Characterization of an α-L-Arabinofuranosidase GH51 from the Brown-rot Fungus Gloeophyllum trabeum

Abstract

Woody biomass is anticipated to be a resource for a decarbonized society, but the difficulty of isolating woody components is a significant challenge. Brown-rot fungi, a type of wood rotting fungi, decompose hemicellulose particularly efficiently. However, there are few reports on the hemicellulases from brown-rot fungi. An α-L-arabinofuranosidase belonging to glycoside hydrolase family 51 (GH51) from the brown-rot fungus Gloeophyllum trabeum (GtAbf51A) was cloned and characterized in the present study. Analyses of the phylogeny of GH51 enzymes in wood rotting fungi revealed the existence of two groups, intercellular and extracellular enzymes. After deglycosylation, the recombinant GtAbf51A produced by Pichia pastoris appeared on SDS-PAGE as approximately 71,777 daltons, which is the expected molecular weight based on the amino acid sequence of GtAbf51A. Maximum enzyme activity occurred between pH 2.2 and 4.0 and at 50 °C, while it was stable between pH 2.2 and 10.0 and up to 40 °C. Due to the presence of a signal peptide, GtAbf51A was thought to hydrolyze polysaccharide containing arabinose. However, the hydrolysis rate of arabinosyl linkages in polysaccharides was only 3-5 % for arabinoxylan and 18 % for arabinan. GtAbf51A, in contrast, efficiently hydrolyzed arabinoxylooligosaccharides, particularly O-α-L-arabinofuranosyl-(1→3)-O-β-D-xylopyranosyl-(1→4)-β-D-xylopyranose, which is the principal product of GH10 β-xylanase. These data suggest that GtAbf51A cooperates with other xylan-degrading enzymes, such as β-xylanase, to degrade xylan in nature.

Abbreviations

Araf, L-arabinofuranoside; A₁X₂, O-α-L-arabinofuranosyl-(1→3)-O-β-D-xylopyranosyl-(1→4)-β-D-xylopyranose; A₁X₃, O-β-D-xylopyranosyl-(1→4)-[O-α-L-arabinofuranosyl-(1→3)]-O-β-D-xylopyranosyl-(1→4)-β-D-xylopyranose; CAZy, Carbohydrate-Active enZymes; GH, glycoside hydrolase; GH51, glycoside hydrolase family 51; GtAbf51A, α-L-arabinofuranosidase from Gloeophyllum trabeum belonging glycoside hydrolase family 51; HPAEC-PAD, high-performance anion-exchange chromatography with pulsed amperometric detection; Xyl, D-xylopyranoside.

INTRODUCTION

Woody biomass, a renewable and abundant natural resource, is being promoted for industrial use in various fields in order to realize a decarbonized society.¹⁾,²⁾ Wood is primarily composed of polysaccharides such as cellulose and hemicellulose, as well as the polymeric aromatic compound lignin. These wood components are tightly bound to one another, making the decomposition of cellulose and hemicellulose into monosaccharides difficult. Therefore, the “decomposition” is the greatest challenge in biorefinery.²⁾,³⁾

Wood rotting fungi, a type of filamentous fungi, are among the few organisms capable of causing the complete decomposition of wood. Based on morphological differences in decayed wood, wood rotting fungi are classified as white-rot fungi or brown-rot fungi.⁴⁾ Their wood degradation systems differ. White-rot fungi degrade all lignocellulose components, whereas brown-rot fungi only degrade polysaccharides, leaving lignin as a polymer.⁴⁾ In the early stages of wood decay, brown-rot fungi preferentially degrade hemicellulose.⁵⁾,⁶⁾ Furthermore, molecular evolutionary phylogenetic analysis suggested that brown-rot fungi evolved from white-rot fungi, with the loss of several genes encoding wood cell wall degrading enzymes.⁷⁾ Although brown-rot fungi have lost enzymes that degrade crystalline cellulose and lignin, a series of enzymes that degrade hemicellulose remain. For example, xylan degradation system of brown-rot fungi contains β-xylanase, which acts on the main chain, and α-L-arabinofuranosidase and α-glucuronidase, which are involved in side chain degradation.

Despite evidence that brown-rot fungi have evolved specifically for hemicellulose degradation, there are few reports on the hemicellulose degrading enzymes from brown-rot fungi, and there are still many unknowns regarding the hemicellulose degradation system of brown-rot fungi. Therefore, we focused on Gloeophyllum trabeum, one of the most prevalent and extensively studied brown-rot fungi. In this study, an α-L-arabinofuranosidase (GtAbf51A) was selected from the genomic information of G. trabeum. We produced GtAbf51A as a recombinant enzyme and analyzed its substrate specificity to determine its role in the degradation of cell wall polysaccharides.

MATERIALS AND METHODS

Substrates.

p-Nitrophenyl-α-L-arabinofuranoside (PNP-α-L-Araf) was purchased from Megazyme Co., Ltd. (Wiscow, Ireland). In addition, oat spelts xylan was purchased from Fluka Biochemika (Weinheim, Germany). Corn hull arabinoxylan was procured from Nihon Shokuhin Kako Co., Ltd. (Tokyo, Japan). Sugar beet arabinan and arabinoxylooligosaccharides such as O-α-L-arabinofuranosyl-(1→3)-O-β-D-xylopyranosyl-(1→4)-β-D-xylopyranose (A₁X₂) and O-β-D-xylopyranosyl-(1→4)-[O-α-L-arabinofuranosyl-(1→3)]-O-β-D-xylopyranosyl-(1→4)-β-D-xylopyranose (A₁X₃) were prepared as described previously.⁸⁾,⁹⁾

Strains.

G. trabeum NBRC 6430 was obtained from the National Institute of Technology and Evaluation (Kazusa, Japan). The hosts used for subcloning and heterologous production of the recombinant protein were Escherichia coli strain Top10 (Thermo Fisher Scientific Inc., Waltham, MA, USA) and Pichia pastoris strain KM71H (Thermo Fisher Scientific), respectively.

Plasmid construction and expression of recombinant GtAbf51A.

G. trabeum was grown on Highley's medium¹⁰⁾ containing 0.2 % (w/v) xylan and 1 % (w/v) avicel as the carbon source. Total RNA was extracted from frozen mycelia using RNeasy plant mini kit (Qiagen, Venlo, Netherlands), and mRNA was purified using Oligotex^TM-dT30 Super (Takara Bio Inc., Kusatsu, Japan). cDNA was synthesized from the mRNA using a GeneRacer kit with SuperScript™ III reverse transcriptase (Thermo Fisher Scientific). The gene encoding GtAbf51A was amplified using the forward and reverse primrs 5'-AGGGGTATCTCTCGAGAAAAGAGTACTGACAGCTCCCAATGCTC and 5'-AGAAAGCTGGCGGCCGCTTAGCTGCCTATACCCGCCG, respectively. Using KOD-Plus-Neo (Toyobo Co., Ltd., Osaka, Japan), a polymerase chain reaction was conducted at 30 cycles of 94 °C for 2 min, 56.65 °C for 30 s, and 68 °C for 90 s. The amplified fragment was ligated into the pPICZαA vector (Thermo Fisher Scientific) using the In-Fusion^® HD cloning kit (Takara Bio). The DNA of the expression plasmid was linearized using Bgl II (Takara Bio). Electroporation and transformant selection was performed according to the instruction of the EasySelect^TM Pichia expression kit (Thermo Fisher Scientific). The selected clone was cultured for 24 h at 28 °C in a growth medium (1 % (w/v) yeast extract, 2 % (w/v) high polypeptone, and 1 % (w/v) glycerol). The cells were harvested by centrifugation, suspended in an induction medium (1 % (w/v) yeast extract, 2 % (w/v) high polypeptone, and 1 % (v/v) methanol), and then induced at 28 °C for 4 d. Every 24 h, 0.8 % (v/v) methanol was added.

Purification of recombinant protein.

The culture was centrifuged at 4 °C with 4,500 × G for 5 min. The supernatant (10 mL) was dialyzed against 20 mM sodium acetate buffer (pH 4.0) three times, filtered through a 0.22 μm filter, and applied to a Toyopearl^® SP-650S (Tosoh Co., Tokyo, Japan) column (5 × 50 mm) preequilibrated with 20 mM sodium acetate buffer (pH 4.0). The column was washed with the same buffer to remove unbound substances, and bound proteins were eluted with a linear gradient of sodium chloride (0-1 M) at a flow rate of 1 mL/min. SDS-PAGE was used to confirm the purity of GtAbf51A, and the final fraction obtained was used as the purified enzyme.

Enzyme assay and enzymatic properties.

The enzyme assay mixture contained 25 μL 2 mM PNP-α-L-Araf, 20 μL McIlvaine buffer (pH 3.0, 0.2 M Na₂HPO₄/0.1 M citric acid), and 5 μL of the enzyme solution. The reactions were carried out at 50 °C for 10 min before being stopped with 50 μL of 0.2 M Na₂CO₃. At 400 nm (an extinction coefficient of 18,700 M^-1 cm^-1), the amount of PNP released was determined. Under these conditions, one unit of enzyme activity is defined as the amount of enzyme that releases 1 µmol of PNP per min from PNP-α-L-Araf.

Using glycine-HCl buffer (pH 1.5-2.6, 0.1 M), McIlvaine buffer (pH 2.6-7.6, 0.2 M Na₂HPO₄/0.1 M citric acid) and Atkins-Pantin buffer (pH 7.6-11.0, 0.2 M H₃BO₃•KCl/0.2 M Na₂CO₃), the pH dependence and stability were determined. For pH stability, the mixture containing 10 μL enzyme solution and 40 μL of each buffer was incubated at 30 °C for 1 h, and the residual activity was determined. Temperature dependence was determined from 10 °C to 81.4 °C, and temperature stability from 10 °C to 60 °C. For temperature stability, the mixture of 6 μL enzyme solution and 24 μL of McIlvaine buffer (pH 3.0) was incubated at each temperature for 1 h, and the residual activity was determined.

Reactivity and substrate specificity.

Reactivity of the polysaccharide was determined by incubating the mixture containing 2 mg of substrate, 80 μL McIlvaine buffer (pH 3.0), 110 μL distilled water, and 10 μL enzyme solution (0.45 mU) at 30 °C for 20 h. The reaction was terminated by heating at 100 °C for 30 min, and the residue was collected by ethanol precipitation. Next, the residue was dried and dissolved in 500 μL 0.2 M trifluoroacetic acid solution and hydrolyzed via incubation at 121 °C for 2 h. After cooling and drying, re-dissolution with 300 μL isopropanol and drying were repeated six times. The saccharide composition was determined with high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) (Thermo Fisher Scientific) using a Dionex CarboPac^TM PA-1 (Thermo Fisher Scientific) column (4 × 250 mm), and the Ara ratio was calculated. At a flow rate of 1 mL/min, the column was eluted with 0.1 M NaOH (0-5 min), followed by a linear gradient (5-35 min) of sodium acetate (0-0.2 M).¹¹⁾,¹²⁾ Oligosaccharide substrate specificity was determined by incubating 25 μL of 1.2 mM substrate, 20 μL of McIlvaine buffer (pH 3.0), 2.5 μL of enzyme solution (0.15 mU for A₁X₂ and 0.3 mU for A₁X₃) and 2.5 μL of 0.015 % L-fucose as an internal standard at 30 °C for 10 min. The reaction was stopped with 75 μL of 0.2 M Na₂CO₃, and the amount of arabinose (Ara) released was determined by HPAEC-PAD.¹¹⁾,¹²⁾

Nucleotide sequence accession number.

The nucleotide sequence of cDNA encoding GtAbf51A is available in the DDBJ database under the accession number LC731959.

RESULTS AND DISCUSSION

Although brown-rot fungi are highly capable of decomposing hemicellulose, the information of the hemicellulases is limited. Arabinose is widely distributed in plant cell wall as arabinoxylan, arabinan, arabinogalactan, and arabinogalactan protein,¹³⁾ and thus we chose GH51 α-L-arabinofuranosidases.

There are five GH families for α-L-arabinofuranosidases: GH3, GH43, GH51, GH54, and GH62. G. trabeum possesses eleven GH3, six GH43, and four GH51 genes. Among them, only four GH51 genes are predicted to encode α-L-arabinofuranosidases. According to a phylogenetic analysis of the GH51 genes in wood rotting fungi, extracellular and intracellular enzymes can be broadly categorized into two distinct groups (Fig. 1). The extracellular enzymes appear to act on polysaccharides and are directly linked to the degradation of polysaccharides in wood cell walls; therefore, GtAbf51A was chosen (JGI protein ID: 134804). The GtAbf51A cDNA sequence consisted of 2,022 bp excluding the stop codon and encoded 674 amino acids. The SignalP program (https://services.healthtech.dtu.dk/service.php?SignalP-5.0) was used to predict N-terminus 23 amino acid residues as signal peptides.¹⁴⁾ The mature protein contained 651 amino acids, weighed 71,777 daltons, and had a predicted pI of 4.58. The sequence was similar to MgGH51 from Meripilus giganteus (97 % query cover and 46.73 % identity; pdb code 6ZPS),¹⁵⁾ according to BLAST searches of the Protein Data Bank (pdb). The alignment of GtAbf51A with MgGH51 and the 3-D structure model of GtAbf51A predicted by AlphaFold 2¹⁶⁾ are depicted in Figs. 2 and 3, respectively. The catalytic residues of MgGH51 (E351 and E429 (E360 and E438 in GtAbf51A)) and the Ara-binding residues of MgGH51 (E23, N231, N350, and Y402 (E30, N234, N359, and Y411 in GtAbf51A)) were fully conserved when the structures of GtAbf51A and MgGH51 were superimposed (Figs. 2 and 3A).

Fig. 1. Phylogenetic analysis of the GH 51 gene in wood-decay fungi.

　The sequences were obtained from JGI (https://mycocosm.jgi.doe.gov/programs/fungi/index.jsf;). The outgroup was Dichotomocladium robustum. Signal peptides were predicted using Signal P 5.0¹⁴⁾ and then deleted, followed by multiple alignments using Clustal W.²⁵⁾ The numbers are JGI protein ID. Asterisks (*) indicate intracellular enzymes. Phylogenetic trees were generated using Jalview and the neighbor-joining method.²⁶⁾

Fig. 2. Alignment of GtAbf51A and MgGH51.

　Gray represents identical amino acid residues, black represents functional residues, and asterisks (*) represent catalytic residues.

Fig. 3. GtAbf51A model structure.

　The model structure was constructed with AlphaFold2.²⁶⁾ A, GtAbf51A Ara-binding pocket. The GtAbf51A model (green) was superimposed on the Ara-binding structure of MgGH51 (pink); B, Structure around subsite +1 of MgGH51with A₁X₃ derived from 2VRQ; C, Structure around subsite +1 of GtAbf51A with A₁X₃ derived from 2VRQ; D, Surface model of substrate binding site of GtAbf51A with A₁X₃ derived from 2VRQ; E, Surface model of substrate binding site of GtAbf51A with A₁X₃ derived from 1V6V; F, Superimposing of A₁X₃ from 2VRQ (white) and 1V6V (green). ^*Ara of A₁X₃ was superimposed on Ara of the Ara-binding structure of MgGH51.

GtAbf51A was successfully expressed and purified using cation exchange chromatography. Purified GtAbf51A had an unexpectedly high molecular weight of 100,000 daltons (Fig. 4, lane 1), but after endoglycosidase H treatment, GtAbf51A appeared at 74,000 daltons on SDS-PAGE (Fig. 4, lane 2). The optimal pH for the recombinant enzyme was between 2.2 and 4.0 (Fig. 5A), while the optimal temperature was 50 °C (Fig. 5B); it was stable from pH 2.2 to 10.0 (Fig. 5C) and up to 40 °C (Fig. 5D). GH51 α-L-arabinofuranosidases from Penicillium subrubescens and Pleurotus ostreatus had an optimal temperature range of 40-50 °C and an optimal pH range of 4.0.¹⁷⁾,¹⁸⁾ GtAbf51A demonstrated greater activity on the more acidic side than the enzymes of previous reports. GtAbf51A had nearly the same thermal stability as α-L-arabinofuranosidase from P. subrubescens, and was more stable in the acidic pH range of 2.2 to 4.0.

Fig. 4. SDS-PAGE analysis of recombinant GtAbf51A.

　On 10 % polyacrylamide gel, each sample was applied. M, molecular weight markers; lane 1, recombinant GtAbf51A; lane 2, recombinant GtAbf51A treated with endoglycosidase H; lane 3, endoglycosidase H.

Fig. 5. The enzymatic properties of recombinant GtAbf51A.

　A, effect of pH for enzyme activity; B, effect of temperature for enzyme activity; C, pH stability; D, thermostability. The following symbols are used: triangle for the glycine-HCl buffer, circle for McIlvaine buffer, and square for Atkins-Pantin buffer.

Next, reactivity and substrate specificity of GtAbf51A was investigated. Since this enzyme is a secreted protein, it is likely involved in the degradation of polysaccharides in plant cell walls. Thus, the activity of GtAbf51A toward polysaccharides containing Ara was examined (Table 1). After 20 h of incubation with GtAbf51A, the amount of Ara in sugar beet arabinan, oat spelts xylan, and corn hull arabinoxylan was reduced by 18.2, 3.1, and 4.5 %, respectively, following enzymatic digestion. The α-L-arabinofuranosidases of filamentous fungi are classified as ArafurB, which acts on polysaccharides, and ArafurA, which does not,¹⁹⁾ and it has been suggested that ArafurA corresponds to GH51 and ArafurB corresponds to GH54, respectively.²⁰⁾ Since GtAbf51A had low polysaccharide activity, the enzyme activity towards oligosaccharides was evaluated. Brown-rot fungi have only GH10 xylanases and no GH11 xylanases, although endo-β-xylanases are primarily GH10 and GH11. These two families of xylanases differ in their patterns of cleavage of heteroxylans.²¹⁾,²²⁾ Consequently, the substrate specificity of arabinoxylooligosaccharides A₁X₂ and A₁X₃, which are products of GH10 xylanase,²³⁾ were analyzed. GtAbf51A hydrolyzed both substrates, releasing Ara at a rate of 0.473 µmol/min from A₁X₂ and 0.148 µmol/min from A₁X₃, indicating that it works in cooperatively with GH10 xylanase. The rate of deterioration of A₁X₂ was more than three times that of A₁X₃. The structural model of GtAbf51A was utilized to comprehend the cause.

Table 1.　Reactivity of recombinant GtAbf51A toward Ara containing polysaccharides.

		Corn hull arabinoxylan	Oat spelts xylan	Sugar beet arabinan
Ara/Xyl	No digestion	0.67	0.096
	After digestion	0.64	0.093
Ara/Gal	No digestion			3.3
	After digestion			2.7

The number in the table is the average of two experiments and represents the ratio of Ara calculated with Xyl or Gal as 1. The experiments were repeated twice.

Superimposition of the A₁X₃ obtained from the arabinofuranosidase (TxGH51) complex structure (pdb code 2VRQ)²⁴⁾ of Thermoacillus xylanilyticus, MgGH51(pdb code 6ZPY),¹⁵⁾ and the constructed GtAbf51A model structure predicted the substrate binding manner of GtAbf51A (Figs. 3B and 3C). MgGH51 and GtAbf51A differed on the + side of the subsite structure. In MgGH51, it is hypothesized that two Phe residues (F441, F354) trap the middle arabinose-linked Xyl at subsite +1 by hydrophobic interaction in order to sandwich the Xyl,¹⁵⁾ but F354 is not in the position of stacking interaction with the middle Xyl (and direction is not face to the xylose side). Y300 adjacent to F363, which corresponds to F354 in MgGH51, is considered to stack the Xyl in GtAbf51A. On the opposite side of the cleft of GtAbf51A, there is no aromatic amino acid corresponding to F441 in MgGH51; as a result, L450 fills the space and forms a cleft-like structure (Fig. 3C). The surface model suggested that A₁X₃ appears to fit neatly on the cleft of GtAbf51A (Fig. 3D). In contrast, when the Ara residue of A₁X₃ obtained from the GH10 xylanase (SoXyn10A) complex structure (pdb code 1V6V)²²⁾ of Streptomyces olivaceoviridis was superimposed on the Ara of A₁X₃ obtaind from the TxGH51-A₁X₃ structure, the non-reducing end Xyl came into steric conflict with W263 and could not fit into the substrate binding site of GtAbf51A (Fig. 3E). This model suggests only the non-reducing terminus of the xylan chain is expected to bind to the substrate-binding site of GtAbf51A. The reactivity (Table 1) and substrate specificity for oligosaccharides suggests that both prediction (Figs. 3D and 3E) are accurate. Figure 3F depicts a superposition of the A₁X₃ binding structures of SoXyn10A-A₁X₃ and TxGH51-A₁X₃ at the middle arabinose-linked xylose. The position of the side chain Ara and the reducing end side Xyl did not significantly change, whereas the position of the non-reducing end Xyl was highly variable.

Thus, the xylan main chain is disordered, making it difficult for it to fit into the catalytic site of the GH51 enzyme, resulting in low GH51 enzyme activity for polysaccharides and A₁X₃. In the GH62 α-L-arabinofuranosidase which is specific to arabinoxylan, there is a hydrogen bond to Xyl at the non-reducing end, and it exhibits high activity toward A₁X₃.²⁰⁾ GtAbf51A has apparently distinct mechanism from GH62 α-L-arabinofuranosidase, and is not good at recognizing xylose residue present on the non-reducing end side from the arabinosyl side chain. Another typical xylanase, GH11 xylanase, which brown-rot fungi lack, produce a product with one more Xyl residue at the non-reducing end than A₁X₃, which would be more difficult for GtAbf51A to recognize. Therefore, brown-rot fungi might evolve to use GH10 enzymes for xylan degradation.

CONCLUSION

In this study, we characterized GtAbf51A for the first time as the α-L-arabinofuranosidase from brown-rot fungi. This enzyme demonstrated low activity for polysaccharides like arabinan (18.2 %) and arabinoxylan (3.1 and 4.5 %, respectively). In contrast, GtAbf51A efficiently hydrolyzed the products of GH10 xylanases, indicating that it works cooperatively with β-xylanase. G. trabeum's genome contains three GH10 xylanases; analysis of the synergistic effect of each enzyme is a future concern. In addition, G. trabeum possesses numerous accessory enzymes, including four GH51 α-L-arabinofuranosidases and two GH115 glucuronidases. In order to understand the xylan-degrading system of brown-rot fungi, it is necessary to characterize these enzymes and their synergistic effects.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.