- Email: [email protected]
Acceptor recognition of kojibiose phosphorylase from Thermoanaerobacter brockii: Syntheses of glycosyl glycerol and myo-inositol
JOURNAL OF BIOSCIENCE AND BIOENGINEERING Vol. 101, No. 5, 427–433. 2006 DOI: 10.1263/jbb.101.427
© 2006, The Society for Biotechnology, Japan
Acceptor Recognition of Kojibiose Phosphorylase from Thermoanaerobacter brockii: Syntheses of Glycosyl Glycerol and myo-Inositol Takuo Yamamoto,1* Hikaru Watanabe,1 Tomoyuki Nishimoto,1 Hajime Aga,1 Michio Kubota,1 Hiroto Chaen,1 and Shigeharu Fukuda1 Glycoscience Institute, Research Center, Hayashibara Biochemical Laboratories, Inc., 675-1 Fujisaki, Okayama 702-8006, Japan1 Received 16 December 2005/Accepted 20 February 2006
The glucosyl transfer reaction of kojibiose phosphorylase (KP; EC 2.4.1.230) was examined using glycerol or myo-inositol as an acceptor. In the case of glycerol, KP produced two main transfer products: saccharides A and B. The structure of saccharide A was O-α-D-glucopyranosyl-(1→1)glycerol and that of saccharide B was O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→1)glycerol. These results show that KP transferred a glucose residue to the hydroxyl group at position 1 of glycerol. On the other hand, when myo-inositol was used as an acceptor, KP produced four transfer products: saccharides 1–4. The structures of saccharides 1 and 2 were O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→5)-myo-inositol, respectively; those of saccharides 3 and 4 were O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→5)-myo-inositol, respectively. KP transferred a glucose residue to the hydroxyl group at position 1 or 5 of myo-inositol. On the basis of the structures of their glucosyl transfer products, glycerol and myo-inositol were found to have a common structure with three hydroxyl groups corresponding to the hydroxyl group of the glucose molecule at positions 2, 3 and 4. The conformation of these three hydroxyl groups in the structure is equatorial. This structure is the substrate recognition site of KP. It has been suggested that KP strictly recognizes the structures of glycerol and myo-inositol, and catalyzes the transfer reaction of a glucose residue to the hydroxyl group at position 1 in glycerol, and at position 1 or 5 in myo-inositol, corresponding to position 2 in glucose. [Key words: acceptor recognition, kojibiose phosphorylase, glycerol, myo-inositol]
inositol were also acceptors, which suggests that glycerol and myo-inositol have a common structure corresponding to that of the glucose molecule. Cellobiose phosphorylase (CP; EC 2.4.1.20) reversibly phosphorolyzes cellobiose (O-β-D-glucopyranosyl-(1→4)D-glucopyranose) into α-D-glucose 1-phosphate (α-G1P) and D-glucose (5). The acceptor specificity of this enzyme in the glucosyl transfer reaction has been investigated in detail by Kitaoka et al. (6). In their study, it was found that CP strictly recognizes the hydroxyl groups of the glucose molecule at positions β-1, β-3 and β-4, and catalyzes a transfer reaction to the hydroxyl group at position β-4. As such, several heterooligosaccharides have been synthesized on the basis of those specificities (7–11). Sucrose phosphorylase (SP; EC 2.4.1.7) catalyzes the reversible phosphorolysis of sucrose (O-β-D-fructofuranosyl-(2→1)-α-D-glucopyranoside) to α-G1P and fructose (12). Kitao and coworkers analyzed the acceptor specificity of SP using various acceptors with phenolic or alcoholic hydroxyl groups (13–16). Because the regiospecificities of phosphorylases are very strict, it is expected that KP strictly recognizes the structures of substrates for the transfer reaction of a glucose residue.
Kojibiose phosphorylase (KP; EC 2.4.1.230) catalyzes the reversible phosphorolysis of kojibiose (O-α-D-glucopyranosyl-(1→2)-D-glucopyranose) with the inversion of the anomeric configuration as follows. β-D-glucose 1-phosphate (β-G1P) + D-glucose kojibiose + inorganic phosphate (Pi) This enzyme also catalyzes the glucosyl transfer reaction using β-G1P as glucosyl donor to the appropriate acceptor (1). We have already examined the transglucosylations to sucrose, maltose, and the cyclic tetrasaccharide, cyclo-{→6)O-α-D-glucopyranosyl-(1→3)-O-α-D-glucopyranosyl-(1→6)O-α-D-glucopyranosyl-(1→3)-O-α-D-glucopyranosyl-(1→}, and determined the structures of the transfer products (2–4). From the acceptor specificity and structures of the transfer products, KP catalyzed the transfer reaction of a glucose residue to the hydroxyl group of the glucose molecule at position 2. As a result of further investigation of the acceptor specificity in this study, it was found that glycerol and myo* Corresponding author. e-mail: [email protected] phone: +81-(0)86-276-3142 fax: +81-(0)86-276-8670 427
428
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
In this study, we synthesized transfer products using glycerol and myo-inositol as acceptors, and determined their structures. Furthermore, on the basis of the results obtained, we discussed the substrate recognition of KP for acceptors. MATERIALS AND METHODS Materials Reagent-grade glycerol and myo-inositol were purchased from Wako Pure Chemicals (Osaka) and Sigma-Aldrich (St. Louis, MO, USA), respectively. β-G1P disodium salt was prepared in our laboratory as described previously (4). Other chemicals and reagents were of analytical or commercial grade. Enzyme preparation KP was extracted from recombinant Escherichia coli cells in which the KP gene was cloned from Thermoanaerobacter brockii ATCC35047 (17). Assay of KP activity The substrate solution contained 0.1% (w/v) kojibiose and McIlvaine buffer (pH 5.5; Pi concentration, 0.1 M) in a total volume of 2.0 ml. The enzyme solution (0.2 ml) was then added to start the reaction. After the incubation at 60°C for 30 min, the reaction was stopped by boiling for 10 min. The amount of glucose released was measured by the glucose oxidaseperoxidase method (18). One unit of enzyme activity was defined as the amount of enzyme that liberates glucose at 1 µmol/min under the conditions described above. Acceptor specificity assay The substrate solution contained a 111 mM acceptor, 29 mM β-G1P, and 50 mM acetate buffer (pH 5.5) in a total volume of 2.0 ml. The enzyme solution (0.2 ml) was added, and the reaction was started. After incubating this mixture at 60°C for 30 min, the reaction was stopped by boiling for 10 min. The amount of Pi released was measured by the method of Fiske and Subbarow (19). One unit of enzyme activity was defined as the amount of enzyme that liberates Pi at 1 µmol/min under the conditions described above. Preparation of glucosyl glycerol using α-glucosidase A mixture of three glucosyl glycerols, O-α-D-glucopyranosyl-(1→1)glycerol ((2S)-1-α-D-glucosyl glycerol), O-α-D-glucopyranosyl(1→3)-glycerol ((2R)-1-α-D-glucosyl glycerol), and O-α-D-glucopyranosyl-(1→2)-glycerol was synthesized using α-glucosidase (20), and used as the standard for the glucosyl glycerols prepared from KP. A reaction mixture (5 ml) containing 0.14 unit/ml α-glucosidase (TGaseL-Amano; Amano Enzyme, Aichi), 4.07 M glycerol , and 0.15 M maltose in 10 mM sodium acetate buffer (pH 5.0) was incubated at 40°C for 24 h. After heating at 100°C for 10 min, glucoamylase (3.4 units/mmol for maltose) was added to the reaction mixture, which was then incubated at 50°C for 24 h. After the hydrolysis reaction, the solution was heated to 100°C for 10 min to inactivate the enzymes. Furthermore, the resultant reaction mixture was filtered and desalted, and then a sample was analyzed by gas liquid chromatography (GLC). High-performance liquid chromatography (HPLC) The composition and purity of transfer products in samples were determined by HPLC. Samples were filtered using a filter kit, KC prep dura (0.45 µm; Katayama Chemical, Osaka), and deionized using a microacilyzer G0 (Asahi Chemical, Tokyo). HPLC was performed with an LC-10AD pump, an RID-10A refractive index monitor, and a C-R7A data processor (Shimadzu, Kyoto) equipped with an MCI GEL CK04SS column (10 mm i.d. × 200 mm × 2; Mitsubishi Chemical, Tokyo) using water as an eluent at a flow rate of 0.5 ml/min at 80°C, or an ODS-AQ AQ-303 column (4.6 mm i.d. ×250 mm; YMC) using water at a flow rate of 0.5 ml/min at 40°C. Yield of the transfer products Zero to 200 µg of the acceptor was analyzed by HPLC and a calibration curve of the peak area vs the amount of the acceptor was constructed. Then, the amounts of the transfer product and acceptor were confirmed from their peak areas using the calibration curve. The yield of the transfer
product was calculated as follows: (the amount of the transfer product/the amount of the acceptor) × 100 (%). GLC Samples of saccharides were dried in a decompression chamber, and then subjected to trimethylsilylation (21). GLC was performed using GC-14 (Shimadzu) on a DB-5 capillary column (J & W, Folsom, CA, USA) at 160–320°C (5°C/min). The carrier gas was helium at a flow rate of 1.0 ml/ min. Myo-inositol was used as an internal standard. Detection was performed using a flame ionization detector (FID). NMR measurement NMR spectral data were recorded for 1–5% solutions in D2O at 27°C with a JNM-AL300 spectrometer (1H 300.4 MHz, 13C 75.45 MHz: JEOL, Tokyo). Chemical shift was expressed in ppm downfield from the signal of 3-(trimethylsilyl)-1-propane-sulfonic acid sodium salt (TPS) that was used as an internal standard.
RESULTS Acceptor specificity of KP Acceptor specificity was examined using β-G1P as a glucosyl donor and various alcohols as acceptors. As shown in Table 1, KP acted only on glycerol, myo-inositol and D-glucose. The relative activities of KP toward D-glucose, glycerol and myo-inositol were 100%, 0.27% and 6.2%, respectively. KP did not act on other acceptors. Preparation and isolation of glycosyl glycerol A reaction mixture (400 ml) containing KP (34.5 units/mmol for β-G1P), β-G1P (29 mM), and glycerol (217 mM) in 50 mM sodium acetate buffer (pH 5.5) was incubated at 50°C for 72 h. A sample (50 µl) of the reaction mixture was analyzed by HPLC. As shown in Fig. 1, KP produced several transfer products. Two main products were designated saccharides A (HPLC retention time (TR) = 45.1 min) and B (TR = 44.1 min). The yields of saccharides A and B were 5.0% and 7.2%, respectively. After the enzyme reaction was stopped by heating in a boiling water bath for 20 min, the mixture was centrifuged, and then the resultant supernatant was desalted by passing it through ion-exchange resins (50 ml of Diaion SK1B [Mitsubishi Chemical], and 100 ml of Amberlite IRA441S [Japan Organo, Tokyo]). The eluent was concentrated to 60 ml by evaporation at 40°C. Twenty milliliters of the saccharide solution was analyzed by repeated preparative HPLC on a TSK gel Toyopearl HW-40S column (70 mm i.d.×1290 mm; Tosoh, Tokyo). Saccharides A and B were TABLE 1. Acceptor specificity of KP Acceptor Relative activity (%) Glucose 100 Ethanol n.d. Ethyleneglycol n.d. 1-Propanol n.d. Propyleneglycol n.d. 1,3-Propanediol n.d. Glycerol 0.26 1-Butanol n.d. 1,2-Butanediol n.d. myo-Inositol 6.2 The activity of the enzyme was examined using various acceptors under the conditions described in Materials and Methods. Values are shown as a percentage of the activity using glucose as an acceptor taken as 100%. n.d., Not detected (less than 0.001%).
VOL. 101, 2006
TABLE 2.
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
429
13
C NMR chemical shift data for saccharides Aa and Ba
Carbon Saccharide Saccharide Kojibiose Glycerolc atomb A B αc I-1 71.2 71.3 65.2 I-2 74.5 74.5 74.7 I-3 65.2 65.2 65.2 II-1 101.0 98.2 90.8 II-2 73.1 77.9 77.1 II-3 75.7 74.4 73.1 II-4 72.2 72.2 71.1 II-5 74.2 74.0 73.1 II-6 63.2 63.2 62.0 III-1 98.7 96.9 III-2 73.0 72.8 III-3 75.4 74.1 III-4 72.1 71.0 III-5 74.0 72.9 III-6 63.0 61.8 a NMR spectra data were recorded for solutions in D2O at 27°C. Chemical shift is expressed in ppm downfield from the signal of 3-(trimethylsilyl)-1-propane-sulfonic acid sodium salt (TPS), which was used as an internal standard. b Roman numerals indicate the positions of the glycerol and glucose residues in saccharides A and B shown in Fig. 3. c Data taken from Bradbury and Jenkins (23).
separated from each other by elution with water as an eluent at a flow rate of 2.5 ml/min at 25°C. The fractions containing saccharides A and B were separately collected and then evaporated at 40°C. The amounts of purified saccharides A and B were 191.1 mg (purity = 99.0%), and 138.0 mg (purity = 95.7%), respectively. Characterization of glycosyl glycerol Saccharide A To confirm the structure of saccharide A, NMR measurement was carried out. The 13C-NMR spectrum contained 9 signals (Table 2). Because 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide A that was confirmed by the C-1 signal of this residue at 101.0 ppm (δ1H, 4.9 ppm), it was expected that saccharide A has a structure containing one glucose residue attached to glycerol by a (1→1) linkage. Furthermore, the GLC retention time of saccharide A was identical to that of (2S)-1-α-D-glucosyl glycerol (Fig. 2). From these results, saccharide A was deduced to be O-α-D-glucopyranosyl(1→1)-glycerol ((2S)-1-α-D-glucosyl glycerol) (Fig. 3a). Saccharide B The 13C-NMR spectrum of saccharide B contained 15 signals (Table 2). From the 1H-NMR and 13CNMR spectra, it was expected that saccharide B contains saccharide A in its structure (data not shown). Moreover, a downfield shift (4.8 ppm) of the C-2 signal of the glucose residue in saccharide A was observed in the spectra (Table 2), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide B that was confirmed by the C-1 signal of this residue at 98.7 ppm (δ1H, 5.1 ppm). From these results, saccharide B was deduced to be O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl(1→1)-glycerol (kojibiosyl glycerol) (Fig. 3b). Preparation and isolation of glycosyl myo-inositol A reaction mixture (400 ml) containing KP (34.5 units/mmol for β-G1P), β-G1P (29 mM), myo-inositol (111 mM) in 50 mM sodium acetate buffer (pH 5.5) was incubated at 50°C for 72 h. A sample (50 µl) of the reaction mixture was ana-
FIG. 1. HPLC profile of reaction products of KP produced using mixture of β-G1P (as the glucosyl donor) and glycerol (as the acceptor).
FIG. 2. GLC profile of reaction products of α-glucosidase and KP. (a) Glucosyl glycerols produced by α-glucosidase are indicated by line. Peaks 1, 2, and 3 are O-α-D-glucopyranosyl-(1→2)-glycerol, O-α-D-glucopyranosyl-(1→3)-glycerol ((2R)-1-α-D-glucosyl glycerol), and O-α-D-glucopyranosyl-(1→1)-glycerol ((2S)-1-α-D-glucosyl glycerol), respectively (20). Glucosyl glycerol (saccharide A) produced by KP is indicated by a broken line. The retention time of peak 3 corresponds to that of saccharide A. (b) GLC profile of mixture of glucosyl glycerols from α-glucosidase and saccharide A from KP. The retention time of peak 2 does not correspond to that of saccharide A.
FIG. 3. Structures of saccharides A (a) and B (b) produced by KP.
430
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
FIG. 4. HPLC profile of reaction products of KP produced using mixture of β-G1P (as the glucosyl donor) and myo-inositol (as the acceptor).
lyzed by HPLC. As shown in Fig. 4, KP produced four transfer products, saccharides 1 (TR = 51.7 min), 2 (TR = 49.6 min), 3 (TR = 48.1 min), and 4 (TR = 45.0 min). The yields of saccharides 1–4 at a reaction time of 72 h were 20.2%, 5.2%, 8.7% and 1.5%, respectively. After the enzyme reaction was stopped by heating in a boiling water bath for 20 min, the mixture was centrifuged, and then the resultant supernatant was desalted. This desalting was performed as described in Preparation and isolation of glucosyl glycerols. The eluent was concentrated to 100 ml by evaporation at 40°C. Twenty milliliters of the saccharide solution was analyzed repeated preparative HPLC on a TSK gel Toyopearl HW-40S column (70 mm i.d.×1290 mm; Tosoh). Saccharides 1–4 were separated from myo-inositol by elution with water as an eluent at TABLE 3. Carbon atomb
Saccharide 1
Saccharide 3
a flow rate of 2.5 ml/min at 25°C. The fractions containing saccharides 1 and 2, and saccharides 3 and 4 were collected separately, and then evaporated at 40°C. To separate saccharides 1 and 2, column chromatography using an ODS-AQ AQ-303 column was performed. The fractions containing saccharides 1 and 2 were separately collected and then evaporated. Saccharides 3 and 4 were separated using the same column chromatography. The amounts of purified saccharides 1–4 were 20.2 mg (purity = 90.5%), 8.0 mg (purity = 99.4%), 18.6 mg (purity = 99.6%), and 9.2 mg (purity = 94.9%), respectively. Characterization of glycosyl myo-inositol Saccharide 1 To confirm the structure of saccharide 1, NMR measurement was carried out. The 13C-NMR spectrum contained 12 signals (Table 3). 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide 1 that was confirmed by the C-1 signal of this residue at 98.1 ppm (δ1H, 5.1 ppm). Furthermore, because the chemical shift data of O-α-D-galactopyranosyl-(1→1)myo-inositol was identical to that of saccharide 1 (Table 3), it was suggested that the glucose residue in saccharide 1 attached to myo-inositol by a (1→1) linkage. From these results, saccharide 1 was deduced to be O-α-D-glucopyranosyl-(1→1)-myo-inositol (Fig. 5a). Saccharide 2 The 13C-NMR spectrum of saccharide 2 contained 12 signals (Table 3). 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide 2 that was confirmed by the C-1 signal of this residue at 102.0 ppm (δ1H, 5.3 ppm). Because a large downfield shift (7.4 ppm) of the C-5 signal of the myo-inositol was observed (Table 3), saccharide 2 was concluded to be O-α-Dglucopyranosyl-(1→5)-myo-inositol (Fig. 5b). Saccharide 3 The 13C-NMR spectrum contained 18
13
Saccharide 2
C NMR chemical shift data for saccharides 1–4a Saccharide 4
5-O-(α-D1-O-(α-DGalactopyranosyl)- Glucopyranosyl)myo-inositolc myo-inositold 96.0 72.0 68.9 72.7 70.0 72.0 69.8 72.0 71.7 82.9 61.6 73.7 76.3 100.1 68.8 72.6 73.0 73.9 71.6 70.3 74.9 72.8 71.6 61.3
Kojibiose αe
I-1 78.5 78.5 73.8 74.0 I-2 71.0 70.7 74.5 74.4 I-3 75.0 75.0 73.8 74.0 I-4 73.6 73.6 73.6 73.7 I-5 77.0 77.0 84.4 83.9 I-6 73.7 73.7 75.6 75.5 II-1 98.1 99.6 102.0 99.3 90.8 II-2 74.1 78.4 74.5 78.4 77.1 II-3 75.6 73.7 75.6 73.9 73.1 II-4 72.2 71.9 72.1 72.0 71.1 II-5 74.5 74.5 74.5 74.5 73.1 II-6 63.1 63.0 63.0 63.0 62.0 III-1 95.2 98.9 96.9 III-2 73.4 73.5 72.8 III-3 75.5 75.5 74.1 III-4 72.2 72.0 71.0 III-5 74.3 74.5 72.9 III-6 63.1 62.8 61.8 a NMR spectra data were recorded for solutions in D2O at 27°C. Chemical shift is expressed in ppm downfield from silyl)-1-propane-sulfonic acid sodium salt (TPS), which was used as an internal standard. b Roman numerals indicate the positions of the myo-inositol and glucose residues in saccharides 1–4 shown in Fig. 5. c Data taken from Schweizer and Horman (24). d Data taken from Sato et al. (25). e Data taken from Bradbury and Jenkins (23). f Measurement in this study.
α-DGlucosee
myoInositolf 75.0 (72.9c) 74.8 (72.7c) 75.0 (72.9c) 73.8 (71.6c) 77.0 (74.8c) 73.8 (71.6c)
92.9 72.5 73.8 70.6 72.3 61.6
the signal of 3-(trimeth-
VOL. 101, 2006
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
431
FIG. 5. Structures of saccharides 1 (a), 2 (b), 3 (c) and 4 (d) produced by KP.
signals (Table 3). From the 1H-NMR and 13C-NMR spectra, it was expected that saccharide 3 contains saccharide 1 in its structure (data not shown). A downfield shift (4.3 ppm) of the C-2 signal of the glucose residue in saccharide 1 was observed in the spectra (Table 3), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide 3 that was confirmed by the C-1 signal of this residue at 95.2 ppm (δ1H, 5.3 ppm). From these results, saccharide 3 was deduced to be O-α-D-glucopyranosyl-(1→2)O-α-D-glucopyranosyl-(1→1)-myo-inositol (kojibiosyl(1→1)-myo-inositol) (Fig. 5c). Saccharide 4 The 13C-NMR spectrum contained 18 signals (Table 3). From the 1H-NMR and 13C-NMR spectra, it was expected that saccharide 4 contains saccharide 2 in its structure (data not shown). A downfield shift (3.9 ppm) of the C-2 signal of the glucose residue in saccharide 2 was observed in the spectra (Table 3), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide 4 that was confirmed by the C-1 signal of this residue at 98.9 ppm (δ1H, 5.5 ppm). From these results, saccharide 4 was concluded to be O-α-D-glucopyranosyl-(1→2)O-α-D-glucopyranosyl-(1→5)-myo-inositol (kojibiosyl(1→5)-myo-inositol) (Fig. 5d). DISCUSSION KP catalyzes the glucosyl transfer reaction using β-G1P as a glucosyl donor to the appropriate acceptor (1). From its acceptor specificity, this enzyme catalyzes the transfer reaction of a glucose residue to the hydroxyl group of the glucose molecule at position 2. Further investigation of the acceptor specificity in this study showed that glycerol and myo-inositol are also acceptors (Table 1). When glycerol was used as an acceptor, several glucosyl products were synthesized (Fig. 1). To determine the structure of these products, we prepared three types of glucosyl glycerol, O-α-D-glu-
copyranosyl-(1→1)-glycerol, O-α-D-glucopyranosyl-(1→3)glycerol, and O-α-D-glucopyranosyl-(1→2)-glycerol using α-glucosidase, because it is difficult to distinguish between the structures of O-α-D-glucopyranosyl-(1→1)-glycerol and O-α-D-glucopyranosyl-(1→3)-glycerol by NMR measurement without standards. Takenaka et al. chemically synthesized three types of glucosyl glycerol, and identified the structures of these glucosyl glycerols synthesized by α-glucosidase using GLC (20). In this study, we used these glucosyl glycerols as standards. The two main products from them were O-α-D-glucopyranosyl-(1→1)-glycerol and kojibiosyl glycerol. Because KP also catalyzes the extension reaction of the side chain of O-α-D-glucopyranosyl-(1→1)glycerol, the minor products (TR =41.0 min and TR =39.9 min) are kojitriosyl and kojitetraosyl glycerol. When glucose was used as an acceptor, kojiheptaose was synthesized by the extension reaction from the side chain of kojibiose (2). From this observation, kojipentaosyl, kojihexaosyl and kojiheptaosyl glycerols are also expected to be synthesized under the conditions used in this study. When myo-inositol was used as an acceptor, O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→5)-myo-inositol were synthesized by the transfer reaction of KP. Kojibiosyl-(1→1)- and kojibiosyl-(1→5)-myo-inositol were also synthesized by the extension reaction by KP. On the basis of the structures of these transfer products obtained, we discussed the correlation of the structures of glucose, glycerol and myo-inositol. As shown in Table 1, KP acted only on glycerol, myo-inositol and D-glucose. Glycerol, having only three carbons, is the smallest acceptor of KP. KP transferred a glucose residue to the hydroxyl groups at position 1 in glycerol, and at position 1 or 5 in myo-inositol. Interestingly, in glycerol and myo-inositol molecules, there is a common structure with three hydroxyl groups corresponding to the hydroxyl groups of the glucose molecule at positions 2, 3 and 4. The conformation of these three hydroxyl groups in the structure is
432
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
cosidic linkage. Namely, oligosaccharides with specific glycosyl linkages can be synthesized using phosphorylases. The phosphorylases that have been reported to date are classified into fourteen kinds by the type of substrate phosphorolyzed (22), and only KP can specifically synthesize an α-1,2 glycosidic linkage. Therefore, novel oligosaccharides having an α-1,2 glycosidic linkage can be synthesized by KP. Furthermore, we demonstrated the substrate recognition of KP in this study, which will enable the prediction of the structure of glucosyl transfer products using the location or the number of the minimum structure in acceptors. ACKNOWLEDGMENTS FIG. 6. Stereochemical comparison among three acceptors for KP. Each circle indicates the structure with three hydroxyl groups that is recognized by KP. myo-Inositol-(b) represents rotated myo-inositol-(a). Arrows indicate the positions of transglucosylation.
equatorial. This common structure is the minimum structure for the substrate recognition of KP. As shown in Fig. 6, glycerol has one minimum structure for the substrate recognition of KP: the hydroxyl groups at positions 1, 2 and 3. On the other hand, myo-inositol has two minimum structures for the substrate recognition of KP: the hydroxyl groups at positions 1, 6 and 5, and at positions 5, 4 and 3. The fact that glycerol and myo-inositol have these structures strongly suggests that KP strictly recognizes this structure, and therefore catalyzes the transfer reaction of a glucose residue to the hydroxyl group at position 1 in glycerol, and at position 1 or 5 in myo-inositol corresponding to position 2 in glucose. Using L-sorbose as an acceptor, Chaen et al. determined the structure of glucosyl-L-sorbose (3). The structure of this saccharide was concluded to be O-α-D-glucopyranosyl-(1→5)-α-L-sorbopyranose, because the configuration of the hydroxyl groups at positions 5, 4, and 3 of L-sorbopyranose was the same as those at positions 2, 3, and 4 of D-glucopyranose. In this study, the percentage of glucosyl glycerol (saccharide A, 5.2%) in the reaction mixture was lower than those of glucosyl myo-inositols (saccharides 1 and 2, 25.4%), and the relative activity of KP toward glycerol was 0.04-fold higher than that toward myo-inositol (Table 1). It is assumed that the minimum structure in myo-inositol is rigid, because it is a pyranose-like ring. In contrast to myo-inositol, glycerol does not have a cyclic structure; therefore, the minimum structure in glycerol is more flexible, so KP has more difficulty in recognizing the minimum structure of glycerol than it has recognizing that of myo-inositol. Moreover, when myo-inositol was used as an acceptor, the amount of O-α-Dglucopyranosyl-(1→1)-myo-inositol was 2.5-fold higher than that of O-α-D-glucopyranosyl-(1→5)-myo-inositol. This difference indicates that the structure adjacent to the minimum structure influences the catalytic reaction of KP. Kitaoka et al. have investigated the acceptor specificity and substrate recognition of CP in detail (6). Using this enzyme in the transglucosyl manner, several heterooligosaccharides have been synthesized (7–11). The regiospecificities of phosphorylases are very strict and these enzymes phosphorolyze or synthesize only their specific type of gly-
This work was supported by a grant from the Research and Development Program for New Bio-industry Initiatives of the Biooriented Technology Research Advancement Institution (BRAIN).
REFERENCES 1. Chaen, H., Yamamoto, T., Nishimoto, T., Nakada, T., Fukuda, S., Sugimoto, T., Kurimoto, M., and Tsujisaka, Y.: Purification and characterization of a novel phosphorylase, kojibiose phosphorylase, from Thermoanaerobium brockii. J. Appl. Glycosci., 46, 423–429 (1999). 2. Chaen, H., Nishimoto, T., Nakada, T., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of kojioligosaccharides using kojibiose phosphorylase. J. Biosci. Bioeng., 92, 173–176 (2001). 3. Chaen, H., Nishimoto, T., Nakada, T., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of novel oligosaccharides from L-sorbose, maltose, and sucrose using kojibiose phosphorylase. J. Biosci. Bioeng., 92, 177– 182 (2001). 4. Watanabe, H., Higashiyama, T., Aga, H., Nishimoto, T., Kubota, M., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of a 2-O-α-D-glucopyranosyl cyclic tetrasaccharide by kojibiose phosphorylase. Carbohydr. Res., 340, 449–454 (2005). 5. Kitaoka, M., Sasaki, T., and Taniguchi, H.: Phosphorolytic reaction of Cellvibrio gilvus cellobiose phosphorylase. Biosci. Biotechnol. Biochem., 56, 652–655 (1992). 6. Kitaoka, M., Sasaki, T., and Taniguchi, H.: Synthetic reaction of Cellvibrio gilvus cellobiose phosphorylase. J. Biochem., 112, 40–44 (1992). 7. Kitaoka, M., Taniguchi, H., and Sasaki, T.: Production of glucosyl-xylose using Cellvibrio gilvus cells and its properties. Appl. Microbiol. Biotechnol., 34, 178–182 (1990). 8. Tariq, M. A., Hayashi, K., Tokuyasu, K., and Nagata, T.: Synthesis and structural analysis of disaccharides of 4-O-β-Dglucopyranosyl-D-glucosamine and 4-O-β-D-glucopyranosyl2-deoxy-D-glucose. Carbohydr. Res., 275, 67–72 (1995). 9. Tariq, M. A. and Hayashi, K.: Synthesis of three hetero disaccharides, 4-O-β-D-glucopyranosyl-6-deoxy-D-glucose, 4-O-β-D-glucopyranosyl-D-mannosamine, and 4-O-β-D-glucopyranosyl-D-mannose, and confirmation of their structures by C-13 NMR and MS. Biochem. Biophys. Res. Commun., 214, 568–575 (1995). 10. Percy, A., Ono, H., Watt, D., and Hayashi, K.: Synthesis of β-D-glucopyranosyl-(1→4)-D-arabinose, β-D-glucopyranosyl-(1→4)-L-fucose and β-D-glucopyranosyl-(1→4)-D-altrose catalyzed by cellobiose phosphorylase from Cellvibrio gilvus. Carbohydr. Res., 305, 543–548 (1997). 11. Percy, A., Ono, H., and Hayashi, K.: Acceptor specificity of cellobiose phosphorylase from Cellvibrio gilvus: synthesis of three branched trisaccharides. Carbohydr. Res., 308, 423–429
VOL. 101, 2006
(1998). 12. Mieyal, J. J. and Abeles, R. H.: Disaccharide phosphorylases, p. 515–532. In Boyer, P. D. (ed.), The enzymes, vol. 7, 3rd ed. Academic Press, New York (1972). 13. Kitao, S. and Sekine, H.: Transglucosylation catalyzed by sucrose phosphorylase from Leuconostoc mesenteroides and production of glucosyl-xylitol. Biosci. Biotechnol. Biochem., 56, 2011–2014 (1992). 14. Kitao, S., Ariga, T., Matsudo, T., and Sekine, H.: The syntheses of catechin-glucosides by transglycosylation with Leuconostoc mesenteroides sucrose phosphorylase. Biosci. Biotechnol. Biochem., 57, 2010–2015 (1993). 15. Kitao, S. and Sekine, H.: α-D-Glucosyl transfer to phenolic compounds by sucrose phosphorylase from Leuconostoc mesenteroides and production of α-arbutin. Biosci. Biotechnol. Biochem., 58, 38–42 (1994). 16. Kitao, S., Matsudo, T., Sasaki, T., Koga, T., and Kawamura, M.: Enzymatic synthesis of stable, odorless, and powdered furanone glucosides by sucrose phosphorylase. Biosci. Biotechnol. Biochem., 64, 134–141 (2000). 17. Yamamoto, T., Maruta, K., Mukai, K., Yamashita, H., Nishimoto, T., Kubota, M., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Cloning and sequencing of kojibiose phosphorylase from Thermoanaerobacter brockii ATCC35047. J. Biosci. Bioeng., 98, 99–106 (2004). 18. Lloyd, J. B. and Whelan, W. J.: An improved method for
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
19. 20. 21.
22. 23. 24. 25.
433
enzymatic determination of glucose in the presence of maltose. Anal. Biochem., 30, 467–470 (1969). Fiske, C. H. and Subbarow, A.: The colorimetric determination of phosphorus. J. Biol. Chem., 66, 375–400 (1925). Takenaka, F., Uchiyama, H., and Imamura, T.: Identification of α-glucosylglycerol in sake. Biosci. Biotechnol. Biochem., 64, 378–385 (2000). Hagiwara, K., Tsuda, A., Ichikawa, T., Kanaya, K., Yoneyama, S., Sudo, H., and Iwao, H.: Quantative analyses of fructooligosaccharides by gas liquid chromatography. Shokuhin Eiseigaku Zasshi, 24, 558–562 (1983). (in Japanese) Kitaoka, M. and Hayashi, K.: Carbohydrate-processing phosphorolytic enzymes. Trends Glycosci. Glycotechnol., 14, 35–50 (2002). Bradbury, J. H. and Jenkins, G. A.: Determination of the structures of trisaccharides by 13C-n.m.r. spectroscopy. Carbohydr. Res., 126, 125–156 (1984). Schweizer, T. F. and Horman, I.: Purification and structure determination of three α-D-galactopyranosylcyclitols from soya bean. Carbohydr. Res., 95, 61–71 (1981). Sato, M., Nakamura, K., Nagano, H., Yagi, Y., and Koizumi, K.: Synthesis of glucosyl-inositol using a CGTase, isolation and characterization of the positional isomers, and assimilation profiles for intestinal bacteria. Biotechnol. Lett., 14, 659–664 (1992).
© 2006, The Society for Biotechnology, Japan
Acceptor Recognition of Kojibiose Phosphorylase from Thermoanaerobacter brockii: Syntheses of Glycosyl Glycerol and myo-Inositol Takuo Yamamoto,1* Hikaru Watanabe,1 Tomoyuki Nishimoto,1 Hajime Aga,1 Michio Kubota,1 Hiroto Chaen,1 and Shigeharu Fukuda1 Glycoscience Institute, Research Center, Hayashibara Biochemical Laboratories, Inc., 675-1 Fujisaki, Okayama 702-8006, Japan1 Received 16 December 2005/Accepted 20 February 2006
The glucosyl transfer reaction of kojibiose phosphorylase (KP; EC 2.4.1.230) was examined using glycerol or myo-inositol as an acceptor. In the case of glycerol, KP produced two main transfer products: saccharides A and B. The structure of saccharide A was O-α-D-glucopyranosyl-(1→1)glycerol and that of saccharide B was O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→1)glycerol. These results show that KP transferred a glucose residue to the hydroxyl group at position 1 of glycerol. On the other hand, when myo-inositol was used as an acceptor, KP produced four transfer products: saccharides 1–4. The structures of saccharides 1 and 2 were O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→5)-myo-inositol, respectively; those of saccharides 3 and 4 were O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl-(1→5)-myo-inositol, respectively. KP transferred a glucose residue to the hydroxyl group at position 1 or 5 of myo-inositol. On the basis of the structures of their glucosyl transfer products, glycerol and myo-inositol were found to have a common structure with three hydroxyl groups corresponding to the hydroxyl group of the glucose molecule at positions 2, 3 and 4. The conformation of these three hydroxyl groups in the structure is equatorial. This structure is the substrate recognition site of KP. It has been suggested that KP strictly recognizes the structures of glycerol and myo-inositol, and catalyzes the transfer reaction of a glucose residue to the hydroxyl group at position 1 in glycerol, and at position 1 or 5 in myo-inositol, corresponding to position 2 in glucose. [Key words: acceptor recognition, kojibiose phosphorylase, glycerol, myo-inositol]
inositol were also acceptors, which suggests that glycerol and myo-inositol have a common structure corresponding to that of the glucose molecule. Cellobiose phosphorylase (CP; EC 2.4.1.20) reversibly phosphorolyzes cellobiose (O-β-D-glucopyranosyl-(1→4)D-glucopyranose) into α-D-glucose 1-phosphate (α-G1P) and D-glucose (5). The acceptor specificity of this enzyme in the glucosyl transfer reaction has been investigated in detail by Kitaoka et al. (6). In their study, it was found that CP strictly recognizes the hydroxyl groups of the glucose molecule at positions β-1, β-3 and β-4, and catalyzes a transfer reaction to the hydroxyl group at position β-4. As such, several heterooligosaccharides have been synthesized on the basis of those specificities (7–11). Sucrose phosphorylase (SP; EC 2.4.1.7) catalyzes the reversible phosphorolysis of sucrose (O-β-D-fructofuranosyl-(2→1)-α-D-glucopyranoside) to α-G1P and fructose (12). Kitao and coworkers analyzed the acceptor specificity of SP using various acceptors with phenolic or alcoholic hydroxyl groups (13–16). Because the regiospecificities of phosphorylases are very strict, it is expected that KP strictly recognizes the structures of substrates for the transfer reaction of a glucose residue.
Kojibiose phosphorylase (KP; EC 2.4.1.230) catalyzes the reversible phosphorolysis of kojibiose (O-α-D-glucopyranosyl-(1→2)-D-glucopyranose) with the inversion of the anomeric configuration as follows. β-D-glucose 1-phosphate (β-G1P) + D-glucose kojibiose + inorganic phosphate (Pi) This enzyme also catalyzes the glucosyl transfer reaction using β-G1P as glucosyl donor to the appropriate acceptor (1). We have already examined the transglucosylations to sucrose, maltose, and the cyclic tetrasaccharide, cyclo-{→6)O-α-D-glucopyranosyl-(1→3)-O-α-D-glucopyranosyl-(1→6)O-α-D-glucopyranosyl-(1→3)-O-α-D-glucopyranosyl-(1→}, and determined the structures of the transfer products (2–4). From the acceptor specificity and structures of the transfer products, KP catalyzed the transfer reaction of a glucose residue to the hydroxyl group of the glucose molecule at position 2. As a result of further investigation of the acceptor specificity in this study, it was found that glycerol and myo* Corresponding author. e-mail: [email protected] phone: +81-(0)86-276-3142 fax: +81-(0)86-276-8670 427
428
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
In this study, we synthesized transfer products using glycerol and myo-inositol as acceptors, and determined their structures. Furthermore, on the basis of the results obtained, we discussed the substrate recognition of KP for acceptors. MATERIALS AND METHODS Materials Reagent-grade glycerol and myo-inositol were purchased from Wako Pure Chemicals (Osaka) and Sigma-Aldrich (St. Louis, MO, USA), respectively. β-G1P disodium salt was prepared in our laboratory as described previously (4). Other chemicals and reagents were of analytical or commercial grade. Enzyme preparation KP was extracted from recombinant Escherichia coli cells in which the KP gene was cloned from Thermoanaerobacter brockii ATCC35047 (17). Assay of KP activity The substrate solution contained 0.1% (w/v) kojibiose and McIlvaine buffer (pH 5.5; Pi concentration, 0.1 M) in a total volume of 2.0 ml. The enzyme solution (0.2 ml) was then added to start the reaction. After the incubation at 60°C for 30 min, the reaction was stopped by boiling for 10 min. The amount of glucose released was measured by the glucose oxidaseperoxidase method (18). One unit of enzyme activity was defined as the amount of enzyme that liberates glucose at 1 µmol/min under the conditions described above. Acceptor specificity assay The substrate solution contained a 111 mM acceptor, 29 mM β-G1P, and 50 mM acetate buffer (pH 5.5) in a total volume of 2.0 ml. The enzyme solution (0.2 ml) was added, and the reaction was started. After incubating this mixture at 60°C for 30 min, the reaction was stopped by boiling for 10 min. The amount of Pi released was measured by the method of Fiske and Subbarow (19). One unit of enzyme activity was defined as the amount of enzyme that liberates Pi at 1 µmol/min under the conditions described above. Preparation of glucosyl glycerol using α-glucosidase A mixture of three glucosyl glycerols, O-α-D-glucopyranosyl-(1→1)glycerol ((2S)-1-α-D-glucosyl glycerol), O-α-D-glucopyranosyl(1→3)-glycerol ((2R)-1-α-D-glucosyl glycerol), and O-α-D-glucopyranosyl-(1→2)-glycerol was synthesized using α-glucosidase (20), and used as the standard for the glucosyl glycerols prepared from KP. A reaction mixture (5 ml) containing 0.14 unit/ml α-glucosidase (TGaseL-Amano; Amano Enzyme, Aichi), 4.07 M glycerol , and 0.15 M maltose in 10 mM sodium acetate buffer (pH 5.0) was incubated at 40°C for 24 h. After heating at 100°C for 10 min, glucoamylase (3.4 units/mmol for maltose) was added to the reaction mixture, which was then incubated at 50°C for 24 h. After the hydrolysis reaction, the solution was heated to 100°C for 10 min to inactivate the enzymes. Furthermore, the resultant reaction mixture was filtered and desalted, and then a sample was analyzed by gas liquid chromatography (GLC). High-performance liquid chromatography (HPLC) The composition and purity of transfer products in samples were determined by HPLC. Samples were filtered using a filter kit, KC prep dura (0.45 µm; Katayama Chemical, Osaka), and deionized using a microacilyzer G0 (Asahi Chemical, Tokyo). HPLC was performed with an LC-10AD pump, an RID-10A refractive index monitor, and a C-R7A data processor (Shimadzu, Kyoto) equipped with an MCI GEL CK04SS column (10 mm i.d. × 200 mm × 2; Mitsubishi Chemical, Tokyo) using water as an eluent at a flow rate of 0.5 ml/min at 80°C, or an ODS-AQ AQ-303 column (4.6 mm i.d. ×250 mm; YMC) using water at a flow rate of 0.5 ml/min at 40°C. Yield of the transfer products Zero to 200 µg of the acceptor was analyzed by HPLC and a calibration curve of the peak area vs the amount of the acceptor was constructed. Then, the amounts of the transfer product and acceptor were confirmed from their peak areas using the calibration curve. The yield of the transfer
product was calculated as follows: (the amount of the transfer product/the amount of the acceptor) × 100 (%). GLC Samples of saccharides were dried in a decompression chamber, and then subjected to trimethylsilylation (21). GLC was performed using GC-14 (Shimadzu) on a DB-5 capillary column (J & W, Folsom, CA, USA) at 160–320°C (5°C/min). The carrier gas was helium at a flow rate of 1.0 ml/ min. Myo-inositol was used as an internal standard. Detection was performed using a flame ionization detector (FID). NMR measurement NMR spectral data were recorded for 1–5% solutions in D2O at 27°C with a JNM-AL300 spectrometer (1H 300.4 MHz, 13C 75.45 MHz: JEOL, Tokyo). Chemical shift was expressed in ppm downfield from the signal of 3-(trimethylsilyl)-1-propane-sulfonic acid sodium salt (TPS) that was used as an internal standard.
RESULTS Acceptor specificity of KP Acceptor specificity was examined using β-G1P as a glucosyl donor and various alcohols as acceptors. As shown in Table 1, KP acted only on glycerol, myo-inositol and D-glucose. The relative activities of KP toward D-glucose, glycerol and myo-inositol were 100%, 0.27% and 6.2%, respectively. KP did not act on other acceptors. Preparation and isolation of glycosyl glycerol A reaction mixture (400 ml) containing KP (34.5 units/mmol for β-G1P), β-G1P (29 mM), and glycerol (217 mM) in 50 mM sodium acetate buffer (pH 5.5) was incubated at 50°C for 72 h. A sample (50 µl) of the reaction mixture was analyzed by HPLC. As shown in Fig. 1, KP produced several transfer products. Two main products were designated saccharides A (HPLC retention time (TR) = 45.1 min) and B (TR = 44.1 min). The yields of saccharides A and B were 5.0% and 7.2%, respectively. After the enzyme reaction was stopped by heating in a boiling water bath for 20 min, the mixture was centrifuged, and then the resultant supernatant was desalted by passing it through ion-exchange resins (50 ml of Diaion SK1B [Mitsubishi Chemical], and 100 ml of Amberlite IRA441S [Japan Organo, Tokyo]). The eluent was concentrated to 60 ml by evaporation at 40°C. Twenty milliliters of the saccharide solution was analyzed by repeated preparative HPLC on a TSK gel Toyopearl HW-40S column (70 mm i.d.×1290 mm; Tosoh, Tokyo). Saccharides A and B were TABLE 1. Acceptor specificity of KP Acceptor Relative activity (%) Glucose 100 Ethanol n.d. Ethyleneglycol n.d. 1-Propanol n.d. Propyleneglycol n.d. 1,3-Propanediol n.d. Glycerol 0.26 1-Butanol n.d. 1,2-Butanediol n.d. myo-Inositol 6.2 The activity of the enzyme was examined using various acceptors under the conditions described in Materials and Methods. Values are shown as a percentage of the activity using glucose as an acceptor taken as 100%. n.d., Not detected (less than 0.001%).
VOL. 101, 2006
TABLE 2.
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
429
13
C NMR chemical shift data for saccharides Aa and Ba
Carbon Saccharide Saccharide Kojibiose Glycerolc atomb A B αc I-1 71.2 71.3 65.2 I-2 74.5 74.5 74.7 I-3 65.2 65.2 65.2 II-1 101.0 98.2 90.8 II-2 73.1 77.9 77.1 II-3 75.7 74.4 73.1 II-4 72.2 72.2 71.1 II-5 74.2 74.0 73.1 II-6 63.2 63.2 62.0 III-1 98.7 96.9 III-2 73.0 72.8 III-3 75.4 74.1 III-4 72.1 71.0 III-5 74.0 72.9 III-6 63.0 61.8 a NMR spectra data were recorded for solutions in D2O at 27°C. Chemical shift is expressed in ppm downfield from the signal of 3-(trimethylsilyl)-1-propane-sulfonic acid sodium salt (TPS), which was used as an internal standard. b Roman numerals indicate the positions of the glycerol and glucose residues in saccharides A and B shown in Fig. 3. c Data taken from Bradbury and Jenkins (23).
separated from each other by elution with water as an eluent at a flow rate of 2.5 ml/min at 25°C. The fractions containing saccharides A and B were separately collected and then evaporated at 40°C. The amounts of purified saccharides A and B were 191.1 mg (purity = 99.0%), and 138.0 mg (purity = 95.7%), respectively. Characterization of glycosyl glycerol Saccharide A To confirm the structure of saccharide A, NMR measurement was carried out. The 13C-NMR spectrum contained 9 signals (Table 2). Because 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide A that was confirmed by the C-1 signal of this residue at 101.0 ppm (δ1H, 4.9 ppm), it was expected that saccharide A has a structure containing one glucose residue attached to glycerol by a (1→1) linkage. Furthermore, the GLC retention time of saccharide A was identical to that of (2S)-1-α-D-glucosyl glycerol (Fig. 2). From these results, saccharide A was deduced to be O-α-D-glucopyranosyl(1→1)-glycerol ((2S)-1-α-D-glucosyl glycerol) (Fig. 3a). Saccharide B The 13C-NMR spectrum of saccharide B contained 15 signals (Table 2). From the 1H-NMR and 13CNMR spectra, it was expected that saccharide B contains saccharide A in its structure (data not shown). Moreover, a downfield shift (4.8 ppm) of the C-2 signal of the glucose residue in saccharide A was observed in the spectra (Table 2), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide B that was confirmed by the C-1 signal of this residue at 98.7 ppm (δ1H, 5.1 ppm). From these results, saccharide B was deduced to be O-α-D-glucopyranosyl-(1→2)-O-α-D-glucopyranosyl(1→1)-glycerol (kojibiosyl glycerol) (Fig. 3b). Preparation and isolation of glycosyl myo-inositol A reaction mixture (400 ml) containing KP (34.5 units/mmol for β-G1P), β-G1P (29 mM), myo-inositol (111 mM) in 50 mM sodium acetate buffer (pH 5.5) was incubated at 50°C for 72 h. A sample (50 µl) of the reaction mixture was ana-
FIG. 1. HPLC profile of reaction products of KP produced using mixture of β-G1P (as the glucosyl donor) and glycerol (as the acceptor).
FIG. 2. GLC profile of reaction products of α-glucosidase and KP. (a) Glucosyl glycerols produced by α-glucosidase are indicated by line. Peaks 1, 2, and 3 are O-α-D-glucopyranosyl-(1→2)-glycerol, O-α-D-glucopyranosyl-(1→3)-glycerol ((2R)-1-α-D-glucosyl glycerol), and O-α-D-glucopyranosyl-(1→1)-glycerol ((2S)-1-α-D-glucosyl glycerol), respectively (20). Glucosyl glycerol (saccharide A) produced by KP is indicated by a broken line. The retention time of peak 3 corresponds to that of saccharide A. (b) GLC profile of mixture of glucosyl glycerols from α-glucosidase and saccharide A from KP. The retention time of peak 2 does not correspond to that of saccharide A.
FIG. 3. Structures of saccharides A (a) and B (b) produced by KP.
430
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
FIG. 4. HPLC profile of reaction products of KP produced using mixture of β-G1P (as the glucosyl donor) and myo-inositol (as the acceptor).
lyzed by HPLC. As shown in Fig. 4, KP produced four transfer products, saccharides 1 (TR = 51.7 min), 2 (TR = 49.6 min), 3 (TR = 48.1 min), and 4 (TR = 45.0 min). The yields of saccharides 1–4 at a reaction time of 72 h were 20.2%, 5.2%, 8.7% and 1.5%, respectively. After the enzyme reaction was stopped by heating in a boiling water bath for 20 min, the mixture was centrifuged, and then the resultant supernatant was desalted. This desalting was performed as described in Preparation and isolation of glucosyl glycerols. The eluent was concentrated to 100 ml by evaporation at 40°C. Twenty milliliters of the saccharide solution was analyzed repeated preparative HPLC on a TSK gel Toyopearl HW-40S column (70 mm i.d.×1290 mm; Tosoh). Saccharides 1–4 were separated from myo-inositol by elution with water as an eluent at TABLE 3. Carbon atomb
Saccharide 1
Saccharide 3
a flow rate of 2.5 ml/min at 25°C. The fractions containing saccharides 1 and 2, and saccharides 3 and 4 were collected separately, and then evaporated at 40°C. To separate saccharides 1 and 2, column chromatography using an ODS-AQ AQ-303 column was performed. The fractions containing saccharides 1 and 2 were separately collected and then evaporated. Saccharides 3 and 4 were separated using the same column chromatography. The amounts of purified saccharides 1–4 were 20.2 mg (purity = 90.5%), 8.0 mg (purity = 99.4%), 18.6 mg (purity = 99.6%), and 9.2 mg (purity = 94.9%), respectively. Characterization of glycosyl myo-inositol Saccharide 1 To confirm the structure of saccharide 1, NMR measurement was carried out. The 13C-NMR spectrum contained 12 signals (Table 3). 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide 1 that was confirmed by the C-1 signal of this residue at 98.1 ppm (δ1H, 5.1 ppm). Furthermore, because the chemical shift data of O-α-D-galactopyranosyl-(1→1)myo-inositol was identical to that of saccharide 1 (Table 3), it was suggested that the glucose residue in saccharide 1 attached to myo-inositol by a (1→1) linkage. From these results, saccharide 1 was deduced to be O-α-D-glucopyranosyl-(1→1)-myo-inositol (Fig. 5a). Saccharide 2 The 13C-NMR spectrum of saccharide 2 contained 12 signals (Table 3). 1H-1H and 1H-13C COSYs showed the α-configuration of the glucose residue in saccharide 2 that was confirmed by the C-1 signal of this residue at 102.0 ppm (δ1H, 5.3 ppm). Because a large downfield shift (7.4 ppm) of the C-5 signal of the myo-inositol was observed (Table 3), saccharide 2 was concluded to be O-α-Dglucopyranosyl-(1→5)-myo-inositol (Fig. 5b). Saccharide 3 The 13C-NMR spectrum contained 18
13
Saccharide 2
C NMR chemical shift data for saccharides 1–4a Saccharide 4
5-O-(α-D1-O-(α-DGalactopyranosyl)- Glucopyranosyl)myo-inositolc myo-inositold 96.0 72.0 68.9 72.7 70.0 72.0 69.8 72.0 71.7 82.9 61.6 73.7 76.3 100.1 68.8 72.6 73.0 73.9 71.6 70.3 74.9 72.8 71.6 61.3
Kojibiose αe
I-1 78.5 78.5 73.8 74.0 I-2 71.0 70.7 74.5 74.4 I-3 75.0 75.0 73.8 74.0 I-4 73.6 73.6 73.6 73.7 I-5 77.0 77.0 84.4 83.9 I-6 73.7 73.7 75.6 75.5 II-1 98.1 99.6 102.0 99.3 90.8 II-2 74.1 78.4 74.5 78.4 77.1 II-3 75.6 73.7 75.6 73.9 73.1 II-4 72.2 71.9 72.1 72.0 71.1 II-5 74.5 74.5 74.5 74.5 73.1 II-6 63.1 63.0 63.0 63.0 62.0 III-1 95.2 98.9 96.9 III-2 73.4 73.5 72.8 III-3 75.5 75.5 74.1 III-4 72.2 72.0 71.0 III-5 74.3 74.5 72.9 III-6 63.1 62.8 61.8 a NMR spectra data were recorded for solutions in D2O at 27°C. Chemical shift is expressed in ppm downfield from silyl)-1-propane-sulfonic acid sodium salt (TPS), which was used as an internal standard. b Roman numerals indicate the positions of the myo-inositol and glucose residues in saccharides 1–4 shown in Fig. 5. c Data taken from Schweizer and Horman (24). d Data taken from Sato et al. (25). e Data taken from Bradbury and Jenkins (23). f Measurement in this study.
α-DGlucosee
myoInositolf 75.0 (72.9c) 74.8 (72.7c) 75.0 (72.9c) 73.8 (71.6c) 77.0 (74.8c) 73.8 (71.6c)
92.9 72.5 73.8 70.6 72.3 61.6
the signal of 3-(trimeth-
VOL. 101, 2006
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
431
FIG. 5. Structures of saccharides 1 (a), 2 (b), 3 (c) and 4 (d) produced by KP.
signals (Table 3). From the 1H-NMR and 13C-NMR spectra, it was expected that saccharide 3 contains saccharide 1 in its structure (data not shown). A downfield shift (4.3 ppm) of the C-2 signal of the glucose residue in saccharide 1 was observed in the spectra (Table 3), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide 3 that was confirmed by the C-1 signal of this residue at 95.2 ppm (δ1H, 5.3 ppm). From these results, saccharide 3 was deduced to be O-α-D-glucopyranosyl-(1→2)O-α-D-glucopyranosyl-(1→1)-myo-inositol (kojibiosyl(1→1)-myo-inositol) (Fig. 5c). Saccharide 4 The 13C-NMR spectrum contained 18 signals (Table 3). From the 1H-NMR and 13C-NMR spectra, it was expected that saccharide 4 contains saccharide 2 in its structure (data not shown). A downfield shift (3.9 ppm) of the C-2 signal of the glucose residue in saccharide 2 was observed in the spectra (Table 3), and 1H-1H and 1H-13C COSYs showed the α-configuration of the other glucose residue in saccharide 4 that was confirmed by the C-1 signal of this residue at 98.9 ppm (δ1H, 5.5 ppm). From these results, saccharide 4 was concluded to be O-α-D-glucopyranosyl-(1→2)O-α-D-glucopyranosyl-(1→5)-myo-inositol (kojibiosyl(1→5)-myo-inositol) (Fig. 5d). DISCUSSION KP catalyzes the glucosyl transfer reaction using β-G1P as a glucosyl donor to the appropriate acceptor (1). From its acceptor specificity, this enzyme catalyzes the transfer reaction of a glucose residue to the hydroxyl group of the glucose molecule at position 2. Further investigation of the acceptor specificity in this study showed that glycerol and myo-inositol are also acceptors (Table 1). When glycerol was used as an acceptor, several glucosyl products were synthesized (Fig. 1). To determine the structure of these products, we prepared three types of glucosyl glycerol, O-α-D-glu-
copyranosyl-(1→1)-glycerol, O-α-D-glucopyranosyl-(1→3)glycerol, and O-α-D-glucopyranosyl-(1→2)-glycerol using α-glucosidase, because it is difficult to distinguish between the structures of O-α-D-glucopyranosyl-(1→1)-glycerol and O-α-D-glucopyranosyl-(1→3)-glycerol by NMR measurement without standards. Takenaka et al. chemically synthesized three types of glucosyl glycerol, and identified the structures of these glucosyl glycerols synthesized by α-glucosidase using GLC (20). In this study, we used these glucosyl glycerols as standards. The two main products from them were O-α-D-glucopyranosyl-(1→1)-glycerol and kojibiosyl glycerol. Because KP also catalyzes the extension reaction of the side chain of O-α-D-glucopyranosyl-(1→1)glycerol, the minor products (TR =41.0 min and TR =39.9 min) are kojitriosyl and kojitetraosyl glycerol. When glucose was used as an acceptor, kojiheptaose was synthesized by the extension reaction from the side chain of kojibiose (2). From this observation, kojipentaosyl, kojihexaosyl and kojiheptaosyl glycerols are also expected to be synthesized under the conditions used in this study. When myo-inositol was used as an acceptor, O-α-D-glucopyranosyl-(1→1)- and O-α-D-glucopyranosyl-(1→5)-myo-inositol were synthesized by the transfer reaction of KP. Kojibiosyl-(1→1)- and kojibiosyl-(1→5)-myo-inositol were also synthesized by the extension reaction by KP. On the basis of the structures of these transfer products obtained, we discussed the correlation of the structures of glucose, glycerol and myo-inositol. As shown in Table 1, KP acted only on glycerol, myo-inositol and D-glucose. Glycerol, having only three carbons, is the smallest acceptor of KP. KP transferred a glucose residue to the hydroxyl groups at position 1 in glycerol, and at position 1 or 5 in myo-inositol. Interestingly, in glycerol and myo-inositol molecules, there is a common structure with three hydroxyl groups corresponding to the hydroxyl groups of the glucose molecule at positions 2, 3 and 4. The conformation of these three hydroxyl groups in the structure is
432
J. BIOSCI. BIOENG.,
YAMAMOTO ET AL.
cosidic linkage. Namely, oligosaccharides with specific glycosyl linkages can be synthesized using phosphorylases. The phosphorylases that have been reported to date are classified into fourteen kinds by the type of substrate phosphorolyzed (22), and only KP can specifically synthesize an α-1,2 glycosidic linkage. Therefore, novel oligosaccharides having an α-1,2 glycosidic linkage can be synthesized by KP. Furthermore, we demonstrated the substrate recognition of KP in this study, which will enable the prediction of the structure of glucosyl transfer products using the location or the number of the minimum structure in acceptors. ACKNOWLEDGMENTS FIG. 6. Stereochemical comparison among three acceptors for KP. Each circle indicates the structure with three hydroxyl groups that is recognized by KP. myo-Inositol-(b) represents rotated myo-inositol-(a). Arrows indicate the positions of transglucosylation.
equatorial. This common structure is the minimum structure for the substrate recognition of KP. As shown in Fig. 6, glycerol has one minimum structure for the substrate recognition of KP: the hydroxyl groups at positions 1, 2 and 3. On the other hand, myo-inositol has two minimum structures for the substrate recognition of KP: the hydroxyl groups at positions 1, 6 and 5, and at positions 5, 4 and 3. The fact that glycerol and myo-inositol have these structures strongly suggests that KP strictly recognizes this structure, and therefore catalyzes the transfer reaction of a glucose residue to the hydroxyl group at position 1 in glycerol, and at position 1 or 5 in myo-inositol corresponding to position 2 in glucose. Using L-sorbose as an acceptor, Chaen et al. determined the structure of glucosyl-L-sorbose (3). The structure of this saccharide was concluded to be O-α-D-glucopyranosyl-(1→5)-α-L-sorbopyranose, because the configuration of the hydroxyl groups at positions 5, 4, and 3 of L-sorbopyranose was the same as those at positions 2, 3, and 4 of D-glucopyranose. In this study, the percentage of glucosyl glycerol (saccharide A, 5.2%) in the reaction mixture was lower than those of glucosyl myo-inositols (saccharides 1 and 2, 25.4%), and the relative activity of KP toward glycerol was 0.04-fold higher than that toward myo-inositol (Table 1). It is assumed that the minimum structure in myo-inositol is rigid, because it is a pyranose-like ring. In contrast to myo-inositol, glycerol does not have a cyclic structure; therefore, the minimum structure in glycerol is more flexible, so KP has more difficulty in recognizing the minimum structure of glycerol than it has recognizing that of myo-inositol. Moreover, when myo-inositol was used as an acceptor, the amount of O-α-Dglucopyranosyl-(1→1)-myo-inositol was 2.5-fold higher than that of O-α-D-glucopyranosyl-(1→5)-myo-inositol. This difference indicates that the structure adjacent to the minimum structure influences the catalytic reaction of KP. Kitaoka et al. have investigated the acceptor specificity and substrate recognition of CP in detail (6). Using this enzyme in the transglucosyl manner, several heterooligosaccharides have been synthesized (7–11). The regiospecificities of phosphorylases are very strict and these enzymes phosphorolyze or synthesize only their specific type of gly-
This work was supported by a grant from the Research and Development Program for New Bio-industry Initiatives of the Biooriented Technology Research Advancement Institution (BRAIN).
REFERENCES 1. Chaen, H., Yamamoto, T., Nishimoto, T., Nakada, T., Fukuda, S., Sugimoto, T., Kurimoto, M., and Tsujisaka, Y.: Purification and characterization of a novel phosphorylase, kojibiose phosphorylase, from Thermoanaerobium brockii. J. Appl. Glycosci., 46, 423–429 (1999). 2. Chaen, H., Nishimoto, T., Nakada, T., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of kojioligosaccharides using kojibiose phosphorylase. J. Biosci. Bioeng., 92, 173–176 (2001). 3. Chaen, H., Nishimoto, T., Nakada, T., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of novel oligosaccharides from L-sorbose, maltose, and sucrose using kojibiose phosphorylase. J. Biosci. Bioeng., 92, 177– 182 (2001). 4. Watanabe, H., Higashiyama, T., Aga, H., Nishimoto, T., Kubota, M., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Enzymatic synthesis of a 2-O-α-D-glucopyranosyl cyclic tetrasaccharide by kojibiose phosphorylase. Carbohydr. Res., 340, 449–454 (2005). 5. Kitaoka, M., Sasaki, T., and Taniguchi, H.: Phosphorolytic reaction of Cellvibrio gilvus cellobiose phosphorylase. Biosci. Biotechnol. Biochem., 56, 652–655 (1992). 6. Kitaoka, M., Sasaki, T., and Taniguchi, H.: Synthetic reaction of Cellvibrio gilvus cellobiose phosphorylase. J. Biochem., 112, 40–44 (1992). 7. Kitaoka, M., Taniguchi, H., and Sasaki, T.: Production of glucosyl-xylose using Cellvibrio gilvus cells and its properties. Appl. Microbiol. Biotechnol., 34, 178–182 (1990). 8. Tariq, M. A., Hayashi, K., Tokuyasu, K., and Nagata, T.: Synthesis and structural analysis of disaccharides of 4-O-β-Dglucopyranosyl-D-glucosamine and 4-O-β-D-glucopyranosyl2-deoxy-D-glucose. Carbohydr. Res., 275, 67–72 (1995). 9. Tariq, M. A. and Hayashi, K.: Synthesis of three hetero disaccharides, 4-O-β-D-glucopyranosyl-6-deoxy-D-glucose, 4-O-β-D-glucopyranosyl-D-mannosamine, and 4-O-β-D-glucopyranosyl-D-mannose, and confirmation of their structures by C-13 NMR and MS. Biochem. Biophys. Res. Commun., 214, 568–575 (1995). 10. Percy, A., Ono, H., Watt, D., and Hayashi, K.: Synthesis of β-D-glucopyranosyl-(1→4)-D-arabinose, β-D-glucopyranosyl-(1→4)-L-fucose and β-D-glucopyranosyl-(1→4)-D-altrose catalyzed by cellobiose phosphorylase from Cellvibrio gilvus. Carbohydr. Res., 305, 543–548 (1997). 11. Percy, A., Ono, H., and Hayashi, K.: Acceptor specificity of cellobiose phosphorylase from Cellvibrio gilvus: synthesis of three branched trisaccharides. Carbohydr. Res., 308, 423–429
VOL. 101, 2006
(1998). 12. Mieyal, J. J. and Abeles, R. H.: Disaccharide phosphorylases, p. 515–532. In Boyer, P. D. (ed.), The enzymes, vol. 7, 3rd ed. Academic Press, New York (1972). 13. Kitao, S. and Sekine, H.: Transglucosylation catalyzed by sucrose phosphorylase from Leuconostoc mesenteroides and production of glucosyl-xylitol. Biosci. Biotechnol. Biochem., 56, 2011–2014 (1992). 14. Kitao, S., Ariga, T., Matsudo, T., and Sekine, H.: The syntheses of catechin-glucosides by transglycosylation with Leuconostoc mesenteroides sucrose phosphorylase. Biosci. Biotechnol. Biochem., 57, 2010–2015 (1993). 15. Kitao, S. and Sekine, H.: α-D-Glucosyl transfer to phenolic compounds by sucrose phosphorylase from Leuconostoc mesenteroides and production of α-arbutin. Biosci. Biotechnol. Biochem., 58, 38–42 (1994). 16. Kitao, S., Matsudo, T., Sasaki, T., Koga, T., and Kawamura, M.: Enzymatic synthesis of stable, odorless, and powdered furanone glucosides by sucrose phosphorylase. Biosci. Biotechnol. Biochem., 64, 134–141 (2000). 17. Yamamoto, T., Maruta, K., Mukai, K., Yamashita, H., Nishimoto, T., Kubota, M., Fukuda, S., Kurimoto, M., and Tsujisaka, Y.: Cloning and sequencing of kojibiose phosphorylase from Thermoanaerobacter brockii ATCC35047. J. Biosci. Bioeng., 98, 99–106 (2004). 18. Lloyd, J. B. and Whelan, W. J.: An improved method for
ACCEPTOR RECOGNITION OF KOJIBIOSE PHOSPHORYLASE
19. 20. 21.
22. 23. 24. 25.
433
enzymatic determination of glucose in the presence of maltose. Anal. Biochem., 30, 467–470 (1969). Fiske, C. H. and Subbarow, A.: The colorimetric determination of phosphorus. J. Biol. Chem., 66, 375–400 (1925). Takenaka, F., Uchiyama, H., and Imamura, T.: Identification of α-glucosylglycerol in sake. Biosci. Biotechnol. Biochem., 64, 378–385 (2000). Hagiwara, K., Tsuda, A., Ichikawa, T., Kanaya, K., Yoneyama, S., Sudo, H., and Iwao, H.: Quantative analyses of fructooligosaccharides by gas liquid chromatography. Shokuhin Eiseigaku Zasshi, 24, 558–562 (1983). (in Japanese) Kitaoka, M. and Hayashi, K.: Carbohydrate-processing phosphorolytic enzymes. Trends Glycosci. Glycotechnol., 14, 35–50 (2002). Bradbury, J. H. and Jenkins, G. A.: Determination of the structures of trisaccharides by 13C-n.m.r. spectroscopy. Carbohydr. Res., 126, 125–156 (1984). Schweizer, T. F. and Horman, I.: Purification and structure determination of three α-D-galactopyranosylcyclitols from soya bean. Carbohydr. Res., 95, 61–71 (1981). Sato, M., Nakamura, K., Nagano, H., Yagi, Y., and Koizumi, K.: Synthesis of glucosyl-inositol using a CGTase, isolation and characterization of the positional isomers, and assimilation profiles for intestinal bacteria. Biotechnol. Lett., 14, 659–664 (1992).