research papers
A GH115 α-glucuronidase structure reveals dimerization-mediated substrate binding and a proton wire potentially important for catalysis
aDepartment of Biotechnology and Biomedicine, Technical University of Denmark, Søltofts Plads 224, 2800 Kongens Lyngby, Denmark, and bSão Carlos Institute of Physics, University of São Paulo, Avenida Trabalhador Sãocarlense 400, 13566-590 São Carlos, SP, Brazil
*Correspondence e-mail: cwil@dtu.dk, premo@dtu.dk
Xylan is a major constituent of plant cell walls and is a potential source of biomaterials, and the derived α-D-glucuronic acid, which is linked to the O2′ position of the β-1,4 D-xylopyranoses composing the main chain of xylans. The xylan-specific α-glucuronidases from glycoside hydrolase family 115 (GH115) specifically catalyse the removal of α-D-glucuronic acid (GlcA) or methylated GlcA (MeGlcA). Here, the molecular basis by which the bacterial GH115 member wtsAgu115A interacts with the main chain of xylan and the indirect involvement of divalent ions in the formation of the Michaelis–Menten complex are described. A at 2.65 Å resolution of wtsAgu115A originating from a metagenome from an anaerobic digester fed with wastewater treatment sludge was determined in complex with xylohexaose, and Asp303 was identified as the likely general acid. The residue acting as the general base could not be identified. However, a proton wire connecting the active site to the metal site was observed and hence a previous hypothesis suggesting a Grotthuss-like mechanism cannot be rejected. Only a single molecule was found in the However, wtsAgu115A forms a dimer with a symmetry-related molecule in the The xylohexaose moieties of the xylohexaose are recognized by residues from both protomers, thus creating a xylohexaose at the dimer interface. The dimer was confirmed by analytical in solution. Kinetic analysis with aldouronic acids resulted in a Hill coefficient of greater than 2, suggesting cooperativity between the two binding sites. Three Ca2+ ions were identified in the wtsAgu115A structures. One Ca2+ ion is of particular interest as it is coordinated by the residues of the loops that also interact with the substrate. Activity studies showed that the presence of Mg2+ or Mn2+ resulted in a higher activity towards aldouronic acids, while the less restrictive coordination geometry of Ca2+ resulted in a decrease in activity.
have been shown to have prebiotic effects. Xylans can be highly substituted with different sugar moieties, which pose to the xylanases that catalyse the hydrolysis of the xylan backbone. One such substituent isKeywords: α-glucuronidases; dimerization; glycoside hydrolase family 115; metal dependence; Grotthuss mechanism.
PDB references: GH115 α-1,2-glucuronidase, 7pxq; complex with xylopentaose, 7pug
1. Introduction
Xylans are the main hemicellulosic constituents found in the lignified secondary cell walls of trees in particular and can account for up to a third of the dry weight of the plant cell wall (Scheller & Ulvskov, 2010). Enzymes that degrade and modify xylan are considered to be attractive for use in biorefining industries, where xylans have shown promise as biomaterials, for example in hydrogels for wound dressing and drug delivery and in packaging materials (Naidu et al., 2018), and the derived have been shown to have prebiotic effects (Santibáñez et al., 2021). A significant xylan component is glucuronic acid, a substituent that is found on many different types of xylan, where the O1 of α-D-glucuronic acid (GlcA) is linked to the O2′ position of the β-1,4-linked D-xylopyranose constituting the main chain of the xylan. The GlcA unit of xylan often carries methyl substitutions at the 4-O position (MeGlcA), which are believed to form interaction points for lignin and cellulose (Peña et al., 2016; Bromley et al., 2013). Xylan-specific α-glucuronidases (EC 3.2.1.131) belong to glycoside hydrolase family 115 (GH115) in the Carbohydrate Active enZyme database (CAZy; https://www.cazy.org; Drula et al., 2021). Xylan-specific α-glucuronidases specifically act on the α-D-glucuronic acid (GlcA or MeGlcA) substituents present in polymeric and oligomeric xylan (Ryabova et al., 2009; Rogowski et al., 2014; Chong et al., 2011), while other accessory enzymes of the GH115 family act on different substituents such as ferulic acid, arabinose and acetyl groups (Yan et al., 2021; Holck, Fredslund et al., 2019; Kmezik et al., 2020). These substituents make xylans recalcitrant to enzymatic degradation (Lyczakowski et al., 2017; Kmezik et al., 2020; Couturier et al., 2018) and hence should be hydrolysed before β-1,4-xylanases (EC 3.2.1.8) and β-1,4-xylosidases (EC 3.2.1.3) can completely degrade the xylan (Holck, Djajadi et al., 2019; Kmezik et al., 2020). Thus, GlcA or MeGlcA substituents present a bottleneck in the degradation of xylan, and as such are crucial targets for biorefineries relying on the enzymatic degradation and utilization of unsubstituted xylan in biomaterials.
Four crystal structures of GH115 α-glucuronidases have been determined to date (Yan et al., 2021; Rogowski et al., 2014; Aalbers et al., 2015; Wang et al., 2016). However, only the structure of the GH115 member from Bacteroides ovatus (BoAgu115A) was obtained in a complex with a ligand, D-guluronic acid. In combination with comprehensive mutational analysis, the specific location of D-guluronic acid was identified to be the active site (Rogowski et al., 2014; Wang et al., 2016). The crystal structures of BoAgu115A and other GH115 orthologs from Saccharophagus degradans (SdeAgu115A) and Amphibacillus xylanus (AxyAgu115A) are considered to be active dimers (Yan et al., 2021; Rogowski et al., 2014; Wang et al., 2016). In BoAgu115A, a structural and mutational study identified the residues at the dimer interface that are involved in binding the xylan (Rogowski et al., 2014). Only the GH115 ortholog from Bacteroides thetaiotaomicron (BtGH115A) is believed to function as a monomer (Aalbers et al., 2015). The importance of dimerization in the catalytic mechanism remains elusive, in part due to a lack of structural information on complexes with xylo-oligosaccharides that mimic the xylan backbone.
BtGH115A is the only characterized GH115 member that shows activity on arabinogalactan, from which it can release α-GlcA, according to Aalbers et al. (2015); however, GlcA is linked to the galactose in arabinogalactan via a β-linkage (Tryfona et al., 2012). It does not show activity on GlcA or MeGlcA linked to xylan (Aalbers et al., 2015). Should BtGH115A be active on a β-linkage this would be highly unusual, as families in CAZy typically display activity on either α-linkages or β-linkages (The CAZypedia Consortium, 2018).
Several metal ions have been observed in the crystal structures of GH115 members, Na+ in the of BoAgu115A, Ni2+ in that of BtGH115A and Na+ in that of SdeAgu115A (Rogowski et al., 2014; Aalbers et al., 2015; Wang et al., 2016), but these have received little or no attention as their potential role in the catalytic mechanism has not previously been investigated.
In this study, we report crystal structures of a bacterial GH115 ortholog (wtsAgu115A) identified in a metagenome from an anaerobic digester fed with wastewater treatment sludge (Wilkens et al., 2017). Crystal structures with and without bound xylohexaose are presented. The xylohexaose-bound forms a Michaelis–Menten-like complex, which reveals how the dimer interface of the enzyme orients xylan in the active site guided by residues from each protomer of the homodimer. Further, wtsAgu115A displayed increased activity in the presence of divalent metal ions. In the a Ca2+ ion was identified in proximity to the active site, which coordinates some of the residues interacting with the xylohexaose and is also part of a proton wire passing through the active-site cavity. While Ca2+ resulted in a slightly decreased activity, the addition of Mg2+ and Mn2+ to the enzymatic assays resulted in increased activity, indicating the advantages of a more rigid coordination geometry around the divalent metal site for optimal activity. We hypothesize that the proton wire potentially delivers the proton needed for catalysis.
2. Materials and methods
2.1. Sequence analysis
The gene encoding wtsGH115A (GenBank Accession No. BK059344) was identified on contig MTKX01000768.1, base pairs 45336–47131 in a metagenomic study of anaerobic digesters Randers (Whole Genome Shotgun Accession No. MTKX00000000) fed with surplus sludge from municipal wastewater treatment, where it was predicted to originate from the bacterial genus Sulfurovum (Wilkens et al., 2017). A 19-amino-acid signal peptide was identified by SignalP 4 (Petersen et al., 2011). The molecular mass and pI were predicted by Compute pI (Gasteiger et al., 2005): the pI was 4.7 and the molecular mass was 97.1 kDa. The theoretical molar extinction coefficient, calculated using ProtParam (Gasteiger et al., 2005), was 184 040 M−1 cm−1.
2.2. Multiple alignment and phylogenetic analysis
The identity of wtsGH115A to other GH115 members was calculated with Clustal Omega (Li et al., 2015), which was also used for the multiple sequence alignment of all GH115 members listed in CAZy from which fragments were removed. The multiple alignment was used to build the LGmaximum likelihood phylogenetic tree using RaxML-HPC BlackBox (version 8.2.10; Stamatakis et al., 2008) at the CIPRES Science Gateway version 3.3 (Miller et al., 2010). RAxML stopped the rapid bootstrap search after 250 replicates with the MRE-based bootstopping criterion.
2.3. Cloning, expression and purification
A codon-optimized mature gene (Supplementary Fig. S1) for Escherichia coli encoding wtsAgu115A was purchased and cloned into pET-28a using the NcoI and XhoI restriction sites (GenScript). The resulting plasmid was transformed into E. coli strain BL21 (DE3) pLysS (Novagen). Transformants were grown in LB medium (2 × 500 ml, 37°C, 18 h), harvested at 2000g for 10 min at room temperature (RT), resuspended in 100 ml M9 medium (see below) and used to inoculate M9 medium (4 l) supplemented with 1%(w/v) glucose, 2 ml l−1 trace elements (Holme et al., 1970) and 50 µg ml−1 kanamycin in a 5 l bioreactor (Sartorius Biostat Aplus). Glucose [50%(w/v)] supplemented with 10 ml l−1 trace elements (Holme et al., 1970) was used as a carbon source during the fed-batch phase. The culture was propagated at 37°C until the OD600 reached 12, whereupon the temperature was decreased to 15°C and expression was induced by the addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 1 mM. The pH was maintained at 7 by adding 28%(w/v) ammonia; the temperature and dissolved oxygen tension were controlled automatically. After 16–18 h, the cells were harvested (2000g, 20 min, 4°C) and stored at −20°C. The cells were resuspended in buffer A (50 mM Tris–HCl, 0.5 M NaCl, 20 mM imidazole pH 7.5) using 10 ml per gram of wet weight cells, lysed (Pressure Cell Homogeniser; Stansted Fluid Power) at RT, treated with benzonase nuclease (Invitrogen) for 30 min at RT and centrifuged (20 000g, 30 min, 4°C). The supernatant was filtrated (0.45 µm Durapore membrane filters; Millipore), applied onto a 5 ml HisTrap HP column (GE Healthcare) at a flow rate of 3 ml min−1 and eluted with a linear imidazole gradient (0.02–0.5 M) in buffer A. Fractions containing wtsGH115A were pooled, dialysed [SnakeSkin (3.5 kDa), Thermo Scientific] into 10 mM sodium acetate pH 6 and applied onto a 6 ml Resource Q column (GE Healthcare) equilibrated with 10 mM sodium acetate pH 6 at a flow rate of 3 ml min−1 and eluted with a linear 0–0.5 M NaCl gradient (30 column volumes). The fractions containing wtsGH115A were pooled, concentrated [Vivaspin (10 kDa), Sartorius] and then applied at a flow rate of 0.5 ml min−1 onto a HiLoad 16/60 Superdex G200 column (GE Healthcare) equilibrated with 10 mM sodium acetate, 150 mM NaCl pH 6. The fractions containing wtsAgu115A were concentrated to 139 µM [Vivaspin (10 kDa), Sartorius] and stored at 4°C. All chromatographic steps were carried out at RT. The purity was checked on 12% SDS–PAGE gels (Supplementary Fig. S2). The concentration of the protein samples was measured as the absorption at 280 nm (A280) using an extinction coefficient of 184 040 M−1 cm−1.
An inactive mutant, D303A, was constructed using CloneAmp polymerase (Takara, Kusatsu, Japan), a set of mutagenic primers (Supplementary Table S1) and pET-28a/GH115 WT as a template. The template was digested by DpnI at 37°C overnight and the PCR products were purified using the Illustra GFX PCR DNA and Gel Band Purification Kit (GE Healthcare Life Sciences, Chicago, Illinois, USA). Purified pET-28a/GH115 D303A was transformed into E. coli DH5α for replication and plated on LB plates supplemented with kanamycin (50 µg ml−1). Positive transformants were selected and the corresponding plasmids were extracted using a GeneJET Plasmid Miniprep Kit (ThermoFisher Scientific). All constructs were checked by sequencing (Macrogen Europe, Amsterdam, Netherlands). The D303A mutant was expressed and purified as described for the wild type.
2.4. Analytical size-exclusion chromatography
Analytical gel filtration was performed on a HiLoad 16/60 Superdex G200 column equilibrated with aldolase (158 kDa), conalbumin (75 kDa) and ovalbumin (44 kDa) (all from GE Healthcare) dissolved in 10 mM sodium acetate, 150 mM NaCl pH 6 at a flow rate of 0.5 ml min−1 with 10 mM sodium acetate, 150 mM NaCl pH 6 as the running buffer.
2.5. Temperature optima and pH
A stock solution of an aldouronic acids mixture (Megazyme; 20 µl) was diluted to 1% in 40 mM Britton–Robinson universal buffers pH 2–10, 0.005% Triton X-100, mixed with 5 µl 24 µM wtsGH115A and incubated for 10 min at 37°C to determine the pH optimum. Prior to mixing, the reaction mixture and wtsAgu115A were pre-incubated for 2 min at 37°C. The reaction was stopped by adding 25 µl 1 M trichloroacetic acid (TCA). According to the manufacturer's recommendations, the release of GlcA was quantified by the D-Glucuronic/D-Galacturonic Acid Assay Kit (Megazyme). In brief, D-glucuronic acid is oxidized to D-glucarate by uronate dehydrogenase in the presence of NAD+, resulting in the formation of NADH, which can be detected at A340. The formation of NADH is stoichiometric with D-glucuronic acid. One activity unit (U) was defined as the amount of enzyme releasing 1 µmol GlcA per minute. The temperature optimum was determined at pH 6 in the range 20–60°C as above. All experiments were performed in triplicate.
2.6. Temperature stability and pH
wtsGH115A (20 µl at 24 µM) was incubated at RT in 80 µl 40 mM Britton–Robinson universal buffers pH 2–10 for five days and the residual activity was quantified as described in Section 2.5. Similarly, the temperature stability was assessed by incubating 24 µM wtsGH115A in 10 mM sodium acetate pH 6 at 37 and 50°C, respectively, and the residual activity was quantified as above.
2.7. Effect of divalent metal ions
The effect of divalent ions on −1 aldouronic acids, 2 mM divalent ions (CaCl2, MgCl2, ZnCl2, MnCl2, NiCl2 or FeCl2) and 2.5 µM GH115 at 37°C in 50 mM sodium acetate, 0.005% Triton X-100 pH 6. The reaction was stopped at regular time intervals by adding one volume of 1 M TCA. The release of GlcA was quantified as above. Prior to analysis, the enzyme was first incubated with 1 mM EDTA for 10 min to exclude the possibility of any metal being bound and EDTA was then removed by dialysis at 1:200(v:v) overnight twice.
was investigated by measuring the activity of the enzyme using 8.5 mg mlThe effect of MgCl2 on the was further investigated under standard assay conditions using MgCl2 concentrations varying from 0 to 10 mM.
2.8. Specific activity
For M sodium acetate, 0.005% Triton X-100 pH 6 and was mixed with 5 µl 24 µM GH115 to initiate the reaction. The reaction mixture was left to incubate for 10 min at 37°C. The reactions were stopped by adding 25 µl 1 M TCA. The release of GlcA was quantified as above.
studies, the following substrates were prepared: 1% potato pectic fibre rhamnogalacturonan I, linear arabinan, potato galactan, soybean pectic fibre rhamnogalacturonan, beechwood xylan, sugar beet arabinan, rye arabinoxylan (high viscosity) and wheat arabinoxylan (insoluble), aldouronic acids, tamarind xyloglucan (all from Megazyme), larch wood arabinogalactan, xanthan gum, birchwood xylan, oat spelt xylan, acacia tree gum arabic (all from Sigma) and corncob xylan (Carbosynth). The mixture was dissolved or suspended depending on the solubility of the compound in 20 µl 50 m2.9. Kinetic parameters
To determine the kinetic parameters of the enzyme, initial velocities were quantified as in Section 2.5 for aldouronic acid mixtures with concentrations varying from 0.5 to 6 mg ml−1 in 50 mM sodium acetate, 0.005% Triton X-100 pH 6 at 37°C using 2.5 µM of the enzyme. The release of GlcA was quantified as described above. Kinetic parameters were determined by plotting initial velocities against initial substrate concentration and fitting the Hill equation,
using the Origin 2019 software (OriginLab), where V0 is the initial velocity, [S0] is the initial substrate concentration, Vmax is the maximum rate, n is the Hill coefficient and k is the half-maximal concentration constant.
2.10. Crystallization, data collection and data processing
The proteins wtsAgu115A and wtsAgu115A D303A (10–20 mg ml−1) in 10 mM sodium acetate, 150 mM NaCl pH 6 were crystallized in SuperClear plates (Jena Bioscience) by mixing 1 µl at 10 mg ml−1 with 1 µl reservoir solution consisting of 0.2 M CaCl2, 0.1 M HEPES pH 7.5, 28% PEG 400. Crystals containing xylohexaose (Megazyme) were obtained by adding a few xylohexaose crystals to the drops, which were then allowed to equilibrate overnight. The crystals were cryoprotected with 20% PEG 400 in reservoir solution for 5 s and cryocooled in liquid nitrogen. Data were collected on BioMAX at the MAX IV Laboratory (Ursby et al., 2020) and on P14 at PETRA III. Data collection on BioMAX was performed using MxCUBE3 (Mueller et al., 2017) and the data were processed and scaled with xia2 (Winter, 2010) to 2.30 Å resolution for the unbound protein and 2.65 Å resolution for the xylohexaose complex (Table 1). The xylohexaose geometry was validated using Privateer (Agirre et al., 2015).
|
2.11. Phasing and refinement
The structure of unbound wtsAgu115A D303A was determined by Phaser (McCoy et al., 2007) from the Phenix package (Liebschner et al., 2019) using BoAgu115A (PDB entry 4c90) as the search model. This was the closest homolog from the PDB (https://www.wwpdb.org/) identified at the time through PDB-BLAST (Altschul et al., 1997), with a coverage of ∼98% and an identity of ∼67%. The initial structure was built with phenix.autobuild (Terwilliger et al., 2008) and then refined iteratively with phenix.refine (Afonine et al., 2012) and manual model rebuilding in Coot (Emsley et al., 2010) to a final Rwork and Rfree of 0.20 and 0.22, respectively. The wtsGH115A–xylohexaose structure (PDB entry 7pug) was determined using unbound wtsAgu115A D303A (PDB entry 7pxq) as the search model and was refined to a final Rwork and Rfree of 0.17 and 0.21, respectively.
with3. Results and discussion
3.1. Overall structure and comparison to other GH115 structures
The structure of the inactive mutant wtsAgu115A D303A in its unbound form was determined at 2.33 Å resolution (Table 1) with a single protomer in the The structures of wtsAgu115A and BoAgu115A show an equivalent tertiary architecture (Figs. 1a and 1b) comprised of four consecutive domains, with the N-terminal domain A comprised of two perpendicular β-sheets and three α-helices (residues 1–169), domain B composed of a (β/α)8 TIM barrel (residues 170–468), domain C composed of an α-helix bundle (residues 474–617) and the C-terminal domain D composed of a canonical β-sandwich fold that includes two β-sheets of five antiparallel β-strands (residues 656–838) (Fig. 1a). Domains B and C are connected by a five-residue loop and domains C and D are connected by a 38-residue loop (Fig. 1a). The root-mean-square deviation (r.m.s.d.) for the Cα atomic coordinates of wtsAgu115A and BoAgu115A was 0.49 Å (642 atom pairs). The GH115 members SdeAgu115A and AxyAgu115A have an additional domain, C+, which is not present in BoAgu115A and wtsAgu115A, and also show a different domain organization (Fig. 1c; Wang et al., 2016; Yan et al., 2021).
Both the unbound wtsAgu115A D303A structure and the wtsAgu115A–xylohexaose structure (2.65 Å) were refined to excellent crystallographic and geometric statistics, as summarized in Table 1. Three calcium ions were assigned to high electron-density sites found on the surface of domain B and domain C and connected to the active site in domain B (Fig. 1a). Overall, the electron density is well defined in both unbound and xylohexaose-bound wtsAgu115A except for residues 301–318 and 307–317, respectively, suggesting that these residues constitute a flexible loop. The corresponding loops are also suggested to be flexible in BoAgu115A, AxyAgu115A and SdeAgu115A and are believed to interact with the substrate (Wang et al., 2016; Rogowski et al., 2014; Yan et al., 2021).
3.2. The Michaelis–Menten-like complex between wtsAgu115A and xylohexaose
wtsAgu115A displayed the highest ), demonstrating that wtsAgu115A is a xylan-specific α-glucuronidase. Apparent and extended electron density for five β-1,4-linked D-xylopyranose moieties was observed in the of wtsAgu115A soaked with xylohexaose, which was positioned in the central cavity of domain B (Figs. 2a and 2b). Superimposition with BoAgu115A in complex with GlcA (PDB entry 4c91; Rogowski et al., 2014) indicates that O2 of the +1 D-xylopyranose moiety would form the linkage to GlcA in the active-site pocket (Fig. 3), and thus this is subsite +1 according to McKee et al. (2012). The D-xylopyranose moieties adopt the stable 4C1 chair conformation (Stortz, 2010). However, the xylohexaose deviates from the 31-fold helical screw conformation observed for xylan in solution and the 21-fold helical screw found on the surface of cellulose (Martínez-Abad et al., 2017; Busse-Wicher et al., 2014). The conformation of the xylan backbone is commonly expressed by two dihedral angles (φ and ψ) that refer to the rotations around the two C—O bonds that form the glycosidic linkages. The sum of φ (O5′—C1′—O4—C4) and ψ (C1′—O4—C4—C3) is indicative of the and is ∼50° for a right-handed twisted 31-fold helical screw conformation and ∼190° for the left-handed conformation (Busse-Wicher et al., 2014). The sum of φ and ψ for the glycosidic linkages between the xylose moiety pairs at +2NR and +1NR, at +1NR and +1, at +1 and +1R and at +1R and +2R are 47°, 27°, 57° and 13°, respectively, which indicates that the xylan backbone is forced out of conformation upon forming the Michaelis–Menten complex. A structural comparison between unbound wtsAgu115A D303A and wtsAgu115A–xylohexaose yields an overall r.m.s.d. of 0.17 Å for Cα atoms, which suggests that the binding of xylohexaose does not impose any major domain conformational changes in wtsAgu115A.
towards an aldouronic acids mixture, followed by beechwood and birchwood xylan, while no activity was observed on arabinogalactan, xanthan gum or gum arabic (Table 2
|
A particular D-xylopyranose moiety engages in CH–π stacking with Trp633 (Fig. 2c). The tryptophan is fully conserved among structurally determined GH115 xylan-specific α-glucuronidases, but not among all GH115 members. However, this particular area of multiple sequence alignment is poor, making it difficult to decide which residues have replaced Tyr633 (see Supplementary File S1). In wtsAgu115A Trp633 ensures that the D-xylopyranose moiety at +1R is perpendicular to the active-site pocket (Fig. 2c). The D-xylopyranose moiety at +2R forms an additional CH–π stacking with Tyr771 from the other protomer in the dimer (Figs. 2a, 2b and 2c). In BoAgu115A, an alanine mutant of the equivalent tyrosine resulted in a fivefold reduction in catalytic efficiency (Rogowski et al., 2014), which demonstrates the importance of the dimerization observed for BoAgu115A (Rogowski et al., 2014) and also wtsAgu115A (Fig. 2c). However, Tyr771 is not conserved in all GH115 xylan-specific α-glucuronidases (see Supplementary File S1), and structurally they deviate significantly from the equivalent in SdeAgu115A and AxyAgu115A that contain the C+ domain (Figs. 1a–1c). Tyr771 is conserved within the clades of the phylogenetic tree containing wtsAgu115A and BoAgu115A (Fig. 3 and Supplementary File S1). Furthermore, wtsAgu115A eluted from a gel-filtration column at a volume consistent with a dimer in solution (Supplementary Fig. S3), which suggests that wtsAgu115A is a dimer in its active form similar to other GH115 xylan-specific α-glucuronidases (Wang et al., 2016; Yan et al., 2021; Rogowski et al., 2014). The Hill coefficients of slightly above 2 (Table 3) suggest that there is cooperation between the two active sites of the dimer.
formed by the dimer is located at position +1NR; the
|
Asp303 Oɛ2 is in position to form a hydrogen bond to O2 of the +1 D-xylopyranose moiety (Fig. 2c). This places Asp303 Oɛ2 close to the anomeric C atom and scissile bond of GlcA (we estimate that the distance is ∼2.7 Å; Fig. 2c), which suggests that Asp303 acts as the catalytic acid, a role that was assigned to Asp206 in BoAgu115A (corresponding to Asp177 in wtsAgu115A; Rogowski et al., 2014).
In BoAgu115A, it was speculated that Arg299 (Arg328 in BoAgu115A) could play a role in substrate binding, which would have required the enzyme to undergo a conformational change (Rogowski et al., 2014). The structural data presented for wtsAgu115A do not confirm this claim. An alternative role for Arg299 is in stabilizing the conformation of the putative catalytic acid Asp303, a residue that has also been suggested to be the catalytic acid in BoAgu115A, thus explaining the almost complete loss of activity when this alanine was substituted by an arginine in BoAgu115A (Rogowski et al., 2014).
Additionally, multiple residues interact with the xylohexaose, which may ensure its optimal orientation, enabling cleavage of the GlcA linkage to O2 of the +1 D-xylopyranose moiety. These include potential hydrogen bonds between Asp221 Oɛ1 and O2 or O3 of the +2R D-xylopyranose, between Tyr408 Oη and O3 of the +1 D-xylopyranose and between Trp220 O and O3 of the +2R D-xylopyranose, which also stacks against +1NR (Fig. 2b).
3.3. A divalent ion is involved in the formation of the Michaelis–Menten complex and possibly in catalysis
Sites for Na+ and Ni2+ metal ions have been identified in the crystal structures of BoAgu115A and BtAgu115A, respectively (Aalbers et al., 2015; Rogowski et al., 2014). However, the influence of the metal ions on the activity of the enzyme was not investigated.
Three Ca2+ ions were modelled in wtsAgu115A D303A in the unbound form. However, two of these are found on the surface of the enzyme (Fig. 1a) and are likely to originate from the crystallization conditions, which contained 0.2 M CaCl2. Ca2+ site 1 is buried within the structure near the active site (Figs. 1a and 4) in a position equivalent to Ni2+ in BtAgu115A and Na+ in BoAgu115A. Increased activity on an aldouronic acids mixture consisting of (Me)GlcA-xylooligosaccharides of different lengths was observed for EDTA-treated wtsAgu115A in 2 mM Mg2+ (∼40%) and Mn2+ (∼50%), while the presence of 2 mM Ca2+ resulted in a ∼10% decrease in activity (Supplementary Fig. S4). A kinetic analysis using an aldouronic acids mixture as a substrate surprisingly only resulted in a modest increase in kcat,app (14%) in the presence of 2 mM Mg2+ after EDTA treatment compared with EDTA-treated wtsAgu115A (Table 3 and Supplementary Fig. S5). Unfortunately, we were unable to obtain structures in complex with (Me)GlcA and aldouronic acids. Superimposition with GlcA from BoAgu115A (PDB entry 4c91) suggests that Asp177, which is involved in the coordination of Ca2+ in wtsAgu115A, could interact with (Me)GlcA. Thus, it can be speculated that the smaller Ca2+ results in increased distances between the substrate and the residues that interact with it, leading to a slightly decreased activity.
Ca2+ site 1 is found to have an octahedral coordination, which in the first coordination sphere is coordinated by Lys416 O and Asp448 Oδ2 together with four water molecules (water molecules 4, 7, 8 and 16; Fig. 4). The aforementioned water molecules can form hydrogen bonds to Asp177 O (3.1 Å) and Asp407 Oδ1 (2.6 Å) or Oδ2 (2.9 Å) (water molecule 4), Lys416 O (3.0 Å) and Asp448 Oδ2 (3.4 Å) (water molecule 7), Asp407 Oδ2 (2.6 Å) and Lys416 O (2.8 Å) (water molecule 8), and Asp448 Oδ2 (3.5 Å), Trp417 O (2.8 Å) and Lys416 O (3.0 Å) (water molecule 16) in the second coordination sphere (Fig. 4). The residues involved in Ca2+ site 1 are structurally conserved in BoAgu115A, but the most striking difference in BtAgu115A is that Ser395 is located at the position of Asp407 and the water molecules involved in the coordination of the two different metal ions naturally also differ. Ni2+ in BtAgu115A is pentahedrally coordinated (Aalbers et al., 2015). In structurally determined GH115 members consisting of five domains the metal-binding sites are not structurally conserved, which corresponds to the lack of metal ion dependency of the activity for these proteins (Yan et al., 2017; Wang et al., 2016).
Two of the three loops (residues 175–179 and 406–421) involved in the coordination of Ca2+ site 1 are also involved in substrate binding with Tyr408 Oη, potentially forming a hydrogen bond to O3 of the +1 D-xylopyranose moiety, which seems to be essential for positioning the scissile bond in the vicinity of the general acid (suggested to be Asp303). Asp177, which was suggested to be the general acid in BoAgu115A (Rogowski et al., 2014), is involved in coordination of Ca2+ site 1 (Fig. 4) and most likely also interacts with the GlcA moiety. Thus, Ca2+ site 1 is directly involved in orientating some of the residues that play a central part in forming the Michaelis–Menten complex.
It was suggested that the residue in BoAgu115A (Asp332) corresponding to Asp303 of wtsAgu115A could be the catalytic base (Rogowski et al., 2014). However, the structure of wtsAgu115A in complex with xylohexaose suggests, as mentioned, that Asp303 is the catalytic acid, and Asp177 could therefore be the catalytic base. The distance between the scissile bond and Asp177 Oδ1 is 6.4 Å (Fig. 2c), which is within the distance of 6–11 Å normally observed in inverting glycoside (McCarter & Withers, 1994). However, from Fig. 4 it seems more plausible that Asp177 would form hydrogen bonds to GlcA. Further, an alanine mutant of the equivalent to Asp177 in BoAgu115A (Asp206) resulted in a 300-fold reduction in catalytic efficiency (Rogowski et al., 2014), which would be expected to be greater if Asp177 were the catalytic base. For BoAgu115A it was also mentioned that GH115 xylan-specific α-glucuronidases could employ a Grotthuss-like mechanism (Rogowski et al., 2014): a proton-transport mechanism in which a proton diffuses through a hydrogen-bond network of water molecules (von Grotthuss, 1806; Agmon, 1995). If this were the case, a distant residue could activate the active-site water which was also suggested to be the case for cellulases in glycoside hydrolase family 6 (Brás et al., 2011; Mayes et al., 2016). The wtsAgu115A–xylohexaose structure revealed a proton wire leading from Ca2+ site 1 to Tyr403 through the active-site cavity via water molecules w427, w446 and w484 (Fig. 5; see Supplementary Fig. S6 for an omit map) and likely beyond Ca2+ site 1. Thus, the proton required for catalysis could be delivered through this proton wire. Interestingly, upon formation of the Michaelis–Menten-like complex, Tyr403 moved nearly 170° and the nearby histidine residues His246 and His247 also underwent conformational changes (Fig. 5), indicating the importance of Tyr403 in particular in catalysis. This is supported by the 10−3 loss in activity observed for an alanine mutant in BoAgu115A (Rogowski et al., 2014). In both BoAgu115A and BtAgu115A the corresponding tyrosine is in a conformation similar to that observed in wtsAgu115A in complex with xylohexaose; thus, the movement is not necessarily dependent on the substrate interaction. Further studies are needed to determine how the proton travels via the proton wire and thus determine the exact role of Tyr403 in catalysis. It would be interesting to see how Mg2+ and Mn2+ affect the divalent metal-binding site and thus substrate binding and the proton wire.
3.4. GH115 phylogeny and metal-binding sites
A phylogenetic tree was constructed to allow an analysis of the grouping of GH115 members in relation to their metal dependence. The xylan α-1,2-glucuronidases are grouped together in the lower clade (Fig. 3), while BtGH115A from B. thetaiotaomicron shown to hydrolyse (Me)GlcA on arabinogalactans is found in the left clade (Fig. 3). wtsAgu115A is found in a clade distant from other xylan α-1,2-glucuronidases together with BoAgu115A. BoAgu115A and wtsAgu115A segregate on neighbouring clades (Fig. 3). Most of the interactions with Ca2+ site 1 in wtsAgu115A are with backbone atoms (Fig. 4). However, for Asp407 it is the side-chain atoms that coordinate the water molecules around Ca2+ site 1, which together with the fact that in the Ni+-containing BtAgu115A Asp407 is replaced by Ser395, suggests that the presence of aspartic acid in this position could be an indication of a divalent metal-binding site. This is the case for about two thirds of the CAZy GH115 members (Fig. 3), but unfortunately the metal dependence was not tested for BoAgu115A (Rogowski et al., 2014) or SpAgu115A from Streptomyces pristinaespiralis (Fujimoto et al., 2011), which are suggested to have a divalent metal-binding site (Fig. 3).
Although BoAgu115A and wtsAgu115A are found in neighbouring clades, an exciting difference between the two clades involves the loop harbouring the suggested general acid Asp303. The 15 sequences constituting the wtsAgu115A clade all share a unique 12-amino-acid insert (K-E-G-E-D-D/H-L/K-Y/F-V-P/S-R/S-D) in the loop mentioned above (Fig. 3). A single sequence (GenBank Accession No. AVM57875.1) from Prevotella heparinolyticus that segregates from the wtsAgu115A and BoAgu115A clades (Fig. 3) contains an insert very similar to that found for the wtsAgu115A clade members (E-E-G-K-D-H-E-Y-V-A-R-Y; Fig. 3). Yan et al. (2021) also observed variance in both the amino-acid composition and the length of the aforementioned loop and suggested that it determines the substrate specificity of GH115 members and indicates that GH115 members with longer flexible loops can accommodate highly substituted xylans. Further studies involving such substrates are needed to confirm this hypothesis.
4. Biochemical characterization
The pH optimum of wtsAgu115A using aldouronic acids as a substrate was pH 6 (Supplementary Fig. S7a). After incubation for five days, wtsAgu115A retained activity in the pH range 6–11, while no activity was detected in the pH range 2–5 after five days of incubation at room temperature (Supplementary Fig. S7b).
The temperature optimum of wtsAgu115A using aldouronic acids was determined to be close to 40°C (Supplementary Fig. S7c). This is not surprising as wtsAgu115A was identified in the metagenome from a mesophilic anaerobic digester (Wilkens et al., 2017). 90 ± 4% of the activity of wtsAgu115A was retained while incubating at 37°C for 23.3 h; the activity then decreased to 67 ± 4% after 78.1 h (Supplementary Fig. S7d). wtsAgu115A retained 4 ± 0.7% of its activity after 5 min incubation at 50°C, while no activity was detected after 10 min incubation at 50°C.
Supporting information
PDB references: GH115 α-1,2-glucuronidase, 7pxq; complex with xylopentaose, 7pug
Supplementary File S1. Alignment file for glycoside hydrolase family 115. DOI: https://doi.org/10.1107/S2059798322003527/lp5057sup1.txt
Supplementary Figures. DOI: https://doi.org/10.1107/S2059798322003527/lp5057sup2.pdf
Acknowledgements
We acknowledge MAX IV Laboratory for time on the BioMAX beamline under Proposal MX20190334. Research conducted at MAX IV, a Swedish national user facility, is supported by the Swedish Research council under contract 2018-07152, the Swedish Governmental Agency for Innovation Systems under contract 2018-04969 and Formas under contract 2019-02496. We acknowledge synchrotron beamline P14 operated by EMBL Hamburg at the PETRA III storage ring (DESY, Hamburg, Germany), on which data were also collected.
Funding information
We thank the Danish Agency for Science, Technology and Innovation for funding the instrument centre DanScatt, which supported our usage of beam time at the synchrotrons, and the Technical University of Denmark is acknowledged for providing funding for the study.
References
Aalbers, F., Turkenburg, J. P., Davies, G. J., Dijkhuizen, L. & Lammerts van Bueren, A. (2015). J. Mol. Biol. 427, 3935–3946. CrossRef CAS PubMed Google Scholar
Afonine, P. V., Grosse-Kunstleve, R. W., Echols, N., Headd, J. J., Moriarty, N. W., Mustyakimov, M., Terwilliger, T. C., Urzhumtsev, A., Zwart, P. H. & Adams, P. D. (2012). Acta Cryst. D68, 352–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Agirre, J., Iglesias-Fernández, J., Rovira, C., Davies, G. J., Wilson, K. S. & Cowtan, K. D. (2015). Nat. Struct. Mol. Biol. 22, 833–834. Web of Science CrossRef CAS PubMed Google Scholar
Agmon, N. (1995). Chem. Phys. Lett. 244, 456–462. CrossRef CAS Web of Science Google Scholar
Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Nucleic Acids Res. 25, 3389–3402. CrossRef CAS PubMed Web of Science Google Scholar
Brás, J. L. A., Cartmell, A., Carvalho, A. L. M., Verzé, G., Bayer, E. A., Vazana, Y., Correia, M. A. S., Prates, J. A. M., Ratnaparkhe, S., Boraston, A. B., Romão, M. J., Fontes, C. M. G. A. & Gilbert, H. J. (2011). Proc. Natl Acad. Sci. USA, 108, 5237–5242. PubMed Google Scholar
Bromley, J. R., Busse-Wicher, M., Tryfona, T., Mortimer, J. C., Zhang, Z., Brown, D. M. & Dupree, P. (2013). Plant J. 74, 423–434. CrossRef CAS PubMed Google Scholar
Busse-Wicher, M., Gomes, T. C. F., Tryfona, T., Nikolovski, N., Stott, K., Grantham, N. J., Bolam, D. N., Skaf, M. S. & Dupree, P. (2014). Plant J. 79, 492–506. CAS PubMed Google Scholar
Chong, S.-L., Battaglia, E., Coutinho, P. M., Henrissat, B., Tenkanen, M. & de Vries, R. P. (2011). Appl. Microbiol. Biotechnol. 90, 1323–1332. CrossRef CAS PubMed Google Scholar
Couturier, M., Ladevèze, S., Sulzenbacher, G., Ciano, L., Fanuel, M., Moreau, C., Villares, A., Cathala, B., Chaspoul, F., Frandsen, K. E., Labourel, A., Herpoël-Gimbert, I., Grisel, S., Haon, M., Lenfant, N., Rogniaux, H., Ropartz, D., Davies, G. J., Rosso, M.-N., Walton, P. H., Henrissat, B. & Berrin, J.-G. (2018). Nat. Chem. Biol. 14, 306–310. CrossRef CAS PubMed Google Scholar
Drula, E., Garron, M.-L., Dogan, S., Lombard, V., Henrissat, B. & Terrapon, N. (2021). Nucleic Acids Res. 50, D571–D577. CrossRef Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Web of Science CrossRef CAS IUCr Journals Google Scholar
Fujimoto, Z., Ichinose, H., Biely, P. & Kaneko, S. (2011). Acta Cryst. F67, 68–71. CrossRef IUCr Journals Google Scholar
Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, A., Wilkins, M., Appel, R. D. & Bairoch, A. (2005). The Proteomics Protocols Handbook, edited by J. M. Walker, pp. 571–607. Totowa: Humana Press. Google Scholar
Grotthuss, C. J. D. T. von (1806). Ann. Chim. 58, 54–73. Google Scholar
Holck, J., Djajadi, D. T., Brask, J., Pilgaard, B., Krogh, K. B. R. M., Meyer, A. S., Lange, L. & Wilkens, C. (2019). Enzyme Microb. Technol. 129, 109353. CrossRef PubMed Google Scholar
Holck, J., Fredslund, F., Møller, M. S., Brask, J., Krogh, K. B. R. M., Lange, L., Welner, D. H., Svensson, B., Meyer, A. S. & Wilkens, C. (2019). J. Biol. Chem. 294, 17339–17353. CrossRef CAS PubMed Google Scholar
Holme, T., Arvidson, S., Lindholm, B. & Pavlu, B. (1970). Process Biochem. 5, 62–66. CAS Google Scholar
Kmezik, C., Bonzom, C., Olsson, L., Mazurkewich, S. & Larsbrink, J. (2020). Biotechnol. Biofuels, 13, 60. CrossRef PubMed Google Scholar
Letunic, I. & Bork, P. (2021). Nucleic Acids Res. 49, W293–W296. CrossRef CAS PubMed Google Scholar
Li, W., Cowley, A., Uludag, M., Gur, T., McWilliam, H., Squizzato, S., Park, Y. M., Buso, N. & Lopez, R. (2015). Nucleic Acids Res. 43, W580–W584. Web of Science CrossRef CAS PubMed Google Scholar
Liebschner, D., Afonine, P. V., Baker, M. L., Bunkóczi, G., Chen, V. B., Croll, T. I., Hintze, B., Hung, L.-W., Jain, S., McCoy, A. J., Moriarty, N. W., Oeffner, R. D., Poon, B. K., Prisant, M. G., Read, R. J., Richardson, J. S., Richardson, D. C., Sammito, M. D., Sobolev, O. V., Stockwell, D. H., Terwilliger, T. C., Urzhumtsev, A. G., Videau, L. L., Williams, C. J. & Adams, P. D. (2019). Acta Cryst. D75, 861–877. Web of Science CrossRef IUCr Journals Google Scholar
Lyczakowski, J. J., Wicher, K. B., Terrett, O. M., Faria-Blanc, N., Yu, X., Brown, D., Krogh, K. B. R. M., Dupree, P. & Busse-Wicher, M. (2017). Biotechnol. Biofuels, 10, 224. CrossRef PubMed Google Scholar
Martínez-Abad, A., Berglund, J., Toriz, G., Gatenholm, P., Henriksson, G., Lindström, M., Wohlert, J. & Vilaplana, F. (2017). Plant Physiol. 175, 1579–1592. PubMed Google Scholar
Mayes, H. B., Knott, B. C., Crowley, M. F., Broadbelt, L. J., Ståhlberg, J. & Beckham, G. T. (2016). Chem. Sci. 7, 5955–5968. CrossRef CAS PubMed Google Scholar
McCarter, J. D. & Withers, G. S. (1994). Curr. Opin. Struct. Biol. 4, 885–892. CrossRef CAS PubMed Web of Science Google Scholar
McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. Web of Science CrossRef CAS IUCr Journals Google Scholar
McKee, L. S., Peña, M. J., Rogowski, A., Jackson, A., Lewis, R. J., York, W. S., Krogh, K. B. R. M., Viksø-Nielsen, A., Skjøt, M., Gilbert, H. J. & Marles-Wright, J. (2012). Proc. Natl Acad. Sci. USA, 109, 6537–6542. CrossRef CAS PubMed Google Scholar
Miller, M. A., Pfeiffer, W. & Schwartz, T. (2010). 2010 Gateway Computing Environments Workshop (GCE), pp. 1–8. Piscataway: IEEE. Google Scholar
Mueller, U., Thunnissen, M., Nan, J., Eguiraun, M., Bolmsten, F., Milàn-Otero, A., Guijarro, M., Oscarsson, M., de Sanctis, D. & Leonard, G. (2017). Synchrotron Radiat. News. 30, 22–27. CrossRef Google Scholar
Naidu, D. S., Hlangothi, S. P. & John, M. J. (2018). Carbohydr. Polym. 179, 28–41. CrossRef CAS PubMed Google Scholar
Peña, M. J., Kulkarni, A. R., Backe, J., Boyd, M., O'Neill, M. A. & York, W. S. (2016). Planta, 244, 589–606. PubMed Google Scholar
Petersen, T. N., Brunak, S., von Heijne, G. & Nielsen, H. (2011). Nat. Methods, 8, 785–786. CrossRef CAS PubMed Google Scholar
Rogowski, A., Baslé, A., Farinas, C. S., Solovyova, A., Mortimer, J. C., Dupree, P., Gilbert, H. J. & Bolam, D. N. (2014). J. Biol. Chem. 289, 53–64. CrossRef CAS PubMed Google Scholar
Ryabova, O., Vršanská, M., Kaneko, S., van Zyl, W. H. & Biely, P. (2009). FEBS Lett. 583, 1457–1462. Web of Science CrossRef PubMed CAS Google Scholar
Santibáñez, L., Henríquez, C., Corro-Tejeda, R., Bernal, S., Armijo, B. & Salazar, O. (2021). Carbohydr. Polym. 251, 117118. PubMed Google Scholar
Scheller, H. V. & Ulvskov, P. (2010). Annu. Rev. Plant Biol. 61, 263–289. Web of Science CrossRef CAS PubMed Google Scholar
Stamatakis, A., Hoover, P. & Rougemont, J. (2008). Syst. Biol. 57, 758–771. CrossRef PubMed Google Scholar
Stortz, C. A. (2010). J. Phys. Org. Chem. 23, 1173–1186. CrossRef CAS Google Scholar
Terwilliger, T. C., Grosse-Kunstleve, R. W., Afonine, P. V., Moriarty, N. W., Zwart, P. H., Hung, L.-W., Read, R. J. & Adams, P. D. (2008). Acta Cryst. D64, 61–69. Web of Science CrossRef CAS IUCr Journals Google Scholar
The CAZypedia Consortium (2018). Glycobiology, 28, 3–8. Web of Science CrossRef PubMed Google Scholar
Tryfona, T., Liang, H. C., Kotake, T., Tsumuraya, Y., Stephens, E. & Dupree, P. (2012). Plant Physiol. 160, 653–666. CrossRef CAS PubMed Google Scholar
Ursby, T., Åhnberg, K., Appio, R., Aurelius, O., Barczyk, A., Bartalesi, A., Bjelčić, M., Bolmsten, F., Cerenius, Y., Doak, R. B., Eguiraun, M., Eriksson, T., Friel, R. J., Gorgisyan, I., Gross, A., Haghighat, V., Hennies, F., Jagudin, E., Norsk Jensen, B., Jeppsson, T., Kloos, M., Lidon-Simon, J., de Lima, G. M. A., Lizatovic, R., Lundin, M., Milan-Otero, A., Milas, M., Nan, J., Nardella, A., Rosborg, A., Shilova, A., Shoeman, R. L., Siewert, F., Sondhauss, P., Talibov, V. O., Tarawneh, H., Thånell, J., Thunnissen, M., Unge, J., Ward, C., Gonzalez, A. & Mueller, U. (2020). J. Synchrotron Rad. 27, 1415–1429. Web of Science CrossRef CAS IUCr Journals Google Scholar
Wang, W., Yan, R., Nocek, B. P., Vuong, T. V., Di Leo, R., Xu, X., Cui, H., Gatenholm, P., Toriz, G., Tenkanen, M., Savchenko, A. & Master, E. R. (2016). J. Biol. Chem. 291, 14120–14133. CrossRef CAS PubMed Google Scholar
Wilkens, C., Busk, P. K., Pilgaard, B., Zhang, W.-J., Nielsen, K. L., Nielsen, P. H. & Lange, L. (2017). Biotechnol. Biofuels, 10, 158. CrossRef PubMed Google Scholar
Winter, G. (2010). J. Appl. Cryst. 43, 186–190. Web of Science CrossRef CAS IUCr Journals Google Scholar
Yan, R., Vuong, T. V., Wang, W. & Master, E. R. (2017). Enzyme Microb. Technol. 104, 22–28. CrossRef CAS PubMed Google Scholar
Yan, R., Wang, W., Vuong, T. V., Xiu, Y., Skarina, T., Di Leo, R., Gatenholm, P., Toriz, G., Tenkanen, M., Stogios, P. J. & Master, E. R. (2021). New Biotechnol. 62, 49–56. CrossRef CAS Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.