The gene, IDSGLUC9-4 (GenBank accession no. WUV 41754.1), was annotated using metatranscriptomic data obtained previously [13], then the sequence was codon optimized and synthesized by GenScript (Nanjing, China). Escherichia coli (E. coli) BL21(DE3) and the expression vector pET30a(+) were produced in our laboratory [14]. The substrates, barley β-glucan (CAS Number 9041-22-9; purity>95%), Icelandic moss lichenan (CAS Number 1402-10-4; purity>75%), tamarind xyloglucan (CAS Number 37294-28-3; purity>95%), konjac gum glucomannan (CAS Number 11078-31-2; purity >98%), cellotriose (CAS Number 33404-34-1; purity>95%), cellotetraose (CAS Number 38819-01-1; purity>90%) and cellopentaose (CAS Number 2240-27-9; purity>95%) were from Megazyme (Wicklow, Ireland), cellobiose (CAS Number 528-50-7; purity>98%) was from Sigma-Aldrich (Darmstadt, Germany). Culture media, kanamycin, 6× His-tagged Ni-NTA resin and isopropyl-thio-β-D-galactopyranoside (IPTG) were from Sangon Biotech (Shanghai, China).

Isoelectric point and molecular weight (pI, Mw) were predicted using the Expasy online tool (https://web.expasy.org/compute_pi/). Signal peptide and functional domain prediction were performed using InterPro (https://www.ebi.ac.uk/interpro/). SWISS-MODEL was employed for comparative modeling of tertiary structures (https://swissmodel.expasy.org/interactive/). Multiple sequence alignment and phylogenetic analysis were conducted using MEGA software (v.11.0; Kumer Lab, Temple University, Philadelphia, PA, USA). Phylogenetic analysis used the Maximum Likelihood statistical method, employing the WAG correction model. The phylogeny assessment used the bootstrap method with 500 bootstrap replications. MEGA and SWISS-MODEL were used for this analysis.

The transformation of the plasmid pET30a-IDSGLUC9-4 was carried out by heat shocking it into E. coli BL21(DE3) competent cells and streaking the mixture onto Luria-Bertani (LB) plates (5 g/L yeast extract, 10 g/L tryptone, 10 g/L sodium chloride, 18 g/L agar, 100 μg/mL kanamycin). Colonies containing the IDSGLUC9-4 gene were confirmed through colony-PCR and plasmid sequencing, and designated as BL21/pET30a-IDSGLUC9-4. The recombinant strain was added to LB medium (500 mL) with a 1% inoculum, then cultured at 37°C with continuously shaking at 180 rpm until the OD₆₀₀ reached 0.6 to 1.0. Then, IPTG (250 μL, 1 mol/L) was added to induce protein expression at 16°C with gentle agitation at 100 rpm for 16 h. The cultured cells were harvested by centrifugation at 4°C and 6,000×g for 10 min, then resuspended in 1× phosphate-buffered saline (50 mL). A 15-min sonication at 65% power, with a 4 s pause every 2 s, was performed using an ultrasonic disruptor (TZL-1200; Perwell, Suzhou, China) to yield the crude enzyme. The crude enzyme solution was treated with Affinity Ni-NTA agarose resin at a ratio of 1:50, then supplemented with 1 mL of 1 mol/L imidazole. Following this, the mixture was gently agitated at 100 rpm for 1 hour at 0°C. The mixture was gently loaded into a chromatography column and the target protein was eluted stepwise with phosphate buffer, containing 20, 50, or 250 mmol/L imidazole. The purified protein was used for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and analysis of substrate selectivity and product profile. The enzyme substrate selectivity was tested on agar plates containing 0.2% (w/v) barley β-glucan, Icelandic moss lichenan, or konjac gum. Approximately 20 U of enzyme activity was spotted onto the plates, which were incubated for 16 h at 25°C, stained with a 0.1% (w/v) Congo red solution for 20 min, then destained with 1 mol/L NaCl until the clear areas became visible.

Enzyme activity was assessed by the 3,5-dinitrosalicylic acid (DNS) reducing sugar assay [15], and the protein concentration was determined using the Bradford assay [16]. One unit (U) of glucanase was defined as the amount of enzyme needed to generate 1 μmol of reducing sugar per min. Briefly, enzyme solution (40 μL, ~0.32 μg) and substrate (40 μL, 5 mg/mL of barley β-glucan, Icelandic moss lichenan, locust bean gum, or konjac glucomannan), were mixed and reacted for 15 min under the optimum conditions (40°C, pH 6.0), then the amount of reducing sugar produced was quantified with DNS; all assays were performed in triplicate.

Enzyme solution (3 mL, ~24 μg) was combined with polysaccharide substrate solution (6 mL; 5 mg/mL of barley β-glucan, Iceland moss lichenan; 0.25 mg/mL of oligosaccharides, G1, glucose; G2, cellobiose; G3, cellotriose; G4, cellotetraose; G5, cellopentaose) and incubated at 37°C. Samples were collected at 0, 20 s, 5, 10, 20, and 30 min, and 1, 3, 6, 12, 24, and 48 h; the reaction was terminated by heating at 95°C for 15 min and the samples centrifuged at 4°C and 10,000×g. The supernatant was used for product profile analysis in three parallel experiments. The hydrolysis products generated by IDSGLUC9-4 were analyzed by thin-layer chromatography (TLC) with n-butanol/acetic acid/water (5:2:3, v/v), as described previously [17], with an oligosaccharide standard mixture (G1–G5) for comparison.

A more in-depth analysis of the hydrolysis products was performed by high-performance liquid chromatography (HPLC), with an LC-1200 instrument (Agilent Technologies, Santa Clara, CA, USA), fitted with an Asahipak NH2P-50 4E chromatography column (Shodex, Tokyo, Japan) and an RID-20A refractive index detector. The mobile phase was 65% aqueous acetonitrile, with a column temperature of 40°C and a flow rate of 1.0 mL/min. In addition, hydrolysis samples taken at 3 and 48 h were analyzed by Ultraflextreme MALDI-TOF/TOF (Bruker, Billerica, MA, USA).

The experiments conducted in this study, including protein concentration determination, DNS colorimetric reaction in enzymatic property studies, and quantification of oligosaccharide concentration in HPLC experiments, were all performed in triplicate. Image preparation and data analysis, including standard deviation and mean calculation, were conducted using GraphPad Prism v.8.0 (San Diego, CA, USA). Data were presented as the mean±standard deviation (n = 3). The section on “Effects of Metal Ions and Organic Reagents on Glucosidase Activity” involved single-factor comparisons. Statistical significance was analyzed using the Student’s t-test integrated into GraphPad Prism (* p< 0.05; ** p<0.01; *** p<0.001). Significant differences were assessed by comparing each experimental group with the control group treated without inhibitors, with the control group set as 100%.

Ruminants, such as sheep and cattle, are unable to digest cellulose and other plant cell wall polysaccharides, so they obtain most of their nutrition indirectly from the action of cellulolytic microorganisms in their rumens [18]. Therefore, the rumen microbiome is a rich source of industrial CAZymes with high activity and useful properties. In this study, a GH9 β-glucanase gene, IDSGLUC9-4 (GenBank accession no. WUV41754.1), was isolated from sheep rumen microbes, using transcriptomic data obtained previously [13]. The IDSGLUC9-4’s open reading frame spanned 1,716 bp, encoding 571 amino acids with a theoretical molecular mass of 63.6 kDa and an isoelectric point of 6.12. Multiple sequence alignment suggested that the IDSGLUC9-4 gene shared high homology (>99%) with two other nucleotide sequences which annotated as GH family 9 protein and cellulase (GenBank accession no. MBR6110719.1 and MBE6326145.1), located in a metagenome-assembled genome annotated as Paludibacteracea bacterium and Bacteroidales bacterium. However, neither of these homologous enzymes has been functionally characterized. Phylogenetic analysis revealed that IDSGLUC9-4 and another enzyme (GenBank accession no. MBR6110719.1) from the Paludibacteraceae are both in the GH9 family (Figure 1). The catalytic mechanism of GH9 enzymes involves an inversion of anomeric stereochemistry. Cel9A, a processive endoglucanase from Thermobifida fusca, is active against bacterial cellulose and is the only known cellulase capable of independently degrading crystalline regions in bacterial cellulose, although it prefers amorphous regions [19]. A closely related cellulase from Clostridium phytofermentans is the only GH9 cellulase encoded in its genome and is essential for cellulose degradation by the organism. Notably, this is the only documented instance of a single cellulase being essential for growth on cellulose [20]. Detailed sequence alignment suggests that IDSGLUC9-4 and six other GH9 enzymes have five shared amino acid residues, three catalytic residues (Asp182, Asp184, and Glu560) and two conserved aromatic residues (Trp436 and Tyr545) (Figure 2) [21–24].

To investigate the biochemical properties of IDSGLUC9-4, the gene was expressed heterologously in E. coli to produce the recombinant enzyme. Following 6× His-tagged affinity purification, a prominent band at ~69 kDa was observed by SDS-PAGE (Figure 3A), consistent with its theoretical molecular weight of 63.4 kDa plus the expression vector backbone sequence of 5.3 kDa. Zymogram analysis indicated that IDSGLUC9-4 was active towards barley β-glucan, Icelandic moss lichenan, xyloglucan, and konjac gum (Figure 3B to 3E).

The GH9 family protein derived from the rumen of Hu sheep exhibits the capability to hydrolyze mixed-linkage glucans. Comparative analysis with previously characterized GH9 family glucanases reveals the presence of similar catalytic residues, indicating the successful isolation of a novel, uncharacterized glucanase from the sheep rumen. This enzyme represents a common exogenous enzyme preparation in livestock production. In-depth enzymatic and hydrolytic property studies will facilitate a comprehensive understanding of its functionality.

The optimum pH of IDSGLUC9-4 was 6.0, with >75% of the maximum catalytic activity between pH 5.0 to 7.0 (Figure 4A). IDSGLUC9-4 was relatively stable (> 70%) between pH 4.0 to 6.0 (Figure 4B) and most stable at pH 6.0, but relatively unstable above pH 7.0. Notably, after preincubation for 1 h at pH 4.0 and 5.0, the residual activities were 78.22% and 89.53%, respectively, indicating that the enzyme and its unknown producing microorganism were well-adapted to the acidic environment of the rumen [17,18]. The optimum temperature for IDSGLUC9-4 was 40°C (Figure 4C) and the enzyme was much less stable at higher temperatures. Thermostability assays revealed that the IDSGLUC9-4 was sensitive to heat-challenge; after 1 h preincubation at 30°C and 40°C, the enzyme retained 75.10%±0.43% and 56.33%±1.14% of its initial activity, respectively (Figure 4D), and above 50°C, the activity decreased rapidly. These findings suggested that IDSGLUC9-4 was a relatively acid-resistant and mesophilic enzyme, which is consistent with the properties of other gastrointestinal tract-derived CAZymes reported previously [25–28].

Substrate selectivity analysis revealed that IDSGLUC9-4 could hydrolyze barley β-glucan, Icelandic moss lichenan, konjac glucomannan and tamarind xyloglucan; barley β-glucan was the substrate with the highest catalytic activity, 109.59± 3.61 μmol/mg min (Table 1). IDSGLUC9-4 was inactive towards beechwood xylan, galactomannan, guar gum, arabinan, Laminaria digitata laminaran, locust bean gum and arabinoxylan. The effects of metal ions and organic compounds on the activity of the enzyme, with β-glucan as substrate, were also determined at 40°C (Table 2). All the metal ions inhibited the enzyme, with Mn²⁺ the strongest inhibitor, as did all the organic compounds, with propanol and methanol the strongest inhibitors. Previously reported β-glucanases have diverse behaviors under the influence of metal ions and organic compounds. The activity of an exo-β-1,3-glucanase from the moose rumen microbiome [29] more than doubled in the presence of zinc ions, retaining normal activity in the presence of EDTA, propanol, and butanol. In contrast, the activity of IDSGLUC9-4 slightly decreased in the presence of 20 mM EDTA (p<0.05) but lost >50% of its activity in the presence of DMSO and propanol.

Enzymatic property studies revealed that this novel glucanase exhibits a specific activity of 109.59±3.61 μmol/mg min towards β-glucan. It shows stability within the temperature range of 30°C to 40°C. Additionally, it demonstrates tolerance to acidity, maintaining over 75% activity after 1 hour of incubation at pH 4.0 to 5.0. These preferences for moderate temperature and tolerance to weak acidity align with its origin from the rumen environment. Moreover, its activity towards four different polysaccharide substrates suggests promising application potential for IDSGLUC9-4. Further investigation into its hydrolytic mechanism is required to comprehensively elucidate its enzymatic properties.

To investigate the mechanism of action of IDSGLUC9-4, TLC, and HPLC were used to monitor the time-course profiles of oligosaccharides released from barley β-glucan, Icelandic moss lichen polysaccharide, tamarind xyloglucan, and oligosaccharides (Figures 5, 6, and 7). IDSGLUC9-4 initially liberated oligosaccharides with a degree of polymerization (DP) >5 from barley β-glucan and Icelandic moss lichenan; subsequently, these intermediates were further cleaved into smaller oligosaccharides as final products (Figures 5A, 5B, 6), indicating that IDSGLUC9-4 was an endo-acting glucanase (EC 3.2.1.4) [17]. IDSGLUC9-4 initially hydrolyzed barley β-glucan into G3, G4, and G5 (Figure 6); after 3 h of hydrolysis, the concentrations of G3, G4, and G5 constituted 1.10%, 49.53%, and 49.38% of total reducing sugars, respectively (Figures 6A, 6B). Matrix assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF) analysis at this time point confirmed the presence of G3–G7 and G9 (Figure 6C) and at longer times, the concentrations of G3 and G4 continued to increase, whereas G5 decreased with time., G2 was barely detectable up to 6 h, then its concentration rapidly increased up to 12 h, after which it was stable. IDSGLUC9-4 initially hydrolyzed Icelandic moss lichenan into G2, G3, and G4 (0 to 6 h); the product composition stabilized after 12 h and remained essentially constant until 48 h (Figures 6E, 6F). IDSGLUC9-4 produced mainly G2-G4 from Icelandic moss lichenan, whereas it produced G2–G5 from barley β-glucan. The total reducing sugar concentration after 48 h was 5,036.92±108.21 and 3,532.72±82.45 nmol/L from β-glucan and lichenan, respectively (Figures 6B, 6F). Mass spectrometry confirmed the initial presence of G3, G4, G5, G6, G7, and G9 during hydrolysis of both β-glucan and lichenan, whereas, after 48 h reaction, only the G3 and G4 oligosaccharides remained (Figures 7G, 6H). G3 and G4 were the main end-products from barley β-glucan, accounting for 36.86% and 50.56%, respectively, of the total reducing sugars, whereas G3 (53.37%) was the dominant end-product from Icelandic moss lichenan.

Although both barley β-glucan and Icelandic moss lichenan contain a mixture of β-1,3- and β-1,4-glycosidic linkages, the major products from IDSGLUC9-4 differed, probably because of different distributions of the β-1,3- and β-1,4-linkages within their main chains [27]. The β-glucan and lichen polysaccharide are mixed-linkage glucans (MLGs) with characteristic ratios of 1,4- to 1,3-β-D linkages. Considering the structure and product profile obtained from these MLG substrates, IDSGLUC9-4 appears to function as an endo-β-1,4-glucanase (EC 3.2.1.4), randomly cleaving β-1,4 linkages between glucose residues anywhere in the polysaccharide chain. HPLC detected the appearance of a peak around 12 min, corresponding to the presence of G6, with mass spectrometry detecting the production of higher DP oligosaccharides (G6, G7, G9) during the initial stage of the reaction. The concentrations of G2, G3, and G4 remained stable in both barley β-glucan and Icelandic moss lichenan reactions between 24 and 48 hours, indicating that the reaction had reached its endpoint. Notably, the enzyme had no activity against β-1,3-glucan, indicating its specificity for β-1,4-glycosidic linkages in its target substrates. This was very similar to the substrate specificity of the cellulases Cel5 and Cel5D from Ruminococcus albus [30]. IDSGLUC9-4 had a very low activity against tamarind xyloglucan, a mixed-linkage, highly branched glucan, consisting of xylose, galactose, arabinose and fucose [31]. IDSGLUC9-4, because of its specificity towards Glc1→4Glc linkages and inability to hydrolyse xylan or arabinan (Table 1), only cleaved tamarind xyloglucan into large oligosaccharides with DP >5 throughout the reaction (Figure 5C).

To further analyze the product profiles of IDSGLUC9-4, mono- and oligosaccharides were used as hydrolysis substrates (Figures 5D, 7D). TLC (Figure 5D) clearly showed that both G5 and G4 were effectively hydrolyzed by IDSGLUC9-4, generating G2 and G3 from G5 and G2 as the final product from G4. Quantitative determination by HPLC (Figures 7) showed that G5 was initially cleaved into equivalent amounts of G3 and G2, then the G3 concentration began to decrease at 1 h, whereas the G2 concentration increased, indicating hydrolysis of G3 into G2 (Figures 7A, 7A’). After 48 h, the products derived from G4 were G2 (2,007.05± 31.11 nmol/L) and G3 (151.40±5.13 nmol/L), whereas those from G5 were G2 (1,425.18±49.28 nmol/L) and G3 (461.40 ±40.38 nmol/L). IDSGLUC exclusively cleaved G4 into two molecules of G2 (Figures 7B, 7B’), but G2 was highly resistant to further hydrolysis; its concentration decreased only slightly from that at 3 h, even after 48 h (Figures 7D, 7D’). Hydrolysis of G3 into G2 (Figure 7C, 7C’) was incomplete and glucose was not detected, indicating that the glycosylation activity of the enzyme partially reversed the G3 hydrolysis and consumed all the glucose produced [9]. Notably, unlike the complete degradation of G4 and G5, the residual G3 concentration after 48 h was 304±15.27 nmol/L (29.3% of the initial amount), suggesting that IDSGlUC9-4 preferentially cleaves high-DP cello-oligosaccharides. In addition, glucose as substrate decreased from 3.070±0.085 to 0.866±0.014 (28.2% of the initial amount) after 48 h (Figure 7E and 7E’), indicating the formation of oligosaccharides and consistent with the glycosylation activity of IDSGlUC9-4.

Similarly, most characterized endo-glucanases hydrolyze polysaccharides into cello-oligosaccharides, such as G2, G3, G4, and G5 [17,32–34]. However, some glucanases have product profiles unlike normal endo-acting glucanases. For example, Cel5A-h38 from the sheep rumen has dual activities, i.e., endo-β-1,3-1,4-glucanase (EC 3.2.1.73) and exo-cellobiohydrolase (EC 3.2.1.91), and exclusively generates glucose and G2 from lichenan [27]. Another bifunctional glucanase/mannanase from Prevotella sp., IDSGH5-14, produces high-DP oligosaccharides from glucan- and mannan-like substrates [35]. In this study, IDSGlUC9-4 was active towards glucose (Figure 5D, 7E, and 7E’), apparently exhibiting glycosylation activity. Glycosylation activity is also commonly observed with β-glucosidases (EC3.2.1.21), which cleave alkyl- and aryl-β-glycosidic linkages, releasing glucose [9,36,37]; glycosylation may reduce feedback inhibition at high glucose concentrations. A recently reported GH3 aryl-β-glucosidase, GluLm, has not only normal β-glucosidase activity, but can also hydrolyze glycosylated phenolic compounds [37]. In the final stages of reactions using polysaccharides and higher molecular weight cello-oligosaccharides as substrates, various polymerization levels of cello-oligosaccharides coexist. This occurs even when compounds like G2, which can serve as final products in G3 and G4 reactions. Transglycosylation and hydrolysis may mutually influence each other, with transglycosylation potentially counteracting substrate hydrolysis as substrate concentrations increase. Initially, hydrolysis and transglycosylation occur at similar rates, producing glucose and oligosaccharides with higher DP. As the reaction progresses with increased substrate consumption, hydrolysis becomes predominant over transglycosylation. In the final phase, characterized by minimal substrate concentration, only hydrolysis is observed [38]. The modification of glycoside hydrolytic enzymes through glycine mutations, such as β-glucosidase, holds promise for advanced research in engineered microbes. This approach seeks to broaden substrate specificity and enhance transglycosylation activity. The glucotolerant β-glucosidase, BGL-1, derived from Talaromyces amestolkiae, shows lower catalytic efficiency in cello-oligosaccharide hydrolysis, potentially limiting its applicability in saccharification processes. The engineered glycosynthase variant, BGL-1-E521G, serves as a versatile tool for regioselective β-1,2 transglycosylation, displaying heightened efficiency in glycoside synthesis [39].

The analysis of the degradation mechanism of polysaccharides and oligosaccharides by IDSGLUC9-4 reveals its typical endo-glucanase activity, capable of hydrolyzing polysaccharide substrates containing β-1,4-glycosidic bonds to produce oligosaccharides with DP>2, with predominant hydrolysis products being G2, G3, and G4. Its action on fibrous oligosaccharides G1–G5 demonstrates the ability to hydrolyze oligosaccharides with DP≥3 into G2 and G3. The hydrolytic potential of polysaccharides and oligosaccharides suggests suitability as a feed enzyme preparation. It can hydrolyze cellulose, which is difficult for monogastric animals to digest, into oligosaccharides that are readily absorbable or utilizable by intestinal probiotics. The substrate spectrum of the enzyme includes β-glucan, lichenin, xylan, and laminarin, indicating its ability to degrade not only β-glucans with mixed linkages of 1,3–1,4 but also polysaccharides with other linkage forms. Future efforts may focus on enhancing the activity of recombinant glucanases through molecular evolution engineering, for application in the development of enzyme preparations for agricultural waste treatment or animal feed.