TcXyn30B is a bifunctional xylanase with endo‐glucuronoxylanase and exo‐xylobiohydrolase activities. The present study determined the crystal structure of the TcXyn30B‐product complex. A 4‐O‐methyl‐α‐d‐glucuronyl moiety‐binding pocket required for glucuronoxylanase activity and Asn‐93 (which is likely involved in xylobiohydrolase activity) was structurally characterized. The present study provides insight into the mechanisms of substrate recognition of a bifunctional xylanase.
subfamily 7 of the glycoside hydrolase family 30
subfamily 8 of the glycoside hydrolase family 30
Xylan is the major component of hemicellulose in plants. Xylan is composed of a linear backbone of β‐d‐xylopyranosyl residues linked by β‐1,4‐glycosidic bonds, which are further decorated with side‐chain residues, such as α‐1,2‐ and/or α‐1,3‐linked‐l‐arabinofuranose, and α‐1,2‐linked‐4‐O‐methyl‐d‐glucuronic acid (MeGlcA). Glucuronoxylanase (https://www.qmul.ac.uk/sbcs/iubmb/) is an appendage‐dependent endoxylanase that must recognize an α‐1,2‐linked MeGlcA common to glucuronoxylans for hydrolysis. Glucuronoxylanase cleaves the glucuronoxylan main chain at the second glycosidic linkage from the MeGlcA substituent toward the reducing end to produce 22‐MeGlcA‐xylooligosaccarides (XnU4m2X, n ≥ 0). The enzyme is classified into glycoside hydrolase family (GH) 30 subfamilies 7 and 8 (GH30‐7 and 30‐8) in the CAZy database (http://www.cazy.org) .
Typically, subfamily 8 of the glycoside hydrolase family 30 (GH30‐8) glucuronoxylanases primarily occur in bacteria [2, 3, 4, 5]. Crystal structures of GH30‐8, such as EcXynA from Dickeya chrysanthemi (formerly Erwinia chrysanthemi) and BsXynC from Bacillus subtilis, have revealed that the enzymes consist of a (β/α)8 ‐barrel with an obligatory side‐associated, nine‐stranded, aligned β‐sandwich [1, 6]. This side β‐sandwich structure is tightly associated with the (β/α)8‐barrel catalytic core domain. Studies of ligand‐bound GH30‐8 xylanase structures have identified the role of the β7–α7 and β8–α8 loop regions in the specific coordination of the MeGlcA substituent through a salt bridge established between the C‐6 carboxylate of the MeGlcA and an arginine (Arg‐293 of EcXynA and Arg‐272 of Bs XynC) that extends from the β8–α8 loop (Fig. 1) [7, 8, 9].
GH30‐7 glucuronoxylanases have been found in fungi, such as XYN VI from Trichoderma reesei, Xyn30B from Talaromyces cellulolyticus (TcXyn30B) and Xyn30A from Thermothelomyces thermophila (Tt Xyn30A) [10, 11, 12]. Unlike the GH30‐8 enzyme, these enzymes act on unsubstituted xylan and xylooligosaccharides. Especially, TcXyn30B and TtXyn30A have been reported as bifunctional xylanases possessing both glucuronoxylanase and xylobiohydrolase activities, which release xylobiose from non‐reducing ends of XnU4m2X (n ≥ 0) produced by glucuronoxylanase activity [11, 12].
We recently determined the 3D‐structure of Tc Xyn30B as the first structure of a GH30‐7 xylanase . The overall structure of TcXyn30B is basically similar to GH30‐8 enzymes. In addition, TcXyn30B has unique structural features, which are probably conserved in other GH30‐7 enzymes. They include a Cys‐pair (cis ‐Cys‐241 and Cys‐242); a β8‐sheet consisting of strands β8, β8A and β8B; and no α6 helix (Fig. 1) . X‐ray crystallography and mutational analysis of Tc Xyn30B without any ligands have suggested that Arg‐46 from the β1‐α1 region conserved in GH30‐7 endoxylanases plays a critical role in recognizing MeGlcA for glucuronoxylanase activity . We also predict that Asn‐93 in the β2–α2 loop may contribute to xylobiohydrolase activity, using the TcXyn30B structure that was superimposed on the GH30‐8 EcXynA model complexed with 22‐MeGlcA‐xylotriose (XU4m2 X) . However, structural factors for substrate recognition cannot be fully explained because the amino acid sequence identity between TcXyn30B and EcXynA is low (24%). Especially, residues involved in the recognition of MeGlcA of GH30‐8 enzymes are not conserved in GH30‐7 enzymes including TcXyn30B. It is also unclear how Asn‐93 in the loop actually interacts with the substrate. In the present study, the crystal structure of TcXyn30B complexed with 22‐MeGlcA‐xylobiose (U4m2X) is determined. U4m2X is a minimum product obtained by glucuronoxylanase activity and an appropriate ligand for understanding the recognition mechanism for MeGlcA and xylobiose. Structural analysis of TcXyn30B‐U4m2X provides valuable insights into the catalytic properties of GH30‐7 bifunctional glucuronoxylanase and xylobiohydrolase.
Recombinant TcXyn30B was expressed in Pichia pastoris using the Pichia Expression Kit (Thermo Fisher Scientific, Waltham, MA, USA). The pPIC9K plasmid (Thermo Fisher Scientific) was used to construct an expression plasmid to produce TcXyn30B. Escherichia coli DH5α (TaKaRa Bio, Kyoto, Japan) was used for the DNA procedures. The TcXyn30B gene excluding signal sequence (residues 1–22) was synthesized. The xyn30B gene coding residues 23–474 was amplified using the forward primer, 5'‐GAATTCCAGATTAATGTGGATCTGCAAGCTCGC‐3', with the EcoRI site (underlined) and the reverse primer, 5'‐GCGGCCGCTCATTCGCATTCGGTCACAAAGCTGG‐3', with the NotI site (underlined). The expression plasmid, pPIC9K‐TcXyn30B, was constructed by ligating the xyn30B fragment that had been digested with EcoRI/NotI into the corresponding site of pPIC9K. The presence of the ligated gene fragment and its location were confirmed by DNA sequencing.
Recombinant TcXyn30B with eight His‐tag at the C‐terminal (TcXyn30B‐His) and its mutant, TcXyn30B‐His N93A, were expressed using the same procedure as described above. The expression plasmid, pPIC9K‐TcXyn30B‐His, was constructed by site‐directed mutagenesis of pPIC9K‐TcXyn30B using the KOD ‐plus‐ Mutagenesis kit (Toyobo, Osaka, Japan). The forward primer 5'‐CATCATCACCATCACCACCATCACTGAGCGGCCGCGAATTAATTCGC‐3' (insertion region underlined) and the reverse primer, 5'‐TTCGCATTCGGTCACAAAGCTGGTCA‐3', were used for PCR. The expression plasmid, pPIC9K‐TcXyn30B‐His N93A, was constructed by site‐directed mutagenesis of pPIC9K‐TcXyn30B‐His. The forward primer 5'‐GCTTTAATGAACAGCATTGAGCCGTTTAGC‐3' (mutation site underlined) and the reverse primer, 5'‐GCTGGTGCTGCTATTGCTGCTGCCGATGCC‐3', were used for PCR. The presence of all ligated gene fragments and their locations were confirmed by DNA sequencing.
The pPIC9K‐TcXyn30B, pPIC9K‐TcXyn30B‐His and pPIC9K‐TcXyn30B‐His N93A were linearized by SacI and transformed into P. pastoris GS115 (Thermo Fisher Scientific) by electroporation. The strains producing TcXyn30B, TcXyn30B‐His, and TcXyn30B‐His N93A were selected based on the amount of recombinant protein in culture supernatant as visualized by SDS/PAGE using NuPage 4–12% Bis‐Tris gels (Invitrogen, Carlsbad, CA, USA). To produce recombinant proteins, the selected strains were cultured in a BMMY medium (1% yeast extract, 2% peptone, 100 mm potassium phosphate, pH 6.0, 1.34% yeast nitrogen base, 4 × 10−5% biotin and 0.5% methanol) as described in the manufacturer’s instructions for the Pichia Expression Kit (Thermo Fisher Scientific).
Purification of TcXyn30B, TcXyn30B‐His and TcXyn30B‐His N93A was performed using an ÄKTA purifier chromatography system (GE Healthcare, Little Chalfont, UK) at room temperature. A culture supernatant including TcXyn30B was filtered through a 0.22‐μm polyethersulfone membrane and the filtrate protein was concentrated and changed to 20 mm 2‐(N‐morpholino) ethanesulfonic acid (pH 6.0) using a Vivaspin 20‐10K centrifugal concentrator (Sartorius, Göttingen, Germany). The sample was applied to a HitrapQ anion‐exchange column (5 mL; GE Healthcare) that had been equilibrated with the same buffer, and protein peaks were eluted with a linear gradient of 0–0.5 m NaCl (20 column volumes) at a flow rate of 2 mL·min–1. Fractions containing the target proteins were confirmed by SDS/PAGE and pooled. (NH4)2SO4 was added to a final concentration of 2.0 m and then the samples were subjected to ResourceISO (6 mL; GE Healthcare) hydrophobic interaction chromatography using a 2.0–0 m (NH4)2SO4 gradient (20 column volumes) in 20 mm sodium acetate buffer (pH 4.0) at a flow rate of 1 mL·min−1. The fractions containing target protein were pooled and concentrated by ultrafiltration using a Vivaspin 20‐5K centrifugal concentrator. The sample was applied to a Superdex 200 Increase 10/300 GL size exclusion chromatography column (GE Healthcare) that had been equilibrated with 0.15 m NaCl in 20 mm sodium acetate buffer (pH 4.0).
Culture supernatants including TcXyn30B‐His and TcXyn30B‐His N93A were mixed with Tris‐HCl (pH 8.0) at a final concentration of 50 mm and then filtered through a 0.22‐μm polyethersulfone membrane. The samples were applied to a HisTrap FF Ni‐affinity column (10 mL; GE Healthcare) that had been equilibrated with 20 mm imidazole in 20 mm Tris‐HCl (pH 7.5) and the column was washed using 40 mm imidazole. Protein peaks were eluted with a linear gradient of 40–300 mm imidazole (20 column volumes) at a flow rate of 4 mL·min−1. Fractions containing the target proteins were confirmed by SDS/PAGE and pooled. (NH4)2SO4 was added to final concentration of 2.0 m and the samples were then subjected to HiTrap Butyl HP (5 mL; GE Healthcare) hydrophobic interaction chromatography using a 2.0–0 m (NH4)2SO4 gradient (20 column volumes) in 20 mm sodium acetate buffer (pH 4.0) at a flow rate of 4 mL·min−1.
All purified enzymes were preserved in a 20 mm sodium acetate buffer (pH 4.0) at 4 °C. Protein concentration was determined by monitoring A280.
The molecular weight of the purified Tc Xyn30B was evaluated by MALDI time‐of‐flight MS with a Spiral TOF JMS‐S3000 (JEOL, Tokyo, Japan) as described previously . The purified sample was applied to the MALDI target plate after dilution into a mixture containing 0.5% (w/v) sinapinic acid, 0.1% trifluoroacetic acid and 25% acetonitrile.
Purified TcXyn30B was concentrated to 10 mg·mL−1 for crystallization by ultrafiltration using a Vivaspin 20‐5K centrifugal concentrator. Crystals were obtained with the hanging‐drop vapor diffusion method at 20 °C for 1 week. The drop was comprised 1.0 µL of protein solution mixed with 1.0 µL of reservoir solution containing 25% poly(ethylene glycol) 3350, 0.1 m Hepes‐sodium hydroxide (pH 7.5) and 200 mm magnesium chloride and was equilibrated against 500 µL of reservoir solution. In the case of the co‐crystallization with a ligand, the 2.0‐µL drops were prepared by mixing the protein, ligand and precipitant solutions at a volume ratio of 0.9 : 0.1 : 1. A mixture of aldouronic acids (Megazyme, Wicklow, Ireland) containing a mixture of U4m2X, 23‐MeGlcA‐xylotriose (U4m2XX) and 24‐MeGlcA‐xylotetraose (U4m2 XXX) at a ratio of 2 : 2 : 1 was used as the ligand solution. The abbreviations used to describe the xylooligosaccharides have been reported previously . The structures of ligands are shown in Fig. S1. A mixture containing 25% poly(ethylene glycol) 3350, 0.1 m Hepes‐sodium hydroxide (pH 7.3) and 200 mm magnesium chloride was used as a precipitant solution for co‐crystallization.
The crystals of TcXyn30B and the enzyme complexed with the mixture of aldouronic acids were soaked with the reservoir solution supplemented with 25% (v/v) glycerol and 10% (w/v) poly(ethylene glycol) 3350 as cryo‐protectants, respectively, and then flash cooled in liquid nitrogen. X‐ray diffraction data of crystals of TcXyn30B and TcXyn30B complexed with U4m2X were collected to resolutions of 1.60 and 1.65 Å at 100 K at the SPring‐8 beamline BL44XU (Hyogo, Japan). Diffraction images were checked with adxv (http://www.scripps.edu/tainer/arvai/adxv.html) and integrated and scaled with xds (version: 15 March 2019) . Phasing was performed using molrep, version 11.6, in ccp4, version 7.0, with TcXyn30B coordinates (PDB ID: http://6IUJ) as the model [15, 16]. The model was manually completed using coot , version 0.8.9 , and refined using Phenix.refine  in phenix , version 1.12 , and refmac , version 5.8 . Model quality was verified using molprobity , version 4.4 . Superpositioning of protein models and calculation of their rmsd were conducted using LSQKAB program in ccp4 program package . Molecular figures were generated with pymol, version 1.8 (Schrödinger, LLC, New York, NY, USA).
All assays were performed in triplicate. Glucuronoxylanase activity was measured by assaying the reducing sugars released after the enzyme reaction with 10 mg·mL–1 beechwood glucuronoxylan (Megazyme) using 3,5‐dinitrosalicylic acid. The enzyme reaction was performed under conditions of 50 mm sodium acetate buffer (pH 4.0) at 40 °C for 15 min. One unit of glucuronoxylanase activity was defined as the amount of protein that could yield 1 μmol of reducing sugar per minute from the hydrolysis of beechwood glucuronoxylan.
Xylobiohydrolase activity was measured in a reaction mixture containing 2 mm xylotriose (X3; Megazyme) in 50 mm sodium acetate (pH 4.0). The reaction was carried out at 40 °C for 15 min. The released xylose was analyzed by high‐performance anion‐exchange chromatography with pulsed amperometric detection using a Dionex ICS‐3000 ion chromatography system (Dionex, Sunnyvale, CA, USA) . One unit of xylobiohydrolase activity for X3 was defined as the amount of protein that could release 1 μmol xylose·min–1.
Determination of the kinetic parameters of TcXyn30B‐His and TcXyn30B‐His N93A was performed using 3.6–48 mg·ml−1 beechwood glucuronoxylan and 1–16 mm X3. The reaction was performed at 40 °C in 50 mm sodium acetate buffer (pH 4.0). Kinetic constants for beechwood glucuronoxylan were determined using the nonlinear least‐squares data fitting method in excel , version 2016 (Microsoft Corp., Redmond, WA, USA) . The initial slopes of the progress curves were used to determine the catalytic efficiency (kcat/Km) of X3. All assays were carried conducted in triplicate.
The TcXyn30B protein was overexpressed and secreted extracellularly by P. pastoris expression system. Tc Xyn30B was purified to homogeneity (Fig. S2). The average molecular mass of TcXyn30B from P. pastoris was determined as 62 182 Da by time‐of‐flight MS. This value was significantly higher than that of TcXyn30B (56 354 Da) produced using the T. cellulolyticus homologous expression system , meaning that glycosylation patterns between two proteins are different. The glycosylation patterns of TcXyn30B from P. pastoris used in the present study were assigned by X‐ray crystallography, as described below. Crystals of ligand‐free TcXyn30B were obtained by hanging‐drop vapor diffusion. Crystals of the TcXyn30B–U4m2 X complex were obtained by a co‐crystallization method under almost the same conditions as those used for the ligand‐free crystals (Fig. S3).
Both ligand‐free and ligand‐complexed TcXyn30B crystals belonged to the P212121 space group. Diffraction data statistics are shown in Table 1. The crystal structures of ligand‐free and ligand‐complexed enzymes were determined at resolutions of 1.60 and 1.65 Å, respectively, by molecular replacement, using TcXyn30B from T. cellulolyticus as the search model (PDB ID: http://6IUJ). One protein molecule was contained in an asymmetric unit. Amino acid residues numbered 20–473 and 18–473 for TcXyn30B without and with ligand were assigned with the electron density map, respectively. Amino acid residues 18‐22 (AYVEF) are from DNA sequence included in pPIC9K vector, whereas residues 23 or later are numbered as with native protein. U4m2 X was modeled at later stages of refinement, when the electron density was unambiguous (Fig. 2A). The overall structure of ligand‐free TcXyn30B from P. pastoris was almost the same as that of the TcXyn30B‐U4m2 X complex (0.131 Å rmsd over 447 Cα atoms) by a least‐squares superposition method . Similarly, there was no difference between the ligand‐free 3D‐structures of TcXyn30B from P. pastoris and T. cellulolyticus (0.550 Å rmsd over 447 Cα atoms).
|TcXyn30B||TcXyn30B with U4m2X|
|Resolution range (Å)||43.14–1.60 (1.66–1.60)a||33.41–1.65 (1.71–1.65)|
|a, b, c (Å)||63.34, 78.77, 117.84||63.25, 78.70, 118.42|
|Total reflections||526 004 (50 937)||317 528 (31 543)|
|Unique reflections||78 001 (7458)||71 087 (7027)|
|Multiplicity||6.7 (6.8)||4.5 (4.5)|
|Completeness (%)||99.6 (96.6)||99.5 (99.0)|
|Mean I/σ(I)||11.78 (2.00)||17.18 (2.55)|
|R‐merge||0.093 (0.801)||0.046 (0.458)|
|R‐pim||0.039 (0.332)||0.024 (0.239)|
|CC1/2||0.996 (0.660)||0.999 (0.847)|
|Reflections used in refinement||77 992 (7458)||71 083 (7027)|
|Reflections used for R‐free||3900 (373)||3554 (351)|
|R‐work||0.180 (0.334)||0.170 (0.242)|
|R‐free||0.202 (0.320)||0.193 (0.267)|
|CC (work)||0.957 (0.743)||0.957 (0.827)|
|CC (free)||0.950 (0.743)||0.961 (0.853)|
|Number of non‐hydrogen atoms||4175||4306|
|Sugar chains and ligands||187||249|
|Sugar chains and ligands||31.7||34.1|
In the electron density maps, N‐glycosylation of TcXyn30B from P. pastoris is observed at Asn‐60, Asn‐88, Asn‐334, Asn‐346 and Asn‐412 (Figs 2B and S4), whereas TcXyn30B from T. cellulolyticus is glycosylated at Asn‐60, Asn‐88, Asn‐215, Asn‐334, Asn‐346 and Asn‐412 . This is probably a result of differences in glycosylation mechanisms in the expression hosts. Comparison of the length of the sugar chain suggests that protein expressed by P. pastoris tends to possess a larger degree of polymerization than that expressed by T. cellulolyticus, although all sugar chains could not be assigned by electron density maps.
A clear density map for U4m2 X is observed in the active cleft (Fig. 2A). Two xylose units modeled in subsites ‐1 and ‐2a are named Xyl ‐1 and Xyl ‐2, respectively. MeGlcA is bound in subsite ‐2b (Fig. 2C). The subsite ‐2b is composed of seven amino acid residues (Fig. 3A). The side chains of five amino acid residues are concentrated near the C‐6 carboxyl group and the 4‐O‐methyl group of the MeGlcA substituent. The C‐6 carboxyl group of MeGlcA is suggested to form hydrogen bonds with Glu‐345 and Ser‐351. Arg‐46 appears to form salt bridge with the C‐6 carboxyl group, similar to an Arg residue conserved in GH30‐8 glucuronoxylanase (Arg‐293 of Ec XynA) (Fig. 3A,B) [7, 8], in agreement with our previous prediction using a superimposed model structures of Xyn30B based on the EcXynA model complexed with XU4m2 X .
The side chains of Glu‐345, Ser‐347, Thr‐349 and Ser‐351 from β8 and a β8‐β8A loop are located near the 4‐O ‐methyl group of MeGlcA (Figs 2C and 3A). The distances between the C‐atom of the 4‐O ‐methyl group and the O‐atoms of Glu‐345, Ser‐347, Thr‐349 and Ser‐351 are 3.5, 3.5, 3.9 and 3.3 Å, respectively, suggesting that a part of these residues and the methyl group may form C‐H…O type of hydrogen bonds, which is a common but underappreciated interaction in biomolecules and molecular recognition (Fig. 3A) . Glu‐345, which corresponded to Arg‐293 of Ec XynA, and Ser‐351 are highly conserved in other GH30‐7 endoxylanases (Fig. 1, highlighted in red) and are considered to play an important role in the recognition of both the C‐6 carboxyl and 4‐O‐methyl groups of MeGlcA. Moreover, Ser‐347 and Thr‐349 of TcXyn30B are partially conserved with polar residues in TtXyn30A, XYN VI and Penicillium purpurogenum XynC endoxylanase (Pp XynC) (Fig. 1, highlighted in red). By contrast, EcXynA and BsXynC, which lack a β8‐sheet structure composed of β8, β8A and β8B, have no structure involved in the recognition of a 4‐O ‐methyl‐group [7, 8]. EcXynA displays almost equivalent activity towards beechwood xylan and 4‐deoxy‐hexenuronosyl beechwood xylan, in which the methyl esters on the 4‐O ‐methyl glucuronic acid substituents are removed . These observations suggest that a methyl‐group recognition pocket is a unique feature of GH30‐7 endoxylanases with the β8‐sheet structure.
The orientation of the MeGlcA moiety bound to TcXyn30B is different from that of the moiety bound to Ec XynA (Fig. S5). The shifts of the C‐6 carboxyl groups and the methyl‐groups between MeGlcA moieties in two enzymes are 2.6 and 2.0 Å, respectively (Fig. S5), indicating that interactions of TcXyn30B with two functional groups significantly influence the substrate position.
The subsite ‐1 of Tc Xyn30B is composed of Trp‐141, Asn‐201, Glu‐202, Tyr‐209, Tyr‐279, Glu‐297, Leu‐301 and Trp‐341 (Fig. 4A). All of these residues except Leu‐301 are conserved in both GH30‐7 and GH30‐8 (Fig. 1, highlighted in green).
At subsite ‐2a of TcXyn30B, the Xyl ‐2 residue takes part in the stacking interaction with the aromatic ring of Tyr‐209 and hydrophobically interacts with Phe‐44 and Trp‐341, similarly to the Xyl ‐2 residue in the Ec XynA that takes part in the interaction with Tyr‐172, Trp‐55 and Trp‐289 (Fig. 4B,C). Asn‐93 in the β2‐α2 loop is a notable residue that is not observed in GH30‐8 xylanases (Figs 2C and 4B,C). The distances between the O3 and O4 atoms of Xyl ‐2 and the Nδ atom of Asn‐93 in Tc Xyn30B are 3.0 and 3.2 Å, respectively (Fig. 4B). This suggests that the xylobiohydrolase activity found in TcXyn30B can be attributed to the interaction between Xyl ‐2 at the non‐reducing end and Asn‐93. Xylobiohydrolase activity has also been reported in Tt Xyn30A . Pp XynC endoxylanase releases xylobiose from linear xylooligosaccharides . These two enzymes have Asp and Asn residues, respectively, corresponding to Asn‐93 of Tc Xyn30B (Fig. 1) and these residues may play a similar role to Asn‐93 of TcXyn30B with respect to the release of xylobiose. On the other hand, Asn‐93 does not appear to be conserved in Bispora sp. MEY‐1 XYLD endoxylanase, T. cellulolyticus Xyn30A exoxylanase (Tc Xyn30A) and XYN VI glucuronoxylanase [10, 23, 28]. Xyn30A and XYN VI possess shorter β2‐α2 loops than that of Tc Xyn30B (Fig. 1). The β2‐α2 loop of Tc Xyn30A was predicted to not protrude into the active site by homology modeling .
To evaluate the role of Asn‐93 in TcXyn30B, His‐tagged TcXyn30B (TcXyn30B‐His) and its mutant enzyme whose Asn‐93 was replaced by Ala (TcXyn30B‐His N93A) were prepared. The glucuronoxylanase activities of TcXyn30B‐His and TcXyn30B‐His N93A for beechwood xylan were 10.9 ± 0.5 and 12.0 ± 0.2 U·mg−1, respectively. By contrast, the xylobiohydrolase activities of TcXyn30B‐His and TcXyn30B‐His N93A for X3 were 0.290 ± 0.006 and 0.0987 ± 0.004 U·mg−1, respectively. The kinetic parameters for each enzyme activity of TcXyn30B‐His and Tc Xyn30B‐His N93A are shown in Table 2. The Km and kcat values for the xylobiohydrolase activity could not be determined because the initial rate of xylose production from X3 was not saturated even at a substrate concentration of 16 mm. However, the three‐fold reduction of the kcat/Km value in the N93A mutant suggests that Asn‐93 could contribute to increased catalytic efficiency for xylobiohydrolase activity but is not essential. These results also support the hypothesis that Asn‐93 interacts with the non‐reducing end of xylose residue at the subsite ‐2.
The glucuronoxylanase activity that releases XnU4m2X from xylan requires the binding of substrate at subsite ‐3 onward. When the TcXyn30B model was superimposed on the EcXynA model complexed with XU4m2X, a steric clash between non‐reducing end of XU4m2X and Asn‐93 of Tc Xyn30B was observed (Fig. 5A). On the other hand, the structural analysis of TcXyn30B‐ligand complex revealed that O4‐atom of Xyl ‐2 bound in TcXyn30B is located at a different position from that bound in Ec XynA, oriented toward a groove between Asn‐93 and Tyr‐209 (Fig. 5A,B, red‐dashed circle, and Fig. S5). This suggests that Xyl ‐3 is likely to fit in the groove formed, as predicted previously . However, the distance between Nδ of Asn‐93 and C of Tyr‐209 is calculated to only be 6.2 Å at the narrowest point. Because the van der Waals radii of Cβ of Tyr, C of xylose ‐3 and Oδ of Asn can be considered as 2.0, 1.7 and 1.6, respectively, the distance of the groove should be at least 7.0 Å for binding of Xyl ‐3 . Thus, the groove is too small for Xyl ‐3 to enter spontaneously. From these observations, we propose that a structural change of the groove, such as the flipping of Asn‐93 (Fig. 5B, indicated by an arrow) or a conformational change of the loop, will occur for the binding of substrate with a high degree of polymerization and plays an important role in the switching between xylobiohydrolase and glucuronoxylanase activity. Such a structural change to bind glucuronoxylan may also be facilitated by the strong recognition and orientation of the MeGlcA substituent at ‐2b.
In the present study, we demonstrated the crystal structure of TcXyn30B complexed with U4m2X and the unique mechanism for substrate recognition in GH30‐7. The structure revealed that TcXyn30B recognizes not only the C‐6 carboxyl group, but also the 4‐O‐methyl group of MeGlcA, unlike GH30‐8 enzymes. Residues interacting with these two functional groups are conserved in GH30‐7 endoxylanases. The enzyme–ligand complex model and site‐directed mutagenesis indicated that the interaction between Asn‐93 on the β2‐α2 loop and Xyl ‐2 residue is partially involved in xylobiohydrolase activity. Our results provide structural insight with respect to substrate recognition in GH30‐7 glucuronoxylanases and xylobiohydrolases.
YN and HI designed the study and mainly contributed to writing the manuscript. YN was responsible for the preparation and crystallization of the proteins. YN and MW performed X‐ray diffraction analysis and processed the data. YN was responsible for modeling and refinement of the crystal structures. AM and HI supervised the study. All authors read and approved the final manuscript submitted for publication.
This work was performed at the BL44XU synchrotron beamline at SPring‐8 under the Collaborative Research Program within the Institute for Protein Research at Osaka University (Hyogo, Japan; Proposal numbers 2018B6863 and 2019B6930). We thank the beamline staff (Drs Eiki Yamashita, Kenji Takagi and Keisuke Sakurai) for their assistance with the data collection. This work was supported by a Basic Research Funding on the National Institute of Advanced Industrial Science and Technology.