Characterization of a novel bifunctional endo-β-1,4-glucanase/mannanase from Aspergillus oryzae Authors: Akira Watanabe, Tomohiko Matsuzawa, Seiryu Ujiie, Takahiro Shintani, Katsuya Gomi Published: 2026 ABSTRACT AbstractThe present study involved a global search of the Aspergillus oryzae genome database. As a result, the gene (gene ID: AO090011000715) encoding a putative endo-β-1,4-glucanase of the glycoside hydrolase family 5 possessing a carbohydrate-binding module was identified. Overexpression of this gene in A. oryzae, followed by the purification of the gene product, revealed that the protein exhibits endoglucanase activity with carboxymethyl cellulose as a substrate. Notably, further enzymatic characterization showed that this enzyme could also hydrolyze glucomannan and galactomannan with relatively lower activity than that of endoglucanase, indicating that the enzyme is likely a bifunctional endoglucanase/mannanase. Detailed analyses of the substrate specificity revealed that the enzyme could hydrolyze cello-oligosaccharides larger than cellotetraose, whereas manno-oligosaccharides could not be hydrolyzed unless they were larger than mannopentaose. Therefore, this enzyme has different modes of action toward cello- and manno-oligosaccharides. To the best of our knowledge, this is the first study to purify and characterize a bifunctional endo-β-1,4-glucanase/mannanase from the Aspergillus genus. The ability of the enzyme to degrade both mannan and cellulose could be used for more efficient degradation of lignocellulosic biomass, indicating that the enzyme shows promise for applications in biomass utilization and the food industry.Key points • A novel bifunctional endoglucanase/mannanase, EgmA, is identified from Aspergillus oryzae. • EgmA is the second bifunctional endo-β-1,4-glucanase/mannanase of fungal origin. • EgmA has different mechanisms of action toward cello- and manno-oligosaccharides. Supplementary InformationThe online version contains supplementary material available at 10.1007/s00253-026-13838-6. Introduction Cellulose is degraded by three different types of enzymes (Lynd et al. 2002), namely endo-β-1,4-glucanase (EC 3.2.1.4), cellobiohydrolase, and β-glucosidase (EC 3.2.1.21). Endo-β-1,4-glucanases are endo-type enzymes that cleave the β-(1 → 4)-glycosidic bond within cellulose in the amorphous region but not in the crystalline region of cellulose. Cellobiohydrolases hydrolyze both crystalline and soluble substrates and continue to hydrolyze along the cellulose chain without detaching (Davies and Henrissat 1995). Cellobiohydrolases are divided into two types: one that hydrolyzes cellulose to cellobiose from the non-reduced end of cellulose (EC 3.2.1.91) and the other that acts from the reduced end (EC 3.2.1.176). β-Glucosidases hydrolyze cellobiose or β-glucoside to release glucose. A large number of bacteria and fungi (Glasgow et al. 2020; Li et al. 2024; Miao et al. 2021) can produce cellulases. Among these, cellulases from fungi are considered suitable for the enzymatic saccharification of lignocellulosic biomass because fungi can secrete multiple types and large amounts of cellulases (Ikeda et al. 2007). For example, Trichoderma reesei has long been studied and is known to produce cellulases with high enzymatic activity (Sheir-Neiss and Montenecourt 1984). Genome data for this fungus (https://www.ncbi.nlm.nih.gov/gdv/browser/genome/?id=GCF_000167675.1) predicts that there are two cellobiohydrolases, 11 endo-β-1,4-glucanases, and ten β-glucosidases, and some of them have already been characterized enzymatically (Sheir-Neiss and Montenecourt 1984; Saloheimo et al. 1988; Karlsson et al. 2001). The hydrolytic activity of lignocellulosic biomass by enzymes is influenced by enzyme cocktail. Since some enzyme activities such as β-glucosidase are generally weak in T. reesei, enhancing these enzymes has been effective for efficient hydrolysis of cellulosic biomass (Nakazawa et al. 2012; Ellilä et al. 2017). The filamentous fungus Aspergillus oryzae secretes large amounts of various useful hydrolytic enzymes, including amylases, proteases, and cellulases, into the culture medium (Kobayashi et al. 2007; Machida et al. 2008). Genome sequencing (Machida et al. 2005) and expressed sequence tag (ESTs) analyses (Akao et al. 2007) of this fungus have predicted that several gene sources, including cellulases, exist in the genome. Two endo-β-1,4-glucanases (CelA and CelB) (Kitamoto et al. 1996) and seven β-glucosidases (BglC, Bgl1, Bgl3, Bgl5, BglA, BglF, and BglH) (Langston et al. 2006; Kaya et al. 2008; Kudo et al. 2015; Watanabe et al. 2016) from A. oryzae have already been characterized. In addition, the enzymes involved in xyloglucan degradation have already been analyzed (Matsuzawa et al. 2021). Endo-β-1,4-glucanases have attracted the particular attention of researchers (Miao et al. 2021) because these enzymes act at the first stage of the cellulose degradation process. In the carbohydrate-active enzymes database (Drula et al. 2022) (CAZy; http://www.cazy.org/), endo-β-1,4-glucanases are found in multiple glycoside hydrolase (GH) families, GH5, GH7, GH9, GH12, GH44, GH45, and so on. The GH5 family includes several cellulases and hemicellulases. Currently, there are 57 subfamilies within this family, including endo-β-1,4-glucanase (EC 3.2.1.4) as well as endo-β-1,4-xylanase (EC 3.2.1.8), β-glucosidase (EC 3.2.1.21), and β-mannosidase (EC 3.2.1.25). GH5 enzymes play important roles in lignocellulosic biomass degradation (Payne et al. 2015). Endo-β-1,4-glucanases used for biomass degradation require characteristics such as high enzyme activity and specificity, microcrystalline cellulose degradation activity, and thermostability (Wohlgemuth et al. 2018). In addition to these characteristics, there is growing interest in carbohydrate-active enzymes with multifunctional properties (Miao et al. 2021), particularly regarding their catalytic mechanisms. In this study, we identified a gene (gene ID: AO090011000715) that encodes a putative endo-β-1,4-glucanase (accession no. in GenBank: BAE65168), which possesses a carbohydrate-binding module 1 (CBM1), by genome mining from A. oryzae genome database. We then overexpressed this gene (hereafter designated egmA) in A. oryzae and purified the gene product for enzymatic characterization. EgmA belongs to the GH5 family and exhibits endo-β-1,4-glucanase activity. In addition, this enzyme exhibits mannanase activity, indicating that it is a novel bifunctional endoglucanase/mannanase. Materials and methods Strains and media A. oryzae mutant deficient in nitrate reductase gene (niaD−), derived from an industrial strain A. oryzae AOK11 (Akita Konno Shoten, Akita, Japan) (Sato et al. 2011), was used as a host for the overproduction of EgmA. A. oryzae RIB40 (Machida et al. 2005) was used as the DNA donor. A. oryzae transformants were cultured in YPM medium (0.5% Bacto yeast extract (Thermo Fisher Scientific K.K., Tokyo, Japan), 1% Bacto peptone (Becton–Dickinson, Tokyo, Japan), 1% maltose (Fujifilm Wako Pure Chemical Industries, Ltd., Osaka, Japan)) at 30 °C. Czapek–Dox (CD) medium, consisting of 0.6% NaNO3, 0.052% KCl, 0.152% KH2PO4, 0.2% MgSO4, 0.8 M NaCl, trace amounts of FeSO4, ZnSO4, CuSO4, MnSO4, (NH4)6Mo7O24, Na2B4O7, and 1% sugar supplemented with appropriate nutrients, was used as a minimal medium for A. oryzae. Escherichia coli DH5α (F−, Φ80dlacZΔM15, Δ(lacZYA-argF)U169, deoR, recA1, endA1, hsdR17(rK−, mK+), phoA, supE44, λ−, thi-1, gyrA96, relA1) was used for the plasmid construction. It was grown in the Luria Bertani medium (1% (w/v) Bacto tryptone (Becton–Dickinson), 0.5% (w/v) Bacto yeast extract, and 0.5% (w/v) NaCl) containing 50 μg/mL ampicillin (Fujifilm Wako Pure Chemical Industries, Ltd.) at 37 °C. Genome mining approach of egmA and phylogenetic tree analysis First novel five genes encoding putative endoglucanases with a signal peptide were identified by screening the Comprehensive Aspergillus oryzae Genome Database (CAoGDX; https://nribf21.nrib.go.jp/CAoGDX/), the CAZy database, and the SignalP 6.0 server (https://services.healthtech.dtu.dk/services/SignalP-6.0/). According to the CAZy database, deduced endoglucanases were grouped to two families of GH; two belonged to the family GH5, other three were in GH12. Since enzymes classified in the GH5 family play an important role in biomass degradation (Payne et al. 2015), we focused on putative GH5 enzymes in A. oryzae genome. The amino acid sequences of GH5 family proteins from Aspergillus nidulans FGSC A4, Aspergillus niger CBS513.88, Neurospora crassa OR74A, T. reesei QM6a, and T. reesei RutC-30, which were selected from the CAZy database and the CAoGDX database were downloaded into a single FASTA file with putative GH5 endoglucanases from A. oryzae RIB40. This FASTA file was uploaded using MEGA 12 software (Kumar et al. 2024). The evolutionary history of these sequences was constructed by selecting the phylogeny tab in the software and using the neighbor-joining method. Plasmid construction for overexpression analysis of putative novel endoglucanase DNA containing the region from 10 bp upstream to 400 bp downstream of an open reading frame of a putative novel endo-β-1,4-glucanase gene was amplified using polymerase chain reaction (PCR) with forward primer (5′-CATAAGCTTGCTCTCCGACAAGCTCGGCATAC-3′) and reverse primer (5′-CCCAAGCTTGGGTTTAGTCGCGAGGTCTGGCC-3′). The genome DNA of A. oryzae RIB40 was used as a template. HindIII restriction enzyme cleavage site (underlined), which was included in the multi-cloning site (MCS) of the expression vector plasmid, pNGA142 (Minetoki et al. 2003), was attached to the 5′-end of each primer. PCR amplification and insertion of the resulting PCR fragments into the MCS of the expression vector plasmid were performed as previously described (Kudo et al. 2015). The transformation of A. oryzae was performed using protoplast and polyethylene glycol as previously described (Gomi et al. 1987). Assays for endoglucanase and mannanase activities The endoglucanase and mannanase activities were assayed using the 3,5-dinitrosalicylic acid (DNS) method (Miller 1959). The assay mixture contained 0.8% sodium carboxymethyl cellulose (CMC [degree of polymerization: approximately 1050]; Tokyo Chemical Industry, Tokyo, Japan), konjac glucomannan (Megazyme International, Bray, Ireland), or carob galactomannan (Megazyme) in 50 mM sodium acetate buffer (pH 4.0). The reaction was started by adding 5 μL of enzyme preparation to 20 μL of assay mixture. The reaction mixture was incubated at 50 °C for 10 min. After incubation, 75 μL of DNS reagent was added to the reaction mixture. This solution was boiled for 5 min, cooled under running tap water, and adjusted to ambient temperature. Milli-Q water (250 μL) was added to 40 μL of the resulting boiled reaction mixture and mixed well, and the absorbance at 500 nm was measured as previously described (Ulpiyana et al. 2024). Several concentrations of glucose and mannose solutions were used to generate a standard curve. One unit of the endoglucanase or mannanase activity was defined as the amount of enzyme that yielded sugar equivalent to 1 μmol of glucose or mannose per min under the assay conditions. The effects of SrCl2, BaCl2, FeCl2, ZnCl2, CuCl2, NiCl2, CoCl2, MnCl2, CaCl2, MgCl2, FeCl3, and EDTA were examined by adding 5 or 10 mM of each metal ion or EDTA to the reaction mixture. The enzyme activity was determined at 50 °C (optimal temperature) and pH 4.0 (optimal pH), and compared with the enzyme activity detected when no metal ions were added to the reaction mixture. Purification of EgmA Culture supernatants of A. oryzae AOK11 transformants overexpressing egmA, grown for 48 h at 30 °C in YPM liquid medium (500 mL), were collected through filtration. Next, the enzyme was purified at 4 °C. Filtrates containing EgmA were applied to a TOYOPEARL Butyl-650 M column (Tosoh Co., Ltd., Tokyo, Japan, flow rate; 2.0 mL/min, column dimension; 1.4 cm by 5.5 cm, fraction volume; 1.5 mL) equilibrated with 50 mM potassium phosphate buffer (pH 5.0) containing 1 M ammonium sulfate. After washing the column with the same phosphate buffer, elution was performed using a linear gradient of ammonium sulfate from 1 to 0 M in 50 mM potassium phosphate buffer. The endoglucanase activities of all the eluted fractions were checked using CMC as a substrate, and fractions with high endoglucanase activity were collected. The collected fractions were dialyzed overnight against 20 mM potassium acetate buffer (pH 5.0) at 4 °C. After dialysis, the dialysate was added to an Avicel affinity column (Fluka BioChemika, Buchs, Switzerland, flow rate; 2.0 mL/min, column dimension; 1.4 cm by 5.5 cm, fraction volume; 1.0 mL) equilibrated with 20 mM potassium acetate buffer (pH 5.0). The column was washed with the same acetate buffer, and elution was performed using a stepwise method, changing from 20 mM potassium acetate buffer to Milli-Q water (Shibata et al. 2017). All eluted fractions were checked to assess their endoglucanase activity levels using CMC as the substrate, and then analyzed using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Fractions with high endoglucanase activity levels were collected and concentrated using Centricut W-10 (Kurabo, Osaka, Japan) to obtain an appropriate concentration for enzyme preparation, which were then used for enzyme characterization. The chromatography procedure was performed using an ÄKTA prime plus chromatography system (GE Healthcare, Tokyo, Japan). SDS-PAGE SDS-PAGE was performed using the method described by Laemmli (1970) as previously described (Kudo et al. 2015). The stacking and the separating gels were 4.5% and 12.5% polyacrylamide, respectively. Proteins were stained with Coomassie Brilliant Blue R-250. The molecular weight standards used were Precision Plus Protein Standards, Unstained, containing ten recombinant protein bands at 10, 15, 20, 25, 37, 50, 75, 100, 150, and 250 kDa (Bio-Rad Laboratories, Hercules, CA, USA). Protein measurement The protein content in the samples was measured using the method of Bradford (1976) with bovine serum albumin (BSA) as a standard. Optimal temperature and pH, and substrate specificity of purified EgmA Optimal temperature of purified EgmA was determined by incubating the enzyme at various temperatures for 10 min at pH 4.0, and the optimal pH of purified enzyme was analyzed by measuring the activity at various pHs for 10 min at 50 °C using CMC and galactomannan as substrates, as described above. The substrate specificity of the purified enzyme was examined by incubating at 50 °C for 10 min at pH 4.0, with various substrates, using the method described in the section on assays for endoglucanase and mannanase activities. The substrates used were CMC, Avicel, beechwood xylan (Sigma-Aldrich, St Louis, MO, USA), laminarin (Sigma-Aldrich), lichenan (Megazyme), glucomannan (Megazyme), and galactomannan (Megazyme). Enzyme activity was quantified by measuring reducing sugar release using the DNS method. The sugar standards used were glucose for CMC, Avicel, laminarin, and lichenan, xylose for beechwood xylan, and mannose for glucomannan and galactomannan. One unit of enzyme activity was defined as the amount of enzyme releasing 1 μmol of reducing sugar per minute in glucose, xylose, or mannose equivalents. Kinetic analysis of EgmA The kinetic parameters, Michaelis constant (Km), and turnover number (kcat) of EgmA toward glucomannan and galactomannan were determined as described above. A 25 μL reaction mixture containing 50 mM sodium acetate buffer (pH 4.0), 0.3 μg of EgmA, and 5.0–15.0 mg/mL of CMC, 0.5–10.0 mg/mL of glucomannan, or 0.5–12.5 mg/mL of galactomannan was incubated at 50 °C for 10 min. The reaction was stopped by adding the DNS reagent, and the levels of reducing sugars produced were measured using the DNS method, as described above. The kinetic constants were calculated by nonlinear regression of the Michaelis–Menten equation (Kitaoka 2023). TLC analysis of hydrolyzed products of CMC, glucomannan, galactomannan, and cello- or manno-oligosaccharides by EgmA Purified EgmA (0.7 μg) was incubated with 0.8% of CMC, glucomannan, or galactomannan, and 4.5 mM cello-oligosaccharides (cellobiose, cellotriose, cellotetraose, cellopentaose, and cellohexaose (Megazyme)) or manno-oligosaccharides (mannobiose, mannotriose, mannotetraose, mannopentaose, and mannohexaose (Megazyme)) in 50 mM acetate buffer (pH 4.0) at 50 °C for 0–24 h (CMC, glucomannan, and galactomannan) or for 20 h (cello-oligosaccharides and manno-oligosaccharides). After incubation, each reaction mixture was stopped by heating at 100 °C for 10 min and cooled, and was placed on thin-layer chromatography (TLC) plates (silica gel 60; Merck, Darmstadt, Germany) and separated with solvents of isopropanol:acetic acid:water = 4:1:1 (CMC, glucomannan, and galactomannan), 1-buthanol:ethanol:water = 10:8:7 (manno-oligosaccharides), or 1-buthanol:acetic acid:water = 2:1:1 (cello-oligosaccharides). Each plate was dried immediately before the separation solvent reached the upper end. The reaction products were detected by spraying a staining reagent of sulfuric acid:ethanol = 5:95 (CMC, glucomannan, and galactomannan), anisaldehyde:sulfuric acid:acetic acid:ethanol = 2.5:3.4:1:93.1 (manno-oligosaccharides and cello-oligosaccharides) to the plate and baking the plate at 110 °C for 10 min. Results Identification of a putative novel endoglucanase gene from A. oryzae genome and its overexpression As described in the “Materials and methods” section, genome mining using the CAZy and CAoGDX databases revealed that A. oryzae RIB40 contains two genes that encode putative endoglucanases belonging to the GH5 family (Fig. 1). The protein (EgmA) encoded by AO090011000715 comprises 490 amino acid residues with a signal peptide and three domains (catalytic domain, Ser/Thr-rich linker domain, and CBM) and has consensus amino acid sequences (VIFDTNNEYH) around the catalytic center region of GH5 (PROSITE accession No. PS00659) (Supplementary Fig. S1). As previously characterized endoglucanases such as CelA and CelB (Kitamoto et al. 1996) belong to GH12 and GH7, respectively, and have no CBM, the enzyme properties of EgmA are of significant interest. Therefore, the egmA gene was overexpressed in A. oryzae using the improved glaA promoter (Minetoki et al. 2003), and the gene product was purified for characterization.Fig. 1Phylogenetic analysis of GH5. Phylogenetic tree of GH5 orthologs from A. nidulans FGSC A4, A. niger CBS513.88, A. oryzae RIB40, Neurospora crassa OR74A, Trichoderma reesei QM6a, and T. reesei RutC-30. Proteins possessing CBM1 are denoted as (CBM1). Previously characterized enzymes are indicated by (C). EgmA are boxed. Evolutionary analyses were conducted using MEGA12. The percentages of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches Phylogenetic analysis of GH5. Phylogenetic tree of GH5 orthologs from A. nidulans FGSC A4, A. niger CBS513.88, A. oryzae RIB40, Neurospora crassa OR74A, Trichoderma reesei QM6a, and T. reesei RutC-30. Proteins possessing CBM1 are denoted as (CBM1). Previously characterized enzymes are indicated by (C). EgmA are boxed. Evolutionary analyses were conducted using MEGA12. The percentages of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches The egmA overexpression strain was grown on a CD plate containing 1% maltose as the carbon source and 0.1% CMC, followed by staining with Congo Red. The strain formed a clear zone around the colony, indicating that EgmA has hydrolytic activity toward CMC (Supplementary Fig. S2). Purification of EgmA EgmA was purified from the culture supernatant of the overexpressing strain using the procedure described in the “Materials and methods” section. The purification of EgmA is summarized in Table 1. Purified EgmA migrated as a single protein band on SDS-PAGE (Fig. 2A) but exhibited a molecular weight different from that predicted based on the amino acid sequence of this enzyme. The N-terminal amino acid sequence of EgmA was determined using the Edman method (Edman and Begg 1967) to be GSGFTWVGVS (Fig. 2B), which corresponds to a region from the 28th to 37th amino acids in the predicted sequence of EgmA, indicating that the target protein was EgmA itself, produced and purified. The SignalP server predicted that the signal peptide would be cleaved between the 18th and 19th amino acid residues, which was inconsistent with the results of the N-terminal amino acid sequence analysis. This discrepancy may have been caused by processing with a Kex2-like protease that recognizes Lys-Arg residues (Fig. 2B). The apparent molecular mass of purified EgmA in the SDS-PAGE analysis was approximately 90 kDa, which was notably larger than the predicted molecular mass (51.6 kDa) from the amino acid sequence of the mature protein. The protein band did not shift from approximately 90 kDa after Endo Hf treatment (Fig. 2C), indicating that the protein is not N-glycosylated but is likely O-linked glycosylated to the S/T-rich region located between the catalytic domain and the CBM. Using the NetOGlyc 4.0 server (https://services.healthtech.dtu.dk/services/NetOGlyc-4.0/), several potential O-glycosylation sites within the Ser/Thr-rich linker domain of EgmA were identified. Table 1Purification of EgmAPurification stepsTotal activity(U)Total protein(mg)Specific activity(U/mg)Purification foldCulture supernatant147061.723.81TOYOPEARL Butyl-650 M2377.332.31.4Avicel chromatography1480.91646.9Fig. 2SDS-PAGE of purified EgmA, amino acid sequence of N-terminus of EgmA predicted from gene, and SDS-PAGE of purified EgmA treated with or without Endo Hf. (A) SDS-PAGE was performed on a slab gel of 12.5% polyacrylamide and proteins were stained with Coomassie Brilliant Blue R-250. Lane 1: standard proteins; lane 2: culture supernatant; lane 3: partially purified EgmA after TOYOPEARL Butyl-650 M chromatography; lane 4: purified EgmA (arrow) after Avicel affinity chromatography. (B) MRIGNLIVAASAASLVHA (underlined) and KR (underlined) show predicted signal sequence of EgmA and processing site by kex2-like protease, respectively. The squared sequence (GSGFTWVGVS) indicates the determined N-terminal amino acid sequence of purified EgmA. (C) Lane 1, molecular weight standards; 2, purified EgmA; 3, purified EgmA treated with Endo Hf Table 1Purification of EgmAPurification stepsTotal activity(U)Total protein(mg)Specific activity(U/mg)Purification foldCulture supernatant147061.723.81TOYOPEARL Butyl-650 M2377.332.31.4Avicel chromatography1480.91646.9 Purification of EgmA SDS-PAGE of purified EgmA, amino acid sequence of N-terminus of EgmA predicted from gene, and SDS-PAGE of purified EgmA treated with or without Endo Hf. (A) SDS-PAGE was performed on a slab gel of 12.5% polyacrylamide and proteins were stained with Coomassie Brilliant Blue R-250. Lane 1: standard proteins; lane 2: culture supernatant; lane 3: partially purified EgmA after TOYOPEARL Butyl-650 M chromatography; lane 4: purified EgmA (arrow) after Avicel affinity chromatography. (B) MRIGNLIVAASAASLVHA (underlined) and KR (underlined) show predicted signal sequence of EgmA and processing site by kex2-like protease, respectively. The squared sequence (GSGFTWVGVS) indicates the determined N-terminal amino acid sequence of purified EgmA. (C) Lane 1, molecular weight standards; 2, purified EgmA; 3, purified EgmA treated with Endo Hf Substrate specificity of EgmA EgmA showed hydrolytic activity toward CMC and lichenan, which are glucose polymers with β-(1 → 4)-linkage and β-(1 → 4)-linkage/β-(1 → 3)-linkage (5:2), respectively. The specific activity of EgmA was 164.2 U/mg for CMC and 1.5 U/mg for lichenan, respectively (Table 2), under optimal pH (pH 4.0) and temperature (50 °C) conditions (refer below). However, EgmA did not hydrolyze laminarin and xylan, which are glucose polymers with β-(1 → 3)-linkage/β-(1 → 6)-linkage (3:1) and a xylose polymer with β-(1 → 4)-linkage, respectively. EgmA showed no hydrolytic activity toward the insoluble microcrystalline cellulose Avicel (Table 2). Furthermore, EgmA also showed hydrolytic activities toward glucomannan and galactomannan, and the specific activity of EgmA was 6.6 U/mg for galactomannan and 79.6 U/mg for glucomannan, under optimal pH (pH 4.0) and temperature (50 °C) conditions (refer below). These results indicated that EgmA was a bifunctional endo-β-1,4-glucanase/mannanase. Table 2Substrate specificity of EgmASubstrateConcentration(mg/mL)Specific activity(U/mg)aCMC2.018.4 ± 4.1*5.096.1 ± 8.4*7.5164.2 ± 3.510.0252.4 ± 13.1*15.0376.1 ± 5.4*25.0432.5 ± 8.6*Avicel (microcrystalline cellulose)7.5N.DXylan7.5N.DLaminarin [(1 → 3)(1 → 6)-β-glucan]7.5N.DLichenan [(1 → 3)(1 → 4)-β-glucan]7.51.5 ± 0.3*Galactomannan7.56.6 ± 0.3*Glucomannan7.579.6 ± 5.0*aMean ± standard deviations (n = 3); N.D., not detected. Statistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to the specific activity at a CMC concentration of 7.5 (mg/mL) Table 2Substrate specificity of EgmASubstrateConcentration(mg/mL)Specific activity(U/mg)aCMC2.018.4 ± 4.1*5.096.1 ± 8.4*7.5164.2 ± 3.510.0252.4 ± 13.1*15.0376.1 ± 5.4*25.0432.5 ± 8.6*Avicel (microcrystalline cellulose)7.5N.DXylan7.5N.DLaminarin [(1 → 3)(1 → 6)-β-glucan]7.5N.DLichenan [(1 → 3)(1 → 4)-β-glucan]7.51.5 ± 0.3*Galactomannan7.56.6 ± 0.3*Glucomannan7.579.6 ± 5.0*aMean ± standard deviations (n = 3); N.D., not detected. Statistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to the specific activity at a CMC concentration of 7.5 (mg/mL) Substrate specificity of EgmA aMean ± standard deviations (n = 3); N.D., not detected. Statistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to the specific activity at a CMC concentration of 7.5 (mg/mL) Optimal pH and temperature on the activity of EgmA The optimal pH for the activity of EgmA was pH 4.0 at 50 °C when CMC (Fig. 3A) or galactomannan (Fig. 3B) was used as the enzyme reaction substrate, and the range in which EgmA showed maximum activity was slightly wider (pH 4.0–5.4) for galactomannan. EgmA retained over 80% of maximum activity between pH 2.9 and 5.4 for CMC, and between pH 3.1 and 6.1 for galactomannan.Fig. 3Optimal pH for CMC (A) and galactomannan (B), and optimal temperature for CMC (C) and galactomannan (D) of EgmA. ■: 50 mM glycine–HCl buffer, ●: 50 mM sodium acetate buffer, ▲: 50 mM potassium phosphate buffer. Error bars: standard deviations (SD; n = 3) Optimal pH for CMC (A) and galactomannan (B), and optimal temperature for CMC (C) and galactomannan (D) of EgmA. ■: 50 mM glycine–HCl buffer, ●: 50 mM sodium acetate buffer, ▲: 50 mM potassium phosphate buffer. Error bars: standard deviations (SD; n = 3) As for the optimal temperature for the activity of EgmA, the maximum activity was obtained at 50 °C at pH 4.0 for both CMC (Fig. 3C) and galactomannan (Fig. 3D) as substrates. TLC analysis of hydrolyzed products of CMC, glucomannan, galactomannan, and cello- or manno-oligosaccharides by EgmA To examine the mode of the hydrolysis reaction of EgmA, the hydrolyzed products of CMC, glucomannan, and galactomannan were analyzed using TLC (Fig. 4A). The hydrolysis of EgmA with CMC almost plateaued at approximately 1 h, and the reaction products were cellobiose and larger oligosaccharides, with cellobiose being the most frequently detected. On the other hand, for galactomannan, which consists of a primary chain of β-D-mannose and side chains of α-D-galactose, the hydrolysis reaction progressed gradually up to 24 h, and the reaction products were sugars larger than disaccharides, with numerous spots of incompletely degraded galactomannan remaining at the origin. For glucomannan, which is a polysaccharide composed of D-glucose and D-mannose units linked in a 1:1.6 ratio through β-1,4 bonds, the hydrolysis reaction progressed up to 24 h, and the reaction was faster than for galactomannan, showing almost no spots remaining at the origin, which is consistent with the results of the specific activity of EgmA toward each substrate shown in Table 2.Fig. 4TLC analysis of hydrolyzed products of (A) CMC, glucomannan, and galactomannan; (B) cello-oligosaccharides; and (C) manno-oligosaccharides through EgmA The hydrolyzed products of (A) CMC, glucomannan, and galactomannan; (B) cello-oligosaccharides; and (C) manno-oligosaccharides produced by EgmA were analyzed using the procedure described in the “Materials and methods” section. Gn: standards containing glucose (G1), cellobiose (G2), cellotriose (G3), cellotetraose (G4), cellopentaose (G5), and cellohexaose (G6). Mn: standards containing mannose (M1), mannobiose (M2), mannotriose (M3), mannotetraose (M4), mannopentaose (M5), and mannohexaose (M6) TLC analysis of hydrolyzed products of (A) CMC, glucomannan, and galactomannan; (B) cello-oligosaccharides; and (C) manno-oligosaccharides through EgmA The hydrolyzed products of (A) CMC, glucomannan, and galactomannan; (B) cello-oligosaccharides; and (C) manno-oligosaccharides produced by EgmA were analyzed using the procedure described in the “Materials and methods” section. Gn: standards containing glucose (G1), cellobiose (G2), cellotriose (G3), cellotetraose (G4), cellopentaose (G5), and cellohexaose (G6). Mn: standards containing mannose (M1), mannobiose (M2), mannotriose (M3), mannotetraose (M4), mannopentaose (M5), and mannohexaose (M6) To further verify whether EgmA hydrolyzes both β-(1 → 4)-glucosidic and β-(1 → 4)-mannosidic bonds, we used cello- or manno-oligosaccharides as substrates. As shown in Fig. 4B, EgmA hydrolyzes cellotetraose (G4), cellopentaose (G5), and cellohexaose (G6), producing cellotriose (G3), cellobiose (G2), and a small amount of glucose (G1). EgmA did not degrade cellobiose (G2) or cellotriose (G3). As shown in Fig. 4C, EgmA hydrolyzes mannopentaose (M5), primarily producing mannotriose (M3), mannobiose (M2), and mannohexaose (M6), which in turn produces mannotetraose (M4), mannotriose (M3), and mannobiose (M2), with a trace amount of mannose (M1). EgmA did not degrade mannobiose (M2), mannotriose (M3), and mannotetraose (M4). From these results, EgmA can hydrolyze β-(1 → 4)-glucosidic bonds as well as β-(1 → 4)-mannosidic bonds, indicating that EgmA is a bifunctional endo-β-1,4-glucanase/mannanase. Furthermore, differences in reactivity toward cellotetraose (G4) and mannotetraose (M4) indicated that EgmA exhibited distinct modes of action toward cello- and manno-oligosaccharides. Kinetic properties of EgmA We analyzed the kinetic properties of EgmA for galactomannan and glucomannan (Table 3). The Km of EgmA for glucomannan was 1.4-fold higher than that for galactomannan. However, the kcat value of EgmA for glucomannan was 14.6-fold higher than that for galactomannan. The catalytic efficiency constant (kcat/Km) of EgmA for glucomannan was 10.7-fold higher than that for galactomannan. These results indicate that glucomannan is a more readily hydrolyzable substrate for EgmA than galactomannan. Even as the CMC concentration in the reaction system increased, the reaction rate did not reach a plateau, making it impossible to calculate the individual parameters for CMC. Therefore, the specific activities of EgmA were compared by setting CMC concentrations ranging from 2 to 25 mg/mL (Table 2). In conclusion, the specific activity results (Table 2) indicated that CMC was more readily hydrolyzed than glucomannan by EgmA. In conjunction with the results of the catalytic constant (kcat/Km) (Table 3), it was concluded that EgmA hydrolyzed the substrates in the following order of ease: CMC, glucomannan, and galactomannan. Table 3Kinetic parameters of EgmA toward galactomannan and glucomannanSubstrateKm (mg mL−1)akcat (s−1)akcat/Km (mL mg−1 s−1)Galactomannan3.9 ± 1.17.9 ± 0.72.0Glucomannan5.4 ± 1.7115.5 ± 16.421.4aMean ± standard deviations (n = 3) Table 3Kinetic parameters of EgmA toward galactomannan and glucomannanSubstrateKm (mg mL−1)akcat (s−1)akcat/Km (mL mg−1 s−1)Galactomannan3.9 ± 1.17.9 ± 0.72.0Glucomannan5.4 ± 1.7115.5 ± 16.421.4aMean ± standard deviations (n = 3) Kinetic parameters of EgmA toward galactomannan and glucomannan aMean ± standard deviations (n = 3) Effects of metal ions and EDTA on EgmA activity The effects of metal ions and EDTA on EgmA activity were examined using the procedure described in the “Materials and methods” section by determining the endoglucanase or mannanase activity in the presence of 5 or 10 mM metal ions or EDTA in the reaction mixture (Table 4). The maximum endoglucanase activity of EgmA was observed in the presence of Co2+, approximately 20% higher than that of the control (no additives in the reaction solution). No other ions induced an increase in EgmA activity. In contrast, the presence of Cu2+, Fe2+, and Fe3+ ions at concentrations of 5 and 10 mM inhibited enzyme activity compared to the control. Moreover, EDTA also inhibited enzyme activity at concentrations of both 5 and 10 mM compared with the control. Other ions (Sr2+ (10 mM), Zn2+ (5 and 10 mM), Ni2+ (5 and 10 mM), and Mn2+ (5 and 10 mM)) showed slightly decreased endoglucanase activity. Similarly, the maximum mannanase activity of EgmA was observed in the presence of 5 mM Co2+, which was approximately 50% higher than that of the control. Besides Co2+, only Mn2+ (5 mM) showed increased enzyme activity, approximately 30% higher than that of the control. In contrast, the presence of Cu2+ (5 and 10 mM), Fe2+ (10 mM), and Fe3+ (5 and 10 mM) inhibited the enzyme activity compared to the control. Table 4Effects of metal ions and EDTA on EgmA activityAdditivesEndoglucanase activity (%)a,bMannanase activity (%)a,c5 mM10 mM5 mM10 mMNone100.0 ± 8.9100.0 ± 6.6100.0 ± 2.5100.0 ± 2.5SrCl298.4 ± 4.983.8 ± 5.5*96.9 ± 2.5100.3 ± 3.9BaCl290.8 ± 7.990.6 ± 4.4100.7 ± 4.3100.3 ± 2.4FeCl254.5 ± 12.4*70.3 ± 2.8*96.8 ± 6.069.0 ± 3.2*ZnCl283.3 ± 2.480.2 ± 5.1*91.1 ± 6.384.9 ± 4.4*CuCl252.9 ± 4.7*46.2 ± 9.0*48.9 ± 2.7*48.5 ± 1.2*NiCl286.4 ± 3.185.5 ± 2.7*95.4 ± 1.590.0 ± 9.0CoCl2122.2 ± 6.0*122.4 ± 10.8*156.6 ± 4.0*139.7 ± 10.2*MnCl280.5 ± 1.7*77.6 ± 5.7*131.2 ± 6.3*87.3 ± 1.8*CaCl288.6 ± 2.987.9 ± 3.793.0 ± 5.2105.7 ± 5.9MgCl289.1 ± 3.389.3 ± 3.991.5 ± 4.9101.1 ± 6.8FeCl346.3 ± 6.8*51.0 ± 3.7*67.6 ± 3.9*46.0 ± 4.1*EDTA80.8 ± 2.7*68.5 ± 6.3*N.DN.DaMean ± standard deviations (n = 3), substrate; bCMC; cgalactomannan; N.D., not determinedStatistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to activities in the absence of additives Table 4Effects of metal ions and EDTA on EgmA activityAdditivesEndoglucanase activity (%)a,bMannanase activity (%)a,c5 mM10 mM5 mM10 mMNone100.0 ± 8.9100.0 ± 6.6100.0 ± 2.5100.0 ± 2.5SrCl298.4 ± 4.983.8 ± 5.5*96.9 ± 2.5100.3 ± 3.9BaCl290.8 ± 7.990.6 ± 4.4100.7 ± 4.3100.3 ± 2.4FeCl254.5 ± 12.4*70.3 ± 2.8*96.8 ± 6.069.0 ± 3.2*ZnCl283.3 ± 2.480.2 ± 5.1*91.1 ± 6.384.9 ± 4.4*CuCl252.9 ± 4.7*46.2 ± 9.0*48.9 ± 2.7*48.5 ± 1.2*NiCl286.4 ± 3.185.5 ± 2.7*95.4 ± 1.590.0 ± 9.0CoCl2122.2 ± 6.0*122.4 ± 10.8*156.6 ± 4.0*139.7 ± 10.2*MnCl280.5 ± 1.7*77.6 ± 5.7*131.2 ± 6.3*87.3 ± 1.8*CaCl288.6 ± 2.987.9 ± 3.793.0 ± 5.2105.7 ± 5.9MgCl289.1 ± 3.389.3 ± 3.991.5 ± 4.9101.1 ± 6.8FeCl346.3 ± 6.8*51.0 ± 3.7*67.6 ± 3.9*46.0 ± 4.1*EDTA80.8 ± 2.7*68.5 ± 6.3*N.DN.DaMean ± standard deviations (n = 3), substrate; bCMC; cgalactomannan; N.D., not determinedStatistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to activities in the absence of additives Effects of metal ions and EDTA on EgmA activity aMean ± standard deviations (n = 3), substrate; bCMC; cgalactomannan; N.D., not determined Statistical significance was verified using a two-tailed t-test for unequal variances. *p < 0.05 compared to activities in the absence of additives Discussion In the present study, we screened putative novel endoglucanase genes from A. oryzae RIB40 genome (Machida et al. 2005). In addition to two endoglucanase genes, celA and celB, previously reported (Kitamoto et al. 1996), we identified additional genes encoding putative endoglucanases in the genome and overexpressed one of them, egmA, encoding the enzyme containing CBM at the C-terminus in A. oryzae. EgmA exhibits endoglucanase activity. From the substrate specificity analysis, interestingly, EgmA appeared to hydrolyze only the β-(1 → 4)-linkage of soluble glucose and mannose polymers because this enzyme showed hydrolytic activities toward CMC, lichenan, galactomannan, and glucomannan but not toward laminarin, xylan, and Avicel (Table 2). Furthermore, EgmA degraded cello-oligosaccharides longer than cellotetraose (Fig. 3B) and hydrolyzed manno-oligosaccharides longer than mannopentaose (Fig. 3C). These results showed that EgmA hydrolyzed β-(1 → 4)-glucosidic bonds as well as β-(1 → 4)-mannosidic bonds, indicating that EgmA is a bifunctional endo-β-1,4-glucanase/mannanase. To the best of our knowledge, this is the first study to characterize a bifunctional endo-β-1,4-glucanase/mannanase from Aspergillus fungi. Enzymes with activity against both glucan- and mannan-based polysaccharides have been reported in a few cases, but are rare among enzymes belonging to the GH5 family (Supplementary Table S1). ApCel5A from Aureobasidium pullulans has recently been reported, which is the first case of a bifunctional endo-β-1,4-glucanase/mannanase from fungal origin (Miao et al. 2021), whereas EgmA in this study is the second one. ManA from Thermoanaerobacterium polysaccharolyticum (Cann et al. 1999), DturCelB from Dictyoglomus turgidum (Fusco et al. 2018), Man/Cel5B from Thermotoga maritima (Sadaqat et al. 2021), and KG51 from the rumen metagenome of black goats (Lee et al. 2018) have been reported. Regarding specific activity, both fungal enzymes, EgmA and ApCel5A, showed higher values toward glucan-based polysaccharides. In contrast, all three thermophilic bacterial enzymes are highly thermostable and show higher specific activity toward mannan-based polysaccharides than glucans. Metagenome-derived KG51 has lower specific activity for both glucan and mannan polysaccharides than the fungal and bacterial enzymes. Structural analyses of cellulases and mannanases belonging to the GH5 family have shown similarities in their active sites (Pereira et al. 2010). This may explain the dual function of T. maritima Man/Cel5B (Sadaqat et al. 2021) and D. turgidum DturCelB (Fusco et al. 2018). Two conserved glutamate residues in the active sites of these enzymes act as acids/bases and nucleophiles during substrate hydrolysis (Sakon et al. 1996; Hilge et al. 1998). In addition, Man/Cel5B uses Glu 139 and Glu 259 as active site residues toward substrate hydrolysis, revealing that the similarity of the active site of GH5 cellulase and GH5 mannanase may be the reason for the duality of Man/Cel5B (Sadaqat et al. 2021). Therefore, further comparisons of the cellulose- and mannan-binding crystal structures of EgmA with those of apoenzymes are required to gain further insight into the dual substrate recognition mechanism of EgmA. The superior ability of EgmA to degrade mannan and cellulose may be preferentially exploited for lignocellulosic biomass degradation. Chen et al. (2012) reported that seven amino acid residues (N20, E23, P53, H95, H96, and E287) determine the substrate specificity of T. maritima Cel5A (Supplementary Fig. S3). In addition, five of these seven amino acids (N20, E23, P53, H96, and E287) are involved in activity against mannans, while one (H95) governed activity against glucans (Chen et al. 2012). Among these, two residues (P53 and H95) were conserved in EgmA. Furthermore, residues H196 and Y198, which are involved in both endoglucanase and mannanase activities (Chen et al. 2012), were also conserved as the corresponding amino acids in EgmA. These facts would support that EgmA is a bifunctional endoglucanase/mannanase. In Fig. 4A, no significant changes in hydrolysis products were observed in the degradation of CMC and galactomannan over a reaction time more than 1 h. Also, as the amount of degradation products increased only slightly over time when glucomannan was used as the substrate, this is likely attributed to enzyme inactivation due to prolonged reaction time at the temperature of 50 °C. Furthermore, CMC is a chemically modified substrate along the cellulose backbone with carboxymethyl groups. Studies on cellulase using CMC as a substrate have reported that highly substituted regions on the cellulose backbone prevent the enzyme from binding to glycosidic linkages, thus limiting the hydrolysis of modified cellulose by endoglucanases to polymer fragments with no or few substituents. Additionally, endoglucanases are unlikely accept substituents on hydroxyl groups in the vicinity of the cleavage site of cellulose derivatives, and even substituents that are not directly involved in enzyme–substrate complex formation may prevent enzyme binding due to steric hindrance (Enebro et al. 2009; Karlsson et al. 2002). As shown in Fig. 4A, the rapid hydrolysis observed in the early stage of the enzyme reaction of CMC is considered to be due to the selective and complete degradation of unsubstituted glucosyl chains. In contrast, the subsequent hydrolysis reaction hardly progressed, suggesting that the remaining intermediate products have a high density of substituents, causing steric hindrance in binding to the enzyme active site, thereby preventing further hydrolysis (Enebro et al. 2009). Similarly, the enzymatic degradation patterns of galactomannan are also presumed to be attributed to the high degree of galactose side chains, which prevents the enzyme from accessing the mannose backbone by steric hindrance. We plan to elucidate the substrate recognition mechanism using structural analysis and other methods to examine the effects of CMC chemical modification and steric hindrance caused by side chains of galactomannan on enzyme activity. EgmA has a CBM (Supplementary Fig. S1), which is classified as a family 1 CBM, along with Aspergillus luchuensis Cel61A (Hara et al. 2004) and T. reesei Cel61A (Karlsson et al. 2001), among others. Family 1 CBMs are exclusively found in fungi and mostly bind to cellulose (Xiao et al. 2001). Although EgmA lacks catalytic activity toward Avicel (Table 2), it likely retains sufficient binding affinity for Avicel via its CBM1. Based on this cellulose-binding property, EgmA was purified using Avicel affinity chromatography. The primary biological function of CBMs is to enhance enzyme activity by facilitating the binding of their catalytic domains to substrates, thereby increasing the local enzyme concentration. In addition, family 1 CBM can disrupt crystalline cellulose structure, thereby increasing substrate access (Shoseyov et al. 2006; Xiao et al. 2001). However, the CBM1 of EgmA could not disrupt the crystalline cellulose structure because of lack of hydrolytic activity toward Avicel of EgmA. The optimal pH for EgmA was acidic (pH 4.0) (Fig. 3A, B). This property has also been observed in previously characterized endoglucanases from fungi (such as A. oryzae CelA and CelB (Kitamoto et al. 1996)), and is suitable for hydrolyzing cellulose biomass because pretreatment of lignocellulosic substrates is generally performed using acids to disrupt the crystalline structure of cellulose before cellulase digestion (Kádár et al. 2007). The relative activity of EgmA increased in the presence of Co2+ when both CMC and galactomannan were used as substrates (Table 4). Such phenomena were also observed for β-mannanase activity of T. maritima Man/Cel5B (Sadaqat et al. 2021), β-mannanases of Phialophora sp. P13 (MAN5AP13) (Zhao et al. 2010), and Bacillus nealsonii PN-11 (Chauhan et al. 2014). Metal ions have been reported to enhance enzymatic activity by assisting in proper substrate binding at the active site (Andreini et al. 2008), and the increase in the relative activity of EgmA by Co2+ is presumed to be because of this. In contrast, a marked decrease in the relative activity of EgmA in the presence of Cu2+, Fe2+, and Fe3+ was observed. A decrease in the relative activity in the presence of Cu2+ has also been reported for Man/Cel5B (Sadaqat et al. 2021) and DturCelB (Fusco et al. 2018). Changes in enzymatic activity in the presence of Cu2+ have been reported to be due to changes in the net charge on the enzyme surface (Zheng et al. 2016). Similar to Man/Cel5B (Sadaqat et al. 2021) and DturCelB (Fusco et al. 2018), EDTA does not affect EgmA activity, indicating that it is not a metalloenzyme. This study indicates that mining of novel cellulases with beneficial activities from microbial genome sequences is highly useful for the identification of enzymes suitable for the degradation of lignocellulosic biomass. Certain potentially useful enzymes for industrial applications are expected to be found among other endoglucanase genes that have not yet been characterized in the A. oryzae genome. In addition to endoglucanases, useful cellobiohydrolases and β-glucosidases, as well as hemicellulases such as xylanase and β-xylosidase, could be screened in the A. oryzae genome. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file 1 (PDF 365 KB) Supplementary file 1 (PDF 365 KB)