AT505263B1 - PROCESS FOR PRODUCING ASCORBIC ACID - Google Patents

Description

teiÄses patcBiamt AT505 263 B1 2009-08-15

Description: The present invention relates to a process for the production of ascorbic acid, in which, in a first step, a carbon source is fermentatively converted to 2,5-diketo-D-gluconic acid, which then in a second step enzymatically with 2,5-diketo D-gluconic acid reductase from Corynebacterium glutamicum is converted to 2-keto-L-gulonic acid.

Ascorbic acid is used in the pharmaceutical and food industries as a vitamin and antioxidant application. L-ascorbic acid is currently produced predominantly according to the process developed by Reichstein (1934). After this manufacturing process of Reichstein obtained vitamin C from D-glucose in a chemical-biocatalytic process. The glucose is first reduced to sorbitol and then oxidized with sorbose bacteria to the carbohydrate L-sorbose. The sorbose is further oxidized with the addition of acetone, wherein after the subsequent cleavage of the acetone and dehydration, the vitamin C is formed. The Reichstein synthesis requires not only high energy and chemical expenditure, but also sophisticated wastewater treatment in order to avoid the release of lindane-like by-products. Furthermore, this process provides only a yield of about 50%.

In recent years, increased efforts have been made to develop biotechnological processes for the production of ascorbic acid. However, the economic breakthrough with such methods has not succeeded in these experiments. The majority of known biotechnological processes are limited to the preparation of intermediates of the ascorbic acid synthesis pathway. Of particular interest are 2,5-diketo-D-gluconic acid (2,5-DKG) and 2-keto-L-gulonic acid (2-KLG). 2,5-DKG is therefore an important intermediate in the production of ascorbic acid because it can be selectively and stereospecifically reduced to the 2-KLG, which in turn is a precursor of ascorbic acid. Subsequently, 2-KLG can easily be converted to ascorbic acid by an acid- or base-catalyzed reaction.

The preparation of the intermediate 2,5-diketo-D-gluconic acid (2.5-DKG) by means of aerobic microorganisms belonging exclusively to the genera Acetobacter, Acetomonas, Gluco-noacetobacter and Pseudomonas, for example, in US Patent 4,316,960 (Pfizer Inc, 1982). However, the use of these microorganisms is unsatisfactory from an industrial point of view because they provide only a low yield of 2,5-DKG at relatively long fermentation times. US Patent 3,790,444 (DAIICHI SEIYAKU CO, Japan, 1974) discloses the production of 2,5-DKG by Acetobacter fragum (ATCC No. 21409) at relatively good yields. Sonoyama et al. (U.S. Patent 4,879,229, Shionogi & Co., Ltd., Japan, 1989) disclose the use of facultative anaerobic strains of the genus Pectobacter to produce 2,5-DKG. In this case, a 2,5-DKG-producing microorganism of the genus Pectobacter or any ingredient thereof (such as enzyme extracts) or immobilized cells is brought into contact with D-glucose.

Non-fermentative, biocatalytic processes for preparing ascorbic acid intermediates have also been described. Thus, EP 1141368B1 (Genencor International, Inc., California, 2006) describes the biocatalytic production of 2-KLG from a carbon source, wherein the co-factor required for the reaction is regenerated. An alternative method is disclosed, for example, in U.S. Patent 4,757,012 (Genentech Inc. California), which discloses the purification and recombinant production of 2,5-diketo-gluconic acid (2,5-DKG) reductase and their use for the reductive conversion of 2,5-diketo-gluconic acid in 2-keto-L-gulonic acid (2-KLG).

The invention is therefore based on the object to provide a method of the type mentioned above, which on the one hand represents an economical alternative to the biotechnological methods described so far and on the other hand, the disadvantages of the Reichstein 1/9 ijitirriiWiSF'fiS patenmamt AT505 263B1 2009- 08-15

Process overcomes.

According to the invention, the invention is achieved in that the carbon source is glucose and the conversion of glucose into 2,5-diketo-D-gluconic acid by the Pectobacter cypripedii strain HeP01 (DSM 12939) is direct.

The advantage of the present invention is that the invention Pectobacter cypripedii strain HEP01 can convert glucose directly into 2,5-diketo-D-gluconic acid (2.5 DKG) and thus represents a further development of the prior art. So far it was by means of Pectobacter spp. it is only possible to ferment gluconic acid, which is considerably more expensive than glucose (gluconic acid costs about twice the amount of glucose), into the intermediate 2.5-DKG. The new Pectobacter strain HEP01 differs from the previous strain described in the prior art (DSMZ 30182) in that it is capable of nitrate reduction, growth on malonic acid, and acid formation from L-rhamnose and D-maltose due to physiological and biochemical characteristics is and also has a urease activity. In contrast, strain HeP01 (DSM 12939) does not have starch hydrolysis activity.

In a preferred embodiment of the invention, in the first step, the air supply can be reduced or adjusted until anaerobic conditions in the nutrient solution, when the formation of 2,5-diketo-D-gluconic acid from the intermediate 2-keto-D -Gluconic acid is absent. This ensures that substantially all of the 2-keto-D-gluconic acid is converted to 2,5-diketo-D-gluconic acid, which is the starting substrate for subsequent reduction to 2-keto-L-gulonic acid, giving a higher total Product yield can be achieved.

In a further embodiment of the invention, the air supply can be controlled via an oxygen probe. By monitoring the oxygen content in the nutrient solution using an oxygen probe, it is achieved that, on the one hand, the oxygen content in the nutrient solution corresponds to the particular optimum of the microorganism used, and, on the other hand, that required for complete conversion of 2-keto-D-gluconic acid into 2,5-diketone. Controlled D-gluconic acid necessary anaerobic conditions.

In yet another embodiment of the invention, the pH can be monitored and / or regulated during the fermentation of the first step. The formation of gluconic acid leads to a lowering of the pH, whereby the growth of microorganisms is suppressed. By controlling the pH, it is ensured that the pH of the nutrient solution is always within the optimum range for the growth of the microorganisms and thus the healthy growth of the microorganisms is maintained.

In an alternative embodiment of the invention, the first step of the method can be carried out in batch mode. In this case, all ingredients are introduced into the fermenter and fermented in a conventional manner by the microorganisms.

Preferably, however, the first step of the process can be carried out in fed-batch mode. With fed-batch operation, a controlled substrate feed avoids an excessive lowering of the educt concentration, so that the microorganisms are always in the exponential growth phase. Furthermore, it is possible to control the growth rate of the microorganisms depending on the chosen delivery strategy. This prevents the formation of growth-associated secondary metabolites.

In a further development of the present invention can be provided that the coenzyme NADPH required in the second step by the glucose dehydrogenase (GDH), which converts glucose into gluconic acid, is regenerated, wherein the resulting by the regeneration of gluconic acid, and optionally remaining Glucose and / or 2,5-diketo-D-gluconic acid of the method according to the invention is returned to the first step. Continued regeneration of the coenzyme NADPH may result in a lower coenzyme concentration than would be required for the complete conversion of 2,5-DKG to KLG without coenzyme regeneration ,

In yet another embodiment of the invention may be used in the second step as a buffer Na citrate or Na acetate. The use of the Na citrate or Na acetate buffer not only reduced the cost of the buffer system, but also increased the enzyme activity compared to the Bis-Tris system described in the literature.

In a further development of the present invention can be provided that the Gluconsäu-re and optionally glucose and / or 2,5-diketo-D-gluconic acid prior to recycling by chromatographic separation methods, such as ion exchange chromatography, of the 2-keto-L Gulonic acid product stream are separated. By separating 2-keto-L-gulonic acid from the product stream, residual glucose or gluconic acid can be recycled to the first step of the process to achieve higher substrate utilization.

In a further embodiment of the invention, the 2,5-diketo-D-gluconic acid reductase from Corynebacterium glutamicum can be produced recombinantly using a pET vector in E. coli.

In yet another embodiment of the invention, for the preparation of the in-sert DNA as a primer for the PCR CglakrPETF and CglakrPETR and by incorporating the coding for the 2,5-diketo-D-gluconic acid reductase gene in the commercial obtained vector pCR®4-TOP® plasmid pPC1 can be used as a template, wherein the amplified fragment is inserted into the expression vector pET-21d. In yet another embodiment of the invention, the E. coli BL21 (DE3) strain transformed with pET-21d can be transfected into a minimal medium MPC-Gly containing 1 g peptone from casein, 10 g glycerol, 0.1 ml 1M CaCl 2, 1 M MgS04 * 7H20, and 200ml MPC-Gly salts, with the MPC-Gly salt solution in one liter of water containing 15g KH2PO4, 2.5g NaCl, and 5g NH4Cl.

In continuation of the present invention, it may be conducted as a hybrid process wherein the reaction mixture is separated into a product stream and a biocatalyst stream by a membrane separation process. This ensures that the enzymes can be performed in the process cycle without major losses.

Finally, in one embodiment of the invention, the enzymes diketo-gluconic acid reductase and glucose dehydrogenase and the coenzymes NADPH and NADP + can be retained by a charged ultrafiltration membrane. The separation of the product stream and the biocatalyst stream through an ultrafiltration membrane results in the enzymes or cofactors being present throughout the process in the enzyme reactor, and need not be constantly replaced by new enzymes or cofactors.

The object of the invention will be described below with reference to a preferred embodiment.

As mentioned above, the production of ascorbic acid by the process of the present invention is carried out in a two-stage process. In a first step, glucose is fermentatively converted to 2,5-diketo-D-gluconic acid (2,5-DKG). For this purpose, a Pectobacter cypripedii strain (HeP01) was isolated from blueberries and deposited with the German Collection of Microorganisms and Cell Cultures GmbH under the Budapest Treaty and received the accession number DSM 12939. Starting from a Stichagarkultur of the organism Pectobacter cypripedii a corresponding number of tryptone Soy Agar (TSA) plates are incubated at 28 ° C for 24 to 48 hours. For the preparation of plates commercial TSA was used, such as Oxoid CM 131 and prepared according to the manufacturer's instructions. The TSA plates were then used to prepare a preculture. For this, 12, 17 g corn steep liquor, 2.76 g glucose monohydrate, 0.27 g KH2PO4, 0.06 g MgSO4 * 7H2 O and 0.27 g NaCl were slurried with tap water to a total volume of about 200 ml , The pH of the medium was adjusted to 7.02 using caustic soda (1 OOg / L). Then the medium was brought to a total volume of 250 ml with tap water. 1.25 g CaCO 3 were weighed out into 300 ml Erlenmeyer flasks and 50 ml each of the medium described above were added and sterilized at 121 ° C. for 20 minutes. After cooling, the Erlenmeyer flasks were in each case inoculated by means of a loop with the Pectobacter cyripedii strain of the TSA plates. The Erlenmeyer flasks were then incubated for 16 hours on a shaking incubator at 200 rpm at 28 ° C. The resulting preculture is used for the subsequent fermentation. The fermentation itself was carried out in a stirred tank fermenter with a total volume of 10 liters, which is equipped with a pH, pO 2, temperature and Gärölelektrode and associated measurement and control electronics. Furthermore, the CO 2 and O 2 content in the exhaust air of the fermenter was monitored.

Both batch and fed-batch fermentation were analyzed by conventional methods. These include the determination of the optical density, the enzymatic determination of glucose and gluconic acid and HPLC. For the determination of the optical density, 1 ml of sample was admixed with 8 ml of citric acid monohydrate solution (40 g / l) and made up to 10 ml. The optical density was measured at 600 nm. The glucose or gluconic acid were determined photometrically using a commercially available enzyme assay at 340 nm (Boehringer, order numbers 716251 and 428191, respectively). For HPLC, an LC 10 system (Shimadzu LC 10) with a Biorad HPX87H (300 mm) separation column was used, with two columns connected in series for better separation efficiency. The column (40) was eluted with 0.005N H 2 SO 4 and the individual fractions (peaks) were measured with UV (Shimadzu SPD-10A) and RI (Shimadzu RID-10A) detectors connected in series. The said HPLC method was used for the analysis of glucose / gluconic acid, 2,5-DKG, 2-keto-D-gluconic acid, lactic acid and acetic acid, respectively.

The fermentation can be carried out both as a batch (Anstellfermentation) and as a fed-batch fermentation (feed fermentation). As a first example, a batch fermentation will be described.

Half of the fermenter is filled with water and sterilized at 101 ° C for 30 minutes. Subsequently, the fermenter was pumped empty and cooled. For the reactor medium, three different solutions (glucose solution, mineral solutions A and B) were prepared and individually sterilized. For the preparation of the glucose solution 1759.8 g of dextrose-dyn (= brand name of the company. Agrana for glucose monohydrate) were dissolved in 1.611 kg of deionized water and sterilized at 121 ° C for 20 minutes. For mineral solution A, 192.2 g of CSL (Com Steep Liquor), 1.35 g of MgSO 4 .7H 2 O and 8.96 g of K 2 HPO 4 were dissolved in 2503 g of tap water and sterilized at 121 ° C. for 20 minutes. The preparation of the mineral solution B was carried out by slurry of 182.0 g of CaCO 3 in 1499.8 g of tap water and subsequent sterilization at 121 ° C for 20 minutes. The sterile media (glucose solution, mineral solutions A and B) were introduced via a sterile funnel into the fermenter and the weight determined by reweighing. The pH of the medium in the fermenter was adjusted to pH 7.0 by the addition of a NaOH solution (300 g / L). Subsequently, the fermenter was inoculated with the contents of two pre-culture Erlenmeyer flasks obtained as described above. The following parameters were set and kept as constant as possible during the entire fermentation period, with the exception of the amount of air, as described later: • Air quantity: • pH: • Temperature: • Dissolved oxygen: 3 l / min pH 7.0 for the first eight fermentation hours, then pH 5.6 to the end of fermentation (controlled by NaOH)

28 ° C 10% saturation (controlled via stirrer speed) After no reaction of 2-keto-D-gluconic acid to 2,5-diketo-D-gluconic acid in the above described fermentation is done more, the air supply is turned off until anaerobic conditions prevail in the fermenter. After a period of 2 to 10 hours, the air supply was turned on again. This action resulted in the complete conversion of 2-keto-D-gluconic acid to 2,5-diketo-D-gluconic acid. 137 g of 2,5-diketo-D-gluconic acid / l could be obtained by the method described above, which corresponds to a yield of 71% (w / w).

The fermentation for the production of 2,5-DKG was performed in another experiment as a fed-batch process, which essentially corresponds to the batch process described above, but with the following changes. The fermentation medium does not contain 200 g / l glucose but only 100 g / l glucose. The sterilized stirred tank was filled with 5.5 liters (without inoculum) medium as described above and inoculated with Pectobacter cypripedii (2 preculture Erlenmeyer flasks). The feed (feed, 50% glucose solution) was started when the carbon content in the exhaust air is above 1% by volume. The feed quantity of the feed solution is controlled via the CO 2 content in the exhaust air, with only so much glucose solution being fed in that the CO 2 content does not drop significantly below 1%. After a fermentation period of 121 hours, 116 g / l of 2,5-diketo-L-gluconic acid were determined in the fermentation medium, which corresponds to a yield of 70% (wt / wt). The 2,5-diketo-gluconic acid (2,5-DKG) prepared in the fermentation described above was purified by precipitating twice with ethanol and once with methanol. The precipitated 2.5 DKG crystals were dried in a desiccator over P 2 O 5. Subsequently, the further purification was carried out via a chromatography step with an Amberlite CG 120-11 (Ca 2+ -form) column, which represents a combination between a gel filtration and an ion exchange chromatography. The crystals were resuspended in 1-2 CV (Column Volumes) 1 M CaCfe and applied to the column equilibrated with buffer A (degassed water) (Amberlite CG 120-11 (Ca 2+ form) , 5-DKG, which corresponds to a yield of 98.3% (wt .- / wt.).

In the second step of the process of the present invention, the purified 2,5-DKG is converted enzymatically into 2-KLG by means of 2,5-diketo-D-gluconic acid reductase. For this purpose, the 2,5-diketo-D-gluconic acid reductase from Corynebacterium glutamicum (DMSZ 20301) was isolated and amplified by PCR. For the PCR reaction, all relevant components were combined in a master mix for the amplification reaction whose composition is shown in Table 1 and then aliquoted into PCR tubes. The primers used for the PCR have the following sequences: Primer 1 (SEQ ID No. 1) 5 'GAT AAG TGG ATC CAA TCT CTG ATG GAT C 3' (SEQ ID NO: 1) [0031 ] Primer 2 (SEQ ID NO: 2) 5 'CAG GGC CTT ACC TTA CTC GAG GTT CAG ATC 3' (SEQ ID NO: 2) TABLE 1

Reagent Capacity for 50 μl sterile water to 50 μΙ 10x PCR Buffer without MgCl 2 5 25 mM MgCl 2 3 10 mM dNTPs 1 10 pmol / μΙ Primer 1 5 10 pmol / μΙ Primer 2 5 Matrices DNA 1 Taq DNA Polymerase (5 U / μΙ The PCR tubes were used in a thermocycler (T3, Biometra (Göttingen, Germany)) and amplified under the following conditions (Table 2). TABLE 2

Temperature [U] time [min] number of cycles initial denaturation 95 2 1 denaturation 94 0.6 annealing 45-65 0.6 25-35 extension 72 2 final extension 72 6 1 The amplified PCR product was purified using cleaned in the prior art known methods. After purification, the DNA insert was digested by the restriction enzymes BamHI and Xhol and cloned into the appropriately pretreated pET 21d vector (Novagen, Darmstadt, Germany) using the Topo TA Cloning Kit (Invitrogen). The pET vectors are based on the T7 expression system, which is characterized by a very high transcription efficiency. The system is derived from bacteriophage T7, which competes successfully with the host's transcriptional machinery through its strong promoters recognized by the phage T7 RNA polymerase. The prepared vector (pET-21 d) was transformed by heat shock transformation into E. coli BL21 (DE3) cells (Invitrogen Carlsbad, USA). The E. coli strain BL21 (DE3) is particularly suitable for the expression of recombinant proteins in pET systems. Under the control of the lacUV-5 promoter, the T7 RNA polymerase is expressed after IPTG addition, and then the genes under the control of the T7 promoter are transcribed. The pET system was therefore chosen because it adds a C-terminal His tag to the desired DNA insert and thus makes very simple purification of the DNA insert possible by means of nickel chelate affinity chromatography. The recombinant cells are then resuspended in a modified minimal medium (MPC-Gly medium) containing in one liter 10 g of peptone from casein, 10 g glycerol, 0.1 ml 1M CaCl 2 .2M MgSO 4 .7H 2 O, and 200 ml MPC-Gly salts. with the MPC-Gly salt solution in one liter of water containing 15 g KH 2 PO 4, 2.5 g NaCl and 5 g NH 4 Cl. After the recombinant cells have been cultured in said medium at 37 ° C to an optical density of 0.6-08, expression of the heterologous gene is induced by the addition of 5 g / L lactose and the cells are allowed, at 251), pH 7.0 and p02 20% until an optical density of 8.0 was reached. The recombinant cells were pelleted by centrifugation (15,000 g) for 20 minutes and then mechanically disrupted by homogenizer (Homogenizer Invensis APV-2000, 1000 bar) or French Pressure Cell (70 bar). As mentioned above, the recombinant 2,5-diketo-D-gluconic acid reductase is purified using immobilized metal ion affinity chromatography (IMAC). In this technique, the medium is first charged with a transition metal ion (Ni 2+) to form a chelate before use. The proteins bind to the medium, depending on the presence of surface histindin residues that have an affinity for the chelated metal ions. The bond strength is basically determined by the metal ion and the pH of the buffer. The bound protein can be eluted by competitive elution with imidazole. The iminodiacetic acid ligand is coupled to the chelating sepharose and thus allows the exchange of immobilized metal ions. The 2,5-diketo-D-gluconic acid reductase (DKR) thus prepared has an activity of about 200 U / l culture. For the enzymatic conversion of 2,5-DKG into 2-KLG, the DKR is used in soluble form. For this purpose, 250 mM (58 g / l) of 2,5-DKG, 2 U / ml of DKR, 250 mM (45 g / l) of glucose, 2 U / ml of glucose dehydrogenase (GDH) and 0.26 mM (0, 21 g / l) NADP + added. The volume in the batch reactor was 500 or 1000 ml. The conversion of 2,5-DKG into 2-KLG requires the coenzyme NADPH, which is generated by the simultaneous oxidation of glucose to 6/9 isfesrisSistke patenuiit AT505 263B1 2009-08-15

Gluconic acid (GA) is regenerated by means of glucose dehydrogenase (GDH: EC 1.1.1.47 from Bacillus cereus, Amano, Japan). The conversion reaction takes place at 25 ° C, pH 7.0, which is kept constant with ammonia with continuous stirring. This reaction gives under the conditions mentioned 56.0 g of 2-keto-L-gulonic acid / l reaction solution (ie buffer, either Na acetate or Na citrate or Bis-Tris) (yield 96.7%) and 42.1 g Gluconic acid / L reaction solution (ie buffer, either Na acetate or Na citrate or Bis-Tris) (yield 93.5%). The 2-keto-L-gulonic acid is subsequently separated from the gluconic acid, glucose and 2.5-DKG by an ion exchanger and used for further reaction in ascorbic acid. The gluconic acid, glucose and 2,5-DKG are recycled in the first reaction step (fermentation of Pectobacter cypripedii).

In an alternative embodiment, the two-stage process described above, which proceeds in two spatially separate and non-interconnected reaction vessels, can be combined to form a so-called hybrid process. For this purpose, the stirred tank, in which the fermentation of Pectobacter cypripedii HEP01 and the concomitant conversion of glucose into 2,5-diketo-D-gluconic acid (steps 1 of the above process) expires, and the enzyme reactor, in which by the 2, 5-diketo-D-gluconic acid reductase catalyzed reaction takes place, directly connected via connecting lines. An advantage of the hybrid method according to the invention is that 2.5-DKG is formed continuously in the fermenter, which, after the microorganisms have been separated from the liquid stream of the fermenter by filtration, is passed into the enzyme reactor. For this purpose, it is necessary to separate the reaction mixture of the enzyme reactor into a product and a biocatalyst, wherein the separation is accomplished by means of a charged ultrafiltration membrane. The enzymes necessary for continuous process control, DKR and GDH, and the coenzymes NADPH and NAPD + are retained in the enzyme reactor, while the unused substrates 2,5-diketo-D-glucose and glucose and the products gluconic acid and 2-keto-L-gulonic acid the membrane can pass unhindered. The enzymes (DKR and GDH) are retained by the ultrafiltration membrane because of their size, whereas the coenzymes (NADPH and NADP +) are prevented from passing because of their negative charge. With regard to the recycling of unused glucose (starting material of the GDH reaction) and of gluconic acid, which on the one hand is the product of GDH catalysis and on the other hand constitutes the starting material for the fermentation of Pectobacter cypripedii, the product of the process according to the invention must be 2-keto-L- Gulonic acid are separated from the product stream. For this purpose, the ion exchange chromatography (strongly basic AI EX) was used to separate the 2-keto-L-gulonic acid from the glucose or gluconic acid. The biocatalyst system (DKR, GDH and NADPH) remains through the ultrafiltration membrane in the enzyme reactor.

The vessels used for the hybrid method according to the invention are known in the art and are on the one hand a conventional stirred tank, which has the necessary measuring and control electronics such as pH electrode, pO 2 electrode, etc., and on the other hand, an enzyme reactor, the one Having semipermeable membrane. The semipermeable membrane used is a charged ultrafiltration membrane (NTR-7430, Nitto Electric Industrial Co. (Japan)). In the stirred tank, the fermentative conversion of gluconic acid into 2.5-DKG by means of Pectobacter cypripedii HEP01 takes place, while in the enzyme reactor, the biocatalytic reaction of 2,5-DKG in 2-KLG proceeds with simultaneous regeneration of the coenzyme NADPH. The sterile stirred tank with a reactor volume of 10 liters was filled with medium, which is described above for the fed-batch fermentation, and inoculated the stirred tank with the contents of two precultilator Erlenmeyer flasks of Pectobacter cypripedii HEP01 strain. The Pectobacter cypripedii strain (HEP01) was fermented at 28 ° C, pH 7 and 5.6, respectively, as described above, at a dissolved oxygen concentration of 10% to give 137 g / l product. A portion of the culture suspension is removed from the stirred tank, filtered, and the supernatant is introduced into the enzyme reactor. The reaction reaction in the enzyme reactor takes place at 25 ° C (since, as mentioned above, two reactors are still used, the 7/9

Claims (13)

ifcftsTisfcistke AT505 263 B1 2009-08-15 interconnected, different temperatures can be used) and a pH value of 6.4 instead, which is kept constant with ammonia. The product of this fermentation, 2,5-diketo-D-gluconic acid, was added to the enzyme reactor in the enzyme reactor with the addition of 2 U / ml of DKR, 2 U / ml and 106 g of glucose / I buffer (either Na-acetate or Na citrate) immediate precursor of ascorbic acid, 2-keto-L-gluonic acid with a yield of 132 g / l converted. The volume in the enzyme reactor is also 10 L. The enzymes (DKR and GDH) and the coenzymes are retained by the ultrafiltration membrane in the enzyme reactor while the product stream (glucose, gluconic acid and 2-keto-L-gulonic acid) is withdrawn from the enzyme reactor passed an ion exchange chromatography column (Amberlite FPA90, Rohm and Haas column material) and eluted with various eluents (H 2 O, methanol and H 2 SO 4) to separate the product stream into its individual components, thereby allowing glucose and gluconic acid from the 2-keto L-gulonic acid, wherein the 2-keto-L-gulonic acid is eluted as methyl ester KLG. The yield of 2-keto-L-gulonic acid was 132 g / l. The glucose or gluconic acid (99 g / l) are returned to the stirred tank and metabolized there by the Pectobacter cypripedii strain HEP01 to form further 2,5-diketo-D-gluconic acid. Thus, the hybrid method described forms a closed circuit in order to achieve the highest possible product yields (2-keto-L-gulonic acid) with simultaneous use of the originally added enzymes or coenzymes. 1. A process for producing ascorbic acid, wherein in a first step, a carbon source is fermentatively converted to 2,5-diketo-D-gluconic acid, which then in a second step enzymatically with 2,5-diketo-D-gluconic acid reductase from Corynebacterium glutamicum to 2-keto-L-gulonic acid, characterized in that the carbon source is glucose and the conversion of glucose into 2,5-diketo-D-gluconic acid by the Pectobacter cypripedii strain HeP01 (DSM 12939) directly he follows. 2. The method according to claim 1, characterized in that in the first step, the air supply is reduced or adjusted until anaerobic conditions in the nutrient solution, if the formation of 2,5-diketo-D-gluconic acid from the intermediate 2-keto D-gluconic acid is absent. 3. The method according to claim 2, characterized in that the air supply is controlled by an oxygen probe. 4. The method according to any one of claims 1 to 3, characterized in that during the fermentation of the first step, the pH is monitored and / or regulated. 5. The method according to any one of claims 1 to 4, characterized in that the first step of the method is carried out in batch mode. 6. The method according to any one of claims 1 to 4, characterized in that the first step of the method is carried out in fed-batch mode. 7. The method according to any one of claims 1 to 6, characterized in that the coenzyme NADPH required in the second step by the glucose dehydrogenase, which converts glucose into gluconic acid, is regenerated, wherein the resulting by the regeneration of gluconic acid, and optionally remaining glucose and / or 2,5-diketo-D-gluconic acid is recycled in the first step. 8. The method according to claim 7, characterized in that in the second step as a buffer Na citrate or Na acetate is used. 8/9 ijitirriiWiSF'fiS patenmamt AT505 263B1 2009-08-15 9. The method according to claim 7 or 8, characterized in that the gluconic acid and optionally glucose and / or 2,5-diketo-D-gluconic acid prior to recycling by chromatographic separation methods, such as ion exchange chromatography, of the 2-keto-L-gulonic acid Product stream is separated. 10. The method according to any one of claims 1 to 9, characterized in that the 2,5-diketo-D-gluconic acid reductase from Corynebacterium glutamicum using a pET vector in E. coli is produced recombinantly. 11. The method according to claim 10, characterized in that for the preparation of the insert DNA as a primer for the PCR SEQ ID NO. 1 and SEQ ID NO. 2 and the plasmid pPC1 is used as a template, wherein the amplified fragment in the Expression vector pET-21 d is used. 12. The method according to claim 11, characterized in that the transformed with pET-21d E. coli BL21 (DE3) strain in a minimal medium MPC-Gly, in a liter of 10g peptone from casein 10g glycerol 0.1 ml 1MCaCI2 1 M MgSO4 * 7H20 contains 200 ml of MPC-Gly salts, the MPC-Gly salt solution in one liter of water containing 15gKH2PO4 containing 2.5 g NaCl containing 5 g NH4CI. 13. The method according to any one of claims 1 to 9, characterized in that it is performed as a hybrid method in which the fermentation of Pectobacter cypripedii HeP01 and the reaction catalyzed by the 2,5-diketo-D-gluconic acid reductase proceed in parallel to each other wherein the reaction mixture is separated into a product stream and a biocatalyst stream by a membrane separation process wherein the enzymes 2,5-diketo-D-gluconic acid reductase and glucose dehydrogenase and the coenzymes NADPH and NADP + are retained by a charged ultrafiltration membrane. For this no sheet of drawings 9/9