Article pubs.acs.org/JAFC
Biocatalytic Synthesis of the Rare Sugar Kojibiose: Process Scale-Up and Application Testing Koen Beerens,† Karel De Winter,† Davy Van de Walle,‡ Charlotte Grootaert,§ Senem Kamiloglu,§ Lisa Miclotte,∥ Tom Van de Wiele,∥ John Van Camp,§ Koen Dewettinck,‡ and Tom Desmet*,† †
Centre for Synthetic Biology, Department of Biochemical and Microbial Technology, ‡Laboratory of Food Technology and Engineering, Department of Food Safety and Food Quality, §Department of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, Ghent B-9000, Belgium ∥ Center for Microbial Ecology and Technology (CMET), Ghent University, Coupure Links 653, Ghent B-9000, Belgium S Supporting Information *
ABSTRACT: Cost-efficient (bio)chemical production processes are essential to evaluate the commercial and industrial applications of promising carbohydrates and also are essential to ensure economically viable production processes. Here, the synthesis of the naturally occurring disaccharide kojibiose (2-O-α-D-glucopyranosyl-D-glucopyranoside) was evaluated using different Bifidobacterium adolescentis sucrose phosphorylase variants. Variant L341I_Q345S was found to efficiently synthesize kojibiose while remaining fully active after 1 week of incubation at 55 °C. Process optimization allowed kojibiose production at the kilogram scale, and simple but efficient downstream processing, using a yeast treatment and crystallization, resulted in more than 3 kg of highly pure crystalline kojibiose (99.8%). These amounts allowed a deeper characterization of its potential in food applications. It was found to have possible beneficial health effects, including delayed glucose release and potential to trigger SCFA production. Finally, we compared the bulk functionality of highly pure kojibiose to that of sucrose, hereby mapping its potential as a new sweetener in confectionery products. KEYWORDS: sucrose phosphorylase, kojibiose, bulk functionality, Caco-2 cells, gut microbiota, fermentable sugars
■
selectively stimulate beneficial gut populations. Indeed, the α1,2 bond is largely resistant to the action of enzymes in the digestive tract but can be cleaved by specific micro-organisms such as lactobacilli.12−14 Moreover, kojibiose has been reported to significantly stimulate the growth of bifidobacteria, a common target for prebiotic development.15 Besides its potential prebiotic properties, kojibiose is not metabolized by common oral bacteria and has therefore attracted attention as a low-calorie sweetener for the prevention of tooth decay.14 Unfortunately, in-depth studies on the health-promoting properties of kojibiose are hampered by its high price and limited availability. Although kojibiose is present in honey,10 the amounts are far too low for practical isolation. Also, the chemical synthesis of kojibiose suffers from low yields and the generation of toxic waste.16 Alternatively, isolation after partial acetolysis of dextran produced by Leuconostoc mesenteroides has been described. However, the latter procedure involves multiple steps requiring chemical reagents among which are acetic anhydride, sulfuric acid, chloroform, and methanol, while the kojibiose yield is limited to 17%.17 Therefore, the enzymatic synthesis of kojibiose has been vastly explored. Examples include the use of α-glucosidases18 and α-glucoamylases,19 glucansucrase,20 kojibiose phosphorylase,21 and a reasonably efficient and ecofriendly procedure based on the dextransu-
INTRODUCTION In light of the growing obesity pandemic, carbohydrates have obtained somewhat of a negative reputation. However, carbohydrates can offer a huge diversity in both structure and function. Palatinose, for example, is a simple isomer of sucrose but is digested much more slowly and does not promote tooth decay.1 In turn, galacto (GOS)- and fructo-oligosaccharides (FOS) are well established as prebiotic ingredients,2,3 while Dtagatose and D-psicose have attracted considerable attention as low-caloric sweeteners.4−6 Unfortunately, only a small number of saccharides are found in nature in sufficient amounts to allow their commercial exploitation. Hence, the International Society of Rare Sugars (ISRS) has classified those scarce in nature as “rare sugars” (monosaccharides and their derivatives that rarely exist in nature7). Rare sugars can be monosaccharides, for which biocatalytic production routes have been summarized before,7,8 or derivatives such as di- and oligosaccharides containing (rare) monosaccharides. Also when linking abundant monosaccharides to form di- and oligosaccharides, some glycosidic linkages are far more abundant than others (e.g., α1,4 in maltose and starch vs α-1,1 or α-1,2 of trehalose9 and kojibiose,10,11 respectively). Nonetheless, this structural diversity at both levels is expected to have effects on the saccharides’ properties and thus their applications. Despite the low availability of certain disaccharides, some studies have indicated that they can be very promising compounds, as is the case for kojibiose (2-O-α-D-glucopyranosyl-D-glucopyranoside). For example, preliminary studies indicated that kojibiose and its derived oligosaccharides could © 2017 American Chemical Society
Received: Revised: Accepted: Published: 6030
May 14, 2017 June 29, 2017 June 30, 2017 June 30, 2017 DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry
industry to produce high-fructose corn syrup starting from glucose. However, due to the reversible nature of the reaction, the enzyme can also be applied to isomerize the released fructose to glucose, which is then used as the substrate for kojibiose formation, hereby improving atom efficiency and reducing waste formation. Prior to its application, the immobilized Sweetzyme IT Extra was allowed to equilibrate for 30 min at room temperature in 20 volumes Milli-Q water and subsequently washed five times with 20 volumes Milli-Q water at room temperature. Purification of Kojibiose. Prior to the yeast treatment, immobilized glucose isomerase was removed by filtration over a Whatman paper, and SP was inactivated by heating the solutions to 90 °C for 10 min. The purification of various reaction mixtures was then performed by adding 30 g/L spray-dried baker’s yeast (Algist Bruggeman) in 50 mL Falcon Tubes with perforated lids. The reactions were incubated at 30 °C and 50 rpm on a thermoshaker (Eppendorf). Unless stated otherwise, no buffers were added, resulting in a gradual pH drop from 7.0 to roughly 5.0 at the end of the yeast treatment. Alternatively, the pH was kept constant at 7.0 or 5.0 throughout the fermentation by performing the yeast treatments in a 50 mM MOPS or sodium acetate buffer. Upon depletion of all contaminating carbohydrates, the yeast was removed by centrifugation (5 000g, 4 °C, 15 min) using a Sorvall RC-6+ centrifuge (Thermo Scientific). Finally, the solution was filtered over a vacuum filtration system with a pore size of 0.22 μm (Corning) to remove any suspended solids. A flow diagram of the downstream process is given in Scheme S.1. Quantification of Reaction Products. The concentration of carbohydrates was followed by high-performance anion-exchange chromatography (HPAEC) (Dionex ICS-3000, Thermo Scientific), using a CarboPac PA20 pH-stable column and pulsed amperometric detection (PAD). Separation of sucrose, glucose, fructose, trehalose, kojibiose, nigerose, maltose, and isomaltose was achieved using a 30 min protocol. After 13 min of isocratic elution with 30 mM NaOH, the concentration was gradually increased to 100 mM over 5 min, kept constant for 3 min, and decreased again to 30 mM within 1 min, followed by an equilibration period of 8 min. The temperature and flow rate were set at 30 °C and 0.5 mL/min, respectively. Alternatively, the purity of the crystallized kojibiose was analyzed by HPLC using an Aminex HPX-87H column (Bio-Rad) equilibrated at 30 °C. The eluent consisted of 5 mM H2SO4 in Milli-Q at a flow rate of 0.6 mL/ min. Adequate detection was achieved with a refractive index detector. The different carbohydrates were quantified using commercial standards. These measurements were used to calculate the following parameters (eqs 1−6), whereby sucrose0 represents the initial sucrose concentration.
crase-catalyzed synthesis of 4-O-β-D-galactopyranosyl kojibiose, resulting in a moderate yield (38%, in weight with respect to the initial amount of lactose used).22 Unfortunately, these processes also suffer from low yields, the generation of side products, or complicated downstream processing strategies. Recently, a number of new sucrose phosphorylase (SP) variants have become available, allowing larger-scale production of the rare disaccharides, kojibiose,11 and nigerose.23 These new enzyme variants are a necessary and important step toward commercial use of these disaccharides, since they opened the door to large-scale production and thus also to the evaluation of the effect of disacccharides’ structural diversity. However, other hurdles still need to be overcome. Although several improved mutants were isolated, and the applicability of the obtained variants for the production of kojibiose was illustrated, further increasing the efficiency of the latter process would significantly increase its industrial applicability and economic viability. Indeed, despite a yield of 74% toward sucrose, less than half of the combined total of sucrose and glucose was converted to kojibiose (mass based),11 thus leaving space for process intensification and improvement of scale-up. Alternatively, the described nigerose production process requires the use of the organic solvent DMSO (dimethyl sulfoxide),23 which should be avoided for food applications. In this work, we describe the optimization and upscaling of a highly efficient process for the synthesis and isolation of kojibiose using recently described sucrose phosphorylase mutants.11 In addition, for the study of the uptake behavior of kojibiose by intesinal epithelial cells, the Caco-2 cell culture model was used, and markers for membrane integrity, mitochondrial respiration, protein content, and oxidative stress were investigated. Furthermore, the fermentation potency of kojibiose by gut microbiota from human origin was assessed by short-term incubation of kojibiose and subsequent monitoring of short-chain fatty acid (SCFA) production. Finally, we compared the bulk functionality of highly pure kojibiose to that of sucrose, hereby mapping its potential as an alternative sweetener in confectionery products.
■
MATERIALS AND METHODS
Chemicals and Enzymes. Sucrose (food grade) was bought from Cargill, Sweetzyme IT Extra was kindly provided by Novozymes, and all other chemicals were analytical grade and purchased from SigmaAldrich, unless otherwise stated. The recombinant SP from Bifidobacterium adolescentis, and the L341I, Q345S, L341IQ345S, L341I_Q345N, and L341I_Y344A_Q345N mutants were routinely grown as previously described.24 The obtained pellets were lysed in 2 mM MOPS buffer at pH 7.0, supplemented with 1 mg/mL lysozyme. After dissolution of the pellets, the mixtures were incubated for 30 min at room temperature, followed by 2 × 2 min sonication (Branson 250 Sonifier, level 3, 50% duty cycle). Next, the enzyme solution was partially purified by 15 min incubation at 58 °C, followed by centrifugation to remove debris (18 000g, 4 °C, 30 min). Protein concentrations were measured with the BCA Protein Assay kit (Pierce) using bovine serum albumin as standard. All assays were performed in triplicate and had a CV of less than 10%. Biocatalytic Synthesis of Kojibiose. The synthesis of kojibiose was performed by dissolving varying amounts of glucose and sucrose in Milli-Q water. Unless stated otherwise, 2 mg/mL of the heatpurified Bifidobacterium adolescentis SP (BaSP) variants was used, and the reaction mixtures were incubated at 55 °C. The pH of this bufferfree system was set at 7.0 and remained constant throughout all conversions. If required, prewashed glucose isomerase (Sweetzyme IT Extra, Novozymes, ≥350 U/g) was added up to 20 g/L after 2 h. Glucose isomerase is mostly known for its application in the food
selectivity (%) =
[kojibiose] *100 [kojibiose] + [maltose] + [isomaltose] (1)
⎛ [sucrose] ⎞ conversion (%) = ⎜1 − ⎟*100 [sucrose] ⎝ 0⎠
(2)
[kojibiose] *100 [sucrose]0
(3)
yield (%) =
hydrolysis (%) [fructose] − [kojibiose] − [maltose] − [isomaltose] = *100 [fructose] (4) g
[kojibiose (M)]*342.3 mol ⎛ g kojibiose ⎞ ⎟= productivity⎜ ⎝ L fermentation*h ⎠ [fermentation (L)]*time (h) (5)
atom efficiency (%) = 6031
mass of atomsdesired product mass of atomsreactants
*100
(6)
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry
Caco-2 Cell Culture. Caco-2 cells (ATCC HTB37) were cultivated in Dulbecco’s Modified Eagle’s Medium (DMEM) with Glutamax (Gibco, Life Technologies, Merelbeke, Belgium), supplemented with 10% fetal bovine serum (FBS, Greiner Bio-One, Wemmel, Belgium), 1% nonessential amino acids, and 1% penicillin/streptomycin (Life Technologies) and were incubated at 37 °C and 10% CO2 in air. For the experiments, cells were trypsinized at 80% confluency and seeded in a 96-well plate at a density of 25 000 cells/well in DMEM. After 2 days, the cells were washed with phosphate saline buffer containing calcium and magnesium (PBSD+) and 200 μL glucose-free Seahorse base XF-medium (Copenhagen, Denmark) with no extra sugar (untreated: U) or with 4.5 g/L cellobiose (C), kojibiose (K), or glucose (G) was added to the cells. After a 1-day incubation, samples were taken for membrane permeability (trypan blue), mitochondrial respiration (MTT), protein content (SRB), reactive oxygen species (ROS) production, and nitric oxide (NO) production. Cell Membrane Permeability Test. Cell-culture medium was removed from the wells and replaced by 200 μL of a 1:1 solution of trypan blue in PBSD+. After 15 min of incubation, cells were washed with PBSD+, and the incorporation of the blue stain into membranedamaged cells as well as cell morphology was assessed microscopically using a Motic AE31 phase-contrast microscope (VWR, Leuven, Belgium). MTT Assay. In the MTT assay, cellular reduction of yellow 3-(4,5dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide (MTT) to blue formazan was used as a measurement for mitochondrial activity. Cells were shielded from light and incubated with 20% MTT solution (5 mg/mL in PBS) in medium for 2 h at 37 °C and 10% CO2. Then, the medium was carefully removed, and the formazan crystals were dissolved in 200 μL of DMSO. The absorbance was measured at 570 nm with an absorbance multireader (Benchmark Plus, Biorad, Temse, Belgium). SRB Assay. The SRB assay was used to determine the cell density by measuring the protein content of the wells. First, the cells were fixated with trichloroacetic acid (50%w/w in Milli-Q water) for 1 h at 4 °C, after which extracellular proteins were gently washed away using tap water. The protein stain sulphurodamine B (50 μL, 0.4% in 1% glacial acetic acid) was subsequently added, and the cells were incubated for 30 min. Next, cells were rinsed with glacial acetic acid (1% in Milli-Q water) to remove unbound stain, and the cells were suspended in 200 μL of Tris buffer (10 mM). The absorbance was measured at 490 nm with a multireader. NO Assay. Nitric oxide production was indirectly assessed by the measurement of nitrite in the cell-culture medium. To this end, 100 μL of cell-culture medium and 100 μL of Griess reagent (Sigma) were added in a 96-well plate, and the absorbance at 540 nm was measured after 30 min. ROS Assay. Intracellular ROS was measured using H2-DCFDA (2,7dichlorodihydrofluorescein diacetate) (Sigma). Intracellular deacetylases convert this nonfluorogenic compound to DCFH, which upon oxidation by ROS is converted to the highly fluorescent 2′,7′dichlorofluorescein (DCF). Cells were washed with PBSD+ and incubated with 20 μM H2-DCFDA in serum-free medium for 30 min. Next, the cells were washed again, and 200 μL PBSD+ was added to each well. Fluorescence was immediately measured on a Spectramax Fluorescent Plate Reader (λex/em = 485/535 nm) (Molecular Devices, CA, USA). Kojibiose Fermentation by Human Gut Microbiota. The fermentability of kojibiose by human gut microbiota was assessed by short-term batch incubations. Fecal microorganisms were isolated from the fecal samples of four healthy individuals, aged 25−31 years. All individuals gave consent for use of stool samples for incubation purposes, and permission for use of human fecal microbiota for incubation purposes was granted by the ethical committee of Ghent University hospital under Belgian registration number BE 670201214538. Each individual stool sample (20 g) was separately brought in 100 mL of a 100 mM phosphate buffer at pH 7 containing 1 g Nathioglycolate/L to maintain anaerobiosis. Individual fecal slurries were separately homogenized in a stomacher for 5 min upon which the fecal
Crystallization of Kojibiose. Prior to the cooling crystallization, the kojibiose solutions were evaporated in vacuo to varying concentrations by means of a rotavapor (Büchi R-200). The temperature and pressure were set at 50 °C and 50 mbar, respectively, while the kojibiose concentration, expressed as Brix (°Bx), was determined by a series of ATAGO hand refractometers. Next, the samples were cooled overnight from 50 °C to room temperature, while shaking on a rotary shaker at 20 rpm. The samples were subsequently cooled to 4 °C over a period of 4 h, and the crystal growth was allowed for another 24 h at 4 °C and 20 rpm. Next, the kojibiose crystals were isolated by filtration over a Whatman paper, washed with ethanol, and dried to the air. Properties of Kojibiose Crystals. The purity of the obtained kojibiose crystals was determined using HPAEC-PAD and HPLC-RI as described above, while the microscopic appearance was evaluated using an Olympus CH30 microscope at 100× and 400× total magnification. The crystallization yield (eq 7), and the whiteness of the crystals, expressed by the ICUMSA score (eq 8), were calculated as follows: crystallization yield (%) =
mass of kojibiose crystals *100 mass of kojibiosein solution before crystallization
(7) ICUMSA 420 Score =
(absorbance420nmof carbohydrate solution*1000) c*b
(8) with c being the carbohydrate concentration (g/mL) and b being the path length (cm). Large-Scale Production of Kojibiose. Large-scale production of kojibiose was performed by dissolving 6 kg sucrose and 350 g glucose in a total volume of 10 L in Milli-Q water. The mixture was supplemented with 1 mg/mL heat-purified B. adolescentis L341I_Q345S SP and incubated at 55 °C and 35 rpm in a 30 L Biostat C reactor (B. Braun Biotech). To save on enzyme cost, we chose to use a lower BaSP concentration, resulting in an extended production time. However, reduction of the process time can easily be achieved by increasing the enzyme’s concentration. After 2 h, 200 g of prewashed Sweetzyme IT Extra (Novozymes) was added, and the formation of kojibiose was continued for 4 days. Next, the immobilized glucose isomerase was conveniently removed by filtration over a fritted column with a hot water jacket at 55 °C. The obtained solution was then heated to 90 °C for 10 min, cooled to 30 °C, and supplemented with 300 g baker’s yeast. Once all of the contaminating carbohydrates were removed, the yeast was separated by vacuum filtration over a Seitz K series depth filter. The residue was then evaporated in vacuo at 50 °C to a Brix of 48, cooled overnight to room temperature, and subsequently incubated for 24 h at 4 °C and 20 rpm. The obtained crystals were washed with 2 L of ethanol over a Seitz K Series Depth Filter, and dried to the air. Upper Gastro-Intestinal Tract Digestion. Simulation of digestion in the upper gastro-intestinal tract lumen was performed according to the internationally accepted protocol published by Minekus and coworkers.25 Briefly, to simulate mouth conditions, a mixture (pH 7) of 5 g of homogenized sample, 3.5 mL of salivary fluid, 0.5 mL of αamylase solution (1500 U/mL), 25 μL of CaCl2 (30 g/L) solution, and 975 μL of distilled water were incubated for 2 min. Next, 6 mL of gastric fluid, 1.28 mL of pepsin (25 000 U/mL) solution, 4 μL of CaCl2 (30 g/L) solution, 0.16 mL of HCl (1 M) solution, and 0.556 mL of distilled water were added to simulate gastric conditions (pH 3), and the mixture was incubated for 2 h. Finally, to simulate small intestinal conditions, the pH was raised to 7, and 7.7 mL of intestinal fluid, 3.5 mL of pancreatin (800 U/mL), 1.75 mL of bile (85.6 mg/ mL), 28 μL of CaCl2 (30 g/L), 0.150 mL of NaOH (1M) solution, and 0.917 mL of distilled water were added, and the sample was incubated for another 2 h. Next, in addition to this protocol described by Minekus, an extra 1.5 h incubation with rat intestinal α-glycosidase (Sigma, 0.27 mg/mL digest) was performed in order to simulate brush border enzymatic breakdown according to Hodoniczky and coworkers.14 All incubations were performed at 37 °C. 6032
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry Table 1. Performance of Various SP Variants for the Synthesis of Kojibiosea
a
enzyme variant
selectivity (%)
conversion (%)
yield (%)
hydrolysis (%)
productivity (g/L/h)b
L341I L341I_Q345S L341I_Q345N L341I_Y344A_Q345N
79.7 97.1 98.3 100.0
95.7 67.8 38.7 28.9
43.5 51.1 23.4 7.4
34.8 30.7 36.0 19.5
28.8 27.1 8.1 1.5
The averages of all values calculated in Table S.1 are reported. bProductivity (g kojibiose/L fermentation/h).
Figure 1. Yeast treatment of reaction mixtures obtained by converting 0.25 (A), 0.5 (B), 1 (C), and 1.5 (D) M sucrose and glucose to equilibrium. Baker’s yeast was added at 30 g/L, and the reactions were incubated at 30 °C and pH 7.0. The concentration of glucose (●), fructose (○), sucrose (▼), kojibiose (Δ), and maltose (■) was followed in time. The plotted points represent the average of three measurements, and the coefficient of variation (cv) was below 10% for all points. SCFA Analysis. SCFA analysis was performed using liquid−liquid extraction with diethyl ether and analyzed by gas chromatography equipped with a flame ionization detector (GC-FID). A detailed procedure can be consulted in Van Herreweghen et al. (2017).27 Bulk Functionality of Highly Pure Kojibiose. The density of kojibiose and sucrose was measured using pycnometry. Hereto, distilled water was used to determine the volume of the pycnometer (roughly 50 cm3). Next, the pycnometer was filled with hexane (Sigma-Aldrich) to measure the density of this solvent. In the final step, 10 g of sweetener was added to the pycnometer that was further filled with hexane. Assuming that the sweetener is completely insoluble in hexane, its density can be calculated. Prior to weighing, the pycnometer holding distilled water, hexane, and samples was incubated for 30 min at a temperature of 25 °C. The moisture content of kojibiose and sucrose was determined according to the Karl−Fischer titration method using the 719 Titrino apparatus (Metrohm),
microbiota were separated from food remnants and solid particles with a 500g centrifugation for 5 min. Pellets were discarded and supernatants were served as inoculum to the batch incubations. To enable assessing the effect of the supplemented carbohydrate, fecal microbiota were inoculated at 10% in sugar-depleted SHIME feed without (control) or with (treatment) kojibiose at 1% w/v in amber serum bottles. Medium composition can be consulted in Van den Abbeele et al.26 Upon introduction of the fecal slurry, serum bottles were closed with butyl rubber stoppers and aluminum caps, and subsequently, the headspace was flushed with N2 gas for 30 min to ensure anaerobic conditions. Incubation was performed for 24 h at 37 °C. Sample (5 mL) was taken at 0, 6, and 24 h for short-chain fatty acid analysis (SCFA). The biological reproducibility of this incubation experiment was tested by incubating the fecal microbiota from four different human individuals. 6033
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry Hydranal titrant 5 (Riedel de Haen, 34801), and Hydranal solvent (Riedel de Haen, 34800). The thermal behavior of kojibiose and sucrose was studied using the Q1000 differential scanning calorimeter (DSC, TA Instruments) equipped with a refrigerated cooling system. Nitrogen was used as purge gas. The cell constant (1.0074) was set with indium (TA Instruments) having a melting enthalpy of 28.57 J/g. In addition to indium (Tm = 156.6 °C), azobenzene (Sigma-Aldrich), and undecane (Acros Organics), which have a melting temperature of 68.5 °C and −26.0 °C, respectively, were used for temperature calibration. Samples (3.0 to 6.0 mg) were hermetically sealed in aluminum cups, and an empty pan was used as a reference. The applied time−temperature protocol was as follows: equilibration at 20.0 °C and holding for 10 min followed by heating at a rate of 5.0 °C/min to 200.0 °C. Melting-peak integration was performed manually using a linear baseline in the Universal Analysis 2000 software version 4.7A (TA Instruments). Peaks were characterized by maximum temperature (Tm, max) and melting enthalpy (ΔHm). After running the integral, onset (Tm, onset) and offset (Tm, offset) temperature (°C) were defined as the temperature at which 1% and 99% of the melting enthalpy is covered. All of the previously mentioned measurements with regard to the bulk functionality of kojibiose and sucrose were done in triplicate.
use of high sucrose and glucose concentrations might hamper the efficient isolation of kojibiose. Indeed, Saccharomyces cerevisiae is known to metabolize all contaminating carbohydrates (sucrose, glucose, fructose, maltose),29 while kojibiose is not consumed by the yeast.22 However, these conversions are typically performed at moderate carbohydrate concentrations, not exceeding 200 g/L total carbohydrates.22,30 Therefore, the reaction mixtures obtained by converting 0.25, 0.5, 1, and 1.5 M sucrose and glucose to kojibiose, were subjected to treatment with 30 g/L baker’s yeast at pH 7.0 and 30 °C (Figure 1). Interestingly, the addition of 30 g/L S. cerevisiae allowed the elimination of all contaminating carbohydrates after 4 and 6 h, when starting from 250 and 500 mM sucrose and glucose, respectively (Figure 1A,B). In contrast to the reaction starting from 500 mM substrate (Figure 1B), baker’s yeast was found to slowly metabolize kojibiose after 6 h when starting from 250 mM sucrose and glucose (Figure 1A). Unfortunately, higher substrate concentrations severely hampered the yeast’s performance. As a result, roughly 70 and 90% of the undesired contaminants remained present after 14 h when using substrate concentrations of 1 and 1.5 M, respectively (Figure 1C,D). Extended incubation of these reaction mixtures (up to 7 days) failed to remove all of the undesired carbohydrates. Similar results were obtained when incubating the baker’s yeast at pH 5.0, indicating the limited effect of the acidity on the performance of S. cerevisiae for the latter conversions (Figure S.1). Although reacting 500 mM sucrose and glucose with the L341I_Q345S variant easily outperforms the most efficient method for the synthesis of kojibiose to date,22 higher kojibiose concentrations and a more economic use of the substrates would be highly desirable. Indeed, when all sucrose is converted, roughly 375 mM kojibiose is formed, while also approximately 500 mM fructose and 200 mM glucose are present (Figure 1). Reducing the amount of the latter side products would allow higher kojibiose titers during the yeast treatment. Therefore, the use of glucose isomerase (GI) was evaluated to convert the stoichiometrically formed fructose to glucose, which can then be used as an acceptor (Figure 2). The latter adaptation was found to allow the use of higher sucrose and lower glucose concentrations, while producing more kojibiose (Table 2). Combining the use of glucose isomerase with high sucrose and low glucose concentrations was found to substantially increase the atom efficiency (Table 2). The latter parameter,
■
RESULTS AND DISCUSSION Selection of SP Variant for the Production of Kojibiose. Although recent research resulted in several sucrose phosphorylase (SP) mutants with enhanced properties toward the synthesis of kojibiose,11 these variants have not been evaluated under production conditions to date. To that end, the most promising variants from our previous study (L341I and L341I_Q345S)11 and two additional new promising variants (L341I_Q345N and L341I_Y344A_Q345N) were studied in detail. The two additional variants L341I_Q345N and L341I_Y344A_Q345N harbor slightly higher selectivities (98.3 and 100.0%, respectively) but reduced specific activities (both around 0.14 U/mg). The substrate concentration was varied between 0.1 and 1 M, while a heat-purified enzyme was added to concentrations between 1 and 4 mg/mL (0.25−1.4 U/mL). From these experiments, a variety of parameters were obtained (Table S.1). The average values for all parameters were then calculated to allow a straightforward comparison of the different variants (Table 1). Triple mutant L341I_Y344A_Q345N displayed the highest selectivity (100%), while hydrolysis was limited to only 19.5%. Unfortunately, this excellent selectivity comes at the expense of a severely reduced productivity, barely reaching 1.5 g of kojibiose per liter of fermentation medium per h. As a result, the average conversion remained below 30%, while a kojibiose yield of only 7.4% was observed. In contrast, single mutant L341I reached nearly 96% conversion, while exhibiting the highest productivity. However, kojibiose accounted for only 80% of the formed glucobioses, while hydrolysis amounted to roughly 35%. Interestingly, both double mutants were found to exhibit intermediate properties. In particular, the L341I_Q345S mutant combines a high selectivity with an enhanced productivity, resulting in the best kojibiose yield of all variants. In addition, the hydrolysis of the latter variant was found to be substantially lower compared to that of the L341I_Q345N mutant. Therefore, mutant L341I_Q345S was identified to be the most promising variant for the production of kojibiose. Optimization of Substrate Concentrations and Downstream Processing: An Integrated Approach. Industrial carbohydrate conversions are preferably run at elevated temperatures with high substrate concentrations.28 Although the L341I_Q345S variant proved to be stable at 55 °C,11 the
Figure 2. Synthesis of kojibiose from sucrose and glucose by the Bifidobacterium adolescentis L341I_Q345S SP. Glucose isomerase (GI) was used to convert the formed fructose to the acceptor substrate glucose. 6034
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry
of 786 mM fructose and glucose, while the reaction starting from 1.5 M sucrose and 0.3 M glucose yielded more than 1 M of the latter carbohydrates. In conclusion, the combined glucose and fructose concentration should be kept well below 1 M. Although further increasing the sucrose concentration to 2 M resulted in an impressive kojibiose concentration approaching 1.7 M, such high concentrations were found to result in kojibiose precipitation during the yeast treatment at 30 °C. Crystallization of Kojibiose. As described above, reacting 1.8 M sucrose with 0.2 M glucose resulted in roughly 1.5 M kojibiose. Although all of the contaminating carbohydrates could be successfully removed by yeast treatment, obtaining the desired product in crystalline form often remains a challenge.31 Indeed, S. cerevisiae is known to produce metabolites (mainly glycerol and/or organic acids), which might hamper the crystallization of kojibiose.22 Nevertheless, we were able to obtain kojibiose crystals by performing an evaporative step at 50 °C to a Brix of 60, followed by cooling crystallization at 4 °C. Remarkably, the addition of glycerol up to 300 mM did not inhibit the crystallization of kojibiose under these conditions. Encouraged by these results, the crystallization of kojibiose was studied in detail. To that end, the concentration of kojibiose prior to the cooling crystallization step was altered between 75 and 42° Bx, corresponding to 2.07 and 1.16 M, respectively (Figure S.2). These mixtures were then cooled overnight from 50 °C to room temperature, and subsequently cooled to 4 °C, after which crystal growth was continued for 24 h. Next, the (microscopic) appearance of the crystals (Figure S.3 and S.4), crystallization yield, whiteness, and purity of the obtained kojibiose crystals were evaluated (Table 3). Although the highest crystallization yield was achieved at Brix 75, the obtained purity of the crystals was limited to 96%. Decreasing the Brix resulted in lower crystallization yields but also substantially improved the purity of the obtained crystals. Interestingly, Brix 48 was found to combine a high crystallization yield, with a purity exceeding 99.8%. Also, the crystals obtained at Brix 48 presented the lowest ICUMSA score, reflecting their whiteness. Moreover, the latter crystals were visible to the naked eye (Figure S.3). Indeed, in contrast to cooling crystallization at higher Brixes, the presence of large
Table 2. Performance of Bifidobacterium Adolescentis L341I_Q345S SP and Yeast Treatment under Various Reaction Conditionsa sucrose (mM)
glucose (mM)
250 500 1000 1500 1500 800 1600 1800 2000
250 500 1000 1500 300 200 200 200 200
GI
reaction time (h)
kojibiose (mM)
atom efficiency (%)
yeast reaction (h)b
-e -e -e -e +f +f +f +f +f
24 24 48 69 69 48 69 69 69
201 374 754 1008 1135 596 1332 1507 1664
52.7 49.0 49.4 44.0 68.5 65.8 78.1 79.1 79.0
4 6 -c -c -c 9 11 12 pd
All reactions were performed at 55 °C in the presence of 2 mg/mL heat-purified L341I_Q345S. bTime required to remove >99% of the contaminating carbohydrates. cThe contaminating carbohydrates could not be removed by yeast treatment. dKojibiose precipitated during yeast treatment. eNo glucose isomerase was added. fGlucose isomerase (20 g/L)was added after 2 h. a
which is a measure for the efficient use of substrates, could be increased to 79%; a tremendous improvement compared to the recently reported 19% for the synthesis from lactose and glucose.22 Moreover, kojibiose concentrations exceeding 1.5 M were readily achieved. We were able to remove all contaminating carbohydrates when reacting 1.8 M sucrose with 0.2 M glucose. In contrast, when using 1.5 M sucrose and 0.3 M glucose, we failed to purify kojibiose. The latter phenomenon was studied in detail by adding kojibiose to 1 M final concentration after the enzymatic conversion of 500 mM sucrose and glucose (Figure 3). Remarkably, the presence of 1 M kojibiose did not significantly influence the performance of S. cerevisiae (Figure 3). Nevertheless, previous experiments demonstrated the inhibitory effect of such elevated carbohydrate concentrations (Figure 1). In contrast to high glucose and fructose concentrations, the presence of kojibiose does not appear to significantly influence the yeast’s performance. Indeed, reacting 1.8 M sucrose with 0.2 M glucose resulted in a combined total
Figure 3. Effect of kojibiose on the performance of Saccharomyces cerevisiae. Yeast treatments were performed on reaction mixtures obtained by converting 500 mM sucrose and glucose to equilibrium (A) and on identical reaction mixtures supplemented with kojibiose to 1 M (B). Baker’s yeast was added at 30 g/L, and the reactions were incubated at 30 °C. The concentration of glucose (●), fructose (○), sucrose (▼), kojibiose (Δ), and maltose (■) was followed in time. 6035
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry Table 3. Cooling Crystallization of Kojibiosea Brix (°Bx)
ICUMSA score
purity HPAECPAD (%)
purity HPLCRI (%)
crystallization yield (%)
75 60 55 50 48 46 44 42
400.0 379.0 217.5 196.5 129.8 168.4 164.9 789.5
96.40 96.66 98.08 98.47 99.81 98.08 98.17 96.12
98.41 94.16 99.87 99.90 99.98 99.95 99.86 89.45
85.5 72.5 74.9 74.2 72.3 58.2 51.7 46.2
bulk carbohydrates, while the atom efficiency was increased to an impressive 79%. In Vitro Gastro-Intestinal Digestion of Kojibiose. To compare the digestibility of kojibiose with that of maltose, an in vitro simulation of the upper gastro-intestinal tract was performed according to the standard Minekus protocol.25 Kojibiose and maltose were not digested in mouth and stomach (Table 4). In the small-intestinal simulation, a relatively small hydrolysis of maltose was observed, whereas no hydrolysis was observed for kojibiose. We therefore conclude that kojibiose is resistant to pancreatic enzymes secreted into the small intestinal lumen. Because it was reported before that kojibiose was cleaved by α-glycosidases,33 we have chosen to expand the conventional digestion protocol with an extra step simulating the brush-border activity. It was observed that both kojibiose and maltose were hydrolyzed by α-glycosidase activity, but the hydrolysis degree of maltose was 10-fold higher than that of kojibiose. This is largely in agreement with earlier findings of Hodoniczky and co-workers14 and corresponds well with previous in vitro digestion assessment using α-glycosidases.33 We therefore conclude that kojibiose is not completely resistant to small intestinal enzymes, but it is less susceptible than maltose. It therefore has potential to delay glucose absorption in the gut, and has a higher chance to reach the more densely microbiota-populated parts of the intestine, where it may be subjected to fermentation and beneficial short-chain fatty acid production. Uptake and Consumption of Kojibiose by Caco-2 Cells. Figure 5 shows the reactivity of the cells to conditions without carbohydrates (U) and after a 1 day treatment with cellobiose (C), kojibiose (K), and glucose (G). The MTT test showed that only glucose was used by the cells as a substrate for respiration. This increase in mitochondrial activity was probably not due to the presence of more cells in the wells, as the SRB values indicating the protein content were not significantly different between the treatments. None of the sugars gave any sign of toxicity, and also, the membrane integrity was not affected as was visually assessed by trypan blue staining. No significant differences in intracellular ROS were observed, indicating that none of the sugars induced oxidative stress to the cells. For glucose, a small but significant increase in NO production was observed, thereby confirming the cellular response to this compound as reported for other cell types
a
The performance at different Brixes was evaluated by determining the crystallization yield and measuring the whiteness and purity of all crystals.
needle-shaped kojibiose crystals was confirmed by microscopic analysis up to Brix 50 (Figure S.4).32 Crystallization at Brix 48 resulted in white crystals with a purity exceeding 99.8% (Figure 4). The corresponding crystallization yield of 72% could be conveniently increased to 96% by performing a second crystallization on the supernatant, yielding 99% pure kojibiose. Alternatively, the sucrose concentration was increased from 1.8 to 2 M, resulting in the spontaneous crystallization of kojibiose upon cooling (Table 2). The latter procedure avoids the use of S. cerevisiae, thereby decreasing the purity of the obtained crystals to 97.5%. Despite the presence of minor amounts of glucose and fructose, this procedure allows the straightforward isolation of kojibiose and was identified a valuable alternative for obtaining bulk kojibiose. Proof of Concept: Production at a 10 L Scale. Finally, the scalability and industrial applicability of the developed process was illustrated by performing the synthesis of kojibiose at a 10 L scale. As a result, roughly 3.3 kg of kojibiose with a purity exceeding 99.8% was obtained, while crystallization of the supernatant resulted in approximately 1.2 kg of 99% pure kojibiose (purity was confirmed by HPLC, HPAEC, and NMR spectroscopy, see Figures S.5 and S.6). In contrast to the current production strategies for kojibiose, we thus developed a sustainable, cost-effective, and scalable biocatalytic process for the production of highly pure kojibiose. Moreover, the substrates sucrose and glucose are cheap and readily available
Figure 4. Purity of kojibiose crystals obtained by reacting 1.8 M sucrose with 0.2 M glucose, followed by cooling crystallization at Brix 48. The crystals were analyzed by HPAEC-PAD (A) and HPLC-RI (B). The small peak at 1 min in the HPAEC-PAD profile corresponds to the injection peak. 6036
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry Table 4. Hydrolysis of Kojibiose and Maltose under Simulated Small Intestinal Conditionsa kojibiose kojibiose (mm) undigested mouth stomach small intestine (SI) SI + α-glucosidase
1125 1141 632 304 141
± ± ± ± ±
140 65 18 9 19
glucose (mm) nd
0 0 nd 0 nd 0 nd 6±0
maltose hydrolysis (%) 0 0 0 0 2.15 ± 0.16
maltose (mm) 1515 1562 666 315 91
± ± ± ± ±
25 42 27 5 17
glucose (mm)
hydrolysis (%)
± ± ± ± ±
0b 0b 0b 0.33 ± 0.01 28.48 ± 7.07
15 15 5 5 73
1 1 2 0 11
Expressed as average ± STD (n = 3). bAlthough background of glucose was present in the maltose substrate, no extra hydrolysis was observed; nd: not detected.
a
Despite the limited number of fecal microbial slurries from human origin (n = 4), clear differences in fermentation activity were noticed between the individual incubations. This was primarily reflected in the most abundant SCFA: acetate, propionate, and butyrate. As expected, acetate was the predominant SCFA at the start of incubation (abundance range: 50−75%) followed by propionate (14−20%) and butyrate (11−20%) (Figure 7). A 24 h incubation did not result in shifts in the SCFA profile for the control samples except for donor 4, which displayed an increase in the proportion of butyrate. In contrast, compared to the control samples, SCFA profiles of kojibiose-incubated samples displayed a high dependence on the origin of the fecal microbiota. Kojibiose incubation with fecal microbiota from donor 1 and donor 4 resulted in a 4% and 7% butyrate increase, respectively. Kojibiose incubation with fecal microbiota from donor 2 was selective for propionate production with 7% proportionality. In sharp contrast, 24 h incubation samples from donor 3 displayed a 22% increase in the proportion of acetate, while proportional propionate and butyrate levels dropped with around 10%. From these data, the possibility of using human fecal microbiota to ferment kojibiose is clear, and the production of a beneficial SCFA-profile could illustrate its possible prebiotic effect if linked to a beneficial shift in microbiota composition. Indeed, previous reports have already illustrated bifidogenic potential for kojibiose.15 It is known that primary carbohydrate degraders are important contributors to general fermentation activity within the gut microbial ecosystem. More importantly, they can trigger the production of propionate, butyrate by other bacteria, or other metabolites in a process called crossfeeding.39 This way, they uphold certain relationships within and lend stability to the human gut microbiota. Fermentation activity seems to be individual-dependent, though. Depending on an individual’s microbiome composition, primary carbohydrate degradation may or may not result in a more diverse SCFA profile. While generalizations cannot be made based on four individual incubations, this small trial indicates that the microbiota from certain individuals are not responsive toward kojibiose in terms of butyrate or propionate production (donor 3), while others display preferential crossfeeding toward butyrate (donor 4), propionate (donor 2), or both (donor 1). On the basis of these four incubations, kojibiose glycolysis seems to be a common feature while cross-feeding toward propionate or butyrate production is more variable. The determinants of these interindividual differences need to be further elucidated, but it indicates that kojibiose has, at least, the potential to trigger the production of SCFA with healthpromoting properties.
Figure 5. Caco-2 cell reactivity after 1 day treatment with 4.5 g/L cellobiose (C), kojibiose (K), and glucose (G) expressed as percentage compared to the untreated (U) condition. MTT, SRB, and NO are the mitochondrial activity, protein content, and nitric oxide production of the cells expressed as % absorbance, whereas ROS is the reactive oxygen species production expressed as % fluorescence. Error bars indicate relative standard deviations and * indicates significantly different values compared to the untreated (U) condition according to a two-tailed t test with unequal variances (p < 0.05, n = 6).
previously.34 Overall, kojibiose did not induce any response in the intestinal Caco-2 cell model. The in vitro cell culture experiments show that kojibiose is, as cellobiose, not used by the cells as a substrate for energy metabolism (MTT) nor for cellular growth (SRB). The incubation time of the cells with the compounds was 24 h, which is at least 6 times greater than the exposure time of sugars in the small intestine. Therefore, we conclude that most probably, kojibiose is not easily taken up by enterocytes as such and therefore, has a high chance of reaching the intestinal microbiota in relatively high concentrations, which is a requirement for prebiotic activity. Yet, it needs to be noted that Caco-2 cells are able to secrete α-glucosidases at their brush-border side35−37 and that kojibiose can be cleaved by rat rodent α-glucosidases.33 In our assay, the α-glucosidase activity of the Caco-2 cells was not measured, and therefore, its concentration may be too low to see effects. In addition, substrate specificity and reaction kinetics may be hostdependent, as mentioned by Hermans and co-workers.38 Kojibiose Fermentation by Human Fecal Microbiota. Analysis of SCFA content in the incubation sample revealed a strong fermentation activity of human fecal microbiota toward kojibiose. Averaging the four individual incubations at time points of 6 and 24 h, a 2.8-fold and a 2-fold greater total SCFA concentration was observed for kojibiose-incubated slurries, respectively, compared to those of the control incubations (Figure 6). The greater increase in SCFA concentration at the early time point of 6 h for all four incubations is indicative of a rapid kojibiose hydrolysis and subsequent fermentation. Kojibiose-incubated samples also displayed lower levels of branched-chain fatty acids for all four donors. As these branched acids are indicative of proteolytic activity, their decrease during kojibiose incubation is reminiscent of the increased carbohydrate fermenting activity as opposed to the control samples. 6037
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry
Figure 6. Concentrations of acetate, propionate, butyrate, branched -chain fatty acids, and total SCFA in the control and kojibiose-incubated fecal microbial slurries.
Figure 7. Proportional distribution of SCFA profiles from control and kojibiose-incubated fecal microbial slurries.
Moreover, a consistent flow behavior is important for weight control during chocolate enrobing and molding processes.41 Moreover, controlling the flow properties is necessary to be able to adjust formulations and processes toward different applications.42 The main factors influencing chocolate-flow behavior are particle-volume fraction, particle-size distribution, particle shape, surface roughness, wetting properties of the suspended particles in the continuous fat phase, interparticle forces, degree of aggregation, and the presence of dispersing agents and fat composition.43 We determined the density, which directly affects the particle volume fraction, and moisture content and solid state, which both influence particle aggregation, of kojibiose in comparison to sucrose (Table 5),
Bulk Functionality of Highly Pure Kojibiose. As a case study, we have analyzed the bulk functionality properties of highly pure kojibiose with a view to its potential application as sweetener in chocolates and chocolate-derived products. Confectionery products such as chocolates are suspensions of solid particles in a continuous fat phase. Their consumer acceptance depends obviously on the taste but also to a large extent on the unique mouthfeel, created by the melting of the fat phase and the release of sweetener (i.e., sugar) particles and other flavoring components. The perceived mouthfeel depends mainly on the particle sizes and flow properties.40 The chocolate composition and processing conditions have to ensure that these important quality attributes meet demands. 6038
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry
experiments remain to be performed, including the determination of caloric content and glycemic index. In a food-application case study, we compared the highly pure kojibiose to sucrose in terms of bulk functionality properties, hereby mapping its potential as an alternative sweetener in chocolates as one example of confectionery products. However, further research is needed here to also map the wider potential of kojibiose in other types of confectionery products, such as soft drinks, marmalades, backery, and all others.
Table 5. Comparison of Moisture Content, Density, and Melting Parameters between Kojibiose and Sucrose parameter
kojibiose
sucrose
density (g/cm3) moisture (%) Tm, onset (°C) Tm, max (°C) Tm, offset (°C) ΔHm (J/g)
1.34 5.8 143.6 161.9 167.6 154.8
1.59 0.5 184.1 191.3 196.9 136.1
■
ASSOCIATED CONTENT
S Supporting Information *
the most common sweetener in chocolate. Different densities for the sampled sugars were measured, indicating that sucrose replacement on a weight basis by kojibiose will result in an increased particle-volume fraction of chocolate, raising its relative viscosity.43 Hereto, we would rather suggest sucrose replacement in chocolate by kojibiose on a volumetric basis.44 Moisture contents of around 0.5% are common to dark chocolate and are controlled by the selection of ingredients meeting certain specifications and conching conditions. Higher levels will result in a strengthening of the network microstructure, inducing nonsmooth textures and poor flow properties in chocolate.45 Indeed, water “glues” the sugar particles, resulting in a secondary particle network that supports the overall structure even when the fat is molten. The relative high moisture content of kojibiose implies an adapted chocolate processing, depending on its solids state. In case of crystalline phase, conching at elevated temperature or for longer time would be recommended to remove as much moisture as possible. Conversely, the presence of kojibiose in a glassy state favors conching below the glass-transition temperature range.46 This would avoid undesirable increases in viscosity and particle aggregation following recrystallization from the metastable rubbery state. Thermal analysis showed that kojibiose is dominantly crystalline, similar to sucrose, as only a first-order transition (melting peak); no second-order transition (glass transition) was observed (Figure S.7). Compared to sucrose, kojibiose melts at lower temperatures. Given a melting temperature beyond 140 °C, this is irrelevant for chocolate production. In conclusion, crystalline kojibiose produced under the given conditions could be applied in fat-based suspensions as chocolates but might require a process adaptation to remove excess moisture. Conversely, in warmer climates, an elevated moisture level is likely to be an advantage toward thermal resistance. In summary, we have evaluated different BaSP variants in view of kojibiose synthesis directly from cheap and abundant sucrose. The L341I_Q345S variant of BaSP was found to efficiently synthesize kojibiose and remain fully active after 1 week of incubation at 55 °C. We achieved, using an integrated process intensification, the biocatalytic synthesis and purification of a total of 4.5 kg of crystalline kojibiose, i.e. two batches of 3.3 and 1.2 kg with 99.8% and 99% purity, respectively. This larger-scale production provided enough kojibiose to do an initial characterization of the potential of kojibiose in food applications. It was found that kojiobiose has potential beneficial effects on glycemic index (e.g., delayed glucose release) and has the potential to trigger the production of SCFA with health-promoting properties. In terms of in-depth characterization and approval for use in industrial food applications (e.g., Novel Food application), a number of extra
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.7b02258. Tables and figures describing the production, purification, and crystallization of kojibiose (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*Tel.: +3292649920; Fax: +3292646032; E-mail: tom.desmet@ ugent.be ORCID
Koen Beerens: 0000-0001-6608-0443 Funding
The authors wish to thank the Fund for Scientific ResearchFlanders (FWO-Vlaanderen, doctoral scholarship for KDW, and SBO project GlycoProFit, grant n° S003617N), the European Commission (FP7-project “Novosides”, grant 265854), COST Chemistry CM1102 (MSMT project LD13042), and the Ghent University (Multidisciplinary Research Partnership “Ghent Bio-Economy”) for financial support. Notes
The authors declare no competing financial interest.
■
ABBREVIATIONS USED SP, sucrose phosphorylase; SCFA, short-chain fatty acid; GOS, galacto-oligosaccharides; FOS, fructo-oligosaccharides; ISRS, International Society of Rare Sugars; DMSO, dimethyl sulfoxide; BCA, bicinchoninic acid; HPAEC, high-performance anion-exchange chromatography; PAD, pulsed amperometric detection; ICUMSA, International Commission for Uniform Methods of Sugar Analysis; MTT, 3-(4,5-dimethylthiazol-2-yl)2,5 diphenyltetrazolium bromide; SRB, sulphurodamine B; ROS, reactive oxygen species; NO, nitric oxide; SHIME, Simulator of the Human Intestinal Microbial Ecosystem; GI, glucose isomerase
■
REFERENCES
(1) O’Donnell, K., Kearsley, M. W., Eds. Sweeteners and sugar alternatives in food technology, 2nd ed.; Wiley−Blackwell: Hoboken, NJ, 2012. (2) Laparra, J. M.; Díez-Municio, M.; Herrero, M.; Moreno, F. J. Structural differences of prebiotic oligosaccharides influence their capability to enhance iron absorption in deficient rats. Food Funct. 2014, 5 (10), 2430−2437. (3) Molis, C.; Flourié, B.; Ouarne, F.; Gailing, M. F.; Lartigue, S.; Guibert, A.; Bornet, F.; Galmiche, J. P. Digestion, excretion, and energy value of fructooligosaccharides in healthy humans. Am. J. Clin. Nutr. 1996, 64, 324−328.
6039
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry (4) Lu, Y.; Levin, G. V.; Donner, T. W. Tagatose, a new antidiabetic and obesity control drug. Diabetes, Obes. Metab. 2008, 10, 109−134. (5) Espinosa, I.; Fogelfeld, L. Tagatose: from a sweetener to a new diabetic medication? Expert Opin. Invest. Drugs 2010, 19 (2), 285−294. (6) Mu, W.; Zhang, W.; Feng, Y.; Jiang, B.; Zhou, L. Recent advances on applications and biotechnological production of D-psicose. Appl. Microbiol. Biotechnol. 2012, 94 (6), 1461−1467. (7) Granströ m, T. B.; Takata, G.; Tokuda, M.; Izumori, K. Izumoring: a novel and complete strategy for bioproduction of rare sugars. J. Biosci. Bioeng. 2004, 97 (2), 89−94. (8) Beerens, K.; Desmet, T.; Soetaert, W. Enzymes for the biocatalytic production of rare sugars. J. Ind. Microbiol. Biotechnol. 2012, 39 (6), 823−834. (9) Walmagh, M.; Zhao, R.; Desmet, T. Trehalose Analogues: Latest Insights in Properties and Biocatalytic Production. Int. J. Mol. Sci. 2015, 16 (6), 13729−13745. (10) Watanabe, T.; Aso, K. Isolation of Kojibiose from Honey. Nature 1959, 183 (4677), 1740. (11) Verhaeghe, T.; De Winter, K.; Berland, M.; De Vreese, R.; D’hooghe, M.; Offmann, B.; Desmet, T. Converting bulk sugars into prebiotics: semi-rational design of a transglucosylase with controlled selectivity. Chem. Commun. (Cambridge, U. K.) 2016, 52 (18), 3687− 3689. (12) Sanz, M. L.; Gibson, G. R.; Rastall, R. A. Influence of disaccharide structure on prebiotic selectivity in vitro. J. Agric. Food Chem. 2005, 53 (13), 5192−5199. (13) Chaen, H.; Nishimoto, T.; Nakada, T.; Fukuda, S.; Kurimoto, M.; Tsujisaka, Y. Enzymatic synthesis of kojioligosaccharides using kojibiose phosphorylase. J. Biosci. Bioeng. 2001, 92 (2), 177−182. (14) Hodoniczky, J.; Morris, C. A.; Rae, A. L. Oral and intestinal digestion of oligosaccharides as potential sweeteners: A systematic evaluation. Food Chem. 2012, 132 (4), 1951−1958. (15) Díez-Municio, M.; Kolida, S.; Herrero, M.; Rastall, R. A.; Moreno, F. J. In vitro faecal fermentation of novel oligosaccharides enzymatically synthesized using microbial transglycosidases acting on sucrose. J. Funct. Foods 2016, 20, 532−544. (16) Wolfrom, M. L.; Thompson, A.; Lineback, D. R. Isopropyl Tetra-O-acetyl-α-D-glucopyranoside; A Synthesis of Kojibiose. J. Org. Chem. 1963, 28 (3), 860−861. (17) Duke, J.; Little, N.; Goldstein, I. J. Preparation of cyrstalline αkojibiose octaacetate from dextran B-1299-S: Its conversion into pnitrophenyl and p-isothiocyanatophenyl β-kojibioside. Carbohydr. Res. 1973, 27 (1), 193−198. (18) Fujimoto, H.; Nishida, H.; Ajisaka, K. Enzymatic Syntheses of Glucobioses by a Condensation Reaction with α-Glucosidase, βGlucosidases and Glucoamylase. Agric. Biol. Chem. 1988, 52 (6), 1345−1351. (19) Cantarella, L.; Nikolov, Z. L.; Reilly, P. J. Disaccharide production by glucoamylase in aqueous ether mixtures. Enzyme Microb. Technol. 1994, 16 (5), 383−387. (20) Monsan, P. F.; Ouarné, F. Oligosaccharides Derived from Sucrose. In Prebiotics and Probiotics Science and Technology; Charalampoulos, D., Rastall, R. A., Eds.; Springer: New York, 2009; pp 293−336. (21) Chaen, H.; Nishimoto, T.; Nakada, T.; Fukuda, S.; Kurimoto, M.; Tsujisaka, Y. Enzymatic synthesis of novel oligosaccharides from L-sorbose, maltose, and sucrose using kojibiose phosphorylase. J. Biosci. Bioeng. 2001, 92 (2), 173−176. (22) Díez-Municio, M.; Montilla, A.; Moreno, F. J.; Herrero, M. A sustainable biotechnological process for the efficient synthesis of kojibiose. Green Chem. 2014, 16 (4), 2219. (23) Kraus, M.; Gorl, J.; Timm, M.; Seibel, J. Synthesis of the rare disaccharide nigerose by structure-based design of a phosphorylase mutant with altered regioselectivity. Chem. Commun. 2016, 52 (25), 4625−4627. (24) De Winter, K.; Soetaert, W.; Desmet, T. An Imprinted CrossLinked Enzyme Aggregate (iCLEA) of Sucrose Phosphorylase: Combining Improved Stability with Altered Specificity. Int. J. Mol. Sci. 2012, 13 (12), 11333−11342.
(25) Minekus, M.; Alminger, M.; Alvito, P.; Ballance, S.; Bohn, T.; Bourlieu, C.; Carrière, F.; Boutrou, R.; Corredig, M.; Dupont, D.; et al. A standardised static in vitro digestion method suitable for food - an international consensus. Food Funct. 2014, 5 (6), 1113−1124. (26) Van den Abbeele, P.; Belzer, C.; Goossens, M.; Kleerebezem, M.; De Vos, W. M.; Thas, O.; De Weirdt, R.; Kerckhof, F.-M.; Van de Wiele, T. Butyrate-producing Clostridium cluster XIVa species specifically colonize mucins in an in vitro gut model. ISME J. 2013, 7 (5), 949−961. (27) Van Herreweghen, F.; Van den Abbeele, P.; De Mulder, T.; De Weirdt, R.; Geirnaert, A.; Hernandez-Sanabria, E.; Vilchez-Vargas, R.; Jauregui, R.; Pieper, D. H.; Belzer, C.; et al. In vitro colonisation of the distal colon by Akkermansia muciniphila is largely mucin and pH dependent. Benefic. Microbes 2017, 8 (1), 81−96. (28) Buchholz, K.; Seibel, J. Industrial carbohydrate biotransformations. Carbohydr. Res. 2008, 343 (12), 1966−1979. (29) Yoon, S.-H.; Mukerjea, R.; Robyt, J. F. Specificity of yeast (Saccharomyces cerevisiae) in removing carbohydrates by fermentation. Carbohydr. Res. 2003, 338 (10), 1127−1132. (30) Li, Z.; Xiao, M.; Lu, L.; Li, Y. Production of nonmonosaccharide and high-purity galactooligosaccharides by immobilized enzyme catalysis and fermentation with immobilized yeast cells. Process Biochem.. 2008, 43, 896−899. (31) Van der Borght, J.; Desmet, T.; Soetaert, W. Enzymatic production of β-D-glucose-1-phosphate from trehalose. Biotechnol. J. 2010, 5 (9), 986−993. (32) Yamauchi, F.; Aso, K. Crystalline α-kojibiose. Nature 1961, 189, 753. (33) Lee, B.-H.; Rose, D. R.; Lin, A. H.-M.; Quezada-Calvillo, R.; Nichols, B. L.; Hamaker, B. R. Contribution of the Individual Small Intestinal α-Glucosidases to Digestion of Unusual α-Linked Glycemic Disaccharides. J. Agric. Food Chem. 2016, 64 (33), 6487−6494. (34) Cosentino, F.; Hishikawa, K.; Katusic, Z. S.; Lüscher, T. F. High glucose increases nitric oxide synthase expression and superoxide anion generation in human aortic endothelial cells. Circulation 1997, 96 (1), 25−28. (35) Klumperman, J.; Fransen, J. A.; Boekestijn, T. C.; Oude Elferink, R. P.; Matter, K.; Hauri, H. P.; Tager, J. M.; Ginsel, L. A. Biosynthesis and transport of lysosomal alpha-glucosidase in the human colon carcinoma cell line Caco-2: secretion from the apical surface. J. Cell Sci. 1991, 100, 339−347. (36) Francí, C.; Egea, G.; Arribas, R.; Reuser, A. J.; Real, F. X. Lysosomal alpha-glucosidase: cell-specific processing and altered maturation in HT-29 colon cancer cells. Biochem. J. 1996, 314, 33−40. (37) Malunga, L. N.; Eck, P.; Beta, T. Inhibition of Intestinal αGlucosidase and Glucose Absorption by Feruloylated Arabinoxylan Mono- and Oligosaccharides from Corn Bran and Wheat Aleurone. J. Nutr. Metab. 2016, 2016, 1−9. (38) Hermans, M. M.; Kroos, M. A.; van Beeumen, J.; Oostra, B. A.; Reuser, A. J. Human lysosomal alpha-glucosidase. Characterization of the catalytic site. J. Biol. Chem. 1991, 266, 13507−13512. (39) Falony, G.; Vlachou, A.; Verbrugghe, K.; Vuyst, L. D. CrossFeeding between Bifidobacterium longum BB536 and AcetateConverting, Butyrate-Producing Colon Bacteria during Growth on Oligofructose. Appl. Environ. Microbiol. 2006, 72 (12), 7835−7841. (40) Bolenz, S.; Manske, A. Impact of fat content during grinding on particle size distribution and flow properties of milk chocolate. Eur. Food Res. Technol. 2013, 236 (5), 863−872. (41) Servais, C.; Ranc, H.; Roberts, I. D. Determination of chocolate viscosity. J. Texture Stud. 2003, 34 (5−6), 467−497. (42) De Graef, V.; Depypere, F.; Minnaert, M.; Dewettinck, K. Chocolate yield stress as measured by oscillatory rheology. Food Res. Int. 2011, 44 (9), 2660−2665. (43) Servais, C.; Jones, R.; Roberts, I. The influence of particle size distribution on the processing of food. J. Food Eng. 2002, 51 (3), 201− 208. (44) Sokmen, A.; Gunes, G. Influence of some bulk sweeteners on rheological properties of chocolate. LWT - Food Sci. Technol. 2006, 39 (10), 1053−1058. 6040
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041
Article
Journal of Agricultural and Food Chemistry (45) Stortz, T. A.; Marangoni, A. G. Heat resistant chocolate. Trends Food Sci. Technol. 2011, 22 (5), 201−214. (46) Zumbé, A.; Bade, A.-M. Milk chocolate and method of making same. US5501865A, 1992.
6041
DOI: 10.1021/acs.jafc.7b02258 J. Agric. Food Chem. 2017, 65, 6030−6041