Identification of the Gene Encoding Isoprimeverose-producing Oligoxyloglucan Hydrolase in Aspergillus oryzae*

Aspergillus oryzae produces a unique β-glucosidase, isoprimeverose-producing oligoxyloglucan hydrolase (IPase), that recognizes and releases isoprimeverose (α-d-xylopyranose-(1→6)-d-glucopyranose) units from the non-reducing ends of oligoxyloglucans. A gene encoding A. oryzae IPase, termed ipeA, was identified and expressed in Pichia pastoris. With the exception of cellobiose, IpeA hydrolyzes a variety of oligoxyloglucans and is a member of the glycoside hydrolase family 3. Xylopyranosyl branching at the non-reducing ends was vital for IPase activity, and galactosylation at a α-1,6-linked xylopyranosyl side chain completely abolished IpeA activity. Hepta-oligoxyloglucan saccharide (Xyl3Glc4) substrate was preferred over tri- (Xyl1Glc2) and tetra- (Xyl2Glc2) oligoxyloglucan saccharides substrates. IpeA transferred isoprimeverose units to other saccharides, indicating transglycosylation activity. The ipeA gene was expressed in xylose and xyloglucan media and was strongly induced in the presence of xyloglucan endo-xyloglucanase-hydrolyzed products. This is the first study to report the identification of a gene encoding IPase in eukaryotes.

Previously, only two IPases had been isolated and characterized as follows: one from the bacterium Oerskovia sp. Y1 (21,22), and the other from the fungus Aspergillus oryzae (20). Oerskovia IPase hydrolyzes XXXG to isoprimeverose and XXG and produces isoprimeverose and glucose as final products (21,22). Oerskovia IPase does not hydrolyze cellobiose, and the branching of xylopyranosyl residues at the non-reducing ends is essential for its activity. In these earlier studies, the gene encoding the Oerskovia IPase was identified and expressed in Escherichia coli (22). Oerskovia IPase consists of 1018 amino acids and has a molecular mass of 105 kDa. Oerskovia IPase belongs to the glycoside hydrolase family 3 (GH3) family and has a carbohydrate-binding module (CBM) family 6 at its C terminus (22). Oerskovia IPase binds to tamarind seed xyloglucan but not to crystalline cellulose, indicating that the CBM6 of Oerskovia IPase can bind to xyloglucan (an Oerskovia IPase substrate). The first reported IPase was purified from an A. oryzae commercial enzyme preparation, and it was found that this IPase was able to release isoprimeverose from XXXG (20). The A. oryzae IPase was not enhanced by addition of cations (such as Ca 2ϩ and Mn 2ϩ ) and was significantly inactivated in the presence of Cu 2ϩ and Hg 2ϩ (20). However, the gene encoding this A. oryzae IPase was not identified.
In this study, we identified the A. oryzae IPase gene, termed ipeA, and expressed it in Pichia pastoris. Although we classified IpeA as a GH3 family member, we found that it had only low similarity to Oerskovia IPase and other characterized GH3 family enzymes. IpeA characteristics, including substrate specificity, kinetic constants for various xyloglucan oligosaccharides, and transglycosylation activity, were determined in our study.
The results from this study illuminate previously uncharacterized eukaryotic GH3 family enzymes.
Purification of A. oryzae IPase and Determination of Its Amino Acid Sequence-A. oryzae IPase was purified as described by Kato et al. (20). The amino acid sequence was analyzed as described previously (13). Briefly, purified A. oryzae IPase was digested with lysyl endopeptidase (Wako Pure Chemical Industries), and the resultant peptides were fractionated using a TSK gel ODS-80Ts QA column (TOSOH Co., Tokyo, Japan). N-terminal amino acid sequences of peptides were analyzed by APRO Life Science Institute (Tokushima, Japan).
Expression and Purification of Recombinant IpeA-The ipeA gene was amplified using primers (5Ј-AGAGAGGCTGAAGC-TCAAGATGAGAAGCCTCGCTACAAG-3Ј and 5Ј-ATGAT-GATGGTCGACCTGCACATAGAAAGTAGCATTGCC-3Ј) and a cDNA template. cDNA was synthesized from A. oryzae total RNA. After amplification, the PCR fragment was integrated into a linearized pGAPZ-␣-A vector using In-Fusion PCR (Takara Bio, Siga, Japan). The linearized pGAPZ-␣-A was amplified using primers (5Ј-AGCTTCAGCCTCTCTTTTCT-CGAGAG-3Ј and 5Ј-GTCGACCATCATCATCATCATCAT-TGAG-3Ј) and pGAPZ-␣-A vector as templates. The resulting pGAPZ-␣-A-ipeA-His 6 vector was transformed into P. pastoris strain X-33. X-33 harboring pGAPZ-␣-A-ipeA-His 6 was grown in YPD medium (1% peptone, 0.5% yeast extract, 2% glucose) containing 100 mM potassium phosphate (pH 6.0) and cultured at 30°C, 100 rpm for 3 days. After cultivation, cells were harvested, and recombinant His-tagged IpeA was purified from culture supernatant using HisTrap FF crude (GE Healthcare, Buckinghamshire, UK). The protein concentration of purified enzyme was measured by UV (280 nm) absorbance using a NanoDrop spectrophotometer (NanoDrop Technologies, Rockland, DE) and revised using ProtParam. The XXXG hydrolysis activity of each fraction was assayed by measuring the production of reducing sugars from XXXGol (21). A fraction of purified IpeA was treated with endogycosidase H (endo H) (New England Biolabs). Briefly, 18 g of purified IpeA was denatured at 99°C for 10 min with 0.5% SDS and 40 mM dithiothreitol (DTT). Then 2000 units of endo H and endo H buffer were added, and the reaction mixture was incubated at 37°C for 60 min.
Effects of pH and Temperature on IpeA Activity-The optimal pH for recombinant IpeA activity toward XXXGol was evaluated in McIlvaine's buffer (pH 3.0 -8.0) (25). The total reaction volume was 50 l. The reaction mixture, containing 2 mM XXXGol, 0.05 g of IpeA and McIlvaine's buffer, was incubated at 40°C for 10 min. The resulting reducing sugars were measured using the 3,5-dinitrosalicylic acid reagent method (26). The optimal temperature for recombinant IpeA toward XXXGol was evaluated in 50-l reaction mixture containing 2 mM XXXGol, 0.05 g of IpeA, and 100 mM sodium acetate buffer (pH 4.5). The reaction mixture was incubated at 40 -65°C for 10 min, and the resulting reducing sugars were measured using the 3,5-dinitrosalicylic acid reagent method.
Substrate Specificity-The substrate specificity of recombinant IpeA for oligosaccharides (oligoxyloglucans, cellobiose, and xylobiose) was determined as follows. A reaction mixture (total reaction volume of 50 l) containing 5 mM oligoxyloglucans (XG, LG, XX, XXG, XXXG, XLXG, XLLG, or XXXGol), 5 mM cellobiose or 5 mM xylobiose, 50 mM sodium acetate buffer (pH 4.5), and 0.05 g of purified recombinant IpeA was incubated at 60°C for 5 min and then at 98°C for 10 min to stop the reaction. The concentration of released isoprimeverose, glucose, or xylose was determined using an Aminex HPX-87H 300 ϫ 7.8-mm column and an HPLC system driven by a pump (PC-2080) (Jasco, Tokyo, Japan) and equipped with a refractive index detector (RI-2031) (Jasco); 5 mM sulfuric acid was used as the column eluent at a flow rate of 0.6 ml/min at 65°C. Substrate specificity for chromogenic substrates was determined in 20 l of reaction mixture containing 10 mM pNP substrates (pNP-␤-D-glucopyranoside, pNP-␣-D-glucopyranoside, pNP-␤-D-cellobioside, pNP-␤-D-xylopyranoside, and pNP-␣-D-xylopyranoside), 50 mM sodium acetate buffer (pH 4.5), and 0.1 g of purified recombinant IpeA and incubated at 60°C for 5 min. The reaction was stopped by the addition of 50 l of 1 M NaHCO 3 , and the concentration of released pNP was determined by absorbance measurement at 405 nm using an Infinite M200 PRO microplate reader (Tecan, Zurich, Switzerland).
Kinetic Analysis of Recombinant IpeA-The kinetic parameters (Michaelis constant (K m ) and turnover number (k cat )) of recombinant IpeA for XG, XX, XXXG, and XLLG were determined at concentrations of 1.25 to 20 mM (XG and XX), 0.5 to 4 mM (XXXG), or 0.5 to 6 mM (XLLG) using 0.05 g of recombinant IpeA. A 50-l reaction mixture containing 50 mM sodium acetate buffer (pH 4.5) was incubated at 60°C for 5 min and then at 98°C for 10 min to stop the reaction. The concentration of released isoprimeverose was determined using an Aminex HPX-87H 300 ϫ 7.8-mm column and an HPLC system as described above. The kinetic constants (K m and V max ) were calculated by a non-linear regression of the Michaelis-Menten equation using GraphPad Prism version 5.0 (GraphPad Software, La Jolla, CA).
Mass Spectrometry-MS spectra were acquired using a MALDI-TOF mass spectrometer (Reflex IV; Bruker Daltonik, Bremen, Germany). Ions were generated by a pulsed 337-nm nitrogen laser and were accelerated to 20 kV. All spectra were obtained in the reflectron mode with a delayed extraction of 200 ns. For sample preparation, 0.5 l of a matrix solution prepared by dissolving sodium 2,5-dihydroxybenzate (0.5 mg/ml) and 2,5-dihydroxybenzoic acid (9.5 mg/ml) in 30% ethanol was spotted onto a target plate (MTP 384 target plate ground steel, Bruker Daltonik) and dried. Subsequently, an aliquot (0.5 l) of the glycan solution was spotted onto the matrix crystal and dried.
The ipeA gene encoding processed IpeA (amino acids 23-779) was cloned into a pGAPZ-␣ vector and expressed in P. pastoris. His-tagged IpeA was purified from a culture supernatant of P. pastoris harboring the IpeA expression vector using a nickel-affinity column (Fig. 2). The yield of the purified recombinant IpeA was ϳ7 mg liter Ϫ1 of cell culture. Although the molecular mass of His-tagged IpeA was estimated to be 84 kDa, the apparent molecular mass of purified His-tagged IpeA was ϳ100 kDa according to SDS-PAGE analysis (Fig. 2, lanes 1 and  2). When IpeA was treated with endo H, the protein band was ϳ85 kDa (Fig. 2, lane 4), indicating that purified IpeA was N-glycosylated. The optimal temperature and pH of recombinant IpeA with XXXGol were 60°C and pH 4.5, respectively.
Substrate Specificity of Recombinant IpeA-The specific activities of recombinant IpeA toward various oligoxyloglucans and chromogenic substrates were examined (Table 1). IpeA exhibited hydrolytic activity toward XG and produced isopri- meverose and glucose, but there was no detectable hydrolytic activity toward LG, cellobiose (GG), and xylobiose. These results indicate that A. oryzae IpeA is an isoprimeverose-producing oligoxyloglucan hydrolase. Xylose branching at the nonreducing ends (subsite Ϫ1Ј) was essential for IpeA activity. Galactosylation of branched xylopyranose at the non-reducing ends (subsite Ϫ1Ј) inhibited IpeA hydrolytic activity. IpeA strictly recognized the isoprimeverose unit at the non-reducing ends of oligoxyloglucans by subsites Ϫ1 and Ϫ1Ј. The specific activity of IpeA toward XX, XXG, and XXXG was ϳ1.4-, 1.9-, and 2.2-fold higher compared with that toward XG, respectively. In contrast, IpeA activity toward XLXG and XLLG was ϳ40% lower compared with that toward XXXG. These results indicate that xylose branching at the second glucopyranosyl residue from the non-reducing end (subsite ϩ1Ј) is not vital for hydrolysis activity but can enhance IpeA activity. Furthermore, galactosylation of the second xylopyranosyl residue from the non-reducing ends (subsite ϩ1Ј) partially inhibited IpeA hydrolysis activity. Based on substrate specificity results, we predicted the IpeA subsites (Fig. 3).
Kinetic Constants of Recombinant IpeA-Next, the kinetic constants (Michaelis constant, K m ; turnover number, k cat ; and catalytic efficiency, k cat /K m ) of recombinant IpeA toward XG, XX, XXXG, and XLLG were examined ( Table 2). The k cat values of IpeA for XX and XXXG were significantly higher than those for XG and XLLG, indicating that the xylopyranosyl side chain of the second glucopyranosyl residue from the non-reducing end (ϩ1 subsite) was important for turnover. The K m value of IpeA for XXXG was ϳ2-fold lower than that for XX. The catalytic efficiency constant (k cat /K m ) for XXXG was much higher than those for XG, XX, and XLLG. These results indicate that, among these substrates, XXXG was the most preferred IpeA substrate.
Transglycosylation Activity-A. oryzae IpeA released isoprimeverose units from oligoxyloglucans at the non-reducing ends. Next, we examined whether IpeA was able to transfer isoprimeverose units to other oligosaccharides (Fig. 4A). IpeA was incubated with a high concentration (80 mM) of XX. Five to 30 min into the reaction, isoprimeverose, XX, XXX, and XXXX were detected. In addition, a small amount of XXXXX was detected midway through the reaction (20 min). Production of XXX, XXXX, and XXXXX was confirmed by mass spectrometry analysis (Fig. 4B). In XX, a small contaminant of a tetrasaccharide consisting of three hexoses and one pentose (m/z 659, [H3P1 ϩ Na] ϩ ), which was not hydrolyzed by IpeA, was detected. These results indicate that IpeA can transfer isoprimeverose to other oligosaccharides. Toward the end of the reaction (60 min), XXX, XXXX, and XXXXX products were hydrolyzed to isoprimeverose.
Expression Pattern of ipeA-Next, the expression pattern of the ipeA gene was examined (Fig. 5). Previously, ipeA gene expression was reported to be controlled by the transcriptional activator XlnR (in that paper, ipeA was referred to as AO090701000274) (28). XlnR mediates the expression of many genes, including genes encoding endo-xylanases (XynF1 and XynG2) and ␤-xylosidase (XylA), in response to xylan and xylose (27,28). However, in both A. oryzae wild type (KBN616) and ⌬xlnR strains, no detectable expression of ipeA was     observed in the glucose or cellobiose medium. The ipeA gene was expressed in the xylose or TXG medium and strongly induced in the presence of TXG endo-xyloglucanase-hydrolyzed products in the wild type strain (Fig. 5). These results indicate that A. oryzae produces IpeA to hydrolyze and utilize oligoxyloglucans as carbon sources. In addition to ipeA, expression of xynF1, xynG2, and xylA was induced not only in the xylose medium but also in the medium of the TXG endo-xyloglucanase-hydrolyzed products (Fig. 5). The induced expression of ipeA, xynF1, xynG2, and xylA in xylose medium was almost completely abolished by deletion of xlnR. However, even in the absence of xlnR, ipeA expression was induced by medium containing TXG or TXG endo-xyloglucanase-hydrolyzed products (Fig. 5).

Discussion
Because the characterized GH3 glycoside hydrolases recognize and release individual glycosyl residues from the non-reducing ends of substrates, chromogenic substrates, such as pNP-monosaccharides, and disaccharides have been widely used for screening and functional analyses of GH3 glycoside hydrolases. However, we previously identified and characterized a unique GH3 glycoside hydrolase (Oerskovia IPase) that can release a disaccharide (isoprimeverose) from the non-reducing ends of oligoxyloglucans (21,22). In A. oryzae, although there are more than 20 genes encoding GH3 family enzymes, including four ␤-glucosidases (EC 3.2.1.21: BglA, BglF, BglH, and BglJ) (31, 32) and one ␤-xylosidase (EC 3.2.1.37: XylA) (29) (CAZy database), the functions of GH3 family enzymes are largely unknown. In this study, the ipeA gene encoding GH3 isoprimeverose-producing oligoxyloglucan hydrolase in A. oryzae was identified. A. oryzae IpeA is the first eukaryotic IPase to be identified. IpeA exhibited low similarity with Oerskovia IPase and other characterized GH3 enzymes. However, putative GH3 enzymes of unknown function (that are homologous to IpeA) conversed in other Aspergillus genus and fungi (Table  3). In addition, enzymes with similarity to IpeA were also found in Pseudocercospora fijiensis, Neofusicoccum parvum, Leptosphaeria maculans, Thielavia terrestris, Marssonina brunnea, and Dactylellina haptotyla. This suggests that IPases are widely distributed in microorganisms. Although some of these enzymes were predicted to be ␤-glucosidases, the functions of these enzymes are unknown. For this reason, we predict that IPases are not able to hydrolyze cellobiose and pNP-␤-D-glucopyranoside, which are well known ␤-glucosidase substrates, and that natural substrates (such as oligoxyloglucan saccharides) are important for the discovery of new enzyme activities. Our identification of an IPase gene in A. oryzae will illuminate the function of putative GH3 enzymes.
The ipeA expression was repressed in medium containing glucose or cellobiose as carbon sources and was induced by xylose, TXG, and TXG endo-xyloglucanase-hydrolyzed products. In particular, TXG endo-xyloglucanase-hydrolyzed products containing XXXG, XXLG, XLXG, and XLLG were strong inducers of ipeA gene expression. Genes encoding endo-xylanases and ␤-xylosidase were also induced in the presence of TXG endo-xyloglucanase-hydrolyzed products in the wild type strain. This induced expression was abolished in the ⌬xlnR strain, indicating that xylose and TXG endo-xyloglucanase-hydrolyzed products induced the expression of these genes in an XlnR-dependent manner. However, ipeA was expressed in the ⌬xlnR strain in the presence of TXG or TXG endo-xyloglucanase-hydrolyzed products. These results suggest that the ipeA gene is expressed in both an XlnR-dependent and -independent manner. Future research will investigate the mechanism by which A. oryzae recognizes oligoxyloglucan and induces ipeA expression. Furthermore, studies on how A. oryzae recognizes lignocellulosic biomass and expresses the corresponding enzymes involved in lignocellulosic biomass degradation will be of utmost importance.
IpeA recognizes isoprimeverose units at the non-reducing ends of oligoxyloglucans and releases isoprimeverose residues. For example, IpeA hydrolyzes XXXG to X (isoprimeverose) and XXG. The specific activity of IpeA toward XXXGol was much higher than that previously reported for Oerskovia IPase (85 mol/min/mg) ( Table 1) (21). Xylose branching at subsite ϩ1 (second glucopyranosyl residue from the non-reducing end) is not essential for IpeA ␤-glucosidase activity but contributes to the V max value and turnover number. Hydrolase activity was completely abolished by galactosylation at branched xylopyranosyl residues at the non-reducing ends (subsite Ϫ1Ј) and was partially blocked by galactosylation at subsite ϩ1Ј. In addition, the K m value for XXXG was much lower than that for XX, indicating IpeA has more than three subsites (Ϫ1, ϩ1, and ϩ2) (Fig. 3). These findings suggest that IpeA has a very unique subsite structure that differs from that of other GH3 family ␤-glucosidases and ␤-xylosidases. Because cellotetraose unitcontaining oligoxyloglucan substrates, such as XXXG, are products of xyloglucan degradation by endo-xyloglucanases, we believe that IpeA and endo-xyloglucanases act in coordination to hydrolyze xyloglucan. Previously, an oligoxyloglucanreducing end-specific xyloglucanobihydrolase (OREX), which belongs to the glycoside hydrolase family 74 (GH74) and releases two glycosyl residue segments (such as XG and LG) from the reducing ends of oligoxyloglucans, was isolated from A. nidulans (19). Enzymes corresponding to A. nidulans OREX have not been identified from A. oryzae. However, synergy among endo-xyloglucanase, OREX, IPase, and other enzymes, such as ␤-galactosidase, might be important for xyloglucan degradation.
IpeA was able to transfer isoprimeverose units to other saccharides, indicating high transglycosylation activity. In this study, we were able to produce unnatural oligosaccharides, XXX, XXXX, and XXXXX, indicating that IpeA has the potential to produce new oligosaccharides by transglycosylation of isoprimeverose units.
Ongoing experiments currently focus on determining the IpeA structure by x-ray crystallography. This analysis will help clarify the mechanism by which IpeA recognizes oligoxyloglu-