| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 22, 23405-23414, May 28, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
Received for publication, October 16, 2003 , and in revised form, February 18, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-D-apiofuranosyl--D-glucopyranoside (acuminoside)) into p-allylphenol and the disaccharide acuminose. We have isolated a cDNA coding for FH from Viburnum furcatum leaves. The open reading frame in the cDNA encoded a 538-amino acid polypeptide including a putative chloroplast transit peptide. The deduced protein showed 64% identity with tea leaf -primeverosidase, which is another disaccharide glycosidase specific to -primeverosides (6-O--D-xylopyranosyl--D-glucopyranosides). The deduced FH also shared greater than 50% identity with various plant -glucosidases in glycosyl hydrolase family 1. The recombinant FH expressed in Escherichia coli exhibited the highest level of activity toward furcatin with a Km value of 2.2 mM and specifically hydrolyzed the -glycosidic bond between p-allylphenol and acuminose, confirming FH as a disaccharide glycosidase. The FH also hydrolyzed -primeverosides and -vicianoside (6-O-
-L-arabinopyranosyl--D-glucopyranoside) but poorly hydrolyzed -gentiobiosides (6-O--D-glucopyranosyl--D-glucopyranosides), indicating high substrate specificity for the disaccharide glycone moiety. The FH exhibited activity toward p-allylphenyl -D-glucopyranoside containing the same aglycone as furcatin but little activity toward the other -D-glucopyranosides. Stereochemical analysis using 1H NMR spectroscopy revealed that FH is a retaining glycosidase. The subcellular localization of FH was analyzed using green fluorescent protein fused with the putative N-terminal signal peptide, indicating that FH is localized to the chloroplast. Phylogenetic analysis of plant -glucosidases revealed that FH clusters with -primeverosidase, and this suggests that the disaccharide glycosidases will form a new subfamily in glycosyl hydrolase family 1. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-glucosides, in which one glucose unit is linked to the hydroxy group of various kinds of aglycones. Glucosylation of secondary products results in enhancement of water solubility, decrease in chemical reactivity, and reduction of biological activity compared with the corresponding free aglycones. The -glucosides are hydrolyzed by specific -glucosidases to release the physiologically active aglycones (1-4). In some cases, the hydroxy group at position C-6 of the glycone moiety (glucose) of -glucosides is further modified by a 6-O-glycosidic bond formation with various monosaccharides such as glucose, xylose, arabinose, rhamnose, and apiose. The resulting disaccharide glycosides are resistant to hydrolysis by single -glucosidases. Stepwise and sequential reactions catalyzed by two independent glycosidases are thought to be required to hydrolyze disaccharide glycosides into two monosaccharide units and aglycones. First, endogenous and/or exogenous enzymes such as -glucosidase and -xylosidase hydrolyze the inter-glycosidic linkage between two sugars, then a second -glucosidase hydrolyzes the resultant -glucosides to release glucose and aglycones (5). For example, the cyanogenic diglucoside amygdalin ((R)-mandelonitrile -gentiobioside (6-O--D-glucopyranosyl--D-glucopyranoside) found in black cherry seeds) is sequentially hydrolyzed by two distinct -glucosidases (amygdalin hydrolase and prunasin hydrolase) (6). On the other hand, some plants contain disaccharide glycosidases, which directly hydrolyze the -glycosidic bond of disaccharide glycosides to release the corresponding disaccharide units and aglycones (Fig. 1). For example, -primeverosidase in Camellia sinensis hydrolyzes -primeverosides (6-O--D-xylopyranosyl--D-glucopyranosides) to liberate a primeverose unit and various aroma alcohols (7, 8). In addition, there are several similar examples for sets of glycosidases and disaccharide glycosides such as furcatin hydrolase in Viburnum furcatum Blume and furcatin (p-allylphenyl -acuminoside (6-O--D-apiofuranosyl--D-glucopyranoside)) (9), rutin hydrolase in Fagopyrum tataricum and rutin (quercetin -rutinoside (6-O--L-rhamnopyranosyl--D-glucopyranoside)) (10, 11), vicianin hydrolases in Vicia augustifolia (12) and Davallia trichomanoides Blume (13) and vicianin (mandelonitrile -vicianoside (6-O--L-arabinopyranosyl--D-glucopyranoside)), and endodiglycosidase in Vitis vinifera L., cv. Muscat of Alexandria (5, 14) and 6-O--L-arabinofuranosyl--D-glucopyranosides. Although some disaccharide glycosidases have been purified from plants (13, 15) and a microorganism (16), biochemical properties such as the substrate specificity and the reaction mechanism are still unclear.
|
-primeverosidase from tea leaves (8). This enzyme hydrolyzes the -glycosidic bond between primeverose and aglycones. It is classified as a family 1 glycosyl hydrolase and shares greater than 50% amino acid sequence identity with various plant -glucosidases. -Primeverosidase shows high substrate specificity for -primeveroside in terms of the glycone moiety and can also hydrolyze disaccharide glycosides such as -vicianoside and -acuminoside to some extent but does not hydrolyze monosaccharide glycosides such as -glucosides and -xyloside at all (17). In recent years, the aglycone specificity of -glucosidases has been investigated in structural and mutational analyses, and several key amino acids interacting with the aglycone moiety were identified (18-20). However, the mechanism of the glycone specificity in various -glycosidases has not been clarified in detail. Comparison studies between disaccharide glycosidases and -glucosidases will provide new insight into the mechanism of glycone recognition and binding.
Leaves of V. furcatum contain a large amount of the disaccharide glycoside furcatin (p-allylphenyl 6-O--D-acuminoside; its incorrect structure was revised (Refs. 21 and 22)). They also contain a unique -glycosidase, furcatin hydrolase (FH),1 which hydrolyzes furcatin into p-allylphenol and the disaccharide acuminose. Previously, FH activity was detected in a crude preparation of V. furcatum leaves (9) but has not yet been purified and characterized. In the present study, FH was partially purified and characterized. To further study the disaccharide glycosidase at the molecular and biochemical level, a cDNA encoding V. furcatum FH was isolated. Its enzymatic properties, expression pattern, and subcellular localization were also investigated.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-D-glucopyranoside, pNP -gentiobioside (6-O--D-glucopyranosyl--D-glucopyranoside), pNP -D-xylopyranoside, pNP -L-arabinopyranoside, pNP -D-galactopyranoside, and prunasin (mandelonitrile -D-glucopyranoside) were purchased from Sigma. Amygdalin (mandelonitrile -gentiobioside) was obtained from Wako Co. (Osako, Japan). 2-Phenylethyl -D-glucopyranoside, 2-phenylethyl -primeveroside, and 2-phenylethyl -gentiobioside were synthesized by the method of Ma et al. (17). pNP -primeveroside was kindly provided by Amano Enzyme Co. (Nagoya, Japan). Vicianin (mandelonitrile -vicianoside) was kindly provided by Dr. Ritsuo Nishida (Department of Applied Life Science, Kyoto University, Kyoto, Japan). p-Allylphenyl -D-glucopyranoside was kindly provided by Prof. Koji Kato (23). Furcatin was isolated from leaves of V. furcatum according to the method of Hattori and Imaseki (21) with slight modifications. Briefly, leaves (3.35 kg) were extracted with boiled water (10 liters x 2). The extract was subjected to chromatography on a charcoal column eluted with a gradient elution of 0-34% (v/v) ethyl acetate in methanol and crystallized from ethyl acetate to give furcatin (14.6 g). p-Allylphenol (chavicol) was synthesized from estragole by the method of Ohigashi and Koshimizu (24). Forty mmol (6 g) of estragole gave 34 mmol (4.35 g) of p-allylphenol in 80% yield. Plant Materials—V. furcatum Blume plants were sampled in the northern mountains of Kyoto, Japan. Leaves, stems, fruits, and roots were immediately frozen in liquid nitrogen and stored at -80 °C.
Preparation of Crude Extracts—Frozen leaves of V. furcatum were finely ground with a pestle in a chilled mortar with liquid nitrogen. Dry ice acetone (-30 °C) was added and filtered in vacuo. The residue was washed with chilled acetone until the filtrate became nearly colorless. The residual acetone powder was placed in vacuo to remove acetone and stored at -40 °C until used. The acetone powder (10 g) was suspended in 300 ml of extraction buffer (20 mM citrate buffer (pH 6.0) containing 10 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride) in the presence of 1.7% (w/v) polyvinylpolypyrrolidone to remove phenolic materials and stirred overnight at 4 °C. The slurry was centrifuged at 14,000 x g for 30 min at 4 °C. The supernatant (crude extract) was used for further purification steps and enzymatic assays.
Partial Purification of FH—All procedures were performed at 0-4 °C. An equal amount of chilled acetone was added to the crude extract with stirring and left overnight. The precipitate obtained by centrifugation at 14,000 x g for 30 min was dissolved in the extraction buffer. Ammonium sulfate was added at up to 50% saturation, and the resultant precipitate was collected by centrifugation at 14,000 x g for 30 min. The precipitate was suspended in the extraction buffer, and the suspension was centrifuged. The supernatant was dialyzed against the extraction buffer overnight. The enzyme preparation was applied to a CM-Toyopearl column (Tosoh, Tokyo, Japan), which was previously equilibrated with 20 mM citrate buffer (pH 6.0). Because the active fraction was not absorbed by the column, the enzyme preparation was applied to a DEAE-Sepharose column (Amersham Biosciences) equilibrated with 20 mM citrate buffer (pH 6.0) and eluted with a linear gradient of NaCl from 0.1 to 0.7 M at a flow rate of 1 ml/min. Fractions were tested for enzyme activity, and active fractions were combined and used for further tests.
SDS-PAGE and Western Blot Analysis—Purification progress and the molecular weight of the protein were checked by SDS-PAGE with a 12% (w/v) polyacrylamide gel. Proteins were detected by staining with Coomassie Brilliant Blue R-250. The preparation of anti--primeverosidase polyclonal antibodies was described previously (8). Western blot analysis using the -primeverosidase antibody was done using alkaline phosphatase and an ECL Western blotting kit (Amersham Biosciences).
Reverse Transcription-Polymerase Chain Reaction—Tissue samples were finely ground in a mortar with a pestle in the presence of liquid nitrogen. Total RNA was prepared essentially as described by Chang et al. (25), except that the extraction buffer consisted of 2% (w/v) cetyltrimethylammonium bromide, 0.1 M Tris-HCl (pH 9.5), 20 mM EDTA, 1.4 M NaCl, and 1% (v/v) -mercaptoethanol. Based on highly conserved regions in the amino acid sequences of plant family 1 -glucosidases including -primeverosidase, the following four degenerate primers were designed; AHN1 (according to the conserved sequence for FG-DRVK(R)H(E/Y)): 5'-TTYGGIGAYAGRGTNAARCA-3', AHN2 (for I(V)WDT(V/N)F(Y)TH(K)): 5'-ATHTGGGAYACITTYACNCA-3', AHN3 (for I(V)TENGM(V/R/I/T)D(G/P/A/N)): 5'-TCCATICCRTTYTCNGTDAT-3', and AHN4 (for GY(F)T(K)V(Q/Y/P/S/E)RF(Y)G): 5'-CCRAAYCTYTGIGTRTANCC-3'. One microgram of total RNA was used to synthesize the first strand cDNA using an RNA PCR Kit (version 2.1; Takara). Reverse transcription was carried out at 42 °C, and the PCR was undertaken for 5 min at 94 °C, followed by 45 cycles of 30 s at 94 °C, 50 s at 45 °C, and 90 s at 72 °C. PCR products were analyzed on a 1% (w/v) agarose gel, and cDNA fragments of expected size were recovered from the agarose gel using a Qiaquick gel extraction kit (Qiagen) and cloned into a pGEM-T easy vector (Promega) according to the instructions from the manufacturer.
Isolation of a Full-length FH cDNA—A cDNA library was constructed from total RNA of V. furcatum leaves. Poly(A)+ RNA was isolated using an OligotexTM-dT30 (Super) kit (Takara). The cDNA was synthesized and inserted in a ZipLox vector using a SuperScriptTM Lambda System for cDNA Synthesis and 
Cloning (Invitrogen).
Approximately 250,000 cDNA-containing phages were screened on nylon filters (Hybond-N, Amersham Biosciences) using an alkaline phosphatase-labeled probe based on a PCR product obtained with degenerate primers. A total of five positive phages were isolated and converted to plasmids. DNA sequence analysis showed that all five clones were identical, except for their length, and the longest clone was selected for further analysis. Because this clone did not contain a full-length cDNA, the 5' end cDNA was obtained by 5'-rapid amplification of cDNA ends (RACE) with a 5'-Full RACE Core Set (Takara) and gene-specific primers. The PCR product containing a start codon was found to overlap with the previously isolated cDNA clone. RT-PCR was performed to obtain a full-length cDNA using the primers FH1 (5'-ATGGCCACCATCACGACACTAG-3'; sense) and FH2 (5'-CTATTTGCCGAAGAGCTTCTTG-3'; antisense), and followed by cloning into the pCR 2.1-TOPO vector (Invitrogen). Sequencing reactions were carried out using a BigDye terminator cycle sequencing kit (PE Applied Biosystems). A DNA sequencer model 377 (PE Applied Biosystems) was used for DNA sequencing. Nucleotide sequences were analyzed using the DNASIS software system (Hitachi).
Heterologous Expression of FH in E. coli—-The FH open reading frame was amplified by PCR using a set of two primers: the N-terminal primer AHN5, 5'-GTCGACATGGCCACCATCACGACACTAG-3', containing a SalI site (underlined) and C-terminal primer AHN6, 5'-GCGGCCGCCTATTTGCCGAAGAGCTTCTTG-3', containing a NotI site (underlined). The amplification product was cloned into the pCR 2.1-TOPO vector, and its sequence was checked for PCR errors. The entire coding region of the FH cDNA was cloned in-frame as a SalI/NotI fragment into the expression vector pGEX-5X-3 (Amersham Biosciences), and the constructed vector, pGEX-FH, was used for production of a fusion protein between glutathione S-transferase (GST) and FH in E. coli strain BL 21. To express the FH, the cells were grown to log phase (A600 = 0.6) at 37 °C, before isopropyl--D-thiogalactopyranoside was added to a final concentration of 1 mM. The culture was incubated at 25 °C for an additional 15 h. The induced culture was harvested by centrifugation, and the cell pellets were sonicated 15 times for 30 s with 30-s intervals and centrifuged at 28,000 x g for 30 min. The recombinant protein in the obtained supernatant was isolated by GST-affinity chromatography (Amersham Biosciences), and the free FH was released by a protease factor Xa (New England Biolabs).
Enzyme Assays—Enzyme activity was determined by measuring the liberation of aglycones from each glycoside. Each reaction mixture (50 µl) contained 10 mM substrate, 20 mM citrate buffer (pH 6.0), and 10 µl of the enzyme solution. A mixture without the enzyme was preincubated at 37 °C, and the reaction was started by adding the enzyme and stopped by the addition of 50 µlof1 M Na2CO3. The activity of a reaction mixture with pNP glycoside was determined spectrophotometrically with liberated p-nitrophenol at 405 nm. In the reaction with furcatin, 20 µl of reaction mixture was subjected to HPLC, and p-allylphenol (retention time, 7.6 min) liberated from the substrate was detected at 277 nm. In the reactions with amygdalin, prunasin, and vicianin, 20 µl of reaction mixture was subjected to HPLC, and benzaldehyde, which is a product of mandelonitrile liberated from the substrates, was detected at 250 nm. The activity toward 2-phenylethyl -glycosides was measured by assaying the amount of 2-phenylethanol liberated from the substrate using HPLC. The protein concentration was determined by the Bradford method (26) using the Coomassie protein assay reagent (Pierce).
HPLC and TLC Analysis—Analysis of liberated p-allylphenol was performed under the following conditions: column, Cosmosil 5C18-AR-II (4.6 x 50 mm) (Nacalai Tesque, Inc., Kyoto, Japan); detection at 277 nm with a Shimadzu SPD-10AUP UV-visible detector; column temperature, 40 °C; mobile phase, 60% (v/v) methanol in water containing 0.1% (v/v) acetic acid; flow rate, 0.75 ml/min. Analysis of liberated benzaldehyde was performed under the following conditions: column, Cosmosil 5C18-AR-II (4.6 x 50 mm) (Nacalai Tesque, Inc.); detection at 250 nm with a Shimadzu SPD-10AUP UV-visible detector; column temperature, 40 °C; mobile phase, 50% (v/v) methanol in water; flow rate, 0.8 ml/min. Analysis of liberated 2-phenylethanol was performed under the following conditions: column, YMC-pack ODS-AQ (4.6 x 250 mm) (YMC Co., Kyoto, Japan); detection at 210 nm with a Waters 996 photodiode array detector; column temperature, 40 °C; mobile phase, 33% (v/v) acetonitrile in water; flow rate, 1.0 ml/min. TLC was carried out on silica gel 60 F254 plates (Merck 5715, 0.25 mm), using a solvent system of butanol-water-acetic acid (3:2:1, v/v/v). Glycosides and sugars were detected by heating at 120 °C after spraying with 0.2% (w/v) naphthoresorcinol in H2SO4:ethanol (1:19, v/v).
1H NMR Analysis for Determination of Stereochemical Outcome— The method was essentially as described by Wong et al. (27). Furcatin (10 mg) was dissolved in 0.7 ml of D2O. FH was lyophilized and redis-solved in 40 µl of D2O. The 1H NMR spectrum of the substrate solution was recorded, and then the enzyme (40 µl) was added to the NMR tube. The tube was incubated at 37 °C and subjected to 1H NMR measurements at time intervals at 25 °C, because the axial proton (Hax) of the -anomer of the released acuminose overlapped with the HDO signal at 37 °C.
Transient Expression of GFP Fusions—Green fluorescent protein (GFP) was used as a reporter protein to examine the subcellular localization of FH. The FH cDNA fragment corresponding to the N-terminal region (amino acid residues 1-65) including a putative signal peptide was amplified by PCR using oligonucleotides containing restriction sites (underlined): FH-SalI (5'-GTCGACATGGCCACCATCACGACACTAGCT-3') and FH-NcoI (5'-CCATGGCTAACCAGTTATCTTTGTTAAAGT-3'). The amplified DNA fragment was ligated into the pGEM-T easy vector (Promega) and digested with SalI and NcoI. The DNA fragment was ligated into SalI-NcoI sites behind the cauliflower mosaic virus 35 S (CaMV35S) promoter of the vector CaMV35S-sGFP-NOS3' kindly provided by Niwa (28). The resulting plasmid was introduced into onion cells and Arabidopsis leaf cells with a particle bombardment device (Biolistic PDS-1000/He, Bio-Rad) with DNA coated with 1.0-µm gold particles (Bio-Rad). After bombardment, cells were incubated for 24 h at 25 °C in a solid Murashige-Skoog medium (sugar-free), before observation by confocal laser scanning microscopy. Excised leaves and onion cells were mounted in phosphate-buffered saline buffer (pH 7.0) under glass coverslips and observed with a confocal laser scanning microscope (Zeiss LSM510). We used a filter set for GFP fluorescence (LP585 and BP505-550 for Arabidopsis cells and BP505-550 for onion cells).
Phylogenetic Analysis—To investigate the evolutionary relationships between FH and its relatives, a multiple alignment of members of glycosyl hydrolase family 1 from various plants was constructed with ClustalW 1.81 (29). A human -glucosidase (AF323990
[GenBank]
) was included in the alignment as an outgroup. Based on the alignment, an unrooted molecular phylogenetic tree was constructed by the neighbor-joining (NJ) method (30). For the analysis, the sites including gaps were excluded from the multiple alignment. The genetic distance between every pair of aligned sequences was calculated as the maximum like-hood estimate (31), using the JTT model (32) for the amino acid substitutions. Based on these distances, an NJ tree was constructed for all of the sequences included in the alignment. The statistical significance of the NJ tree topology was evaluated by bootstrap analysis (33) with 1000 iterative tree constructions. For the phylogenetic analysis, two software packages, PHYLIP 3.5c (34) and MOLPHY 2.3b3 (35), were used. The tree was drawn by TreeView (36).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
|
-primeverosidase activity as well as -acuminosidase activity. Ma et al. (17) reported that -primeverosidase from tea leaves was capable of hydrolyzing -vicianoside and -acuminoside to some extent, suggesting similar properties between FH and -primeverosidase. Thus, to identify the FH protein in the preparation, Western blot analysis was performed using an anti--primeverosidase polyclonal antibody (Fig. 2B). A single clear band, which was identical to the major 56-kDa protein in the final preparation, was detected. In addition, the signal increased during the purification procedures, indicating that a native FH was immunoreactive with the anti--primeverosidase sera. This result suggests that FH will be a 56-kDa protein with a primary structure similar to -primeverosidase. It was therefore expected that FH, like -primeverosidase, would belong to glycosyl hydrolase family 1.
cDNA Cloning of FH—To obtain the cDNA encoding FH, we attempted to isolate cDNAs coding for family 1 glycosidases that are abundantly expressed in V. furcatum leaves. Sequence alignment of -primeverosidase with various plant -glucosidases in family 1 gave several highly conserved regions: FG-DRVK(R)H(E/Y), I(V)WDT(V/N)F(Y)TH(K), I(V)TEN(C)GM(V/R/I/T)D(G/P/A/N), and GY(F)T(K)V(Q/Y/P/S/E)RF(Y)G. Four degenerate PCR primers were constructed, based on these conserved regions (AHN-1, -2, -3, and -4). RT-PCR was performed using total RNA of V. furcatum leaves as templates, and three fragments of expected size were amplified. A primer set of AHN-2 and -3 amplified 720-bp DNA fragments, and 920- and 1100-bp fragments were obtained from sets of AHN-2 and -4, and from AHN-1 and -3, respectively. These cDNA fragments were cloned into a pGEM-T vector, and the obtained clones (18 clones) were sequenced. There were three different glucosidases encoded by the cDNA fragments, and a cDNA library of V. furcatum leaves was screened using these fragments as probes. This procedure led to the isolation of two full-length clones showing high sequence homology to tomato -mannosidase (80%, AF403444
[GenBank]
) and banana -glucosidase (56%, AF321287
[GenBank]
) (data not shown), and one partial clone encoding putative -glucosidases, which showed greatest homology to tea leaf -primeverosidase. Therefore, we performed 5'-RACE PCR to obtain the full-length cDNA of this third clone. Finally, the full-length 1.86-kb cDNA, presumably encoding a FH, was obtained. The cloned FH cDNA consisted of an untranslated sequence of 12 nucleotides at the 5' end, a presumptive coding sequence of 1617 nucleotides, and 229 nucleotides of untranslated sequence including a poly(A) tail at the 3' end. The open reading frame encoded a protein consisting of 538 amino acid residues with a calculated molecular mass of 60,625 Da. PSORT analysis (psort.nibb.ac.jp) also predicted that the deduced protein contains a putative signal peptide for chloroplast transport in the N-terminal region (amino acid residues 1-45).
Characterization of FH Compared with Other Plant -Glucosidases—The amino acid sequence deduced from the FH cDNA showed the highest sequence similarity to -primeverosidase from tea plants (C. sinensis var. sinensis cv. Yabukita) (64% identity, AB088027
[GenBank]
(Ref. 8)) and shared greater than 50% identity with amygdalin hydrolase from Prunus serotina (56.4%, U26025
[GenBank]
(Ref. 6)), prunasin hydrolase from P. serotina (55.9%, AF221526
[GenBank]
(Ref. 6)), indican -glucosidase from Polygonum tinctorium (54.6%, AB003089
[GenBank]
(Ref. 37)), and linamarase from Trifolium repens (53.8%, X56733
[GenBank]
(Ref. 38)). Furthermore, other -glucosidases from various plants were also retrieved from the protein data base, and sequence identities were 40-50%. The deduced FH contained several sequence motifs that are highly conserved among family 1 -glucosidases (Fig. 3). The NEP sequence motif, of which the Glu residue is an acidbase catalyst, was found at residues 237-239, and the sequence ITENG, of which the Glu residue is a catalytic nucleophile of -glucosidases, was also found at residues 445-449. However, the conserved Asn residue in this motif was replaced by Cys-448 in the FH sequence. From this result, FH is classified as a family 1 glycosyl hydrolase. The residues involved in the binding of the glycone (glucose) moiety are highly conserved in all family 1 glycosidases (39-41), and these residues are also conserved at Gln-88, His-192, Asn-237, Glu-238, Glu-501, and Trp-502 in the FH sequence.
|
|
|
-D-thiogalactopyranoside resulted in the appearance of a new 88-kDa protein, which corresponded to the predicted molecular mass of the fusion protein comprising GST and FH. The fusion protein was purified by GST affinity chromatography, and the free FH protein was released by factor Xa cleavage (Fig. 6A). The amount of soluble fusion protein produced by the E. coli cells was dependent on the culture temperature. Analysis of the pellet and soluble fractions of crude lysate showed that induction at 37 °C leads to the formation of inclusion bodies. However, induction at 25 °C produced the soluble fusion protein. Fig. 6B shows that the anti--primeverosidase antibody recognized both the recombinant FH fused with GST and the free FH.
|
-acuminosidase activity toward the natural substrate furcatin. The reaction mixture containing furcatin and the recombinant FH was analyzed by thin-layer chromatography (Fig. 7, lane 5). The spot corresponding to furcatin completely disappeared, and a spot corresponding to a disaccharide, which showed a similar RF value to those of primeverose and cellobiose, was detected. This spot was confirmed to be acuminose by fast atom bombardment mass spectrometric analysis (data not shown). Glucose and apiose were not detected in the reaction mixture. This pattern of hydrolysis was identical to that of a native FH partially purified from V. furcatum leaves (Fig. 7, lane 6). No hydrolysis was detected with lysates from induced cells transformed with the empty vector pGEX-5X. Thus, the FH cDNA isolated in this study was confirmed to encode furcatin hydrolase of V. furcatum. The results also demonstrate that FH is a disaccharide-specific acuminosidase hydrolyzing the -glycosidic bond between p-allylphenol and acuminose without cleaving the inter-glycosidic bond between apiose and glucose. The recombinant FH was active over a pH range of 4-10 and was stable below 50 °C, with optimum activity at 
40 °C and pH 5.
|
-glucosides was measured using the recombinant FH (Table II). The FH showed the highest level of activity toward the natural substrate furcatin. pNP -primeveroside, 2-phenylethyl -primeveroside, and vicianin were also hydrolyzed with a relative activity of 87, 15, and 5%, respectively, compared with furcatin hydrolysis. The apparent Km values of the FH for furcatin and pNP -primeveroside were determined to be 2.2 ± 0.14 and 5.1 ± 0.45 mM, respectively. The FH showed 4% of the activity toward pNP -gentiobioside, whereas its activity was very low compared with that toward pNP -primeveroside. The FH did not hydrolyze 2-phenylethyl and mandelonitrile -gentiobiosides. These results indicate that FH shows high substrate specificity in terms of the disaccharide glycone moiety and also that -gentiobioside is a poor substrate for FH. Interestingly, the FH showed 43% of the activity toward p-allylphenol -D-glucopyranoside, which contains the same aglycone as furcatin. However, it showed only 3% of the activity toward pNP -D-glucopyranoside and was essentially inactive toward prunasin and 2-phenylethyl -D-glucopyranoside. These results indicate that FH also exhibits high substrate specificity for -D-glucosides in terms of the aglycone moiety. The FH did not hydrolyze other monosaccharide glycosides such as pNP -D-xylopyranoside, pNP -L-arabinopyranoside, and pNP -D-galactopyranoside (data not shown).
|
-primeverosidase are known to be retaining glycosidases (8, 40). Catalysis by retaining glycosidases proceeds via a double displacement mechanism, in which a covalent glycosylenzyme intermediate is formed and hydrolyzed via oxocarbenium ion-like transition states (41). The anomeric configuration of the glycone moiety in the substrate (-D-glucopyranoside) is retained in the product (-D-glucopyranose). To investigate the stereochemical outcome for the hydrolysis of a glycosidic bond by FH, the FH protein was incubated with furcatin, and the stereochemistry of enzymatic hydrolysis was analyzed by 1H NMR spectroscopy. The 1H NMR spectra of the reaction mixture revealed that the -anomer (Hax, 
4.65, J = 8.1 Hz) of acuminose was formed first, followed by the -anomer (Heq, 5.23, J = 3.7 Hz) as a consequence of mutarotation (Fig. 8). This result demonstrates that FH is a retaining glycosidase.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-primeverosidase, confirmed that the cDNA actually encodes a disaccharide-specific acuminosidase having FH activity. This is the second example of disaccharide glycosidase in glycosidase family 1. It is certainly apparent that the glycone moiety (e.g. acuminose and primeverose) is important for substrate specificity in disaccharide glycosidases. The amino acid residues involved in the binding of the glycone moiety (glucose) are highly conserved in all family 1 glycosidases (39-41), and the residues are known to form hydrogen bonds with the hydroxy groups of the -D-glucopyranoside ring. These residues are conserved in the FH sequence at Gln-88 (H-bonded to O-4), His-192 (H-bonded to O-2), Asn-237 (H-bonded to O-3), Glu-238 (H-bonded to O-2), and Glu-501 (H-bonded to O-6), and therefore it is expected that FH can bind the glycone moiety in a similar fashion to family 1 glucosidases.
We determined the substrate specificity of FH toward four sets of glycosides containing four kinds of aglycones (Table II). Comparison of the activity between disaccharide and monosaccharide glycosides containing the same aglycones clearly demonstrated that FH showed much greater activity toward disaccharide glycosides (-acuminoside, -primeverosides, and -vicianoside) except for -gentiobiosides than toward monosaccharide -glucosides. This indicates that the presence of a pentose unit on the C-6 hydroxy group of the glucose of disaccharide glycosides is important for substrate recognition and binding by FH. According to this point of view, we propose that the substrate-binding pocket in the active site of FH consists of three subsites, namely subsite -2 and -1 for the apiosyl and glucosyl moieties of the glycone moiety, respectively, and subsite +1 for the aglycone moiety.
Assuming the presence of the three subsites can help us to explain the substrate specificity of FH in terms of the disaccharide glycone moiety. FH exhibited the highest level of activity toward the natural substrate -acuminoside and also hydrolyzed -primeverosides as well as -vicianoside to some extent. These results suggest that subsite -2 in the glycone binding pocket can bind the apiosyl moiety as well as the xylosyl and arabinosyl moieties. On the other hand, FH poorly hydrolyzed -gentiobiosides. The ratio of hydrolytic activity was 21:1 for pNP -primeveroside/-gentiobioside, 37:1 for 2-phenylethyl -primeveroside/-gentiobioside, and 10:1 for mandelonitrile -vicianoside/-gentiobioside. These results clearly indicate that FH has higher specificity toward -primeveroside and -vicianoside than toward -gentiobioside. The second sugar moiety of -gentiobioside is glucose instead of pentose, and glucose has an extra bulky hydroxymethyl group at the C-6 position. The bulky group at C-6 is likely to hinder the glucosyl moiety from binding to subsite -2, and this would be one reason why -gentiobiosides are not hydrolyzed by FH.
A hallmark of disaccharide glycosidases is that they specifically hydrolyze the -glycosidic bond between disaccharides and aglycones to liberate disaccharide units. The two catalytic residues conserved in all family 1 members are also found in the FH sequence (Glu-238 and Glu-447, in Fig. 3), and therefore these residues are likely to be located in between acuminose and p-allylphenol, where the hydrolysis of the -glycosidic bond of furcatin occurs. An important question arises as to why FH prefers to disaccharide glycosides rather than -glucosides. On the recognition of -glucoside by FH, subsite -2 remains vacant and -glucoside is subsequently loosely bound, i.e. not fixed well in the substrate-binding pocket of the active site. This theory would explain why -glucosides are not good substrates for FH. However, p-allylphenyl -D-glucopyranoside, which contains the same aglycone as furcatin, was hydrolyzed by FH (43% of the activity relative to furcatin hydrolysis). This observation is quite different from the results for -primeverosidase, which specifically hydrolyzes -primeverosides but not -glucosides at all (8, 17). The aglycone specificity of tea leaf -primeverosidase is low because the natural substrate -primeverosides in tea leaves contain a variety of aroma alcohols as aglycones (42), suggesting that the aglycone binding site (subsite +1) is only a minor factor for substrate recognition and binding by -primeverosidase. In contrast, FH showed very high aglycone specificity for p-allylphenol and exhibited only weak activity toward pNP -D-glucopyranoside and negligible activity against mandelonitrile and 2-phenylethyl -D-glucopyranosides. Thus, p-allylphenyl -D-glucopyranoside is considered to be recognized by subsite +1 (the aglycone binding site) as well as subsite -1 and therefore fixed well in the FH substrate binding site. This would be the reason why FH hydrolyzes p-allylphenyl -D-glucopyranoside as well as disaccharide glycosides.
The popular name of V. furcatum in Japan is "Mushikari," in reference to the many small traces left by insects on the leaves. The leaves of this plant contain a large amount of furcatin (approximately 1% (w/w) of fresh leaves). In addition, the aglycone p-allylphenol is known to have fungicide, insect-repelling, larva growth-inhibitory, and nematocide activity (8, 43, 44). Many plant -glucosidases (e.g. cyanogenic glucosidases and thioglucosidases) have been known to play a role in defending against pathogen attacks or injury by producing toxic aglycones (45, 46). Most -glucosides accumulate in vacuoles (47), whereas -glucosidases are localized in the ER body, cell wall, protein bodies, or plastid (48-50). When the compartments in plant cells are destroyed by an insect and/or pathogen attack, -glucosidases and the corresponding -glucosides come into contact, and toxic aglycones are released. The physiological function of FH and furcatin seems to be certainly the case described above. In this study, we have demonstrated that FH is localized to the chloroplast. In addition, when wounding was induced either by scratching or by rubbing the leaves, the production of p-allylphenol increased (51). These results strongly suggest that FH and furcatin are located in separate compartments, although the intracellular localization of furcatin is still unknown. Furthermore, furcatin is resistant to hydrolysis by most -glucosidases because of the modification of the glycone moiety with a 6-O-apiosyl group, and FH is the only enzyme to efficiently hydrolyze furcatin in V. furcatum leaves. The timing and place of the release of the toxic compound p-allylphenol in the leaves is likely to be controlled by spatial and biochemical means. Thus, insects stop feeding on the leaves because of the release of the toxic aglycone p-allylphenol during the feeding process. The insects leave behind small feeding traces, thereby leading to the plant name Mushikari. The transcripts of the FH gene are expressed only in the leaves that are readily susceptible to attack by insects. Taken together, these observations suggest that in Mushikari, both FH and furcatin play an important role in defending against insect feeding.
A phylogenetic analysis of FH with various plant -glucosidases in family 1 is shown in Fig. 9. FH clusters together with tea leaf -primeverosidase in the phylogenetic tree. This suggests that disaccharide glycosidases may have evolved from a common monosaccharide -glucosidase to develop the binding site (subsite -2) for the second sugar moiety. Clustering between the two disaccharide glycosidases among many -glucosidases from various plant species suggests that FH and -primeverosidase should form a new subfamily in family 1. The other disaccharide glycosidases such as -vicianosidase and -rutinosidase may also lie within this new subfamily and potentially define a new subfamily in glycosidase family 1.
|
-primeverosidase. By comparing their primary sequences and the three-dimensional models with those of -glucosidases in family 1, one is expected to identify the binding site (subsite -2) and key amino acid residues interacting with the second sugar group at C-6 of the glycone moiety. Together with FH and -primeverosidase, cloning of other disaccharide glycosidases such as vicianin hydrolase and rutin hydrolase will offer further clues as to the molecular mechanism that determines the substrate specificity of these disaccharide glycosidases. |
FOOTNOTES |
|---|
* This work was supported in part by Grant-in-aid (B)(2)13460049 from the Ministry of Education, Science, Sports, and Culture of Japan (to K. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 
To whom correspondence should be addressed. Tel.: 81-774-38-3232; Fax: 81-774-38-3229; E-mail: mizutani{at}scl.kyoto-u.ac.jp.
1 The abbreviations used are: FH, furcatin hydrolase; GFP, green fluorescent protein; GST, glutathione S-transferase; Hax, axial proton; Heq, equatorial proton; pNP, para-nitrophenyl; HPLC, high performance liquid chromatography; RACE, rapid amplification of cDNA ends; TLC, thin layer chromatography; NJ, neighbor-joining; RT, reverse transcription.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
