METHOD FOR OBTAINING 1-KESTOSE

Abstract

The present invention discloses an industrial scale method to obtain 1-kestose by the use of a recombinant fructosyltransferase (FTF), isolated from Festuca arundinacea, expressed constitutively in a non-saccharolytic yeast. In this invention, the recombinant FTF type sucrose:sucrose 1-fructosyltransferase (1-SSTrec) is produced constitutively, stable and at high yield, both in the culture supernatant and in intact cells of the host Pichia pastoris. Hence, the invention additionally provides a method for 1-SST production at industrial scale. The recombinant enzyme is then used for mass production of short-chain fructooligosaccharides (FOS), specifically 1-kestose, from sucrose. The method of the present invention establishes conditions that allow conversion rates where the synthesized FOS constitute above 55% (w/w) of the total sugars in the reaction mixture and the 1-kestose content reaches values higher than 90% (w/w) of the total FOS fraction.

Claims

1. A method for the production of 1-kestose on an industrial scale characterized by the conversion of sucrose into 1-kestose in a bioreactor with the use of a recombinant fructosyltransferase (FTF) from Festuca arundinacea expressed constitutively in a non-saccharolitic yeast.

2. The method of claim 1 wherein the FTF is a sucrose:sucrose 1-fructosyltransferase (1-SST).

3. The method of claim 1 wherein the non-saccharolitic yeast is a Pichia pastoris strain.

4. The method of claim 3 wherein the Pichia pastoris strain contains multiple copies of the gene encoding the 1-SST integrated in the genome.

5. The method of claim 3 wherein the FTF is recovered from the supernatant and/or the cell sediment of the Pichia pastoris culture.

6. The method of claim 1 wherein the sucrose concentration is above 400 g/L.

7. The method of claim 1 wherein the FTF is produced by the recombinant yeast grown in a fermentor with discontinuous, continuous or fed-batch operation.

8. The method of claim 7 wherein the carbon source used for the yeast growth is a compound selected from glycerol, glucose and sucrose of any purity degree.

9. The method of claim 2 wherein the sucrose conversion into 1-kestose is performed by free or immobilized 1-SST.

10. The method of claim 2 wherein the conversion of sucrose into 1-kestose is performed in a bioreactor of the type stirred tank, fixed bed or membrane.

11. The method of claim 10 wherein the membrane bioreactor is operated in a continuous or semicontinuous mode.

12. A crude enzyme preparation for sucrose conversion into 1-kestose at industrial scale comprising the F. arundinacea 1-SST expressed constitutively in a non-saccharolitic yeast.

13. The crude enzyme preparation of claim 12 wherein the non-saccharolitic yeast is a Pichia pastoris strain.

14. The crude enzyme preparation of claim 13 wherein the Pichia pastoris strain contains multiple copies of the gene encoding 1-SST integrated in the genome.

15. The crude enzyme preparation of claim 12 wherein 1-SST is in solid or liquid state and in free or immobilized form.

16. The crude enzyme preparation of claim 13 wherein 1-SST is recovered from the supernatant and/or the cell sediment of the Pichia pastoris culture.

17. The crude enzyme preparation of claim 12 wherein the sucrose concentration is above 400 g/L.

18. (canceled)

19. (canceled)

20. (canceled)

21. A product for human or animal feeding comprising the 1-kestose produced by the method of claim 1.

22. The product of claim 21 which is additionally formulated with probiotic preparations, constituting a symbiotic for its use as a nutraceutical.

Description

BRIEF DESCRIPTION OF FIGURES

[0041] FIG. 1. Graphic representation of the expression cassettes resulting from the insertion of the genes encoding the respective 1-SST enzyme from tall fescue (1-sstf), onion (1-sstc) and blue agave (1-ssta) in the vector pGAPZA,B,C. The corresponding plasmids were named as p1-SSTF (tall fescue 1-SST), p1-SSTC (onion 1-SST), and p1-SSTA (blue agave 1-SST), respectively. P.sub.GAP: GAP promoter, SP: -factor signal peptide of Saccharomyces cerevisiae, His.sub.6: polyhistidine tag, TT: AOX1 transcription terminator of Pichia pastoris.

[0042] FIG. 2. Strategies used in the construction of plasmid p1SSTF6x+6 additional copies of the expression cassette for constitutive high-level expression of the 1-sstf gene in the Pichia pastoris strain GS115. A. Construction with one gene copy. B. Construction with six gene copies, selection by histidine4 gene complementation C. Construction with nine gene and hygromycin resistance. P.sub.GAP: GAP promoter, 5AOX1: promoter of alcohol oxidase, SP: -factor signal peptide of S. cerevisiae, His.sub.6: polyhistidine tag, TT: AOX1 transcription terminator, His4: non-mutated histidine4 gene.

[0043] FIG. 3. Time course of FOS synthesis by the enzyme 1-SSTrec in a batch stirred tank reactor. Reaction conditions: 1-SSTrec 9000 U/L, sucrose 600 g/L in 0.1 M sodium acetate buffer (pH 5.5); temperature 40 C., stirring speed 250 rpm, reaction time 6 h. Legend: 1-kestose (GF.sub.2); nystose (GF.sub.3), total FOS (FOS=sum of GF.sub.2 and GF.sub.3); sucrose (GF); glucose (G); fructose (F).

[0044] FIG. 4. HPLC chromatograms showing the product profile in samples retrieved at different reaction times (t.sub.r). A. A mixture of the standards nystose (GF.sub.3), 1-kestose (GF.sub.2), sucrose (GF), glucose (G), and fructose (F). B. FOS synthesis by the 1-SSTrec at reaction time t.sub.r=0. C. t.sub.r=180 minutes. D. t.sub.r=360 minutes.

[0045] FIG. 5. Schematic representation of the system designed for FOS production in a membrane bioreactor.

[0046] FIG. 6. Time course of 1-kestose synthesis by the enzyme 1-SSTrec in a membrane bioreactor during 3 consecutive cycles of semicontinuous operation. Reaction conditions: 1-SSTrec 9000 U/L, sucrose 600 g/L in 0.1 M sodium acetate buffer (pH 5.5); temperature 40 C., stirring speed 250 rpm. Operation sequence: Cycle no. 1, continuous synthesis reaction for 3 h and discharge (D) for 30 min. Cycles no. 2-3, continuous synthesis reaction for 2.5 h and discharge (D) for 30 min.

[0047] FIG. 7. HPLC chromatograms showing the product profile in samples retrieved after each consecutive cycle of FOS synthesis in a membrane bioreactor with semicontinuous operation. A, Cycle No. 1, B, Cycle No. 2. and C, Cycle No. 3. Legend: Nystose (GF.sub.3), Kestose (GF.sub.2), Sucrose (GF), Glucose (G), Fructose (F).

EXAMPLES

Example 1. Comparative Study of Fructosyltransferase (FTF) Activity Levels Displayed by Three Sucrose:Sucrose 1-Fructosytransferase (1-SST) from Plants Produced in Pichia pastoris

[0048] To compare the FTF activity levels of the three enzymes mentioned above, the cDNAs encoding the 1-SST enzyme from tall fescue (Festuca arundinacea), onion (Allium cepa), and blue agave (Agave tequilana) were isolated from its native hosts via Reverse transcription (RT)-Polymerase Chain Reaction (PCR) using primers previously described in the literature [Vijn et al. 1998, The Plant Journal 11:387-398; Luscher et al. 2000, Plant Physiology 124:1217-1227; Avila-Fernandez et al. 2007, Plant Science 173:478-486].

[0049] The amplified PCRs products corresponding to the DNAs coding for the mature enzyme (Sizes: 1-SST fescue: 1668 bp, 1-SST onion: 1668 bp and 1-SST Agave: 2026 bp) were fused at its 5 end, following the correct reading frame, to the S. cerevisiae a factor signal peptide and at the 3end to the sequences encoding both, the myc epitope and six histidine residues tag present in the commercial vector pGAPZ a C (Invitrogen, Leek, Holland). This commercial vector allows selection of the resulted transformants by resistance to the antibiotic zeocin.

[0050] In the three obtained constructs, called p1-SSTF (fescue 1-SST), p1-1SSTC (onion 1-SST) and p1-SSTA (1-SST agave), the chimeric genes coding for three1-SSTs were placed under the transcriptional control of the GAP promoter and the transcriptional alcohol oxidase 1 terminator (AOX1TT), as shown in FIG. 1.

[0051] The three constructs were digested at the single AvrII restriction site in the GAP promoter and introduced by electroporation into the genome of the X-33 host yeast strain. As a result, about 20 transformants were obtained for each construct after grown on YP medium supplemented with 2% glycerol and Zeocin 100 m g/mL. For the comparative study, 3 Zeocin resistant clones of each of the three variants were grown in a 5 liter (effective volume) fermenter up to reach the cells stationary phase.

[0052] Each fermenter was inoculated with 200 mL of each clone inoculums previously grown in a shaker.

[0053] To maintain the dissolved oxygen values above 20% in the first stage of the fermentation, agitation was increased automatically from 500 to 900 rpm, and aeration was kept at 1 vvm (volume of air/volume of medium/minute). Once increased the value of dissolved oxygen, indicative of glycerol depletion, the second feed stage started.

[0054] To start the second stage, the air flow Increased to 2 vvm and the culture was fed with 1.5 L of increment solution (with the same initial carbon source) at a flow rate between 5 and 7 mL/L/h controlled by variations of dissolved oxygen values. No toxic effects were observed during the 72 hours of cultivation for the recombinant or wild type strains grown under similar conditions.

[0055] Then, the final culture was separated by centrifugation to produce two fractions, a cell pellet (or biomass) and a culture supernatant. Samples of 0.2 mL of both fractions reacted for 30 minutes with a sucrose solution to 300 g/L (0.87 M). The concentration of liberated glucose as a result of the transfructosylation reaction over sucrose was used as indicative of the activity level of the recombinant clones FTF. The intensity of the transfructosylation reaction was established from a calibration curve which relates the color variations in samples to defined amounts of glucose. A relative high activity (sample color turned red with intensity equivalent to glucose concentrations above 5.5 mM) was observed in the 30-min reactions of the intact cells and the culture supernatant samples from the three clones expressing the tall fescue 1-SSTgene. By contrast, none of the clones carrying either the onion 1-SST gene or the blue agave 1-SST gene showed detectable activity in the 30-min reactions. A slight activity was evident (shift of sample color to light pink being equivalent to glucose concentrations in the range 0.5-5.5 mM) only after the longer incubations for 3 and 5 h.

Example 2. Mean Values of the Parameters Analyzed During Fermentation Run of the Three Single 1-sstf Gene Copy Clones with High FTF Activity

[0056] The three clones carrying a single 1-sstf copy Incorporated in its genome were compared at the fermenter level, using the same experimental conditions described in Example 1.

TABLE-US-00001 TABLE 1 Comparison of the evaluated parameters obtained during the fermentation of the three clones with high 1-SSTrec activity Parameter Value Biomass yield (g/L of culture) 366 4 1-SSTrec Intracellular activity (U/g of wet biomass) 4.3 0.2 1-SSTrec Extracellular activity (U/L of culture 3.7 0.1 supernatant) Culture time (Hours) 69 Total 1-SSTrec activity (U/L of culture) 3955 211 Total 1-SSTrec activity in the Biomass (U/L de cultivo 1573.8 25.6 Productivity of biomass Activity (U/L/h) 22.8

[0057] One 1-SSTrec unit (U) represents the amount of enzyme which releases 1 micromol of glucose per minute when react with a 50% (1.46 M) sucrose solution in sodium acetate buffer 0.1 M (pH 5.5) for 30 minutes at 30 C. The data shown in Table 1 represent the average of the evaluated parameters values obtained in the three fermentations corresponding to each tested clone a standard deviation.

[0058] The productivity of these three clones was 2-fold higher when using the constitutive expression system than that obtained with the methanol inducible system due to a higher concentration of cells (366 g/L) was reached in a shorter culture time (69 hours).

Example 3. Increased 1-SSTrec Activity by Integration of Multiple Copies of the 1-Sstf Gene Expression Cassette in the Pichia pastoris AOX1 Locus

[0059] To develop an economical industrial production technology to produce FOS, high levels of 1-SSTrec activity are required therefore, the need to increase the gene dosage in the host yeast Is needed.

[0060] To obtain multiple copies in tandem of the expression cassette, the plasmid p1-SSTF which contains only one copy of the expression cassette containing the gene 1-sstf in its genome was digested, with the restriction enzymes BamI and BglII. The resulted 2.82 kb band containing the expression cassette was isolated from agarose gel and religated using T4 ligase.

[0061] Joining by T4 ligase of BamHI (G-GATCC) and BglI (A-GATCT) restriction sites, generates a hybrid site with GGATCT sequence, which is not recognized by any of these two enzymes. During the ligation reaction these two enzymes, BamHI and BglII, were added to facilitate the connection of different ends. The 5.64 kb band containing two copies of the expression cassette was isolated and treated again with BamHI and BglII enzymes to ensure that the expression cassettes are joined in the same transcriptional direction.

[0062] This sequence was inserted into the same plasmid p1-SSTF BamHI digested and dephosphorylated with alkaline phosphatase. The new built construction (p1-SSTF3x) carries three copies in tandem of the expression cassette. The p1-SSTF3x was digested with the enzymes BamHI and BglII, the 8.46 kb band was isolated and religated obtained as described in the previous step. The 16.92 kb sequence was inserted into the pAO815 vector, BamHI digested and dephosphorylated with alkaline phosphatase.

[0063] This vector allows selection of P. pastoris GS115 transformants by complementation of the his4 auxotrophy. The resulting plasmid (p1SSTF6x) contains six copies of the expression cassette arranged in tandem and in the same transcriptional direction, inserted between the AOX1 promoter and a 3 fragment of the AOX1 locus terminator region of (FIG. 2).

[0064] This plasmid was BglII digested for transformation by electroporation of P. pastoris GS115 strain. With this digestion, two fragments, one yielding a 22.23 kb band carrying in the center six expression cassettes in tandem and the gene complementing auxotrophy generated by his4, at the 5end is the AOX1 promoter and in the 3 end a 3 fragment of the terminator region of the AOX1 locus.

[0065] With this strategy the double homologous recombination that replaces the AOX1 locus is favoured. Colonies of the GS115 strain transformed with plasmid p1-SSTF6x were selected on minimal YNB medium supplemented with 2% glucose. In order to evaluate the ability of the transformants to use sucrose as a carbon source, 93 colonies His4+were grown individually in a 100-well YP agar plate (pH 5.5) supplemented with 5% sucrose and a pH indicator, bromothymol blue 0.025%.

[0066] The GS115/p1-SSTF clone, with a single 1-sst gene copy (PF1x) was used as positive control for this experiment. Two clones named PF6Xb PF6Xa turned the medium color from the initial green (pH 5.5) to yellow (pH6.0), due to the 1-SSTrec transfructosylation reaction over sucrose that yielded lactic acid due to consumption by the microorganism of the released glucose from sucrose. This colour change of the medium occurred more quickly in the multicopy strains that in the strain carrying a single gen. This fact indicates that multicopy clones displays greater enzymatic activity than those carrying a single gene copy.

[0067] To corroborate this result, PF6Xb PF6Xa multicopy clones, and the simple copy PF1x, were grown in 10 mL of liquid YP medium supplemented with 2% glycerol in orbital shaker for 24 hours at 28 C. Glucose released due to the FTF activity of the recombinant enzymes in the fractions corresponding to the pellet (biomass or cells) and the culture supernatant of both, multicopy clones and the single copy clone as well as was determined by the glucose-Trinder (Sigma) kit based on colorimetric reaction of the oxidase/peroxidase/glucose chromogenic enzyme complex in the same way as explained in Example 1.

[0068] The two multicopy clones showed higher enzyme activity than that displayed by the simple copy (Table 2), demonstrating that increased copies of the 1-sstf gene integrated into the P. pastoris genome increased 1-SSTrec activity in these two recombinant strains. The clone PF6Xb showed a 31.4% of 1-SSTrec activity in the culture supernatant greater than the PF6Xa clone and 64.2% greater than the single copy done. In the biomass, clone PF6Xb had a 18% of enzyme activity greater than PF6Xa clone and 36% greater than the single copy clone.

[0069] Table 2 shows the effect of the 1-sstf gene copy number on enzymatic activity of the multicopy P. pastoris clones. As controls the strain GS115, and the single copy clone were used. Different letters to the right of the data indicate significant differences determined by a simple Classification ANOVA using the statistical package StatGraph3. The mean values of enzyme activity (n=3) were compared using the Tukey HSD test (p<0.01).

TABLE-US-00002 TABLE 2 Comparison of 1-SSTrec Activity in single-and multicopy clones Specific enzimatyc activity (10.sup.3 UAE/D.O.sub.600) Strain Biomass (B) Culture supernatant (S) GS115 0.1081 0.0015.sup.d 0.1523 0.0017.sup.d PF1X 2.5813 0.0012.sup.c 4.7830 0.0015.sup.c PF6Xa 4.9717 0.0020.sup.b 9.1545 0.0021.sup.b PF6Xb 7.3548 0.0014.sup.a 13.3512 0.0016.sup.a

[0070] Units of enzyme activity/optical density measured at a wavelength of 600 nm (UAE/D.O.sub.600). Different letters denote significant differences between the enzyme activities compared to each other by using the Tukey HSD test (p<0.01).

[0071] According to the results obtained above, the multicopy clone PF6Xb showed the highest 1-SSTrec activity in both, the biomass and the culture supernatant so that, it was chosen for further expression experiments. From now, this selected clone was re-named as PF6X.

Example 4. Increased 1-SST Activity by Retransformation of the PF6X Multicopy Clone by Insertion of Six Additional Copies of the Expression Cassette in the AOX1 Locus

[0072] To increase the enzymatic activity of clone 1-SST PF6X a new plasmid was built called pALS223. To obtain this new construction, the 16.92 kb sequence containing six copies of the expression cassette in tandem and in the same transcriptional direction used to construct the plasmid p1-SSTF6x, was inserted into the vector pPICHaC AOX1-linker previously digested with BamHI and dephosphorylated with alkaline phosphatase.

[0073] This vector allows single homologous recombination in the P. pastoris AOX1 promoter and further transformants selection with the antibiotic hygromycin. After checking this genetic construct by restriction analysis and DNA sequencing, we proceeded to linearize this new plasmid with the enzyme Hpa I and retransform PF6X clone with this new construct through PF6X cells electroporation. This enzyme cuts in a specific site of the AOX1 promoter, which promotes integration into the yeast genome by simple homologous recombination at the AOX1 locus.

[0074] Transformants with more than 6 copies of the expression cassette inserted in the host yeast genome were selected in solid YP medium supplemented with hygromycin 2% glycerol 0.2 g/L. Enzymatic activity in the biomass and in the culture supernatant to more than 60 hygromycin (HigR) resistant colonies was determined using the colorimetric reaction of the enzyme complex glucose oxidase/peroxidase/chromogen reagent kit glucose-Trinder. In this assay also were included the PF6x and PF1x strains as controls.

[0075] One hygromycin-resistant clone called CIGB 308, showed the highest 1-SSTrec enzyme activity in both, the cell pellet and the culture supernatant. This new clone showed higher enzyme activity in the supernatant (1.87 times) and biomass (1.76 times) than PF6X strain.

[0076] When comparing with the single copy strain, 1-SST activity displayed by clone CIGB 308 was 3.58 times higher in the supernatant, and increased 2.41-fold in the biomass, thus confirming that increasing the copy number of the 1-sstf gene stably integrated in the host, also increases 1-SST activity in the yeast host. Southern blot analysis revealed that in clone CIGB 308 were stably integrated 9 copies of the expression cassette. These results indicate that in the event of retransformation Inserted only 3 copies of the expression cassette instead of 6 as expected.

Example 5. The P. pastoris GIGB 308 Strain has More 1-SST Enzymatic Activity and Display More Productivity than its Predecessors PF1X and PF6X at Fermenters Scale

[0077] The P. pastoris CIGB 308 clone and their precursors, with one and six genomic integrated copies of the expression cassette (PF1X, PF6X) respectively, were grown in 7.5 L fermenters with 5 L working volume at 28 C., pH 5.5, 500-900 rpm, aeration 1.2 vvm, and controlled dissolved oxygen above 20%. Regardless of the different integrated copy number of the expression cassette, the three recombinant strains, showed a similar growth pattern.

[0078] After 19 and 20 hours of growth, the initial glycerol content depleted, while the dissolved oxygen pressure was controlled up to 20% by gradually increasing the agitation from 500 to 900 rpm. With the glycerol depletion, there was a rapid dissolved oxygen rise and then started the culture feeding with 50% glycerol (v/v), during the 72 hours of the fermentation process.

[0079] Under these culture conditions, the overall biomass obtained from the three compared clones was 3588 g/L wet weight, so it can be inferred that the gene dosage, as well as the production and accumulation of the recombinant enzyme, did not affect growth and was no toxic to the yeast host.

[0080] Just after 70-72 hours of culture GIGB 308 clone showed the highest extra- and intracellular 1-SST activity, reaching a maximum of 29.70.2 U/mL of culture and 12.40, 2 U/mL of culture (34 U/g wet weight), respectively. From the overall 42.10.2 U/mL detected after the 308 CIGB growth, 70.6% of the FTF activity was found in the culture supernatant and 29.4% in the cells. From the results obtained in this comparative study we decided to choose the CIGB 308 clone for the mass production of 1-SSTrec.

[0081] At the time of this invention there were no reports in the literature describing the fermentation strategy to obtain a plant 1-SSTrec constitutively expressed from a multicopy P. pastoris clone.

Example 6. Incremented Culture Strategy for the Pichia pastoris CIGB 308 Strain Growth Using Sucrose as a Carbon Source

[0082] The fermentation cost of P. pastoris CIGB 308 strain is reduced using a cheaper carbon source other than glycerol, such as sucrose or glucose. P. pastoris GS115 strain, which is used as host has no invertase activity, so it can not use sucrose as a carbon source. However, due to the new FTF activity acquired by the yeast host glucose is released as consequence of the 1-SSTrec transfructosylation reaction from sucrose and then it is metabolized directly for the growth of the recombinant yeast host. This behaviour of the recombinant yeast strain allow the reduction of the fermentation costs during the production process.

[0083] Sucrose fermentation in a batch increased culture was performed in a 75 L fermenter capacity with 50 L working volume. The adjusted parameters were: Temperature: 28 C., pH 5.5 Agitation: 600 rpm. Aeration: 1.0 vvm. Operating pressure: 0.2 atm. The carbon source used was sucrose at 50 g/L, contained either in refined sugar, raw sugar or honey.

[0084] With the carbon source depletion (detected by increased pH or increasing the pressure of oxygen), at approximately 20 hours after fermentation starting, the increment solution was added (solution of the same carbon source Initially used, but 500 g/L) at a rate of 8 mL/L/h increment by initial culture volume. Then, fermentation parameters were readjusted as follow: Agitation: 800 rpm, aeration 1.5 vvm, oxygen pressure: 0.4 atm. The fermentation was performed during 72 h.

[0085] With these culture conditions the reached biomass yields were similar to those achieved with this same done but grown in glycerol medium. On the other hand, total enzyme activity (within 72 hours of culture) was by far superior to the sucrose-containing media regardless of the used raw material-containing sucrose. Growth results of the P. pastoris 308 CIGB strain, using different carbon sources, are summarized in Table 3.

TABLE-US-00003 TABLE 3 Summary of the results obtained after P. pastoris CIGB 308 strain growth in fermenters using glycerol or sucrose as carbon source. Extracellular Activity (U/mL Intracellular Wet Carbon of cell free activity (U/ml Total U/ XWeight (g/ source supernatant) of culture) mL culture L) Glycerol 29.7 0.2 12.4 0.3 42.1 0.4 361 4 (70.6%) (29.4%) Refined 101.6 9.5 38.9 5.4 102.4 11.3 375 9 sugar (62.02%) (37.98%) Raw sugar 53.9 3.3 39.9 2.9 74.2 3.1 363 4 (46.3%) (53.7%) Honey B 110.7 0.2 49.6 5.1 119.6 6.3 368 14 (58.5%) (41.5%)

[0086] Values in parentheses represent the percentage of intracellular and extracellular enzymatic activity of the P. pastoris CIGB 308 strain after 72 hours of culture. An enzyme unit (U) represents the amount of 1-SSTrec which liberates 1 micromol of glucose per minute at initial velocities of the reaction in a sucrose solution in 1.75 M sodium acetate buffer 0.1M pH 5, 5, to 30 C. The data represent the mean of the fermentations conducted with each of the carbon sources standard deviation.

[0087] According to these results it was concluded that both sucrose and honey are suitable substrates to undertake industrial production of this recombinant FTF.

Example 7. 1-SSTrec Enzyme Production in Continuous Culture

[0088] There are no reports in the literature describing the continuous production of recombinant FTFs expressed at high levels in P. pastoris. Continuous production of 1-SSTrec was performed in a 7.5 L INFORS HT fermenter with total working volume of 5 L. The following parameters were established and recorded throughout the culture, through Iris V 5.0 Software: The temperature was maintained at 28 C., while the pH value of 5.5 was controlled by automatic addition of NH.sub.3OH (28% (v/v)) and H.sub.3PO.sub.4 (40% (v/v)). The dissolved oxygen was maintained throughout the culture above 20% by automatically varying the agitation (between 500 and 900 rpm), air flow (1-2 vvm) and the pressure (0-0.7 atm).

[0089] The initial volume was 3 L fermentation medium containing 22 g/L NH.sub.4SO.sub.4, 18.2 g/L of K.sub.2HPO.sub.4, 7.5 g/L of MgSO.sub.4 7H.sub.2O, 0.5 g/L of CaCl.sub.2 2H.sub.2O; yeast extract 5 g/L, trace salts and vitamins in sufficient amounts plus sucrose 50 g/L. For discontinuous increase stage used 1.5 L of a solution of sucrose 500 g/L. In the stage of continuous culture was used a medium containing 200 g/L sucrose, yeast extract 2.5 g/L, 11 g/L NH.sub.4SO.sub.4, 9.1 g/L of K.sub.2HPO.sub.4, 3.75 g/L of MgSO.sub.4, 7H.sub.2O, 0.25 g/L of CaCl.sub.2 2H.sub.2O; salts trace and vitamins.

[0090] The fermenter was Inoculated with 200 mL of inoculums previously grown in a shaker. Once exhausted the carbon source, the discontinuous increment stage started and the culture was fed at a flow rate ranging between 7 to 30 mL/L-h. With the increment depletion, the continuous culture started by feeding the bioreactor with a 1 day-dilution speed (D). After reaching the steady state, the culture operating was kept for 45 days, with an average activity yield of 705 U/mL and a cell concentration of 35211 g/L wet weight.

Example 8. Determination of the Optimal Reaction Parameters of 1-SSTrec for the Synthesis of 1-Kestose

[0091] The enzyme preparation obtained in the fermenter supernatant was subjected to a filtration process using a Sarticon Slice 200 (Sartorius) filter with a Hydrosart membrane (0.2 m), following the manufacturer's instructions. Subsequently, the filtrated was concentrated 10 times by diafiltration, using the same equipment but with a Hydrosart ultrafiltration membrane (10 kDa) against sodium acetate buffer 0.1M to give a final preparation of 1000 U/mL. Optionally, the filtrate was subjected to a lyophilization process to obtain a solid enzyme preparation with an activity greater than 8500 U/g.

[0092] a) Determination of the Optimal pH for 1-SSTrec Activity:

[0093] The 1-SSTrec activity was examined in a pH range between 4 and 8. The reaction was performed for 1 hour at 30 C. in a 0.87 M sucrose solution and 10 U of enzyme in a final volume of 0.5 mL. For the pH range of 4.0 to 5.5 a sodium acetate buffer 0.1M was used, and for pH between 6.0-8.0 0.1 M phosphate buffer The maximum values of 1-SSTrec enzymatic activity was found at pH values between 5.5 and 6.0.

[0094] b) Temperature and Optimum Substrate Concentration for FOS Synthesis:

[0095] For the determination of these parameters, 60 U of 1-SSTrec reacted in buffer 0.1 M sodium acetate, pH 5.5 with substrate concentrations ranging between 200 and 600 g/L, at 30, 40 and 50 C. respectively, in a final reaction volume of 10 mL at 250 rpm. After 1 hour of reaction, sugars composition in the reaction mixture was determined by HPLC. For this chromatography 20 l of the sample were applied in a Aminex HPX 42-C (BioRad, Richmond) column, with a work flow of 0.6 mL/min, a pressure of about 52 bar and a working temperature of 81 C. The mobile phase used was water and a refractive index detectorKnauer Differential Refractometer was employed. Sugars were quantified using the Biocrom software package, version 3.0, IGBC, 1996-1997.

[0096] Table 4 shows the composition and quantification (%) of sugars determined for different reaction conditions, the rate of 1-kestose synthesis and the relationship between transfructosylation and hydrolytic activity. The maximum rate of 1-kestose synthesis with no hydrolytic activity was reached at 40C and a sucrose concentration of 600 g/L.

[0097] Similar results were obtained when intact cells with 1-SST periplasmic activity, or immobilized cells in calcium alginate or the immobilized enzyme covalently joint to Eupergit (Sigma) were used as enzymatic preparation for sucrose conversion to 1-kestose.

TABLE-US-00004 TABLE 4 Influence of temperature and substrate concentration in the FOS synthesis Temperature G F GF GF.sub.2 GF.sub.3 r.sub.(GF2) R.sub.T/H 200 g/L 30 C. 16.0 0.5 39.6 45.8 0.8 1.0 88 40 C. 13.1 0.9 41.2 43.7 1.1 1.0 50 50 C. 1.5 0.0 91.2 6.7 0.0 0.1 400 g/L 30 C. 11.8 0.1 58.0 29.9 0.3 1.3 585 40 C. 11.0 0.4 48.6 39.3 0.8 1.0 111 50 C. 2.4 0.0 86.3 11.1 0.0 0.5 600 g/L 30 C. 6.9 0.0 67.5 25.5 0.0 1.8 40 C. 9.9 0.0 54.7 34.7 0.8 2.4 50 C. 7.3 0.0 64.8 27.9 0.0 1.9

[0098] Composition of the reaction mixture after one hour of reaction (G), glucose (F), fructose (GF) sucrose, (GF.sub.2) 1-kestose (GF.sub.3) nystose. Reaction parameters (r(GF.sub.2)) Speed synthesis of 1-kestose given in g/min (RT/H) Ratio transfructosylation and hydrolytic activity given by the ratio of 1-kestose and fructose composition.

[0099] c) Half-Life of Free and Immobilized 1-SSTrec:

[0100] Thermal stability was evaluated for free and Eupergit immobilized enzyme and for P. pastoris CIGB 308 cells immobilized in calcium alginate. Both, free or immobilized forms were incubated in 0.1 M acetate buffer, pH 5.5, at 30, 35 and 40 C. Samples were taken from each reaction with 24 hour intervals for 30 C., 1 hour for 35 C. and 20 minutes for 40 C., respectively, to test the residual activity. Subsequently, the half life time was defined as the time at which each of the assayed enzyme preparations had lost 50% of its initial activity. The results in Table 5 show that enzyme preparations containing free 1-SSTrec are much more stable than cells with 1-SST activity immobilized in calcium alginate

[0101] Moreover, unexpectedly, the average life time of the crude extract in solution of 1-SSTrec at 30 C., under non-reactive conditions, is 1432 hours. This time exceeds more than 100 times the half-life times reported to date for plants enzymes 1-SST type.

TABLE-US-00005 TABLE 5 Half-life time of different 1-SSTrec preparations under non-reactive conditions Half life time (hours) Enzymatic preparation 30 C. 35 C. 40 C. Free 1-SSTrec (crude Extract) 1432.0 6.1 0.7 P. pastoris CIGB 308 cells 36.0 4.2 0.3 immobilized in calcium alginate Eupergit immobilized 1-SSTrec 1856.0 12.9 1.6

[0102] A thermal stability test was performed to the lyophilized enzyme preparation. The result was that the solid preparation has a half-life time greater than three years at 30 C.

Example 9. Sucrose Transformation to FOS Catalyzed by the 1-SSTrec in a Batch Reaction Using a Stirred Tank Reactor

[0103] The time course of FOS synthesis catalyzed by 1-SSTrec was conducted at sucrose concentration of 600 g/L, adding an enzymesubstrate weight ratio of 15 U/g in buffer 0.1M sodium acetate, pH 5.5; at 40 C., for 6 hours in a 1 L reactor at 250 rpm. Quantification and composition of the produced sugars was determined in samples picked every 20 minutes by HPLC similarly to Example 8.

[0104] Maximum production of 1-kestose was 320.8 g/L, 53.4% of the total carbohydrates in the mixture and 90.4% of total FOS. FIG. 3 shows that the maximum production of 1-kestose was reached between 2.7 and 3 hours of reaction, when over 70% of the initial sucrose was consumed. At this point matches the maximum concentration of 1-kestose with the onset of the appearance of the tetrasaccharide nystose.

[0105] This fact is advantageous for the use of 1-SSTrec in the large scale production of 1-kestose since not synthesized nystose in the transformation reaction of sucrose appears until 50% of initial sucrose is consumed as seen in FIG. 4. By contrast, the fungal FTF accumulates nystose from the start of the reaction with the synthesis of the first 1-kestose molecules and also synthesize the pentasaccharide fructosylnystose.

[0106] Similar values and behaviour of the course of sucrose transformation into 1-kestose were obtained using cells with periplasmic 1-SSTrec activity, free or immobilized in calcium alginate or 1-SSTrec covalently immobilized to Eupergit (Sigma) under similar reaction conditions. Table 6 shows the FOS concentration obtained by different enzyme preparations.

TABLE-US-00006 TABLE 6 Concentration of synthesized FOS after 3 hours of reaction using different 1-SSTrec enzyme preparations. Cells immobilized in 1-SSTrec Synthesized calcium immobilized on FOS Free 1-SSTrec alginate Eupergit total FOS 354.7 g/L 330.5 g/L 342.2 g/L 1-Kestose 320.8 g/L (90.5%) 322.1 g/L 318.3 g/L (97.4%) (93.0%) Nistose 33.9 g/L (9.5%) 8.4 g/L 23.9 g/L (2.6%) (7.0%)

[0107] The transfructosylation reaction mixture obtained after 3 hours of synthesis, was subjected to a pasteurization process so, the enzyme was inactivated. Subsequently, the syrup was subjected to a polishing process that began with a filtration, followed by demineralization, decolourization, concluding with a chromatographic separation for simulated moving bed (SMB), that after elution yielded a rich FOS stream with more than 90% of 1-kestose. This fact demonstrates the technical feasibility of this procedure to produce a 1-kestose rich syrup, the FOS with the highest prebiotic effect and so, becoming in the most commercial one.

[0108] Technical feasibility of this procedure at industrial scale, was confirmed through the scaled up of the transformation reaction from sucrose to 1-kestose in 30 and 100 L capacity reactors, respectively. The scaled up of this operation was performed by the method based on the Principle of Similarity, from the Information obtained in the tests performed in the 1 L model reactor. The concentration of 1-kestose obtained in the two new scales averaged 3227 g/L. This result shows no significant difference with those obtained in the model reactor. This fact also demonstrated that the conversion reaction from sucrose to 1-kestose catalyzed by 1-SSTrec in stirred tank reactors is reproducible at higher scales.

Example 10. 1-Kestose Synthesis from Sucrose in a Membrane Bioreactor Operated Semicontinuously

[0109] The procedure described in Example 9, for 1-kestose synthesis from sucrose by employing free 1-SSTrec in a stirred tank bioreactor was followed. For this purpose 1 L (working volume) stirred tank bioreactor was coupled in its output with a cartridge type ultrafiltration membrane (Prep/Scale TFF-1 30 kDa, Millipore, with nominal filter area of 0.09 m.sup.2), allowing separation of the reaction products and the enzyme as shown in FIG. 5.

[0110] The parameters used were: 9000 U/L, initial enzyme concentration, initial sucrose concentration 600 g/L in 0.1 M acetate buffer, pH 5.5, temperature 40 C., 250 rpm stirring speed, flow bioreactor output feeding the membrane was 40 mL/min.

[0111] During the synthesis step, both, the retentate and permeate flow at the outlet of the membrane return to the enzymatic bioreactor. Every 30 minutes samples of the permeate stream were analyzed by HPLC in order to determine the conversion ratio of sucrose into 1-kestose as described in Examples 8 and 9.

[0112] After 3 hours of reaction the permeate recirculation valve to the bioreactor was closed, and the valve to the FOS collector tank previously kept closed, was opened. The valve corresponding to the retentate stream was regulated to achieve 30 mL/min of permeate flow.

[0113] After 30 minutes, the 90% of the total reaction volume was discharged and so, the permeate outlet valve to the collection tank was closed. Then the returned back valve to the bioreactor was opened and the retained valve was regulated, so that the returning permeate flow to the bioreactor was established to 5 mL/min. At this point the bioreactor was charged with 900 mL of sucrose 600 g/L, in 0.1 M acetate buffer, pH 5.5. Also a 180 U of 1-SSTrec were added, to keep the same reaction time and the same conversion ratio of 1-kestose, thus beginning the second synthesis cycle.

[0114] After 2 hours and 30 minutes of reaction the discharge of the reaction products proceeded as performed in the first cycle. After complete discharge of the 90% of the reaction volume corresponding to the second cycle, the bioreactor was loaded again in the same way as was done in the second cycle. These reaction-discharge steps are repeated sequentially up to complete 10 operation cycles and then the BRM cleaning step is carried out. FIG. 6 shows the behaviour during the first 3 cycles, and FIG. 7 shows the product profile at the end of each sequential cycle.

[0115] The use of a BRM sequentially operated has similar productivity to that of a stirred tank bioreactor with the same capacity, but consuming 8 times less enzyme by amount of produced 1-kestose.

[0116] Similar 1-kestose concentrations are obtained for other BRM operating conditions semi-continuously operated. Among operation conditions that could be varied without affecting the product profile are the ratio enzyme-substrate (2-40 U/g of sucrose), sucrose concentration 400-800 g/L, temperature (30-50 C.), pH (5.0-6.5), assuming always that the download time is between 10 and 20% of the time in which the maximum production of 1-kestose is reached.

[0117] Download times over this range favour nystose synthesis and fructose production from 1-kestose. Prior to this invention there existed no reports in the literature to describe the production of 1-kestose in a BRM.