Large-scale Preparation of 1,2-β-Glucan Using 1,2-β-Oligoglucan Phosphorylase

Abstract

1,2-β-Glucan was produced enzymatically from 1.0 M sucrose and 0.5 M glucose by the combination of sucrose phosphorylase and 1,2-β-oligoglucan phosphorylase in the presence of 100 mM inorganic phosphate. Accumulation of 1,2-β-glucan in 2 L of the reaction mixture reached over 800 mM (glucose equivalent). Sucrose, glucose and fructose were removed after the reaction by yeast treatment. 1,2-β-Glucan was precipitated with ethanol to obtain 167 g of 1,2-β-glucan from 1 L of the reaction mixture.

Abbreviations

Glc1P, α-D-glucose 1-phosphate; LNB, lacto-N-biose I; GlcNAc, N-acetyl-D-glucosamine; SP, sucrose phosphorylase; GNB, galacto-N-biose; GalRha, D-galactosyl-β-1,4-L-rhamnose; ManGlcNAc, β-1,4-D-mannosyl-N-acetyl-D-glucosamine; OGP, 1,2-β-oligoglucan phosphorylase; DP, degrees of polymerization; Sopns, 1,2-β-glucooligosaccharides; CGS, cyclic β-1,2-glucan synthase; GT, glycosyl transferase family; GH, glycoside hydrolase family; CβG, cyclic β-1,2-glucan; LβG, linear 1,2-β-glucan; TLC, thin layer chromatography; Glc6P, glucose 6-phosphate; G6PDH, glucose 6-phosphate dehydrogenase; PGM, phosphoglucomutase; NMR, nuclear magnetic resonance.

INTRODUCTION

Phosphorylases are exo-type enzymes that phosphorolyze glycosidic bonds at the non-reducing ends to release the corresponding sugar 1-phosphate. The remarkable features of phosphorylases are the reversibility and strict regioselectivity of the reactions, which are beneficial for the synthesis of various oligosaccharides with little byproducts.^{1) 2)} Though sugar 1-phosphate is still an expensive material in terms of practical production, reversibility of the reaction makes it possible to avoid using sugar 1-phosphate as a material through the combined actions of two phosphorylases in a one-pot reaction. Namely, sugar 1-phosphate necessary for the synthesis of target oligosaccharides is supplied by another phosphorylase producing the same sugar 1-phosphate from inexpensive materials. Production of trehalose from maltose was the first reported to involve the combination of phosphorylases.³⁾ Then cellobiose⁴⁾ and laminaribiose⁵⁾ were synthesized using sucrose phosphorylase to supply α-D-glucose 1-phosphate (Glc1P) from sucrose with cellobiose phosphorylase and laminaribiose phosphorylase, respectively. Nigerose was also prepared based on the idea of a combination of phosphorylases.⁶⁾

The method has been extended to combinations of two phosphorylases acting on different sugar 1-phosphates by the addition of enzymes converting sugar 1-phosphate into another one. β-Galactosides such as lacto-N-biose I (D-galactosyl-β-1,3-N-acetyl-D-glucosamine, LNB) have been synthesized from sucrose and N-acetyl-D-glucosamine (GlcNAc) by means of a one-pot enzymatic reaction using the combined actions of sucrose phosphorylase (SP) and lacto-N-biose I phosphorylase together with the conversion of Glc1P into α-D-galactose 1-phosphate using the reactions of both UDP-glucose-hexose 1-phosphate uridylyltransferase and UDP-glucose 4-epimerase in the presence of catalytic amounts of UDP-glucose and inorganic phosphate.⁷⁾ This procedure is applicable to the practical preparation of galacto-N-biose (GNB) and D-galactosyl-β-1,4-L-rhamnose (GalRha).^{8) 9)} Then a β-mannoside, β-1,4-D-mannosyl-N-acetyl-D-glucosamine (ManGlcNAc), was synthesized from sucrose and GlcNAc by means of a one-pot enzymatic reaction using the combined actions of SP and ManGlcNAc phosphorylase accompanied by the multiple enzymatic reaction essential for the conversion of Glc1P into α-D-mannose 1-phosphate.¹⁰⁾

Recently, we found 1,2-β-oligoglucan phosphorylase (OGP), a novel enzyme phosphorolyzing 1,2-β-glucan with degrees of polymerization (DP) of more than 2.¹¹⁾ OGP produces 1,2-β-glucooligosaccharides (Sopns) from sophorose and Glc1P through the reverse reaction. We also succeeded in Gram-scale production of 1,2-β-glucan from glucose and Glc1P despite that glucose is a poor acceptor of OGP. However, the method is not still practical in terms of using Glc1P as a material.

Cyclic β-1,2-glucan synthase (CGS) is another enzyme for producing 1,2-β-glucan. The N-terminal domain of CGS belongs to glycosyl transferase family (GT) 84. CGS has a phosphorylase domain belonging to glycoside hydrolase family (GH) 94 at its C-terminus.¹²⁾ The GT84 domain plays the main role in the synthesis of cyclic β-1,2-glucan (CβG) involving UDP-glucose, while the GH94 domain just phosphorolyzes 1,2-β-glucan produced by the GT84 domain to adjust the DP of the glucan to around 20 before cyclization of the product.^{13) 14) 15)} Some α-proteobacteria produce CβG as symbiotic or virulent factor.^{16) 17) 18) 19)} CβG is involved in adaption to a change in osmotic conditions in Rhizobium meliloti and Agrobacterium tumefaciens.²⁰⁾ Bacterial production of CβG has been developed. A Rhizobium phaseoli mutant produces approximately 1 g of CβG per 1-liter culture medium.²¹⁾ The maximum production yield of CβG on Rhizobium trifolii TA-1 (now Rhizobium leguminosarum bv. trifolii)²²⁾ culture is 10.9 g/L.²³⁾ However, other exopolysaccharides are simultaneously produced during bacterial production and column chromatography is used for purification of the product.²³⁾

Linear 1,2-β-glucan (LβG) is also found in some bacteria as an extracellular or periplasmic glucan. Some strains of Acetobacter produce LβG with a DP of 6‒42.²⁴⁾ Involvement of CGS in the synthesis of LβG is unknown due to the unavailability of genomic DNA sequences of the strains, though strains of Acetobacter with available genomic DNA sequences possess genes homologous to cgs. In Escherichia coli and Pseudomonas syringae, MdoG and MdoH (periplasmic glucan biosynthesis proteins) are involved in production of short LβG with a DP of 5‒13 possessing β-1,6-glucosyl branches.^{25) 26)} Amount of short LβG produced is regulated by osmolarity in E. coli.^{25) 27)}

Practical preparation of 1,2-β-glucan is needed for research associated with the function of 1,2-β-glucan and related enzymes, but application of CGS, MdoG, and MdoH to large-scale production seems difficult, because CGS, MdoG, and MdoH are membrane enzymes. In this paper, we describe the one-pot large-scale production of linear 1,2-β-glucan from sucrose and glucose, inexpensive materials, employing the successive actions of SP and OGP, as shown in Fig. 1.

Fig. 1.

Reaction scheme for synthesis of 1,2-β-glucan in a one-pot reaction.

The enzymes used are boxed. Material sugars are indicated by gray shading. Reaction products are underlined. The intermediate of the reaction is shown in parenthesis.

RESULTS

Optimization of synthetic conditions for 1,2-β-glucan.

First, we optimized the glucose concentration by varying the concentration from 0.1 M to 1.0 M. The velocity of production of 1,2-β-glucan was evaluated by comparing the thickness of the spot at the origin on TLC plates at an early stage of the reaction (Fig. 2 (A)). We determined 500 mM to be optimum because a significant increase in the amount of the reaction product was not observed with over 500 mM. Next, we optimized the concentration ratio of SP and OGP for reducing the amounts of the enzymes used. The velocity of production of 1,2-β-glucan in the presence of various concentration ratios of the enzymes was evaluated as described above. While reduction of the OGP concentration caused a remarkable decrease in the amount of synthesized 1,2-β-glucan, a reduction of the SP concentration to one-fourth of that of OGP was possible without a decrease in the amount of the product (Fig. 2 (B)).

Fig. 2.

Optimization of the glucose concentration (A) and ratio of SP to OGP (B) for 1,2-β-glucan synthesis.

Lane M represents the marker (0.3 μL of 50 mM glucose and 100 mM sucrose). An asterisk represents the origin of a TLC plate. (A) TLC analysis of reaction products in the presence of various concentrations of glucose with fixed concentrations of the other compounds. (B) TLC analysis of reaction products at various concentration ratios of SP and OGP with fixed concentrations of the other compounds. Concentrations of SP and OGP are shown below the TLC plate. The ratios adjacent to the concentrations represent that enzyme concentrations were reduced at the ratio from 0.2 mg/mL of enzymes.

Large-scale preparation of 1,2-β-glucan.

We carried out a 2-liter scale reaction under the optimized conditions. After the reaction had proceeded for 12 days, the reduction of sucrose and the increase of 1,2-β-glucan had each reached a plateau, according to TLC analysis (Fig. 3). No visible precipitant was observed during the reaction. Quantification of components in the reaction mixture showed a reduction of the sucrose concentration to approximately 100 mM and an increase in the 1,2-β-glucan concentration (glucose equivalence) to over 800 mM (Fig. 4). The fructose concentration also increased in a similar manner to that of 1,2-β-glucan, which is consistent with reduction of the sucrose concentration. The concentration of glucose did not change during the reaction since material glucose is used for 1,2-β-glucan synthesis only as reducing end of 1,2-β-glucan while the other moiety is derived from Glc1P. The amount of glucose converted into sophorose by OGP is assumed to be the same as that of glucose released through hydrolysis of sucrose, a minor reaction of SP. The phosphate concentration first decreased rapidly and then increased gradually. The change of the Glc1P concentration corresponded to that of the phosphate concentration (Fig. 4). Then, we analyzed the time course of the DP distribution of 1,2-β-glucan. Interestingly, distribution of the DP of 1,2-β-glucan spread upward time-dependently (Fig. 5). Evaluation of the distribution of 1,2-β-glucan with lower DPs was difficult due to loss of them by dialysis for desalting.

Fig. 3.

TLC analysis of the time course production of 1,2-β-glucan on a large-scale.

Lanes M1 and M2 represent Glc and Suc, and Fru as markers, respectively (1 μL of 0.2% each sugar). An asterisk represents the origin of the TLC plate.

Fig. 4.

Time course of component concentrations on large-scale production of 1,2-β-glucan.

The concentrations of sucrose, glucose, inorganic phosphate, Glc1P, fructose and 1,2-β-glucan are shown by open circles, open triangles, open squares, closed squares, closed triangles and closed circles, respectively.

Fig. 5.

HPLC analysis of DP of 1,2-β-glucan during the reaction.

The reaction time and DP of the product are shown at the right and above, respectively. The samples dialyzed through dialysis membrane with MWCO of 3500 were used.

Purification of 1,2-β-glucan.

Sucrose, fructose, and glucose were successfully removed from the reaction solution by treatment with yeast after removal of the enzymes (Fig. 6), though two-fold dilution of the reaction solution was needed for sufficient removal of the sugars, probably due to the hypertonicity of the reaction mixture. In contrast, efficient separation of 1,2-β-glucan from other sugars by addition of ethanol directly to the reaction mixture was unsuccessful. The synthesized 1,2-β-glucan was so soluble that it dissolved in an approximately 10-fold concentrated supernatant on yeast treatment. The addition of an equal volume of ethanol to the concentrated supernatant resulted in a sticky precipitant, which was then dried. The purified 1,2-β-glucan (167 g) was obtained as an almost white powder from 1 L of the reaction solution. The average DP of the product was 25 according to NMR analysis (Fig. S1).

Fig. 6.

TLC analysis of components at different 1,2-β-glucan preparation steps.

M1 and M2, markers (1 μL of 0.2% each sugar); lanes 1 and 2, 10-fold diluted mixtures before and after enzymatic reaction, respectively; lane 3, 10-fold diluted supernatant after yeast treatment; lane 4, 1% purified 1,2-β-glucan. Lane 1‒4 0.4 μL of the samples were spotted. An asterisk represents the origin of a TLC plate.

DISCUSSION

We are the first to succeed in enzymatic production of 1,2-β-glucan through a one-pot large-scale reaction and purification of 1,2-β-glucan from the reaction solution without using column chromatography. This indicates that the method shown in Fig. 1 is applicable to soluble polysaccharide production. The reaction yield of 1,2-β-glucan using the combination of OGP and SP was over 80%, which is much higher than the production yield of 1,2-β-glucan using Glc1P as a starting material estimated to be approximately 40% according to Nakajima et al.¹¹⁾ In the case of synthesis of LNB using GNB/LNB phosphorylase, and GalRha using GalRha phosphorylase, the reaction yield of each of LNB and GalRha on combination with SP was higher than theoretical reaction yield calculated from the equilibrium constant of phosphorolysis of each of LNB and GalRha alone.^{7) 9)} Thus, the high reaction yield in this study is considered to be due to the high energy of the β-fructofuranosyl linkage in sucrose as in the synthesis of LNB and GalRha.

We examined the distribution of DP of 1,2-β-glucan. Distribution of the DP of the product clearly expanded upward during the reaction, ending in a wide DP distribution, though evaluation of the time course of Sopns with lower DPs was difficult due to dialysis of the samples. This is explainable with the characteristics of OGP. Sophorose is supplied gradually during the whole reaction period due to the poor preference of glucose as an acceptor. This causes elongation of sophorose molecules at a similar velocity for varying times, resulting in diversified DPs of the product. The difference of β-glucan in solubility also appears to be related with the DP distribution. Cellodextrin phosphorylase produces 1,4-β-glucan from glucose and Glc1P through a mechanism similar to the case of OGP but the average DP of the product is 9,²⁸⁾ which is due to the poor solubility of 1,4-β-glucan and the inactivity of CDP toward the precipitant. 1,3-β-D-Glucan phosphorylase also produces insoluble 1,3-β-glucan with an assumed DP of over 30 from laminaribiose and Glc1P.^{29) 30)} In contrast, 1,2-β-glucan remained soluble throughout all steps in this and previous³¹⁾ studies, which means that 1,2-β-glucan is free from the limitation of the DP increase due to precipitation.

Combination of SP in this study enabled us to obtain a large amount of 1,2-β-glucan. The achievement will contribute to further studies on 1,2-β-glucan, and related enzymes and proteins.

EXPERIMENTAL

Materials. SP from Bifidobacterium longum subsp. longum and OGP from Listeria innocua were produced in E. coli BL21(DE3) with the corresponding expression plasmids, and were purified as previously described in the references.^{7) 11)} Glucose and sucrose were purchased from Nacalai Tesque, Inc (Kyoto, Japan).

Optimization of synthetic conditions for 1,2-β-glucan. The synthetic reaction was performed with a small scale volume (240 μL) in the presence of sucrose, glucose, sodium phosphate buffer (pH 7.0), SP, and OGP at 30°C for 1 day. A sample (20 μL) was mixed with 180 μL of water and then heated at 100°C for 5 min to stop the reaction. The products were analyzed by thin layer chromatography (TLC). For optimization of the glucose concentration, varying concentrations of glucose (0.1 M to 1.0 M) and fixed concentrations of sucrose (1.0 M), sodium phosphate buffer (100 mM), SP (0.2 mg/mL), and OGP (0.2 mg/mL) were used. To optimize the concentration ratio of the enzymes, the concentration ratio of SP and OGP was varied between 1:32 and 8:1 by reducing the concentration of one enzyme from 0.2 mg/mL, the concentrations of sucrose, glucose, and sodium phosphate being fixed to 1.0 M, 500 mM, and 100 mM, respectively.

Large-scale preparation of 1,2-β-glucan. A large-scale reaction was performed with a 2-liter reaction mixture comprising 1.0 M sucrose, 0.5 M glucose, 100 mM sodium phosphate (pH 7.0), 0.01 mg/mL SP, and 0.04 mg/mL OGP at 30°C for 12 days. For measurement of the time course concentrations of components, samples (20 μL) taken at intervals of 1 day were each mixed with 180 μL of water and then heated at 100°C for 5 min to stop the reaction.

After the reaction, 10 g of DEAE-cellulose (DE52; Wako Pure Chemical Industries, Osaka, Japan) equilibrated with 50 mM sodium phosphate buffer (pH 7.0) was added to the reaction mixture to adsorb the enzymes. The suspension was stirred at room temperature for 1 h, and then the DEAE-cellulose was removed by filtration. After an equal volume of water had been added to the supernatant, 80 g of baker’s yeast (Oriental Yeast Co., Ltd., Tokyo, Japan) was added to the mixture. The sample was incubated at 30°C with vigorous agitation overnight to eliminate fructose, glucose, and sucrose. Then the yeast cells were removed by centrifugation and filtration with a 0.22 μm filter (Sartorius, Palaiseau, France). The filtrate (2 L) was concentrated approximately 10-fold with an evaporator (EYELA NVC-2000, TOKYO RIKAKIKAI CO., LTD., Tokyo, Japan). 1,2-β-Glucan was precipitated by the addition of an equal volume of ethanol. After the supernatant was removed by decantation, the sample was dried.

Thin layer chromatography. Samples were 10-fold diluted with water and then heat-treated at 100°C for 5 min to stop the reaction. The samples (0.5 μL) were then spotted onto a TLC plate (7.5 cm × 5 cm, Kieselgel 60 F254; Merck KGaA, Darmstadt, Germany), and the TLC plate was developed with acetonitrile/water (3 : 1, v/v). The TLC plate was dipped in a 5% sulfuric acid-ethanol solution briefly and then heated in an oven until the carbohydrates were sufficiently visualized.

Quantification of components during large-scale production of 1,2-β-glucan. The concentrations of glucose, fructose, inorganic phosphate, Glc1P, sucrose, and 1,2-β-glucan were calculated from the absorbance measured in a well of a 96-well microplate (EIA/RIA plate, 96-well half area; Corning Inc., Corning, USA) on a Spectramax 190 (Molecular Devices, Inc., Sunnyvale, USA), as described below.

The glucose concentration was determined using the glucose oxidase/peroxidase method. The absorbance of 510 nm derived from a red-stained compound produced from 4-aminoantipyrine by peroxidase was measured after 180 μL of GOPOD reagent (Megazyme International Ireland Ltd., Wicklow, Ireland) mixed with a 200-fold diluted sample (20 μL) had been incubated at 42°C for 20 min.

The fructose concentration was determined using an F-kit D-glucose/fructose (J.K. International Inc., Tokyo, Japan). First, the reaction mixture (151 μL) containing reagent I (50 μL), reagent II (1 μL), and a 6,000-fold diluted sample (100 μL) was incubated at 25°C for 15 min. During this process, glucose and fructose were phosphorylated by hexokinase and then glucose 6-phosphate (Glc6P) was oxidized by glucose 6-phosphate dehydrogenase (G6PDH). Then the reaction mixture with 0.5 μL of reagent III added was incubated at 25°C for 15 min to convert fructose 6-phosphate into Glc6P, which then successively reacted with G6PDH in reagent II. The increase in absorbance of 340 nm derived from NADPH on the addition of reagent III was measured to determine the fructose concentration.

The method of Lowry and Lopez³¹⁾ was used to determine the concentration of inorganic phosphate in the reaction solution. After 10 μL aliquots of 100-fold diluted samples had been mixed with 80 μL of 0.2 M sodium acetate (pH 4.0), 10 μL of 1% ammonium molybdate containing 25 mM sulfuric acid and 10 μL of 1% ascorbic acid containing 0.05% potassium bisulfate, the mixtures were incubated at 37°C for 1 h. The concentration of inorganic phosphate was determined by measuring the absorbance of 700 nm.

The concentration of Glc1P was determined based on the phosphoglucomutase (PGM)-glucose 6-phosphate dehydrogenases (G6PDH) method.³²⁾ After 50 μL of a 1,000-fold diluted reaction solution mixed with 50 μL of reagent containing 50 mM sodium phosphate (pH 7.5), 2 mM thio-NAD⁺ (Oriental Yeast, Tokyo, Japan), 20 μM glucose 1,6-bisphosphate (Sigma-Aldrich Co., St. Louis, USA), 10 mM MgCl2, 18 U/mL PGM from rabbit muscle (Sigma-Aldrich), and 20 U/mL G6PDH from Leuconostoc mesenteroides (Oriental Yeast) had been incubated at 30°C for 10 min, the absorbance of 400 nm was measured to calculate the concentration of Glc1P.

The concentrations of sucrose and 1,2-β-glucan (glucose equivalent) were determined by subtraction of the concentrations of Glc1P with and without treatment with corresponding phosphorylases to convert them into Glc1P. Quantification of Glc1P was performed as described above. For determination of the sucrose concentration, after treatment of 50 μL of a 1,000-fold diluted sample in the presence of 100 mM sodium phosphate (pH 7.0) with SP (0.246 mg/mL) at 30°C for 2 h, the sample was incubated at 100°C for 5 min and then the Glc1P concentration was measured. For determination of the 1,2-β-glucan concentration, 0.5 mg/mL OGP was used in place of SP.

HPLC analysis. Ten-fold diluted samples (30 μL) were dialyzed against water with an Easy Sep MD-003-50 (MWCO 3500, TOMY, Tokyo, Japan) overnight, and then diluted twice and subjected to high performance ion chromatography using a Dionex DX500 (Thermo Fisher Scientific Inc., Franklin, USA) equipped with a pulsed amperometric detector. The samples (15 μL) were injected into CarboPac PA100 (φ 4 mm × 250 mm; Themo Fisher Scientific) and eluted with a 0‒250 mM sodium acetate convex gradient (curve 3 in section 2.9.4 of the GP50 gradient pump operator’s manual) in 100 mM NaOH at a flow rate of 1 mL/min for 60 min. The DPs of the products were determined by superimposition of peaks of Sopns produced by OGP¹¹⁾ and the samples.

NMR. Purified 1,2-β-glucan powder (20 mg) dissolved in 150 μL of water was dialyzed against water overnight with an Easy Sep MD-003-50 and then was lyophilized to obtain 19.1 mg of the product. A one-dimensional ¹H nuclear magnetic resonance (NMR) spectrum of the product (2 mg) was obtained in D2O (0.7 mL) with 2-methyl-2-propanol as an internal standard using a Bruker Avance 800 spectrometer (Bruker Biospin, Rheinstetten, Germany). Proton signals were assigned based on the ¹H-NMR spectrum of 1,2-β-glucan reported previously.¹¹⁾ The average DP of the product was determined using the peak area of C-6 ¹H of internal glucose units and C-2 ¹H at the non-reducing end as described by Nakajima et al.¹¹⁾

ACKNOWLEDGEMENTS

This work was supported by Noda Institute for Scientific Research GRANT (2013 Young Investigator Research Grant). We thank the staff of the Instrumental Analysis Center for Food Chemistry of the National Food Research Institute for recording the NMR spectra. The Sugawara Laboratory of the Tokyo University of Science kindly helped us to prepare the 1,2-β-glucan.

REFERRENCES

• 1)  M. Kitaoka and K. Hayashi: Carbohydrate-processing phosphorolytic enzymes. Trends Glycosci. Glycotechnol., 14, 35‒50 (2002).
• 2)  H. Nakai, M. Kitaoka, B. Svensson, and K. Ohtsubo: Recent development of phosphorylases possessing large potential for oligosaccharide synthesis. Curr. Opin. Chem. Biol., 17, 301‒309 (2013).
• 3)  S. Murao, H. Nagano, S. Ogura, and T. Nishino: Enzymatic-synthesis of trehalose from maltose. Agric. Biol. Chem., 49, 2113‒2118 (1985).
• 4)  M. Kitaoka, T. Sasaki, and H. Taniguchi: Conversion of sucrose into cellobiose using sucrose phosphorylase, xylose isomerase and cellobiose phosphorylase. J. Jpn. Soc. Starch Sci., 39, 281‒283 (1992).
• 5)  M. Kitaoka, T. Sasaki, and H. Taniguchi: Conversion of sucrose into laminaribiose using sucrose phosphorylase, xylose isomerase and laminaribiose phosphorylase. J. Jpn. Soc. Starch Sci., 40, 311‒314 (1993).
• 6)  T. Nihira, F. Miyajima, K. Chiku, M. Nishimoto, M. Kitaoka, K. Ohtsubo, and H. Nakai: One pot enzymatic production of nigerose from common sugar resources employing nigerose phosphorylase. J. Appl. Glycosci., 61, 65‒80 (2014).
• 7)  M. Nishimoto and M. Kitaoka: Identification of the putative proton donor residue of lacto-N-biose phosphorylase (EC 2.4.1.211). Biosci. Biotechnol. Biochem., 71, 1587‒1591 (2007).
• 8)  M. Nishimoto and M. Kitaoka: One-pot enzymatic production of β-D-galactopyranosyl-(1→3)-2-acetamido-2-deoxy-D-galactose (galacto-N-biose) from sucrose and 2-acetamido-2-deoxy-D-galactose (N-acetylgalactosamine). Carbohydr. Res., 344, 2573‒2576 (2009).
• 9)  M. Nakajima, M. Nishimoto, and M. Kitaoka: Practical preparation of D-galactosyl-β1→4-L-rhamnose employing the combined action of phosphorylases. Biosci. Biotechnol. Biochem., 74, 1652‒1655 (2010).
• 10)  T. Nihira, E. Suzuki, M. Kitaoka, M. Nishimoto, K. Ohtsubo, and H. Nakai: Discovery of β-1,4-D-mannosyl-N-acetyl-D-glucosamine phosphorylase involved in the metabolism of N-glycans. J. Biol. Chem., 288, 27366‒27374 (2013).
• 11)  M. Nakajima, H. Toyoizumi, K. Abe, H. Nakai, H. Taguchi, and M. Kitaoka: 1,2-β-Oligoglucan phosphorylase from Listeria innocua. PLoS One, 9 (3), e92353 (2014).
• 12)  A.E. Ciocchini, L.S. Guidolin, A.C. Casabuono, A.S. Couto, N.I. de Iannino, and R.A. Ugalde: A glycosyltransferase with a length-controlling activity as a mechanism to regulate the size of polysaccharides. Proc. Natl. Acad. Sci. USA, 104, 16492‒16497 (2007).
• 13)  O.A. Castro, A. Zorreguieta, V. Ielmini, G. Vega, and L. Ielpi: Cyclic β-1,2-glucan synthesis in Rhizobiaceae: roles of the 319-kDa protein intermediate. J. Bacteriol., 178, 6043‒6048 (1996).
• 14)  M.S. Roset, A.E. Ciocchini, R.A. Ugalde, and N. Iñón de Iannino: Molecular cloning and characterization of cgt, the Brucella abortus cyclic β-1,2-glucan transporter gene, and its role in virulence. Infect. Immun., 72, 2263‒2271 (2004).
• 15)  M. Hisamatsu: Cyclic (1→2)-β-D-glucans (cyclosophorans) produced by Agrobacterium and Rhizobium species. Carbohydr. Res., 231, 137‒146 (1992).
• 16)  G. Briones, N. Iñón de Iannino, M. Roset, A. Vigliocco, P.S. Paulo, and R.A. Ugalde: Brucella abortus cyclic β-1,2-glucan mutants have reduced virulence in mice and are defective in intracellular replication in HeLa cells. Infect. Immun., 69, 4528‒4535 (2001).
• 17)  X.S. Zhang and H.P. Cheng: Identification of Sinorhizobium meliloti early symbiotic genes by use of a positive functional screen. Appl. Env. Microbiol., 72, 2738‒2748 (2006).
• 18)  T. Dylan, P. Nagpal, D.R. Helinski, and G.S. Ditta: Symbiotic pseudorevertants of Rhizobium meliloti ndv mutants. J. Bacteriol., 172, 1409‒1417 (1990).
• 19)  L.A. Rigano, C. Payette, G. Brouillard, M.R. Marano, L. Abramowicz, P.S. Torres, M. Yun, A.P. Castagnaro, M.E. Oirdi, V. Dufour, F. Malamud, J.M. Dow, K. Bouarab, and A.A. Vojnov: Bacterial cyclic β-1,2-glucan acts in systemic suppression of plant immune responses. Plant Cell, 19, 2077‒2089 (2007).
• 20)  J. Gay-Fraret, S. Ardissone, K. Kambara, W.J. Broughton, W.J. Deakin, and A. Le Quéré: Cyclic-β-glucans of Rhizobium (Sinorhizobium) sp. strain NGR234 are required for hypo-osmotic adaptation, motility, and efficient symbiosis with host plants. FEMS Microbiol. Lett., 333, 28‒36 (2012).
• 21)  T. Higashiura, M. Ikeda, M. Okubo, M. Hisamatsu, A. Amemura and T. Harada: An improved method for preparation of cyclic (1→2)-β-D-glucan using an acidic polysaccharide-negative mutant. Agric. Biol. Chem., 49, 1865‒1866 (1985).
• 22)  M.H. Ramírez-Bahena, P. García-Fraile, A. Peix, A. Valverde, R. Rivas, J.M. Igual, P.F. Mateos, E. Martínez-Molina, and E. Velázquez: Revision of the taxonomic status of the species Rhizobium leguminosarum (Frank 1879) Frank 1889AL, Rhizobium phaseoli Dangeard 1926AL and Rhizobium trifolii Dangeard 1926AL. R. trifolii is a later synonym of R. leguminosarum. Reclassification of the strain. Int. J. Syst. Evol. Microbiol., 58, 2484‒2490 (2008).
• 23)  M.W. Breedveld, L.P. Zevenhuizen, and A.J. Zehnder: Excessive excretion of cyclic β-1,2-glucan by Rhizobium trifolii TA-1. Appl. Env. Microbiol., 56, 2080‒2086 (1990).
• 24)  A. Amemura, T. Hashimoto, K. Koizumi, and T. Utamura: Occurrence of extracellular (1→2)-β-D-glucans and (1→2)-β-D-gluco-oligosaccharides in Acetobacter. J. Gen. Microbiol., 131, 301‒307 (1985).
• 25)  J.P. Bohin: Osmoregulated periplasmic glucans in Proteobacteria. FEMS Microbiol. Lett., 186, 11‒19 (2000).
• 26)  P. Talaga, B. Fournet, and J.P. Bohin: Periplasmic glucans of Pseudomonas syringae pv. syringae. J. Bacteriol., 176, 6538‒6544 (1994).
• 27)  K.J. Miller, E.P. Kennedy, and V.N. Reinhold: Osmotic adaptation by Gram-negative bacteria: possible role for periplasmic oligosaccharides. Science, 231, 48‒51 (1986).
• 28)  M. Hiraishi, K. Igarashi, S. Kimura, M. Wada, M. Kitaoka, and M. Samejima: Synthesis of highly ordered cellulose II in vitro using cellodextrin phosphorylase. Carbohydr. Res., 344, 2468‒2473 (2009).
• 29)  Y. Yamamoto, D. Kawashima, A. Hashizume, M. Hisamatsu, and N. Isono: Purification and characterization of 1,3-β-D-glucan phosphorylase from Ochromonas danica. Biosci. Biotechnol. Biochem., 77, 1949‒1954 (2013).
• 30)  K. Ogawa, J. Tsurugi, and T. Watanabe: The dependence of the conformation of a (1→3)-β-D-glucan on chain-length in alkaline solution. Carbohydr. Res., 29, 397‒403 (1973).
• 31)  O.H. Lowry and J.A. Lopez: The determination of inorganic phosphate in the presence of labile phosphate esters. J. Biol. Chem., 162, 421‒428 (1946).
• 32)  G. Michal: D-Glucose-1-phosphate in Methods of Enzymatic Analysis, H. Bergmeyer, J. Bergmeyer, and M. Grassl, eds., 3rd Ed., Verlag Chemie, Weinheim, pp. 185‒191 (1984).