METHOD FOR PRODUCING PSICOSE

Abstract

A method of preparing D-psicose includes a step of reacting D-fructose as a substrate and an epimerase thereof in microorganisms at a temperature of 40° C. or higher. The method may further includes inducing the microorganisms to have resting cells by culturing the microorganisms in a medium not containing the D-fructose before the reaction. The method of preparing D-psicose may significantly improve the production amount of D-psicose and production rate of D-psicose.

Claims

1. A method of preparing D-psicose, comprising a step of reacting D-fructose as a substrate and an epimerase thereof in microorganisms at a temperature of 40° C. or higher.

2. The method according to claim 1, wherein the microorganisms express the epimerase endogenously or by transformation.

3. The method according to claim 1, further comprising a step of inducing the microorganisms to have resting cells by culturing the microorganisms in a medium not containing the D-fructose before the reaction.

4. The method according to claim 1, further comprising a step of recovering and reusing the microorganisms to convert D-fructose into D-psicose after the reaction.

5. The method according to claim 4, wherein the microorganisms are gram-positive bacteria.

6. The method according to claim 1, wherein the D-fructose is provided from a medium containing only inorganic salts and D-fructose.

7. The method according to claim 6, wherein the inorganic salts are manganese salts or cobalt salts.

8. The method according to claim 1, wherein the reaction temperature is 40 to 90° C.

9. The method according to claim 1, wherein the microorganisms are Escherichia, Bacillus, Corynebacterium, Actinomyces, yeasts, Kluyveromyces or combinations thereof.

10. The method according to claim 1, wherein the microorganisms are transformed with a gene encoding Agrobacterium tumefaciens-derived D-psicose 3-epimerase corresponding to SEQ ID NO: 1 or a gene encoding Anaerostipes caccae-derived D-psicose 3-epimerase corresponding to SEQ ID NO: 2.

11. The method according to claim 1, wherein the microorganisms are transformed with a gene encoding an amino acid sequence of SEQ ID NO: 5.

12. The method according to claim 1, wherein the microorganisms are transformed with a gene encoding an amino acid sequence in which the 32nd amino acid is substituted with leucine or the 196th amino acid is substituted with cysteine in an amino acid sequence of SEQ ID NO: 6.

13. The method according to claim 1, wherein the microorganisms are transformed with a gene encoding Clostridium-derived D-psicose 3-epimerase.

14. The method according to claim 13, wherein the gene is a gene encoding Clostridium bolteae-derived D-psicose 3-epimerase corresponding to SEQ ID NO: 7 or a gene encoding Clostridium hylemonae-derived D-psicose 3-epimerase corresponding to SEQ ID NO: 8.

Description

DESCRIPTION OF DRAWINGS

[0069] FIG. 1 shows the amount of D-psicose produced from a substrate, D-fructose, depending on temperature in a reaction of converting D-psicose 3-epimerase-introduced Corynebacterium glutamicum transformants into resting cells (i.e., a reaction that produces D-psicose by reacting D-fructose and an epimerase).

[0070] FIG. 2 shows the amount of D-psicose produced from a substrate, D-fructose, depending on temperature in a reaction of converting D-psicose 3-epimerase-introduced Escherichia coli MG1655 transformants into resting cells.

[0071] FIG. 3 shows the amount of D-psicose produced in reused microorganisms. A reaction that converts transformants of D-psicose 3-epimerase-introduced Corynebacterium glutamicum and D-psicose 3-epimerase-introduced Escherichia coli MG1655 into resting cells was performed at 60° C. for 3 hours, and then the transformants were recovered and reacted under the same reaction conditions to determine the amount of produced D-psicose.

[0072] FIG. 4 shows the amount of D-psicose produced from D-fructose depending on the composition of a medium for D-psicose production used in a reaction of converting D-psicose 3-epimerase-introduced Corynebacterium glutamicum transformants into resting cells.

[0073] FIG. 5 shows the production amount of D-psicose depending on strains from which D-psicose 3-epimerase is derived and the composition of a medium for D-psicose production when performing a reaction of converting D-psicose 3-epimerase-introduced Corynebacterium glutamicum transformants into resting cells.

[0074] FIG. 6 shows the production amount of D-psicose depending on strains from which D-psicose 3-epimerase is derived and heating time when performing a reaction of converting D-psicose 3-epimerase-introduced Corynebacterium glutamicum transformants into resting cells.

[0075] FIG. 7 is a result of comparing the amino acid sequences of D-psicose 3-epimerases derived from various strains.

[0076] FIG. 8 shows the production amount of D-psicose depending on the number of times of reuse of Corynebacterium glutamicum into which Clostridium-derived D-psicose 3-epimerase was introduced.

MODES OF THE INVENTION

[0077] Hereinafter, the present invention is described in detail with reference to examples.

EXAMPLES

[0078] 1. Changes in D-Psicose Production Rate and Production Amount Depending on Temperature in Process of Producing D-Psicose from D-Fructose Using Corynebacterium glutamicum Strain

(1) Preparation of Recombinant Strains

[0079] pCES208 (J. Microbiol. Biotechnol., 18:639-647, 2008), a shuttle vector for Escherichia coli-Corynebacterium, was modified to produce a pSGT208 shuttle vector in which a terminator and a lac promoter were inserted.

[0080] To produce D-psicose in Corynebacterium glutamicum, the 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), which encodes D-psicose 3-epimerase, was introduced into the prepared pSGT208 shuttle vector.

[0081] Specifically, a dpe gene was amplified from the genome of Agrobacterium tumefaciens using primer 1 of SEQ ID NO: 31 and primer 2 of SEQ ID NO: 32, digested with restriction enzymes, KpnI and BamHI, and inserted into the pSGT208 shuttle vector digested with the same enzymes to produce a pS208-dpe recombinant shuttle vector containing D-psicose 3-epimerase.

[0082] Thereafter, to increase the expression level of D-psicose 3-epimerase in Corynebacterium glutamicum, the lac promoter of the pS208-dpe vector was substituted with a pTrc99a-derived trc promoter. This was named pS208cT-dpe.

[0083] The prepared recombinant vectors, pS208-dpe and pS208cT-dpe, containing D-psicose 3-epimerase and pSGT208 vector as a negative control thereof, were introduced into wild-type Corynebacterium glutamicum ATCC 13032, and the transformed Corynebacterium glutamicum ATCC 13032 was used to produce D-psicose from D-fructose. A transformation method followed a method specified in Handbook of Corynebacterium glutamicum (Lothar Eggeling et. al., ISBN 0-8493-1821-1, 2005 by CRC press).

(2) Cultivation of Recombinant Strains and Production of D-Psicose Using the Same.

[0084] To obtain microorganisms at a high concentration, the above-prepared Corynebacterium glutamicum transformants were inoculated into 5 ml of a LB medium (Difco) containing 20 μg/mL of kanamycin and subjected to seed culture at 30° C. and 250 rpm. After culture, the seed culture was inoculated into a minimal medium (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.45H2O, 50 mg NaCl, 2 g urea, 0.1 mg biotin, and 0.1 mg thiamine per liter) containing 10 g/L of glucose and 20 μg/mL of kanamycin and then subjected to main culture. The main culture was performed in a grooved 500 ml Erlenmeyer flask with a 100 ml volume at 30° C. and 180 rpm for 12 hours to induce sufficient cell mass and sufficient expression of proteins.

[0085] The obtained culture solution was centrifuged to remove a supernatant and recover the microorganisms. The microorganisms were resuspended to an OD.sub.600 value of 40 in the same minimal medium as above described, containing 40% (w/v) D-fructose as a substrate, and then a reaction that converts the microorganisms into resting cells was performed at 25, 30, 37, 50, 60 or 70° C. and 180 rpm.

[0086] High-performance liquid chromatography (HPLC) was used to determine the concentrations of D-fructose and D-psicose. HPLC was performed using SCL-10A (Shimadzu, Japan) equipped with a Kromasil 5NH.sub.2 column (4.6 mm×250 mm), and mobile phases were separated at 40° C. with 75% acetonitrile at a flow rate of 1.5 mL/min and analyzed using a RI (Reflective Index) detector. Under the above conditions, retention times were 5.5 minutes for D-fructose and 4.6 minutes for D-psicose.

[0087] The measurement results are shown in FIG. 1. Referring to FIG. 1, when a conversion reaction was performed on a Corynebacterium glutamicum ATCC13032 strain, in which a pSGT208cT-dpe shuttle vector was introduced, in a medium containing 40% D-fructose, the production rate of D-psicose was remarkably increased and the production amount of D-psicose was increased in proportion to reaction temperature. In particular, in experiment groups reacted at 50, 60, and 70° C., the enzymatic reaction of D-psicose 3-epimerase reached equilibrium within about 3 hours, producing 120 g/L of D-psicose. These results indicated that a conversion rate at which D-psicose 3-epimerase converts D-fructose to D-psicose and the production amount of D-psicose are temperature-dependent.

[0088] The production amount of D-psicose increased rapidly from 50° C., which was significantly higher than the temperature required for conventional enzymatic reactions. It is considered that the reaction between the enzyme and the substrate has changed at relevant temperatures.

2. Changes in D-Psicose Production Rate and Production Amount Depending on Temperature in Process of Producing D-Psicose from D-Fructose using Escherichia Coli

[0089] According to a method described in Example 1 of Korean Patent Registration No. 10-1106253, a pTPE plasmid was prepared by introducing Agrobacterium tumefaciens-derived D-psicose 3-epimerase into a pTrc99A vector, and an E. coli MG1655(ApfkA, als2) strain was transformed with the pTPE plasmid.

[0090] To block the degradation pathway of D-psicose, Escherichia coli MG1655 deficient in pfkA (SEQ ID NO: 11) and als2 (SEQ ID NO: 14, 15, 16, 17, 18 and 19) genes were used.

[0091] To obtain microorganisms at a high concentration, the above-prepared Escherichia coli MG1655 transformants were inoculated into 5 ml of a LB medium (Difco) containing 100 μg/mL of ampicillin and subjected to seed culture at 37° C. and 250 rpm. After culture, the seed culture was inoculated into a 2YT medium containing 10 g/L of glucose and 100 μg/mL of ampicillin and then subjected to main culture. The main culture was performed in a grooved 500 ml Erlenmeyer flask with a 100 ml volume at 37° C. and 180 rpm for 12 hours to induce sufficient cell mass and sufficient expression of proteins.

[0092] The obtained culture solution was centrifuged to remove a supernatant and recover the microorganisms. The microorganisms were resuspended to an OD.sub.600 value of 40 in an Escherichia coli minimal medium, a M9 medium (11.3 g M9 minimal salts (Difco), 0.1 mL 1 M CaCl.sub.2, 2 mL 1 M MgSO.sub.4, 1 mL 100 mM MnSO.sub.45H.sub.2O per liter), containing 40% (w/v) D-fructose as a substrate, and then a reaction that converts the microorganisms into resting cells was performed at 37, 60 or 70° C. and 180 rpm. The concentrations of D-fructose and D-psicose were analyzed according to the method described in Example 1. Measurement results are shown in FIG. 2.

[0093] Referring to FIG. 2, when a conversion reaction was performed on an Escherichia coli MG1655 (ApfkA, als2) strain, in which a pTPE vector was introduced, in a medium containing 40% D-fructose, the production rate and production amount of D-psicose were increased in proportion to reaction temperature. In particular, in experiment groups reacted at 60 and 70° C., as with experiments using Corynebacterium, the enzymatic reaction of D-psicose 3-epimerase reached equilibrium within about 2 hours, producing 120 g/L of D-psicose.

[0094] These results indicated that a conversion rate at which D-psicose 3-epimerase converts D-fructose to D-psicose and the production amount of D-psicose are temperature-dependent.

[0095] When a reaction of converting microorganisms into resting cells was performed at a high temperature on representative gram-positive bacteria, Corynebacterium, and representative gram-negative bacteria, Escherichia coli, since D-psicose 3-epimerase, a glycosyltransferase, was protected from an extreme external environment in cells, the D-psicose 3-epimerase produced D-psicose at a higher rate without heat denaturation at high temperature in comparison with a pure enzyme state. The advantage of such cell conversion reaction is that the reaction may be applied to most microorganisms.

3. Continuous Production of D-Psicose from D-Fructose by Recovering and Reusing Cells in Reaction of Converting Corynebacterium glutamicum Transformants into Resting Cells

[0096] In the results of Examples 1 and 2, when D-psicose 3-epimerase was introduced into Corynebacterium and Escherichia coli transformants and a conversion reaction was performed at a high temperature of 50° C. or more, the production amount of D-psicose reached a maximum within 3 hours and did not increase anymore.

[0097] To confirm whether D-psicose 3-epimerase still retained the activity of converting D-fructose into D-psicose even after 3 hours at which the reaction equilibrium and maximal production of D-psicose were reached, the microorganisms, which were used for D-psicose production in the conversion reaction for 3 hours in the presence of D-fructose, were recovered and reused in another conversion reaction for D-psicose production.

[0098] The reaction of converting the reused microorganisms into resting cells were repeated three times at a temperature of 60° C.The first reaction of converting into resting cells is represented by R0, a reaction of converting into resting cells, in which cells recovered from the previous reaction solution are reused once, is represented by R1, a reaction of converting into resting cells, in which the cells are reused twice, is represented by R2, and a reaction of converting into resting cells, in which the cells are reused three times, is represented by R3. Culture conditions and analysis methods were the same as in Example 1. Results are shown in FIG. 3.

[0099] Referring to FIG. 3, cells may be reused even when the sugar conversion reaction is performed at a high temperature of 60° C. However, as the cells were reused, enzymatic activity decreased to some extent. In addition, when the cells were reused at high temperature, a Corynebacterium glutamicum strain, a gram-positive bacterium, had higher residual enzymatic activity than an Escherichia coli MG1655 strain, a gram-negative bacterium.

4. Production of D-Psicose from D-Fructose in Various Media for Reaction of Converting into Resting Cells

[0100] In Example 1, when a conversion reaction of producing D-psicose from D-fructose in Corynebacterium glutamicum was performed, a minimal medium (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, and 0.1 mg thiamine per liter) containing 40% fructose was used. The components of the minimal medium used in this conversion reaction were minimized to prepare a more economical and convenient medium, and the minimal medium was compared with the medium used in Example 1 in terms of D-psicose productivity.

[0101] Referring to FIG. 4, even with a phosphate-buffered (pH 7) medium containing 40% fructose and 0.1 mM MnSO.sub.4, more simply, a medium containing only 40% fructose and 0.1 mM MnSO.sub.4, there was no significant difference in the final amount of D-psicose production. The above pattern was observed uniformly at reaction temperatures of 30 and 60° C. Without addition of 0.1 mM MnSO.sub.4, the cofactor of D-psicose-3-epimerase, the production amount decreased.

[0102] According to the above example, it can be seen that only D-fructose as a substrate, and MnSO.sub.4, a cofactor of D-psicose-3-epimerase, are required as constituents of a medium used in a conversion reaction for producing D-psicose.

5. Preparation of Recombinant Corynebacterium Glutamicum Containing Polynucleotides Encoding Amino Acid Sequences of D-Psicose 3-Epimerases Derived from Various Strains

[0103] The whole genome of Anaerostipes caccae (Anaerostipes caccae DSM 14662; taxid: 411490) was purchased from DSMZ, Germany. The first PCR was performed using a primer pair of SEQ ID NOS: 33 and 34 to include a gene (AP endonuclease; Sequence ID: gb|EDR98778.1|; GI: 167654649; SEQ ID NO: 4), which is presumed to be D-psicose 3-epimerase, using the purchased whole genome as a template. The second PCR was performed using a primer pair of SEQ ID NOS: 35 and 36, which specifically bind to a D-psicose 3-epimerase gene, using the amplified PCR product as a template.

[0104] The obtained PCR product was digested with restriction enzymes, BamHI and XbaI, and inserted into the restriction sites of pS208cT-dpe (vector described in Example 1 disclosed in Korean Patent Application No. 10-2013-0060703) digested with the same restriction enzymes to produce a recombinant vector, pS208cT-AcDPE vector.

[0105] The resulting pS208cT-AcDPE vector was transformed into wild-type Corynebacterium glutamicum ATCC 13032 and used for the production of D-psicose from D-fructose. A transformation method followed a method specified in Handbook of Corynebacterium glutamicum (Lothar Eggeling et. al., ISBN 0-8493-1821-1, 2005 by CRC press).

[0106] The obtained recombinant Corynebacterium glutamicum strain was stored at −80° C. and used for culture.

[0107] A plasmid containing a gene (hypothetical protein CLOBOL_00069; Sequence ID: gb|EDP19602.1|; GI:15844190; SEQ ID NO: 9), which is presumed to be D-psicose 3-epimerase 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 a D-psicose 3-epimerase gene.

[0108] The obtained PCR product was digested with restriction enzymes, KpnI and XbaI, and inserted into the restriction sites of pS208cT-dpe (vector described in Example 1 disclosed in Korean Patent Application No. 10-2013-0060703) digested with the same restriction enzymes to produce a recombinant vector, pS208cT-CbDPE vector.

[0109] The resulting recombinant pS208cT-CbDPE vector was transformed into wild-type Corynebacterium glutamicum ATCC 13032 using the same method as described above and used for the production of D-psicose from D-fructose. The obtained recombinant Corynebacterium glutamicum strain was stored at −80° C. and used for culture.

[0110] The whole genome of Clostridium hylemonae (Clostridium hylemonae DSM 15053; taxid:553973) was purchased from DSMZ, Germany. The first PCR was performed using a primer pair of SEQ ID NOS: 39 and 40 to include a gene (dolichol monophosphate mannose synthase; Sequence ID:ref|WP_006442985.1—; GI:225161759; SEQ ID NO: 10), which is presumed to be D-psicose 3-epimerase, using the purchased whole genome as a template. The second PCR was performed using a primer pair of SEQ ID NOS: 41 and 42, which specifically bind to a D-psicose 3-epimerase gene, using the amplified PCR product as a template.

[0111] The obtained PCR product was digested with restriction enzymes, BamHI and XbaI, and inserted into the restriction sites of pS208cT-dpe (vector described in Example 1 disclosed in Korean Patent Application No. 10-2013-0060703) digested with the same restriction enzymes to produce a recombinant vector, pS208cT-ChDPE vector.

[0112] The resulting pS208cT-ChDPE vector was transformed into wild-type Corynebacterium glutamicum ATCC 13032 using the same method as described above and used for the production of D-psicose from D-fructose. The obtained recombinant Corynebacterium glutamicum strain was stored at −80° C. and used for culture.

6. Production of D-Psicose from D-Fructose using Recombinant Corynebacterium Glutamicum Strains into Which D-Psicose 3-Epimerases Derived from Various Strains are Introduced

[0113] Corynebacterium glutamicum transformants prepared in Example 5 were used to produce D-psicose from high concentration of D-fructose.

[0114] The transformants were inoculated into a 2YT medium containing 20 μg/mL of kanamycin and subjected to seed culture at 30° C. and 250 rpm. After culture, the seed culture was inoculated into a 2YT medium containing 20 μg/mL of kanamycin and subjected to main culture. The main culture was performed in a grooved 300 ml Erlenmeyer flask with a 60 ml volume at 30° C. and 180 rpm for 7 hours to induce sufficient cell mass and sufficient expression of proteins.

[0115] The obtained culture solution was centrifuged to remove a supernatant and recover the microorganisms. The microorganisms were resuspended in a simple conversion reaction medium containing 20 μg/mL kanamycin, 40% (w/v) D-fructose as a substrate, and 0.1 mM concentration of manganese or cobalt known as a primary cofactor of D-psicose 3-epimerase, and then a reaction that converts the microorganisms into resting cells was performed at 55° C. The concentrations of D-fructose and D-psicose were measured in the same method as described in Example 1. The results are shown in FIG. 5 (AtDPE refers to D-psicose 3-epimerase of existing Agrobacterium tumefaciens).

[0116] Referring to FIG. 5, when using cobalt rather than manganese as a cofactor, the production rate of D-psicose was slightly faster in all types of D-psicose 3-epimerase-introduced recombinant strains.

[0117] In the case of recombinant Corynebacterium glutamicum into which Anaerostipes-derived and Agrobacterium-derived D-psicose 3-epimerases were introduced, respectively, the production amount of D-psicose reached equilibrium at 6 hours. On the other hand, in the case of recombinant Corynebacterium glutamicum into which Clostridium-derived D-psicose 3-epimerase was introduced, the production amount of D-psicose reached equilibrium at 3 hours when manganese was used as a cofactor.

[0118] It was confirmed that the Clostridium-derived D-psicose 3-epimerase produced D-psicose from D-fructose, and furthermore, it was found that the production rate of D-psicose was faster than that of Agrobacterium-derived D-psicose 3-epimerase.

7. Continuous Maintenance of D-Psicose Production Activity of Recombinant Corynebacterium Glutamicum Strains at High Temperature, into Which D-Psicose 3-Epimerases Derived from Various Strains are Introduced

[0119] In the results of Examples 1 and 2, when a conversion reaction was carried out at a high temperature of 50° C. or higher, all D-psicose was produced within 3 hours. Because of the rapid production of D-psicose from D-fructose at high temperature, a stable D-psicose 3-epimerase is required even at constant high temperature when the cells are reused. Accordingly, it was confirmed to what extent the activity of D-psicose 3-epimerases derived from various strains was maintained at high temperature.

[0120] The cells obtained by the method described in Example 1 were suspended in 2YT and the heat of 60° C. was continuously given for 0, 3, 6, 9, 12, and 24 hours using a shaking incubator. The heat-treated cells were collected at each time point and suspended in a simple medium for a conversion reaction containing only 20 μg/mL kanamycin, 0.1 mM manganese, and 40%(w/v) D-fructose, followed by a conversion reaction at 60° C. for 3 hours. The concentrations of D-fructose and D-psicose were measured by the same method as described in Example 1. The results are shown in FIG. 6.

[0121] Referring to FIG. 6, in the case of conventional recombinant Corynebacterium glutamicum into which Agrobacterium-derived D-psicose 3-epimerase and Anaerostipes-derived D-psicose 3-epimerase were introduced, respectively, the production of D-psicose hardly occurred when heat of 60° C. was applied for 3 hours. On the other hand, in the case of recombinant Corynebacterium glutamicum into which Clostridium-derived D-psicose 3-epimerase was introduced, the production of D-psicose was maintained even after heat was applied for 24 hours. These results indicate that Clostridium-derived D-psicose 3-epimerase is better than Agrobacterium-derived D-psicose 3-epimerase in the production of D-psicose in a high temperature process.

[0122] In the case of Agrobacterium tumefaciens-derived D-psicose 3-epimerase, when the 33rd and 213th amino acids in a sequence (Reference 1) important for thermal stability are substituted with leucine and cysteine respectively, the half-life of the enzyme is increased 3.3-fold and 7.2-fold, respectively, at 50° C. Furthermore, it is known that when both of these amino acids are substituted, the half-life of the enzyme is increased 29.9-fold.

[0123] A comparison between the amino acid sequence of Agrobacterium tumefaciens-derived D-psicose 3-epimerase and the amino acid sequence of Clostridium-derived D-psicose 3-epimerase is shown in FIG. 7. As a result, Clostridium-derived D-psicose 3-epimerase with high thermal stability had one or two amino acid sequence characteristics important for thermal stability as described above.

8. Production of D-Psicose from D-Fructose by Cell Recovery and Reuse in Reaction of Converting Recombinant Corynebacterium Glutamicum, into Which Clostridium-Derived D-Psicose 3-Epimerase is Introduced, into Resting Cells

[0124] In Example 7, it was confirmed that Clostridium-derived D-psicose 3-epimerase exhibited high stability at high temperature. Therefore, among two Clostridium-derived D-psicose 3-epimerase-introduced recombinant strains, the effect of reusing cells after a conversion reaction at high temperature was confirmed for recombinant Corynebacterium glutamicum into which Clostridium hylemonae-derived D-psicose 3-epimerase was introduced.

[0125] Cells obtained in the same method as described in Example 3 were subjected to a reaction for converting into resting cells at 60° C. for 3 hours, and then the cells were recovered again and reacted in the same manner. The cells were reused three times in total (experiments were carried out under the same conditions as Example 3). The first reaction of converting into resting cells is represented by R0, a reaction of converting into resting cells, in which cells recovered from the previous reaction solution are reused once, is represented by R1, a reaction of converting into resting cells, in which the cells are reused twice, is represented by R2, and a reaction of converting into resting cells, in which the cells are reused three times, is represented by R3. The results are shown in FIG. 8.

[0126] Referring to FIG. 8, when cells were reused several times at high temperature, recombinant cells, into which Clostridium-derived D-psicose 3-epimerase was introduced, showed no reduction in the production amount of D-psicose. When the results in Example 7 are taken into consideration, it is confirmed that the thermal stability of an enzyme itself is very advantageous not only in the continuous sugar conversion reaction at high temperature but also in reuse of cells.