composed of three different OspC variants, not only protected against the Borrelia strains whose OspC variants were present in the vaccine, but also to an unexpectedly high percentage (80%) against a heterologous Borrelia Tribe. Table 5:

Example 7: Precipitation optimization for rOspC

The optimal precipitation conditions can vary from solvent to solvent, especially as regards the amount of organic solvent added. According to the invention, the optimal concentration of the respective organic solvent is selected as the concentration at which the cellular proteins are essentially completely precipitated, but the Osp protein has not yet or only to a small extent precipitated.

To determine the optimum precipitation concentration of various organic solvents, a recombinantly produced culture liquid of rOspC as the starting solution with methanol, 2-propanol, t-butanol, DMSO, acetonitrile and dioxane as solvent with different concentrations (10%, 20%, 40% and 50%) treated and the result checked by means of gel electrophoresis (see FIGS. 1 and 2).

It was shown that optimal separation of the accompanying proteins from the OsP protein by precipitation with methanol at 40% with 2-propanol at 50%, with t-butanol at 50%, with DSMO at 40%, with acetonitrile 40% and with dioxane with 50%. Example 8: Precipitation optimization for ethanol

An execution solution containing recombinant OspC was treated with 30%, 40%, 45% and 50% ethanol and the solutions obtained were checked by means of SDS gel electrophoresis and Western blot analysis.

The results are shown in FIG. 3 and show that with ethanol between 40% and 50% an optimal precipitation for rOspC can be achieved.

Example 9: Optimization of the adsorption treatment

To optimize the adsorption treatment, the elution of an Osp protein adsorbed by a solid support with a linear ethanol gradient between 45% and 70% was investigated. The results are shown in Fig. 4 and in the table below.

Table 6:

Peak RT type wide area height area

# [min] [min] [mAO x s] [mAO]% 1 - - 1 - 1 1 1. 1.

1 7,129 w 0.080 8,27969 1,62639 0.1677

2 7,666 PV 0.070 25.54446 4 .91349 0.5174

3 7,883 w 0.099 4851.37891 684 .38525 98.2665

4 8,449 VB 0.049 5.95952 1. .73231 0.1207

5 9,556 w 0.118 45.79729 5 .07893 0.9276

Total 2 4936.95996 697. .73639 Example 10: Comparison test to WO 94/25596

To compare the OspC preparations obtained according to WO 94/25596 and the invention, recombinant OspC from strain 0rth / F17 and from strain ACA1 was determined using the "conventional" method according to WO 94/25596 or with the inventive measure method cleaned up. The results are listed in FIGS. 5 and 6 and tables 7 and 8 (for the method according to WO 94/25596) and in FIGS. 7 and 8 and tables 9 and 10 (for the method according to the invention).

It was found that with the conventional cleaning process different degrees of cleaning of the OspC product can be obtained, which can vary between a purity of 56% and 80%. A recombinant OspC from strain 0rth / F17 with an 80% purity and from strain ACA1 with a 56% purity could be obtained using the method according to WO 94/35596.

In contrast, with the cleaning method according to the invention, a significantly improved purity of the preparation of recombinant, non-lipidated OspC can be obtained, which is at least 95% and can even reach 99% (see FIG. 8 and Table 10).

In addition, it should be emphasized that the cleaning method according to the invention can be applied to all previously known OspC variants, with which a uniform degree of purification of the antigens can be obtained for all variants, which was not possible by the previously described methods. By obtaining different OspC variants with a uniformly high degree of purification, it is therefore also possible for the first time to produce preparations with a purity of at least 90%, in particular by mixing at least 3 different serological OspC variants, a vaccine preparation can be produced which also protects against a challenge with a heterologous Borrelia strain. A decisive breakthrough in the production of Borrelia vaccines with a wide protection range is thereby achieved. In contrast, WO 94/25596 was found studies in the Comparison according to that obtained from yeast OspC preparation _l ediglich protects against infection with the homologous strain.

Table 7

Peak RT type wide area height area

# [min] [min] [mAO x s] [mAO]% 1 - - 1 - 1 .1 -_ | | _ 1

1 4,628 PV 0.229 177.69142 10.31256 3.6800

2 5,075 w 0.076 23.25017 3.97309 0.4815

3 5,131 w 0.051 17.73720 4.89392 0.3673

4 5,180 w 0.068 23.47639 4.70628 0.4862

5 5,410 w 0.071 102.99156 20.97543 2.1329

6 6,296 w 0.054 16.66883 4.29893 0.3452

7 6,456 w 0.102 27.91520 3.57525 0.5781

8 7,124 w 0.102 96.15434 12.05207 1.9913

9 7,340 w 0.102 51.24856 6.48721 1.0614

10 7,456 w 0.077 36.35993 6.25955 0.7530

11 7,634 w 0.072 2449.22803 516.16937 50.7232

12 7,721 w 0.092 1456.86633 237.46468 30.1715

13 8,150 w 0.063 32.59999 7.25347 0.6751

14 8,247 w 0.065 69.71961 15.43299 1.4439

15 8,748 w 0.082 37.33817 6.13417 0.7733

16 8,847 w 0.055 110.15652 29.61396 2.2813

17 9,413 w 0.051 35.14756 10.23840 0.7279

18 10.066 BV 0.056 41.27703 10.55653 0.8548

19 10.310 w 0.058 22.78672 5.92479 0.4719

Total 4,828,61377 916.32269

Table 8

Peak RT type wide area height area

# [min] [min] [mAO x s] [mAO]% 1 1- __

1 3,666 w 0.104 76.72591 11.26516 1.8058

2 3,776 w 0.035 13.63044 6.43005 0.3208

3 3,840 w 0.051 32.56386 9.11639 0.7664

4 3,891 w 0.067 44.94176 8.83538 1.0577

5 4,117 w 0.095 136.84679 18.46668 3.2207

6 4,216 w 0.049 32.40656 9.43964 0.7627

7 4,293 w 0.059 42.06274 9.72205 0.9900

8 4,708 w 0.271 1011.13867 49.68585 23.7974

9 5,124 w 0.151 153.73065 13.06296 3.6181

10 5,327 w 0.083 60.29200 10.13262 1.4190

11 5,421 w 0.086 76.35560 11.45869 1.7970

12 5,664 w 0.073 29.88213 5.62164 0.7033

13 6,072 w 0.070 48.29964 9.32081 1.1367

14 6,461 w 0.055 14.11311 3.63805 0.3322

15 6,801 w 0.076 10.21406 2.13184 0.2404

16 7,689 PV 0.078 34.36339 6.54271 0.8087

17 7,854 w 0.086 2392.88062 403.43259 56.3169

18 8,257 w 0.090 23.96358 3.49553 0.5640

19 8,854 w 0.044 9.72101 3.31730 0.2288

20 9,415 BV 0.050 4,82040 1,36065 0.1134

Total _: 4248.95313 596.47662

Table 9

Peak RT type wide area height area

# [min] [min] [mAO xs] [mAO]%

1 4,670 BV 0.197 53.18630 3.46254 1.9235

2 5,028 W 0.034 1.48847e-l 7.31092e-2 0.0054

3 5,423 W 0.036 5.55494 2.34552 0.2009

4 6.457 W 0.073 5.84710 1.07350 0.2115

5 6,629 PV 0.041 9.57081e-l 3.55412e-l 0.0346

6 7,124 W 0.056 3.70759 9.01566e-l 0.1341

7 7.364 BV 0.066 10.94985 2.37225 0.3960

8 7,503 W 0.046 9.87367 3.15355 0.3571

9 7,640 W 0.069 1879.25562 418.27679 67.9650

10 7,760 W 0.087 671.17523 113.45638 24.2737

11 7,982 W 0.116 99.16217 11.69299 3.5863

12 8,259 W 0.057 8,63425 2.11423 0.3123

13 8,855 W 0.046 9.47382 3.07899 0.3426

14 10,070 W 0.068 7.10788 1.47169 0.2571

Total . 2765.03442 563.82855 Table 10

Peak RT type wide area height area

# [min] [min] [mAO x s] [mAO]%

| __ _ j 1 ι_ 1

1 7. 823 PV 0.085 3995.36450 690.66522 99.3560

2 8. 424 VB 0.147 25.89507 2.92927 0.6440

Total: 4021.25952 693.59448

Claims

Patent claims:

1. A process for the production and purification of recombinant Osp protein derived from Borrelia surface lipoproteins from eukaryotic host cells, characterized in that a host cell lysate is mixed with a protein-precipitating organic solvent under conditions in which cellular proteins in the are substantially completely precipitated and recombined Osp protein remains selectively in solution, and the Osp protein is purified from the solution and recovered.

2. The method according to claim 1, characterized in that the precipitated, cellular proteins are separated from the rekom¬ binant Osp protein by sedimentation and recombinant Osp protein remains in solution.

3. The method according to any one of claims 1 or 2, characterized gekenn¬ characterized in that the solution containing Osp protein is brought into contact with a solid carrier under conditions in which the Osp protein selectively binds to the carrier and that it is subsequently with an organic solvent is eluted.

4. A process for the purification of Osp proteins from an Osp-containing starting material, characterized in that the Osp protein is selectively bound to a solid support suitable for reverse phase chromatography, optionally washed, and eluted.

5. The method according to claim 3 or 4, characterized in that the solid carrier is a silicate, preferably Spherisorb 5C18, or a plastic polymer, preferably Source PR15 ™ or Amberchrome®.

6. The method according to any one of claims 3 to 5, characterized gekenn¬ characterized in that the elution is carried out by means of a gradient of an organic solvent.

7. The method according to any one of claims 1 to 6, characterized gekenn¬ characterized in that the organic solvent is selected from the group of protein-precipitating substances, in particular mono- or polyhydric alcohols, short-chain ketones, sulfoxides or nitriles, preferably methanol, n-propanol, 2-propanol and other isomers of propanol, t-butanol, 2-butanol and other isomers of butanol, DMSO, acetonitrile, dioxane or acetone and particularly preferably ethanol.

8. The method according to any one of claims 1 to 7, characterized gekenn¬ characterized in that a mixture of different organic solvents is used.

9. The method according to any one of claims 1 to 8, characterized gekenn¬ characterized in that the organic solvent used in a concentration of 10% to 90%, preferably 20% to 70% and particularly preferably 40% to 60% in aqueous buffer solution becomes.

10. The method according to any one of claims 1 to 9, characterized gekenn¬ characterized in that a eukaryotic host cell from the group of yeasts, preferably Hansenula sp. , Saccharomyces cerevisiae, Schizosaccharomyces pombe and particularly preferably Pichia pastoris, is used.

11. The method according to any one of claims 1 to 10, characterized gekenn¬ characterized in that OspC is obtained as the Osp protein.

12. Recombinant OspC, characterized in that it is non-lipidated and consists only of the protein portion of the B. burgdorferi OspC protein.

13. Recombinant OspC, characterized in that it is non-lipidated and free of cellular proteins, pyrogenic and toxic substances, nucleic acids and lipids and is soluble in a protein-precipitating, organic solvent.

14. Recombinant OspC according to claim 13, characterized in that the protein-precipitating organic solvent is a mono- or polyhydric alcohol, preferably ethanol.

15. Recombinant OspC according to claim 14, characterized in that ethanol is present in an aqueous buffer solution in a concentration between 40% and 60%, preferably between 45% and 55%.

16. Recombinant OspC according to one of claims 12 to 15, characterized in that it has a purity of at least 90%, preferably 95% and particularly preferably 98%.

17. Recombinant OspC according to one of claims 12 to 16 for pharmaceutical use.

18. Vaccine, characterized in that it contains recombinant, non-lipidated OspC, and optionally a physiologically acceptable carrier.

19. Vaccine according to claim 18, characterized in that it additionally contains an antigen selected from the group of

B.burgdorferi antigen OspA, OspB, OspD, OspE or OspF, or of TBE virus, contains.

20. Vaccine, characterized in that it contains recombinant, non-lipidated OspC according to one of claims 12 to 17 and optionally a physiologically acceptable carrier.

21. Use of recombinant, non-lipidated OspC for the production of a vaccine for the immunization of mammals against Lyme disease.

22. OspC preparation, obtainable by a process according to any one of claims 1 to 11.