Method for preparing psicose

Abstract

A method for obtaining psicose from a mixture of fructose and psicose includes adding a mannitol dehydrogenase to a mixture of fructose and psicose to convert fructose into mannitol, and separating the mannitol from the psicose. Accordingly, the psicose can be more easily isolated from the mannitol with high efficiency by using the method.

Claims

1. A method for obtaining psicose from a mixture of fructose and psicose, comprising: adding a mannitol dehydrogenase to a mixture of fructose and psicose to convert fructose into mannitol; and separating the mannitol from the psicose.

2. The method of claim 1, wherein the mannitol dehydrogenase is a mannitol-2-dehydrogenase that is from Leuconostoc pseudomesenteroides ATCC 12291 and has the amino acid sequence of SEQ ID NO: 43, a mannitol-2-dehydrogenase that is from Leuconostoc mesenteroides and has the amino acid sequence of SEQ ID NO: 44, a mannitol-2-dehydrogenase that is from Rhodobacter sphaeroides and has the amino acid sequence of SEQ ID NO: 45, or a mannitol-2-dehydrogenase that is from Pseudomonas fluorescens DSM 50106 and that has the amino acid sequence of SEQ ID NO: 46.

3. The method of claim 1, wherein the conversion of the fructose into mannitol is performed in the presence of NADH sources.

4. The method of claim 3, wherein the NADH sources are formic acid and a formate dehydrogenase.

5. The method of claim 1, wherein the conversion of the fructose into mannitol is performed by a microorganism including a gene encoding a mannitol dehydrogenase.

6. The method of claim 5, wherein the microorganism is a microorganism of the genus Corynebacterium.

7. The method of claim 5, wherein the conversion of the fructose into mannitol is performed in a medium containing formic acid, and the microorganism further includes a gene encoding a formate dehydrogenase.

8. The method of claim 7, wherein the formate dehydrogenase is from Mycobacterium vaccae N10 and has the amino acid sequence of SEQ ID NO: 49.

9. The method of claim 7, wherein the conversion of the fructose into mannitol is performed in a medium containing a buffer solution in an open reaction system.

10. The method of claim 5, wherein the microorganism further includes a gene encoding a glucose transport protein.

11. The method of claim 10, wherein the glucose transport protein is from Zymomonas mobilis and has the sequence of SEQ ID NO: 50.

12. The method of claim 1, wherein the separation of the mannitol is performed by mannitol crystallization according to a difference in solubility between the mannitol and the psicose in a solvent.

13. The method of claim 1, further comprising: preparing the mixture of fructose and psicose by a reaction between the fructose as a substrate and an epimerase thereof.

14. The method of claim 13, wherein the preparation of the mixture of fructose and psicose is performed in a microorganism at 40 to 90 C.

15. The method of claim 1, wherein the separation of mannitol is performed by liquid chromatography.

16. A method for preparing psicose, comprising: converting fructose into psicose; adding a mannitol dehydrogenase to convert unreacted fructose into mannitol.

17. The method of claim 16, further comprising: separating psicose from the mannitol.

Description

BRIEF DESCRIPTION OF THE DRWAINGS

(1) FIG. 1 shows the psicose production amount from fructose, which is a substrate, measured according to a reaction temperature of a resting cell conversion reaction from a Corynebacterium glutamicum transformant into which a psicose-3-epimerase is introduced.

(2) FIG. 2 shows the psicose production amount from fructose, which is a substrate, measured according to a reaction temperature of a resting cell conversion reaction from an E. coli MG1655 transformant into which a psicose-3-epimerase is introduced.

(3) FIG. 3 shows the psicose production amount obtained by recovering bacterial cells after a resting cell conversion reaction for producing psicose is performed at 60 C. for 3 hours using Corynebacterium glutamicum and E. coli MG1655 transformants into which psicose-3-epimerase is introduced, and then performing a reaction again under the same conditions.

(4) FIG. 4 shows the psicose production amount from fructose according to a composition change of a psicose production reaction medium, which is used in a resting cell conversion reaction of a Corynebacterium glutamicum transformant into which a psicose-3-epimerase is introduced.

(5) FIG. 5 shows the psicose production amount according to a psicose-3-epimerase-derived strain and the composition of a psicose production reaction medium in a resting cell conversion reaction of a Corynebacterium glutamicum transformant into which a psicose-3-epimerase is introduced.

(6) FIG. 6 shows the psicose production amount according to a psicose-3-epimerase-derived strain and a heating time in a resting cell conversion reaction of a Corynebacterium glutamicum transformant into which a psicose-3-epimerase is introduced.

(7) FIG. 7 shows the comparison of amino acid sequences of psicose-3-epimerases derived from various strains.

(8) FIG. 8 shows the psicose production amount according to a reuse number of Corynebacterium glutamicum into which Clostridium-derived psicose-3-epimerase is introduced.

(9) FIG. 9 is a diagram showing the mannitol production produced using Corynebacterium recombinant strains into which mannitol dehydrogenases and formate dehydrogenases are derived from various strains are introduced. LpMDH is Leuconostoc pseudomesenteroides-derived mannitol dehydrogenase, LmMDH is Leuconostoc mesenteroides-derived mannitol dehydrogenase, RsMDH is Rhodobacter sphaeroides-derived mannitol dehydrogenase, and PfMDH is Pseudomonas fluorescens-derived mannitol dehydrogenase.

(10) FIGS. 10(A) and 10(B) are diagrams showing an amount of psicose remaining without a reaction by a mannitol dehydrogenase and a production amount of mannitol converted from fructose, when mixed saccharides including psicose and fructose were provided as substrates.

(11) FIGS. 11(A) and 11(B) are diagrams showing a mannitol production amount and a pH change when a conversion medium containing buffer solutions with different pHs is used and when a conversion medium containing water without a buffer solution. W uses water without a buffer solution, SA uses a sodium acetate buffer solution with pH 5, P uses a PIPES buffer solution with pH 6, and SP uses a sodium phosphate buffer solution with pH 6.5 used as conversion media.

(12) FIGS. 12(A) and 12(B) are diagrams comparing a mannitol production amount and a pH change according to the use of a conical tube, which is a closed reaction vessel, and a test tube, which is an open reaction vessel.

(13) FIG. 13 is a diagram showing a mannitol production amount when water or a PIPES buffer solution with pH 6 is used as a conversion medium in a test tube, which is an open reaction vessel.

(14) FIG. 14 is a diagram comparing a mannitol production amount according to whether or not a glucose transport protein (GLF) is introduced.

(15) FIG. 15 is a diagram showing the concentrations of fructose, mannitol and psicose after crystallized mannitol is separated from a mixed saccharide solution for a mannitol conversion reaction according to an amount of added ethanol.

(16) FIG. 16 is a diagram showing the concentrations of fructose, mannitol and psicose after crystallized mannitol is separated as the volume of a mixed saccharide solution for a mannitol conversion reaction is reduced by evaporation.

(17) FIGS. 17(A) to 17(C) are diagrams comparing a resolution patterns of high performance liquid chromatography (HPLC) for mixed saccharides in a reaction solution depending on an additional mannitol conversion reaction after a psicose production process.

DETAILED DESCRIPTION

(18) Hereinafter, the present invention will be fully described in detail with reference to examples.

EXAMPLES

(19) 1. Change in Psicose Production Rate and Production Amount According to Temperature in Process of Producing Psicose from Fructose Using Corynebacterium glutamicum Strain

(20) (1) Preparation of Recombinant Strain

(21) A pSGT208 shuttle vector into which a terminator and a lac prompter were inserted was constructed by modifying an E. coli-Corynebacterium shuttle vector, pCES208 (J. Microbiol. Biotechnol., 18:639-647, 2008).

(22) To produce psicose in Corynebacterium glutamicum, a psicose-3-epimerase was prepared by introducing a dpe gene (AGR_L_260, GI:15890243, SEQ ID NO: 1) of Agrobacterium tumefaciens (Agrobacterium tumefaciens str. C58; taxid:176299; GenBank NID: NC_003062, ATCC33970) into the constructed pSGT208 shuttle vector.

(23) In detail, the dpe gene was amplified from the Agrobacterium tumefaciens genome using a primer 1 of SEQ ID NO: 31 and a primer 2 of SEQ ID NO: 32, cleaved with restriction enzymes KpnI and BamHI to be inserted into the same site of the pSGT208 shuttle vector, thereby constructing a pS208-dpe recombinant shuttle vector including psicose-3-epimerase.

(24) Afterward, to increase the expression level of a psicose-3-epimerase in Corynebacterium glutamicum, in pS208-dpe, a lac promoter was substituted with a trc promoter derived from pTrc99a, and thus the resulting vector was named pS208cT-dpe.

(25) Recombinant vectors pS208-dpe and pS208cT-dpe including the constructed psicose-3-epimerase and a pSGT208 vector, which was a negative control thereof, were introduced into wild-type Corynebacterium glutamicum ATCC 13032 for transformation, and thus were used in production of psicose from fructose. The transformation was performed by a method specified in the Handbook of Corynebacterium glutamicum (Lothar Eggeling et al., ISBN 0-8493-1821-1, 2005 by CRC press).

(26) (2) Culture of Recombinant Strain and Production of Psicose Using the Strain

(27) To ensure a high concentration of bacterial cells, the prepared Corynebacterium glutamicum transformant was seeded into 5 ml of a LB medium (Difco) containing 20 g/ml of kanamycin to perform seed culture at 30 C. and 250 rpm, and then the cultured cells were seeded into a minimal medium (per liter, 1 g K.sub.2HPO.sub.4, 10 g (NH.sub.4).sub.2SO.sub.4, 0.4 g MgSO.sub.47H.sub.2O, 20 mg FeSO.sub.47H.sub.2O, 20 mg MnSO.sub.45H.sub.2O, 50 mg NaCl, 2 g urea, 0.1 mg biotin, 0.1 mg thiamine) containing 10 g/L glucose and 20 g/ml of kanamycin to perform main culture. For the main culture, the transformants were cultured in a 500 ml Erlenmeyer flask with markings to have a volume of 100 ml at 37 C. and 180 rpm for 12 hours, thereby leading to a sufficient cell mass and sufficient expression of a protein.

(28) The obtained culture solution was centrifuged to remove the supernatant and recover the cells, and the cells were resuspended in the above-mentioned minimal medium containing 40% (w/v) fructose as a substrate to have a cell concentration of 40 at OD.sub.600, followed by a resting cell conversion reaction at 25, 30, 37, 50, 60 or 70 C. and 180 rpm.

(29) Concentrations of fructose and psicose were measured using HPLC. For HPLC, SCL-10A (Shimadzu, Japan) equipped with a Kromasil 5NH.sub.2 column (4.6 mm250 mm) was used, and separation was performed by adding 75% acetonitrile as a mobile phase at 1.5 mL/min and 40 C., and analysis was performed using a reflective index (RI) detector. A retention time of fructose under the above conditions was 5.5 minutes, and a retention time of psicose was 4.6 minutes.

(30) Measurement results are shown in FIG. 1. Referring to FIG. 1, a pSGT208cT-dpe shuttle vector-introduced Corynebacterium glutamicum ATCC13032 strain was subjected to a conversion reaction in a medium containing 40% fructose, which showed that, as a reaction temperature increased, a production rate for psicose considerably increased, and a production amount also increased. Particularly, in experimental groups reacted at 50, 60 and 70 C., psicose-3-epimerase reached equilibrium within approximately 3 hours, and approximately 120 g/L of psicose was produced. Such a result can show that the conversion rate of a psicose-3-epimerase from fructose to psicose and the production amount were dependent on a temperature.

(31) Starting from 50 C., the production amount drastically increased, and the temperature was considerably higher than that required for a conventional enzyme reaction, and it was determined that the reaction patterns between an enzyme and a substrate vary at the corresponding temperature.

(32) 2. Change in Production Rate and Production Amount of Psicose According to Temperature in Production of Psicose from Fructose Using E. coli

(33) According to the method described in Example 1 in Korean Patent No. 10-1106253, an E. coli MG1655 (pfkA, als2) strain was prepared by transforming a pTrc99A vector with an Agrobacterium tumefaciens-derived psicose-3-epimerase-introduced pTPE plasmid.

(34) To block a psicose degradation pathway, E. coli MG1655 from which pfkA (SEQ ID NO: 11) and als2 (SEQ ID NOs: 14, 15, 16, 17, 18 and 19) genes were deleted were used.

(35) To ensure a high concentration of cells, the E. coli MG1655 transformant prepared as described above was seeded into 5 ml LB medium (Difco) containing 100 g/ml of ampicillin for seed culture at 37 C. and 250 rpm, and then seeded into a 2YT medium containing 10 g/L of glucose and 100 g/ml of ampicillin for main culture. For the main culture, the transformants were cultured in a 500 ml Erlenmeyer flask with markings to have a volume of 100 ml at 37 C. and 180 rpm for 12 hours, thereby leading to a sufficient cell mass and sufficient expression of a protein.

(36) The obtained culture solution was centrifuged to remove the supernatant and recover the cells, and the cells were resuspended in E. coli minimal medium M9 (per liter, 11.3 g M9 minimal salts (Difco), 0.1 mL 1M CaCl.sub.2, 2 mL 1M MgSO.sub.4, 1 mL 100 mM MnSO.sub.45H.sub.2O) containing 40% (w/v) fructose as a substrate to have a cell concentration of 40 at OD.sub.600, followed by a resting cell conversion reaction at 37, 60 or 70 C. and 180 rpm under respective conditions. Concentrations of fructose and psicose were analyzed according to the method described in Example 1. Measurement results are shown in FIG. 2.

(37) Referring to FIG. 2, the pTPE vector-introduced E. coli MG1655 (pfkA, als2) strains were also subjected to a conversion reaction in a medium containing 40% fructose, which showed that, as the reaction temperature increased, a psicose production rate considerably increased and a production amount also increased. Particularly, in experimental groups reacted at 60 or 70 C., like the experiment using Corynebacterium, psicose-3-epimerase reached equilibrium within approximately 2 hours, and approximately 120 g/L of psicose was produced.

(38) Such a result also showed that the psicose conversion rate of psicose-3-epimerase from fructose and the psicose production amount were dependent on a temperature.

(39) According to a resting cell conversion reaction experiment performed for a representative Gram-positive bacteria Corynebacterium and a representative Gram-negative bacteria E. coli at a high temperature, it can be seen that, since protected from a severe external environment in cells, a saccharide conversion enzyme, psicose-3-epimerase, was not thermally denatured, and psicose was produced at a high rate at a high temperature, rather than in a pure enzyme state. The advantage of such a cell conversion reaction seems to be also applied to most types of microorganisms.

(40) 3. Continuous Production from Fructose to Psicose Through Cell Recovery and Reuse in Resting Cell Conversion Reaction of Corynebacterium glutamicum Transformant Strain

(41) According to the results of Examples 1 and 2, when psicose-3-epimerase was subjected to a conversion reaction in a Corynebacterium or E. coli transformant at a high temperature of 50 C. or more, the production amount of psicose reached the maximum level within 3 hours, and did not increase any more.

(42) That is, even three hours after the reaction reached equilibrium at which the psicose production is in the maximum level, to confirm how much an activity of the psicose-3-epimerase to convert fructose into psicose, cells used in production of psicose through a resting cell conversion reaction performed for 3 hours in the presence of fructose were recovered, and then reused in the resting cell conversion reaction for producing psicose.

(43) The resting cell conversion reaction by the cell reuse was repeated three times at 60 C. The first resting cell conversion reaction was represented as R0, a resting cell conversion reaction by the first reuse of cells recovered from the above-described reaction solution was represented as R1, and a resting cell conversion reaction by the second reuse thereof was represented as R2, and a resting cell conversion reaction by the third reuse thereof was represented as R3. Culture conditions and an analysis method were the same as described in Example 1. Results are shown in FIG. 3.

(44) Referring to FIG. 3, bacterial cells can be reused in a saccharide conversion reaction at a high temperature of 60 C. However, as the cells were used multiple times, the enzyme activity seems to be reduced by a certain level. In addition, it can be seen that, for reuse of cells at a high temperature, the Gram-positive Corynebacterium glutamicum strain exhibits a higher retained enzyme activity than the Gram-negative E. coli MG1655 strain.

(45) 4. Production of Psicose from Fructose in Various Resting Cell Conversion Reaction Media

(46) In Example 1, for a conversion reaction for producing psicose from fructose in Corynebacterium glutamicum, a 40% fructose-contained minimal medium (per liter, 1 g K.sub.2HPO.sub.4, 10 g (NH.sub.4).sub.2SO.sub.4, 0.4 g MgSO.sub.47H.sub.2O, 20 mg FeSO.sub.47H.sub.2O, 20 mg MnSO.sub.45H.sub.2O, 50 mg NaCl, 2 g urea, 0.1 mg biotin, 0.1 mg thiamine) was used. A more economic and simple medium can be prepared by minimizing components for a medium used in the conversion reaction, and the productivity of psicose when this medium is used was compared with that when the medium used in Example 1 was used.

(47) Referring to FIG. 4, although a phosphate buffer (pH 7) medium containing 40% fructose and 0.1 mM MnSO.sub.4, or more simply, a medium containing only 40% fructose and 0.1 mM MnSO.sub.4 was used, there is no significant difference in the final psicose production amount. This result was shown at both reaction temperatures of 30 C. and 60 C. When 0.1 mM MnSO.sub.4 was not added as a cofactor of psicose-3-epimerase, the production amount was reduced.

(48) In this example, it can be seen that components for the medium used in the resting cell conversion reaction for producing psicose only include fructose as a substrate and MnSO.sub.4, which is the cofactor of psicose-3-epimerase.

(49) 5. Preparation of Recombinant Corynebacterium glutamicum Including Polynucleotides Encoding Amino Acid Sequences of Psicose-3-epimerases Derived from Various Strains

(50) The total genome of Anaerostipes caccae (Anaerostipes caccae DSM 14662; taxid: 411490) was purchased from DSMZ (Germany). The first PCR was performed with the purchased total genome as a template using a primer pair of SEQ ID NOs: 33 and 34 to include a psicose-3-epimerase estimated gene (AP endonuclease; Sequence ID: gb|EDR98778.1|; GI: 167654649; SEQ ID NO: 4). The second PCR was performed using the amplified PCR product as a template and a primer pair of SEQ ID NOs: 35 and 36, which specifically bind to a psicose-3-epimerase gene.

(51) The obtained PCR product was inserted into the same enzyme site of pS208cT-dpe (the vector described in Example 1 of Korean Patent Application No. 10-2013-0060703) using restriction enzymes BamHI and XbaI, thereby constructing a recombinant vector pS208cT-AcDPE.

(52) The constructed pS208cT-AcDPE vector was introduced into wild-type Corynebacterium glutamicum ATCC 13032 for transformation, and was used in production of psicose from fructose. The transformation was performed by the method described in the Handbook of Corynebacterium glutamicum (Lothar Eggeling et al., ISBN 0-8493-1821-1, 2005 by CRC press).

(53) The obtained recombinant Corynebacterium glutamicum strain was stored at 80 C. and then cultured.

(54) A plasmid including a psicose-3-epimerase estimated gene (hypothetical protein CLOBOL_00069; Sequence ID: gb|EDP19602.1|; GI:15844190; SEQ ID NO: 9) of Clostridium bolteae (Clostridium bolteae ATCC BAA-613; taxid:411902) was obtained from Korea Yakult Co., Ltd. PCR was performed using a primer pair of SEQ ID NOs: 37 and 38, which specifically bind to the psicose-3-epimerase gene.

(55) The obtained PCR product was inserted into the same enzyme site of pS208cT-dpe (the vector described in Example 1 of Korean Patent Application No. 10-2013-0060703) using restriction enzymes KpnI and XbaI, thereby constructing a recombinant vector pS208cT-CbDPE.

(56) The constructed recombinant vector pS208cT-CbDPE was introduced to transform wild-type Corynebacterium glutamicum ATCC 13032 by the method described above, and was used in production of psicose from fructose. The recombinant Corynebacterium glutamicum strain was stored at 80 C. and cultured.

(57) The total genome of Clostridium hylemonae (Clostridium hylemonae DSM 15053; taxid:553973) was purchased from DSMZ (Germany). The first PCR was performed with the purchased total genome as a template and a primer pair of SEQ ID NOs: 39 and 40 to include a psicose-3-epimerase estimated gene (dolichol monophosphate mannose synthase; Sequence ID:ref|WP_006442985.1|; GI:225161759; SEQ ID NO: 10). The second PCR was performed with the amplified PCR product as a template and a primer pair of SEQ ID NOs: 41 and 42, which specifically bind to a psicose-3-epimerase gene.

(58) The obtained PCR product was inserted into the same enzyme site of pS208cT-dpe (the vector described in Example 1 of Korean Patent Application No. 10-2013-0060703) using restriction enzymes BamHI and XbaI, thereby constructing a recombinant vector pS208cT-ChDPE.

(59) The constructed pS208cT-ChDPE vector was introduced into wild-type Corynebacterium glutamicum ATCC 13032 for transformation, and was used in production of psicose from fructose. The obtained recombinant Corynebacterium glutamicum strain was stored at 80 C. and cultured.

(60) 6. Production of Psicose from Fructose Using Recombinant Corynebacterium glutamicum Strain into which Psicose-3-Epimerases Derived from Various Strains were Introduced

(61) The production of psicose from a high concentration of fructose using the Corynebacterium glutamicum transformant constructed in Example 5 was confirmed.

(62) The transformant was seeded in a 2YT medium containing 20 g/ml of kanamycin for seed culture at 30 C. and 250 rpm, and then cultured again in a 2YT medium containing 20 g/ml of kanamycin for main culture. For the main culture, the transformants were cultured in a 300 ml Erlenmeyer flask with markings to have a volume of 60 ml at 30 C. and 180 rpm for 7 hours, thereby leading to a sufficient cell mass and sufficient expression of a protein.

(63) The obtained culture solution was centrifuged to remove the supernatant and recover the cells, and the cells were subjected to a resting cell conversion reaction at 55 C. using a simple conversion reaction medium containing 20 g/ml kanamycin, 40% (w/v) fructose as a substrate and 0.1 mM manganese or cobalt known as a main cofactor of psicose-3-epimerase. Concentrations of fructose and psicose were measured in the same manner as described in Example 1. Measurement results are shown in FIG. 5 (AtDPE refers to the conventionally used psicose-3-epimerase of Agrobacterium tumefaciens).

(64) Referring to FIG. 5, when cobalt, rather than manganese, was used as a cofactor, it seems that, in all recombinant strains into which various psicose-3-epimerases were introduced, a production rate of psicose was a little faster.

(65) While it seems that psicose production amounts of recombinant Corynebacterium glutamicum into which Anaerostipes and Agrobacterium-derived psicose-3-epimerases were introduced reached equilibrium within 6 hours, a psicose production amount of recombinant Corynebacterium glutamicum into which Clostridium-derived psicose-3-epimerase was introduced seems to reach equilibrium within three hours even when manganese was used as a cofactor.

(66) Therefore, when the Clostridium-derived psicose-3-epimerase was used, the psicose production from fructose can be confirmed, and it can also be known that the psicose production rate was faster than when the Agrobacterium-derived psicose-3-epimerase was used.

(67) 7. Maintenance of Continuous Activity to Produce Psicose at High Temperature in Recombinant Corynebacterium glutamicum Strains into which Psicose-3-Epimerases Derived from Various Strains were Introduced

(68) According to the results obtained in Examples 1 and 2, it was confirmed that when the conversion reaction was performed at a high temperature of 50 C. or more, total psicose production was achieved within three hours. Likewise, since the production of psicose from fructose is fast under such a high temperature condition, for the reuse of bacterial cells, psicose-3-epimerase which is stable at a high temperature for a long time is needed. Therefore, it was confirmed how much psicose-3-epimerases derived from various strains maintained the activity at a high concentration.

(69) The bacterial cells obtained by the method described in Example 1 were suspended in 2YT, and then heat was continuously applied thereto using a shaking incubator at 60 C. for 0, 3, 6, 9, 12, or 24 hours. After heating for each period of time, the bacterial cells were recovered and resuspended in a simple conversion reaction medium containing 20 g/ml kanamycin, and only 0.1 mM manganese and 40% (w/v) fructose to perform a resting cell conversion reaction at 60 C. for 3 hours. Concentrations of fructose and psicose were measured in the same manner as described in the method of Example 1. Measurement results are shown in FIG. 6.

(70) Referring to FIG. 6, it is shown that after heating at 60 C. for 3 hours, the recombinant Corynebacterium glutamicum into which the conventionally used Agrobacterium-derived psicose-3-epimerase and Anaerostipes-derived psicose-3-epimerase were introduced rarely produce psicose. Contrarily, the recombinant Corynebacterium glutamicum into which Clostridium-derived psicose-3-epimerase was introduced seems to maintain psicose production even after heating for 24 hours, and it is considered that the recombinant Corynebacterium glutamicum of the present invention is more advantageous for psicose production at a high temperature process than the Agrobacterium-derived psicose-3-epimerase.

(71) It has been known that, when the amino acids no. 33 or 213 in the sequence (Reference 1) critical for thermal stability of Agrobacterium tumefaciens-derived psicose-3-epimerase was replaced with leucine or cysteine, respectively, the half-lifetime of the enzyme at 50 C. increased 3.3 times or 7.2 times, respectively, or when all of the amino acids were replaced, the half-life of the enzyme at 50 C. increased 29.9 times.

(72) The comparison between the amino acid sequence of the Agrobacterium tumefaciens-derived psicose-3-epimerase and the amino acid sequence of the Clostridium-derived psicose-3-epimerase was shown in FIG. 7, and referring to this, the Clostridium-derived psicose-3-epimerase having a high thermal stability had a sequence corresponding to one or both of the above-described amino acid substitutions.

(73) 8. Production of Psicose from Fructose Through Cell Recovery and Reuse in Resting Cell Conversion Reaction of Recombinant Corynebacterium glutamicum Transformant Strain into which Clostridium-Derived Psicose-3-Epimerase was Introduced

(74) In Example 7, it was confirmed that the Clostridium-derived psicose-3-epimerase has a high stability at a high temperature, a bacterial cell reuse effect was confirmed after the resting cell conversion reaction at a high temperature of the recombinant Corynebacterium glutamicum cells into which Clostridium hylemonae psicose-3-epimerase was introduced, among the two recombinant strains into which Clostridium-derived psicose-3-epimerase was introduced.

(75) The bacterial cells obtained in the same manner as described in Example 3 were subjected to a resting cell conversion reaction at 60 C. for 3 hours, recovered to perform a resting cell conversion reaction again in the same manner, and then reused total three times (the experiment was performed under the same condition as used in Example 3). The first resting cell conversion reaction was represented as R0, a resting cell conversion reaction by the first reuse of bacterial cells recovered from the above-described reaction solution was represented as R1, and a resting cell conversion reaction by the second reuse thereof was represented as R2, and a resting cell conversion reaction by the third reuse thereof was represented as R3. Results are shown in FIG. 8.

(76) Referring to FIG. 8, it can be seen that, in case of the reuse of bacterial cells several times at a high temperature, the recombinant bacterial cells into which Clostridium-derived psicose-3-epimerase was introduced were maintained at a predetermined level without a decrease in psicose production amount. According to the results shown in Example 7, it can be confirmed that the thermal stability of the enzyme itself is very advantageous for the reuse of bacterial cells as well as a continuous saccharide conversion reaction at a high temperature.

(77) 9. Development of Corynebacterium glutamicum Recombinant Strain for Producing Mannitol from Fructose

(78) (1) Development of Recombinant Corynebacterium Strain

(79) The pSGT208 vector (the vector described in Example 1 of Korean Patent Application No. 10-2013-0060703), which was modified to easily clone an E. coli-Corynebacterium shuttle vector, for example, a pCES208 vector was used. A trc promoter was used as a promoter to allow continuous expression by removing a lac operator.

(80) As mannitol producing enzymes, Leuconostoc pseudomesenteroides ATCC 12291 (or KCTC 3652), Leuconostoc mesenteroides, Rhodobacter sphaeroides and Pseudomonas fluorescens DSM 50106-derived enzymes were used.

(81) The Leuconostoc pseudomesenteroides strain was purchased from KCTC and used as a template following purification of the total genome thereof, and PCR was performed using a primer pair of SEQ ID NOs: 47 and 48 as primers to include mannitol dehydrogenase (mannitol 2-dehydrogenase; MDH; GenBank: AJ486977.1, GI: 28865822, SEQ ID NO: 43, LpMDH).

(82) Leuconostoc mesenteroides-derived mannitol dehydrogenase (GenBank: ACT22631.1, GI: 253317413, SEQ ID NO: 44, LmMDH), Rhodobacter sphaeroides-derived mannitol dehydrogenase (GenBank: AAC45771.1, GI: 2338764, SEQ ID NO: 45, RsMDH) and Pseudomonas fluorescens-derived mannitol dehydrogenase (GenBank: AAC04472.1, GI: 2293418, SEQ ID NO: 46, PfMDH) were synthesized from GenScript.

(83) The obtained PCR product and synthesized gene were inserted into the same restriction site of a pSGT208 vector into which a trc promoter without a lac operator was introduced using restriction enzymes KpnI and BamHI, thereby preparing recombinant vectors pS208cT-LpMDH, pS208cT-LmMD1H, pS208cT-RsMDH and pS208cT-PfMDH.

(84) In addition, an enzyme which can help in continuously regenerating NADH used as a coenzyme of the mannitol dehydrogenase was also introduced. As the enzyme that can help in regenerating NADH, a formate dehydrogenase for regenerating NADH by oxidation from formic acid to carbon dioxide was used. Mycobacterium vaccae N10-derived formate dehydrogenase (FDH, GenBank: AB072394.1, GI: 15982576, SEQ ID NO: 49) was used, and a gene was synthesized from GenScript. The synthesized gene was inserted into the same restriction site of the previously-constructed pS208cT-LpMDH, pS208cT-LmMDH, pS208cT-RsMDH or pS208cT-PfMDH vector using the restriction enzymes BamHI and XbaI, thereby preparing recombinant vectors pS208cT-LpMDH-FDH, pS208cT-LmMDH-FDH, pS208cT-RsMDH-FDH or pS208cT-PfMDH-FDH, respectively.

(85) The prepared pS208cT-LpMDH-FDH, pS208cT-LmMDH-FDH, pS208cT-RsMDH-FDH or pS208cT-PfMDH-FDH vector was introduced into wild-type Corynebacterium glutamicum ATCC 13032 (Corynebacterium glutamicum ATCC 13032), and transformation was performed by the method described in Handbook of Corynebacterium glutamicum (Lothar Eggeling et al., ISBN 0-8493-1821-1, 2005 by CRC press). The transformed recombinant Corynebacterium strain was stored at 80 C. before use.

(86) (2) Experiment for Producing Mannitol Using Recombinant Strain

(87) For mannitol production, a resting cell conversion reaction using a high concentration of bacterial cells was used. To ensure the high concentration of bacterial cells, after the prepared Corynebacterium glutamicum recombinant strain was seeded in a 2YT medium containing 20 g/ml of kanamycin and cultured at 30 C. and 250 rpm for seed culture, a 5 to 10% (v/v) seed culture solution was seeded in a 2YT medium containing 20 g/ml of kanamycin and 0.5% (v/v) glucose to perform main culture for inducing a sufficient cell mass and protein expression at 30 C. and 180 rpm.

(88) A large quantity of the bacterial cells obtained by the main culture were used in a conversion reaction after the removal of the medium. The resting cell conversion reaction was carried out for 48 hours under the condition of 30 C. and 250 rpm after the bacterial cells were suspended in a 100 mM sodium phosphate (pH 6.5) buffer solution containing 10% (v/v) fructose and 555 mM formic acid to have a concentration of 40 at OD.sub.600. Concentrations of fructose and mannitol were measured using HPLC. For HPLC, SCL-10A (Shimadzu, Japan) equipped with a Kromasil 5NH.sub.2 column (4.6 mm250 mm) was used, and separation was performed by adding 80% acetonitrile as a mobile phase at 1.5 mL/min and 40 C., and analysis was performed using a reflective index (RI) detector. Measurement results are shown in FIG. 9.

(89) Referring to FIG. 9, while there was no significant difference in mannitol production amounts according to the mannitol dehydrogenases, when the Leuconostoc pseudomesenteroides-derived mannitol dehydrogenase was used, 48 hours after the conversion reaction, approximately 26 g/L of mannitol was produced from 100 g/L of fructose, resulting in the largest mannitol production amount. Accordingly, for a subsequent mannitol conversion reaction, the Leuconostoc pseudomesenteroides-derived mannitol dehydrogenase was continuously used. While not shown in the example, compared to before the formate dehydrogenase was introduced, it was confirmed that the mannitol production amount was increased approximately 86 times, and NADH regeneration was effectively performed using the formate dehydrogenase.

(90) 10. Mannitol Production Using Mixed Saccharide of Psicose and Fructose

(91) As described above, psicose and fructose have very similar physical properties. Accordingly, when the mannitol dehydrogenase is any enzyme which does not have a high specificity to a substrate, psicose as well as fructose may also be recognized as substrates and reduced. Therefore, it was examined whether a strategy for converting only fructose of the mixed saccharide of the psicose and the fructose into mannitol and easily isolating the fructose from the psicose can be applied.

(92) Only a substrate was converted into the mixed saccharide of the psicose and the fructose using the constructed recombinant Corynebacterium strain, and then the mixed saccharide was used in an experiment in the same manner under the same experimental conditions used in Example 9. The mixed saccharide includes the fructose and the psicose at a ratio of 7:3 (v/v), obtained after the psicose conversion reaction from the fructose, and a concentration of the mixed saccharide was 10% (v/v) as shown in Example 9. Results are shown in FIGS. 10(A) and 10(B).

(93) FIG. 10(A) shows a concentration of remaining psicose, which has not been used as a substrate, and during the conversion reaction for 48 hours, the initial concentration of approximately 30 g/L was maintained without a change, and it was confirmed that the psicose was rarely used. FIG. 10(B) shows an amount of mannitol produced from 70 g/L of fructose constituting 100 g/L of the mixed saccharide, and during the conversion reaction for 48 hours at a conversion temperature of 30 C., approximately 20 g/L of mannitol was produced. Accordingly, it can be confirmed that, in a mixed saccharide substrate solution in which a mannitol dehydrogenase was present with psicose and fructose, only fructose can be selectively used as a substrate. Therefore, it was confirmed that the idea of the present invention in which fructose remaining after the psicose conversion reaction from the fructose is converted into mannitol to easily isolate the psicose can be successfully applied.

(94) 11. Comparison of Mannitol Production Using Various Conversion Media

(95) To find the optimal pH condition for enhancing mannitol productivity, mannitol production amounts according to pH were compared. Since it has been known that the mannitol conversion reaction is carried out in a weak acidic condition, a buffer solution which can keep pH in a weak acidic level was selected. In addition, it was also examined whether the mannitol production was efficiently performed when a conversion medium composed of only water without a buffer solution was used.

(96) The mannitol productivity when water (W) without a buffer solution, a sodium acetate buffer solution (SA) with a pH 5, and a PIPES buffer solution (P) with a pH of 6 were used was compared with that when the conventionally used sodium phosphate buffer solution (SP) with a pH of 6.5 was used. Not shown in the example, a convention temperature was 45 C., rather than 30 C. at which the faster conversion reaction was shown, and a conversion medium included 10% (v/v) fructose as a substrate and 555 mM formic acid in water or a buffer solution, as described above, and when a buffer solution was used, a concentration of the buffer solution was 100 mM. Except these, the experiment was carried out under the same conditions as used in Example 9, and with an increased conversion rate, the conversion reaction was carried out only for 24 hours. Results are shown in FIGS. 11(A) and 11(B).

(97) Referring to FIG. 11(A), when PIPES with a pH of 6 was used, approximately 29 g/L of mannitol was produced, and thus the highest mannitol production amount was obtained. In addition, when a conversion medium composed of water was used, there was no difference in mannitol production amount obtained using the conventionally used sodium phosphate buffer solution with a pH of 6.5 as a conversion medium, and therefore it can be confirmed that even with water as a conversion medium, instead of the buffer solution, mannitol can be efficiently produced.

(98) According to the comparison of the mannitol production amounts obtained when different conversion media were used as shown in FIG. 11(A) and pH changes in the conversion reaction of FIG. 11(B), when the pH of the conversion medium was maintained in a range of 6.5 to 7, it seems that mannitol is actively produced. Therefore, it is considered that the optimal pH for the mannitol production under the present conditions was in a range of 6.5 to 7, and this is considered as an eclectic pH between the weakly acidic mannitol production environment with an optimal pH of 5 and the weakly alkaline formate dehydrogenase with an optimal pH of 7.5.

(99) 12. Comparison of Mannitol Production Amounts Depending on Open and Closed Reaction Vessels

(100) In the previous example, a mannitol conversion reaction was performed in a closed environment using a conical tube. However, to confirm that the reaction was influenced by dissolving CO.sub.2 produced in the dehydrogenation of formic acid in the conversion reaction medium without exhaustion, an experiment was performed using a test tube allowing free access of a gas. The experiment was carried out under the same experimental conditions using water as a conversion medium as previously described in Example 11. Experimental results are shown in FIGS. 12(A) and 12(B).

(101) FIG. 12(A) shows mannitol production amounts obtained in a conical tube, which is a closed vessel reaction condition, and a test tube, which is an open vessel reaction condition. It can be confirmed that mannitol production was produced approximately 9 times smaller than that under the open environment using the test tube. However, referring to the result of measuring pH of FIG. 12(B), 24 hours after a mannitol conversion reaction, it can be seen that the pH under an open environment was 9.2, which is much higher than the pH, which was 7.2, under a closed environment. It is assumed that this is because CO.sub.2 is not exhausted but dissolved in the medium in the form of carbonic acid under the closed environment to interrupt an increase in pH and the action of a formate dehydrogenase.

(102) The action of a formate dehydrogenase has an effect of increasing a pH of the medium through the consumption of formic acid. However, it is considered that under an open environment, CO.sub.2 is well exhausted and the action of a formate dehydrogenase becomes active, and thus the pH is highly increased. Under such a high pH condition, the action of a mannitol dehydrogenase is inhibited, and thus a mannitol production amount is reduced.

(103) Therefore, given that the action of a formate dehydrogenase is a reaction which provides NADH required for the action of a mannitol dehydrogenase, only if the increase in pH at more than a level in which a reaction of the mannitol dehydrogenase is inhibited is prevented under an open environment in which the reaction of a formate dehydrogenase is active, it was assumed that high mannitol productivity can be obtained.

(104) 13. Mannitol Production Using Weakly Acidic Conversion Medium in Open Environment

(105) As confirmed in the above-described Example 12, when a test tube was used, NADH was sufficiently provided by the active action of a formate dehydrogenase. Therefore, in addition to the open environment, to be suitable for mannitol production, mannitol production was carried out by compensating pH to be weakly acidic. A PIPES (pH 6) buffer solution was added to a conversion medium to have a final concentration of 300 mM, resulting in pH compensation, and a conversion temperature was 45 C. Results are shown in FIG. 13.

(106) Referring to FIG. 13, when a medium composed of only a substrate and water was used without pH compensation, mannitol production was low as 3.6 g/L after 12 hours of the conversion reaction, but when the pH was compensated with a PIPES buffer solution after 24 hours of the conversion reaction, it can be confirmed that 51 g/L, which was the half of 100 g/L fructose, of the substrate was converted into mannitol. It is considered that this is caused by a synergistic effect of improvement in NADH regeneration rate of the formate dehydrogenase in the open environment and improvement in mannitol production rate of the mannitol dehydrogenase due to the pH compensation by the use of the PIPES buffer solution.

(107) 14. Improvement in Uptake of Substrate by Introduction of Fructose Transport Protein

(108) (1) Development of Recombinant Corynebacterium Strain

(109) A mannitol production amount was to be improved by increasing an uptake amount of fructose, which is a substrate in cells using a glucose transport protein (GLF) well known to uptake the fructose. The glucose transport protein was Zymomonas mobilis (Zymomonas mobilis subsp. mobilis ZM4 or an ATCC 31821)-derived enzyme, and purchased from KCTC (KCTC 1534). PCR for the glucose transport protein of Zymomonas mobilis (GenBank: AAG29864.1; GI: 11095424; SEQ ID NO: 50) was performed using a purified genome as a template and a primer pair of SEQ ID NOs: 51 and 52 as primers. The obtained PCR product was inserted into the same restriction site of pJC1-1-cT-Cmp vector using restriction enzymes BamHI and XbaI, thereby constructing recombinant vector pJC1-1-cT-GLF-Cmp.

(110) The expression vector (pJC1-1-cT-Cmp) used in the construction was prepared by modifying pJC1, which was an E. coli-Corynebacterium shuttle vector, and a pJC1-1 vector was constructed by introducing a promoter variable region and an MCS region of a pSGT208 vector and additionally introducing a terminator region of Corynebacterium. A pJC1-1-cT vector was constructed by substituting a lac promoter region with a lac operator of the pJC1-1 vector with a trc promoter, and to be used in expression with the pSGT208 vector used in the previous example, a kanamycin-resistant gene was substituted with a chloramphenicol-resistant gene, thereby finally constructing a pJC1-1-cT-Cmp vector.

(111) The previously manufactured pJC1-1-cT-GLF-Cmp recombinant vector was introduced into Corynebacterium glutamicum ATCC 13032 transformed with pS208cT-LpMDH-FDH, and the recombinant Corynebacterium strain was stored at 80 C.

(112) (2) Experiment for Mannitol Production Using Recombinant Strain

(113) An experiment for mannitol production was performed using the previously constructed recombinant Corynebacterium strain and a PIPES buffer solution conversion medium under the same conditions used in Example 13, and a chloramphenicol concentration was 5 g/mL. Results are shown in FIG. 14.

(114) Referring to FIG. 14, when a glucose transport protein was not introduced, 24 hours after the conversion reaction, 51 g/L of mannitol was produced, but when a glucose transport protein was introduced to improve the uptake of fructose, with the 24-hour conversion reaction, 95 g/L of mannitol was produced, and thus almost 100% of a conversion ratio from the fructose was shown.

(115) Accordingly, psicose was able to be easily separated from mannitol by converting fructose that is not easily separated from psicose into mannitol, and a mixed saccharide of psicose and mannitol itself was able to serve as an expensive functional saccharide.

(116) 15. Increase in Psicose Content Through Mannitol Crystallization in Mannitol Conversion Reaction Solution

(117) (1) Crystallization of Mannitol Through Ethanol Addition

(118) Conventionally used fructose and psicose are difficult to be separated and purified since they have similar physical properties. However, since mannitol has a large difference in solubilities between fructose and psicose, separation and purification can be easily performed. In a psicose production process, fructose, which is a substrate, is converted into psicose at a conversion ratio of approximately 30%. Therefore, when 400 g/L of fructose is used, in a reaction solution after the reaction was completed, approximately 280 g/L of fructose is mixed with approximately 120 g/L of psicose.

(119) A mannitol conversion reaction from remaining fructose was performed in the reaction solution using the recombinant Corynebacterium strain (pS208cT-MDH-FDH) of Example 13, thereby obtaining a final reaction solution in which fructose (140 g/L), mannitol (140 g/L) and psicose (120 g/L) are mixed. 500 L of. The mixed saccharide solution of fructose (140 g/L), mannitol (140 g/L) and psicose (120 g/L) was dispensed into an Eppendorf tube and 100 to 900 L of ethanol was added to separate mannitol which has a low solubility by crystallization.

(120) Following the addition of ethanol, the tube was stayed at room temperature (25 C.) for 1 hour, and centrifuged (13,000 rpm, 10 min) to separate the crystallized mannitol from the solution. A saccharide concentration in the solution obtained through centrifugation was assessed by HPCL under the conditions used in Example 9. Results are shown in FIG. 15.

(121) Referring to FIG. 15, it can be confirmed that the fructose and psicose concentrations in the mannitol conversion reaction solution are maintained during the mannitol crystallization, but only the mannitol concentration is considerably reduced. Therefore, the solution with an increased psicose content may be easily obtained by separating only mannitol through ethanol addition.

(122) (2) Mannitol Crystallization Through Evaporation

(123) 100 mL of the mixed saccharide solution (fructose 140 g/L, mannitol 140 g/L, psicose 120 g/L) for the mannitol conversion reaction was added to a beaker, and boiled in water while stirring. Due to evaporation of the mixed saccharide solution, a concentration of the total solution increases, and a volume thereof decreases. Here, when the volume of the solution was gradually reduced to 100 mL, 75 mL and 50 mL due to the evaporation, 1 mL of the solution was cooled at room temperature (25 C.) to crystallize mannitol. After the cooled solution was centrifuged (13,000 rpm, 10 min), the supernatant was subjected to HPLC under the conditions as used in Example 9 to measure saccharide concentrations. Results are shown in FIG. 16.

(124) Referring to FIG. 16, it can be confirmed that the fructose and psicose concentrations in the early mixed saccharide solution were concentrated and increased, but the mannitol concentration was reduced. Therefore, when the fructose is converted into mannitol, the psicose ratio in the mixed saccharide may be increased by a simple method for increasing a temperature.

(125) 16. Psicose Separation from Mixed Saccharide Using HPLC

(126) Resolution patterns of mixed saccharides in a reaction solution before and after a mannitol conversion reaction were observed under the HPLC analysis conditions of Example 1. The analyses were performed with a reaction solution for psicose production (psicose 120 g/L, fructose 280 g/L; FIG. 17(A)), a 50% mannitol conversion reaction solution (psicose 120 g/L, fructose 140 g/L, mannitol 140 g/L; FIG. 17(B)), and a 100% mannitol conversion reaction solution (psicose 120 g/L, mannitol 280 g/L; FIG. 17(C)).

(127) Referring to FIG. 17(A), it can be confirmed that the psicose and the fructose have a wide overlapping peak area due to similar retention time. Referring to FIG. 17(B), it is shown that as the fructose is converted 50% into mannitol which has a longer retention time, the overlapping region between the psicose and the fructose is reduced. Referring to FIG. 17(C), when the fructose is all converted into mannitol, it can be confirmed that the HPCL peaks of the psicose and the mannitol are completely separated.

(128) A column used for HPLC in this experiment is a laboratory-level column. Compared to the laboratory-level column, an industrial HPLC column has a larger capacity, can directly separate a high concentration of solution without dilution, but has a lower resolution. Thus, when the industrial HPLC column is used, the overlapping area of the psicose and the fructose will be more obviously shown. Therefore, to efficiently separate psicose from fructose with the industrial HPLC column, it is necessary to convert remaining fructose into mannitol.