D-Allulose Production from D-Fructose by Permeabilized Recombinant Cells of Corynebacterium glutamicum Cells Expressing D-Allulose 3-Epimerase Flavonifractor plautii

A d-allulose 3-epimerase from Flavonifractor plautii was cloned and expressed in Escherichia coli and Corynebacterium glutamicum. The maximum activity of the enzyme purified from recombinant E. coli cells was observed at pH 7.0, 65°C, and 1 mM Co2+ with a half-life of 40 min at 65°C, Km of 162 mM, and kcat of 25280 1/s. For increased d-allulose production, recombinant C. glutamicum cells were permeabilized via combined treatments with 20 mg/L penicillin and 10% (v/v) toluene. Under optimized conditions, 10 g/L permeabilized cells produced 235 g/L d-allulose from 750 g/L d-fructose after 40 min, with a conversion rate of 31% (w/w) and volumetric productivity of 353 g/L/h, which were 1.4- and 2.1-fold higher than those obtained for nonpermeabilized cells, respectively.


Introduction
The International Society of Rare Sugars (ISRS) stated at the 2014 Rare Sugar Symposium that 'D-psicose' should be referred to as 'D-allulose' and use of the 'D-psicose' term should be discontinued because D-allulose is an isomerized product of D-allose. D-Allulose (D-psicose, D-ribo-2hexulose) is a rare sugar that is present in small amounts as a non-fermentable component of commercial carbohydrates [1] and as a free sugar in agricultural products [2]. This sugar has attracted a great deal of attention in the field of functional foods owing to its health benefits. D-Allulose is used as a low-calorie sweetener and as a functional sugar for diabetes because it does not contribute calories and exhibits hypoglycemic, hypolipidemic, and antioxidant activities [3][4][5][6].
D-Allulose has been produced from D-fructose by the reactions of biocatalysts, including Dtagatose 3-epimerases (DTEases) from Pseudomonas cichorii [7] and Rhodobacter sphaeroides [8]; D-allulose 3-epimerases (DAEases) from Agrobacterium tumefaciens [9], Clostridium sp. encoding the putative DAEase was amplified by PCR using F. plautii genomic DNA as a template. The primer sequences used for gene cloning were based on the DNA sequence of the putative DAEase from F. plautii (GenBank accession number EHM40452.1). Forward (5 0 -CA TATGAACCCGATTGGAATGCACTAC-3 0 ) and reverse primers (5 0 -CTCGAGTTACGCG GTCAGCTCCTTGAGG-3 0 ) were designed to introduce the underlined NdeI and XhoI restriction sites, respectively, and were synthesized by Bioneer (Daejon, Korea). The DNA fragment amplified by PCR was purified using a PCR purification kit (Promega, Madison, WI, USA) and inserted into the pET15b vector digested with the same restriction enzymes. The resulting plasmid was transformed into E. coli strain ER2566 using an electroporator (MicroPulser, Bio-Rad, Hercules, CA, USA). The transformed E. coli was plated on Luria-Bertani (LB) agar containing 25 μg/mL ampicillin. An ampicillin-resistant colony was selected, and the plasmid DNA from the transformant was isolated with a plasmid purification kit (Promega). DNA sequencing was carried out at the Macrogen facility (Seoul, Korea). Gene expression was estimated by both SDS-PAGE and enzyme activity assay.
C. glutamicum ATCC 13032 (ATCC, Manassas, USA), and E. coli−C. glutamicum shuttle expression vector pEKEx2 (Juelich Research Centre, Juelich, Germany) were used as the sources of host cells and expression vector, respectively. The DAEase gene from F. plautii was ligated into the expression vector pEKEx2. A ribosomal binding site (rbs) was encoded upstream of the DAEase gene, which was amplified by PCR using the template vector pET15b from E. coli. The primer sequences used for gene cloning were based on the DNA sequence of F. plautii DAEase. Forward (5 0 -CTGCAGAAAGGAGATATAGATGAACCCGATTGGAAT GCACTACGGC-3 0 ) and reverse primers (5 0 -GTCGACTTACGCGGTCAGCTCCTTGAG-3 0 ) were designed to introduce the underlined PstI and SalI restriction sites, respectively, and were synthesized by Bioneer. The amplified DNA fragment obtained by PCR was purified and inserted into the pEKEx2 vector digested with the same restriction enzymes. The resulting plasmid was transformed into C. glutamicum strain ATCC 13032 using an electroporator. The transformed C. glutamicum cells were plated on brain-heart infusion (BHI) agar containing 15 μg/mL kanamycin. A kanamycin-resistant colony was selected, and the plasmid DNA from the transformant was isolated with a plasmid purification kit (Promega). DNA sequencing was carried out at the Macrogen facility.

Culture conditions
Recombinant E. coli containing the DAEase/pET15b gene from F. plautii was cultivated in a 2 L flask containing 500 mL of LB medium and 25 μg/mL ampicillin at 37°C with shaking at 200 rpm. When the optical density of the bacterial culture at 600 nm reached 0.6, 0.1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) as a final concentration was added to the culture to induce DAEase expression, and the culture was then incubated with shaking at 150 rpm at 16°C for 16 h to express the enzyme.
Recombinant C. glutamicum containing the DAEase/pEKEx2 gene from F. plautii was cultivated in a 2 L flask containing 500 mL of BHI medium and 15 μg/mL kanamycin at 30°C with shaking at 200 rpm. When the optical density of bacteria culture at 600 nm reached 0.6, 1.0 mM IPTG and 20 mg/L penicillin as final concentrations were added to the culture to induce DAEase expression and to injure the peptidoglycan layers of C. glutamicum cell wall, respectively. For enzyme expression, the culture was further incubated with shaking at 200 rpm at 30°C for 20 h.

Enzyme preparation
Recombinant E. coli expressing F. plautii DAEase was harvested from the culture broth by centrifugation at 8,000 × g for 30 min at 4°C, and then washed twice with 0.85% NaCl. Washed recombinant cells were resuspended in 50 mM phosphate buffer (pH 7.0) containing 300 mM NaCl and 1 mg/mL lysozyme. Resuspended cells were disrupted using a sonicator on ice for 2 min. Unbroken cells and cell debris were removed by centrifugation at 13000g for 20 min at 4°C, and the supernatant was filtered through a 0.45 μm pore size filter. All purification steps were carried out at 4°C with a Profinia protein purification system (Bio-Rad). The filtrate was loaded onto an immobilized metal ion affinity chromatography (IMAC) cartridge (Bio-Rad, Hercules, CA, USA), which was previously equilibrated with 50 mM phosphate buffer (pH 8.0). The bound protein was eluted with a linear gradient between 10 to 500 mM imidazole at a flow rate of 1 mL/min. The eluent was collected and loaded immediately onto a Bio-Gel P-6 desalting cartridge (Bio-Rad), previously equilibrated with 50 mM piperazine-N,N 0 -bis(2-ethanesulfonic acid) buffer (PIPES) (pH 7.0). The loaded protein was eluted with 50 mM PIPES buffer (pH 7.0) at a flow rate of 1 mL/min. The active fractions were collected and dialyzed against 50 mM PIPES buffer (pH 7.0) for 16 h. The solution resulting from dialysis was used as the purified enzyme.

Determination of molecular mass
The subunit molecular mass of F. plautii DAEase was determined by SDS-PAGE under denaturing conditions, using a ladder of pre-stained proteins (MBI fermentas, Hanover, MD, USA) as references. All protein bands were stained with Coomassie Blue for visualization. The molecular mass of the native enzyme was investigated by gel-filtration chromatography using a Sephacryl S-300 preparative-grade column HR 16/60 (GE Healthcare, Piscataway, NJ, USA). The purified enzyme was loaded onto the column and eluted with 50 mM Tris-HCl (pH 7.5) buffer containing 150 mM NaCl at a flow rate of 1mL/min. The column was calibrated with ferritin (440 kDa), catalase (206 kDa), aldolase (158 kDa), and conalbumin (75 kDa) as reference proteins (GE Healthcare). The molecular mass of the native enzyme was calculated by comparing with the migration length with that of the reference proteins.

Determination of specific activity and kinetic parameters
One unit (U) of enzyme activity for D-fructose was defined as the amount of enzyme required to liberate 1 μmol of D-allulose per min at pH 7.0 and 65°C. The specific activity (U/mg) was defined as the amount of monosaccharide produced per enzyme amount per unit reaction time. Various concentrations of D-fructose and D-allulose (5−80 mM) were used to decide the kinetic parameters for the enzyme. K m (mM) and k cat (1/s) were determined by the Lineweaver-Burk plot derived from the Michaelis-Menten equation. The catalytic constant, k cat , was calculated by dividing the subunit molecular mass by the DAEase concentration.

Preparation of permeabilized cells
To prepare permeabilized C. glutamicum cells, cells were harvested from the culture broth by centrifugation at 13,000 × g for 20 min at 4°C and then washed twice with 0.85% sodium chloride solution. The washed cells were resuspended in solutions containing 2 and 5 mg/L penicillin, and 2, 5, 50, and 100 mg/L ethambutol, ethionamide, and isoniazid as antibiotics; 0.2% and 0.5% (w/v) CTAB and Triton X-100, and 2% and 5% (w/v) Span 20, Span 80, Tween 20, Tween 40, and Tween 80 as detergents; and 10% and 20% (v/v) acetone, DMSO, ethanol, hexane, methanol, 1-propanol, iso-propanol, and toluene as solvents. The cell suspension was incubated at 4°C for 15 min and washed twice with 50 mM PIPES buffer (pH 7.0). The treated cells were used as permeabilized cells for the production of D-allulose from D-fructose. Unless otherwise stated, the enzyme and cell reactions were performed at 65°C for 10 min in 50 mM PIPES buffer (pH 7.0) containing 50 mM D-fructose in the presence of 1 mM Co 2+ with 0.5 U/mL enzyme and 7.5 g/L cells, respectively, as standard conditions.

Effects of temperature, pH, and metal ions on DAEase activity
The effects of temperature and pH on the activity of DAEase, were investigated by varying the temperature from 30 to 70°C at a constant pH of 7.0 and by varying the pH from 6.0 to 8.5 at a constant temperature of 65°C under the standard conditions. To evaluate the effect of metal ions on enzyme activity, the enzyme assay was conducted after treatment with 1 mM ethylenediaminetetraacetic aicd (EDTA) at 4°C for 1 h or after the individual additions of 1 mM MnSO 4 , MnCl 2 , NiSO 4 , CoCl 2 , MgSO 4 , and FeCl 3 .
Effects of temperature, pH, and metal ions on the activity of recombinant cells The effects of temperature and pH on the activity of permeabilized cells were examined by varying the temperature from 50 to 75°C in 50 mM PIPES buffer at a constant pH of 7.5 and by varying the pH from 6.0 to 8.5 at a constant temperature of 65°C, respectively. The effect of metal ions on the production of D-allulose from D-fructose was evaluated by nonpermeabilized and permeabilized C. glutamicum cells expressing DAEase from F. plautii.

Thermal inactivation for the enzyme in permeabilized C. glutamicum cells and the purified enzyme
The influence of temperature on the stability of the enzyme in permeabilized cells was investigated as a function of the incubation time by placing the cell solution at five different temperatures (45, 50, 55, 60, and 65°C) in 50 mM PIPES buffer (pH 7.5). The influence of temperature on enzyme stability was a function of incubation time by placing the enzyme solution at five different temperatures (45, 50, 55, 60, and 65°C) in 50 mM PIPES buffer (pH 7.0). Samples were withdrawn at specific time intervals and the activities were measured under standard conditions. The half-life of the enzyme was calculated using Sigma Plot 9.0 software (Systat Software, San Jose, CA).

Optimization of cell and substrate concentrations, and D-allulose production under the optimized conditions
The optimal concentrations of cells and substrate for increased production of D-allulose from D-fructose by whole permeabilized cells were determined by varying the cell concentration from 1 to 15 g/L at a constant D-fructose concentration of 750 g/L, and by varying the substrate concentration from 50 to 750 g/L at a constant cell concentration of 10 g/L. The production of D-allulose from D-fructose by whole permeabilized cells or nonpermeabilized cells was performed in 50 mM PIPES buffer (pH 7.5) containing 10 g/L cells, 750 g/L D-fructose, and 1 mM Co 2+ at 65°C for 1 h.

Analytical methods
Cell mass was determined using a linear calibration curve of optical density at 600 nm verse dry cell weight. The OD 600 values of culture broth were measured and then converted to the dry cell weight (1 OD 600 = 0.42 g/L dry cell weight). The concentrations of monosaccharides, including D-fructose, D-allulose, D-sorbose, D-tagatose, D-xylulose, and D-ribulose, were determined using a Bio-LC system (Dionex ICS-3000, Sunnyvale, CA, USA) with an electrochemical detector and a CarboPac PA10 column. The column was eluted at 30°C with 200 mM sodium hydroxide at a flow rate of 1 mL/min.

Results and Discussion
Gene cloning, purification, and molecular mass determination of DAEase from F. plautii The gene (885 bp) encoding F. plautii DAEase (GenBank Accession No. EHM40452.1) was cloned and expressed in E. coli. The expressed enzyme was purified as a soluble protein from a crude cell extract by His-trap affinity chromatography. The specific activity of the purified enzyme for D-allulose was 20 U/mg. The expressed protein analyzed by SDS-PAGE showed a single band with a molecular mass of approximately 33 kDa, consistent with the calculated value of 32,987 Da based on the 295 amino acids plus the hexa-histidine tag. The expression level of the protein in C. glutamicum was lower than that of E. coli (Lanes 1 and 4 of Fig 1A). However, the enzyme was expressed as a soluble protein with a similar level between two host cells in crude extracts (Lanes 2 and 5 of Fig 1A). The native enzyme existed as a tetramer with a molecular mass of 131 kDa as determined by gel filtration chromatography ( Fig 1B and S1  Fig). The total molecular masses and association forms of DAEases and DTEases were 125-132 kDa and tetramers, respectively.

Biochemical properties of DAEase from F. plautii
The biochemical properties of F. plautii DAEase are compared with other reported DTEases and DAEases in Table 1. The maximum enzyme activity of the conversion of D-fructose to Dallulose by F. plautii DAEase was observed at 65°C (S3A Fig) and pH 7.0 (S3B Fig). The activities of DAEase from Dorea sp. [15] and DTEase from R. sphaeroides [8], were maximal at pH 6.0 and pH 9.0, respectively. The pH values of other DAEases and DTEases were maximal in the range of 7.0 to 8.0. The lowest and highest temperatures for maximal activity were 40°C for DTEase from R. sphaeroides [8] and 70°C for DAEases from Dorea sp. [15] and T. primitia.  highest activity in presence of Co 2+ . The half-life of DAEase from C. cellulolyticum at 60°C was 408 min [12], which was the highest reported thermal stability among the DAEases and DTEases that was 3.1-fold higher than that of DAEase from F. plautii. DAEase from F. plautii produced 239 g/L D-allulose from 750 g/L D-fructose at pH 7.0 and 65°C after 60 min, with a conversion rate of 32% and a productivity of 239 g/L/h (S4 Fig). This is the highest production of D-allulose ever reported. The conversion rates of D-fructose to D-allulose by DAEases were higher than those by DTEases.
Permeabilization of recombinant C. glutamicum cells expressing DAEase from F. plautii for increased production of D-allulose To increase D-allulose production, recombinant C. glutamicum cells expressing DAEase from F. plautii were permeabilized by treatment with antibiotics, including ethambutol,  , and was 2.5-, 1.7-, and 1.8-fold higher than that by non-treated cells, respectively. Penicillin was used to permeabilize C. glutamicum during glutamic acid fermentation [31] and toluene was an efficient permeabilizer to G. suboxydans for the production of L-sorbose [35].
To determine the synergistic effect in D-allulose production, cells were treated with 20 mg/L penicillin, 3% (w/v) Span 20, and 20% (v/v) toluene in combination. Permeabilized cells treated with 20 mg/L penicillin and 20% (v/v) toluene showed the highest activity among cells permeabilized by the combined treatment, and exhibited 3-fold higher D-allulose production than that by non-treated cells (Fig 5). The increases in D-allulose production by permeabilized cells treated with efficient permeabilizers such as penicillin, Span 20, and toluene were not critical above 2 mg/L, 1% (w/v), and 5% (v/v), respectively, which were all lower concentrations than those used for the maximal production of D-allulose, respectively. Cells were treated with a combination of the permeabilizers using these concentrations (S5 Fig). Permeabilized cells treated with 2 mg/L penicillin and 5% (v/v) toluene also showed the highest activity among cells permeabilized with the combined treatment and lower D-allulose production than that by permeabilized cells treated with 20 mg/L penicillin and 20% (v/v) toluene. Thus, permeabilized cells treated with 20 mg/L penicillin and 20% (v/v) toluene were used for D-allulose production.
Effects of temperature, pH, and metal ions on the production of Dallulose from D-fructose by permeabilized recombinant C. glutamicum cells The maximum activity of permeabilized C. glutamicum cells expressing DAEase from F. plautii for D-allulose production was observed at 65°C (S6A Fig) and pH 7.5 (S6B Fig). The reactions of whole recombinant E. coli cells expressing DAEase from C. bolteae [11], C. cellulolyticum [12], and A. tumerfaciens [20] were performed at pH 6.5 and 55°C, pH 8.0 and 55°C, and pH 8.5 and 60°C, respectively. Mn 2+ and Co 2+ strongly enhanced D-fructose epimerization by DAEases and DTEases. Thus, the effect of metal ions such as Mn 2+ and Co 2+ on the activity of D-allulose production was evaluated in nonpermeabilized and permeabilized C. glutamicum   cells expressing DAEase from F. plautii (Fig 6). The activities of both recombinant nonpermeabilized and permeabilized C. glutamicum cells in the presence of Co 2+ were higher than those in the presence of Mn 2+ . D-Allulose production by permeabilized cells in the presence of Co 2+ was 25-fold higher than that by nonpermeabilized cells. The significantly increased production of D-allulose may be due to the improved contact of metal ions and the enzyme caused by cell permeabilization.

Thermal inactivation for DAEase from F. plautii in permeabilized recombinant C. glutamicum cells and the purified DAEase
The thermal stabilities of DAEase from F. plautii in recombinant permeabilized C. glutamicum cells and the purified enzyme were examined by measuring the activities after incubation at temperatures ranging from 45 to 65°C. Thermal inactivation of the enzyme in cells and the purified enzyme followed first-order kinetics. The half-lives of the enzyme in cells at 45, 50, 55, 60, and 65°C were 9000, 2340, 1098, 211, and 48 min, respectively (Fig 7A), which were 1.8-, 1.0-, 2.3-, 1.4-, and 1.1-fold higher than those of the double-site variant DAEase from A. tumefaciens [20] in recombinant E. coli cells, respectively. The half-lives of the purified enzyme at 45, 50, 55, 60, and 65°C were 5760, 2010, 762, 130, and 40 min (Fig 7B), respectively, which were 1.5-, 1.1-, 1.4-, 1.6-, and 1.2-fold lower than those of the enzyme in recombinant C. glutamicum cells, respectively. Thus, DAEase in cells is more thermally stable than the purified DAEase. Effects of cell and substrate concentrations on D-allulose production by permeabilized recombinant C. glutamicum cells The optimal concentration of recombinant permeabilized cells for D-allulose production was determined by varying the concentration of permeabilized cells from 0.5 to 15 g/L with 750 g/L D-fructose as a substrate for 10 min (Fig 8A). Below 10 g/L cells, D-allulose production increased with increasing cell concentration up to 10 g/L cells. However, above 10 g/L cells, the production reached a plateau. Therefore, the cell concentration for the maximal production of D-allulose was 10 g/L. The production of D-allulose from D-fructose was investigated by varying the substrate concentration from 50 to 750 g/L with 10 g/L cells (Fig 8B). As the D-fructose concentration increased, D-allulose production increased, but the conversion rate decreased. To obtain the highest concentration of D-allulose, the concentration of D-fructose as a substrate was determined to be 750 g/L.
Production of D-allulose from D-fructose by nonpermeabilized and permeabilized recombinant C. glutamicum cells under the optimized conditions The optimal reaction conditions for the production of D-allulose from D-fructose by permeabilized recombinant C. glutamicum cells expressing DAEase from F. plautii were pH 7.5, 65°C, 1  mM Co 2+ , 10 g/L cells, and 750 g/L D-fructose. Under the optimized conditions, the timecourse reactions for D-allulose production were performed using nonpermeabilized and permeabilized cells (Fig 9). The initial production rate within 10 min of D-allulose by permeabilized cells was significantly higher than that by nonpermeabilized cells. Nonpermeabilized cells produced 166 g/L D-allulose after 60 min, with a conversion rate of 22% (w/w), a specific productivity of 16.6 g/g/h, and a volumetric productivity of 166 g/L/h, whereas permeabilized cells produced 235 g/L D-allulose after 40 min, with a conversion rate of 31% (w/w), a specific productivity of 35.3 g/g/h, and a volumetric productivity of 353 g/L/h, which were 1.4-, 2.1-and 2.1-fold higher than those of nonpermeabilized cells, respectively, and the volumetric productivity of permeabilized cells was 1.5-fold higher than that of purified enzyme. Thus, cell permeabilization was an effective method for increasing D-allulose production.
The permeabilized cells of wild-type R. sphaeroides treated with 0.1% (w/v) CTAB produced 6.5 g/L D-allulose from 50 g/L D-fructose after 8 h, a specific productivity of 0.03 g/g/h, and a volumetric productivity of 0.81 g/L/h [18]. The permeabilized cells of wild-type Sinorhizobium sp. treated with 10% (v/v) toluene produced 37 g/L D-allulose from 700 g/L D-fructose after 15 h, a specific productivity of 0.04 g/g/h, and a volumetric productivity of 2.5 g/L/h [19]. Wildtype cells were not suitable for D-allulose production because of the markedly lower specific productivity than that of recombinant cells. The production of D-allulose from D-fructose by whole recombinant E. coli, Bacillus subtilis, or C. glutamicum cells expressing DAEase is presented in Table 3. Whole recombinant E. coli cells expressing the double-site variant (I33L-S213C) DAEase from A. tumefaciens produced 230 g/L D-allulose from 700 g/L D-fructose after 40 min, with a conversion rate of 33% (w/w), a specific productivity of 86.2 g/g/h, and a volumetric productivity of 345 g/L/h [20], which were previously the highest reported production values. The concentration, volumetric productivity, and thermal stability of recombinant permeabilized C. glutamicum cells expressing DAEase from F. plautii in D-allulose production were higher than those of recombinant E. coli cells expressing the double-site variant DAEase from A. tumefaciens, whereas the conversion rate and specific productivity were lower. DAEase from Clostridium scindens was expressed in B. subtilis as a food-grade host and the food-grade recombinant B. subtilis was used for the production of D-allulose [36]. The product concentration, conversion rate, and volumetric productivity of the recombinant C. glutamicum cells in the resent study were higher than those of the recombinant B. subtilis cells, but the specific productivity was lower.
In conclusion, the putative DAEase from F. plautii was cloned and expressed in E. coli. The biochemical properties of the expressed enzyme for the epimerization of D-fructose to D-allulose were characterized. Recombinant C. glutamicum cells were permeabilized to increase the production of D-allulose. The reaction conditions, including pH, temperature, metal ions, and the concentrations of cells and substrate, were optimized for the permeabilized cells. Under the optimized conditions, the production of D-allulose from D-fructose by permeabilized C. glutamicum cells expressing DAEase from F. plautii treated with penicillin and toluene was significantly higher than that by nonpermeabilized cells. Whole permeabilized recombinant C. glutamicum cells produced D-allulose with the highest concentration and volumetric productivity reported to date.
(empty circle), and 65°C (filled circle) for various incubation times. (B) Thermal stability of enzyme activity. The enzymes were incubated at 45°C (filled triangle), 50°C (empty square), 55°C (filled square), 60°C (empty circle), and 65°C (filled circle) for various incubation times. A sample was withdrawn at each time point and the relative activity was measured. Data represent the means of three separate experiments and error bars represent the standard deviation.