D-Cellooligosaccharides [D-cellobiose (C₂) to D-cellohexaose (C₆)], β-glucan from barley (low viscosity), curdlan from Alcaligenes faecalis, xyloglucan, and glucomannan were purchased from Megazyme International Ireland Ltd. (Wicklow, Ireland). All other materials including D-glucose (C₁), sodium carboxymethylcellulose (CMC), Avicel PH-101, chitosan, beechwood xylan, locust bean gum, para-nitrophenyl (pNP)-glucopyranoside, and pNP-cellobioside were provided by Sigma-Aldrich (St. Louis, MO, United States).

Genomic DNA of L. cladoniae PAMC 26568 was prepared from the cells, which were cultivated in R2A broth (BD Difco, Franklin Lakes, NJ, United States) for 14 days at 15°C, using a Mini Tissue DNA kit (Cosmo Genetech Co. Ltd., Seoul, South Korea) and was used as a template in polymerase chain reaction (PCR). With a PCR thermal cycler (TaKaRa, Kyoto, Japan), amplification of the gene encoding recombinant GluL proteins (rGluL) without a signal sequence was performed using the designed two gene-specific primers GluL-F (5′-CATATGGATGCCGATGTCCAGTGGC-3′) and GluL-R (5′-AAGCTTTCATGGCGAATGTCTTTCCGTCC-3′), which had restriction sites NdeI and HindIII, respectively. In this case, the PCR mixture (50 μL) included 2.5 U of FastStart Taq DNA polymerase (Roche, Basel, Switzerland), 250 μM of each dNTP, 2 pmol of each PCR primer, 20 ng of template DNA, and a PCR buffer. The initial template denaturation was carried out for 4 min at 95°C, followed by 35 cycles of 30 s at 95°C, 30 s at 54.5°C, and 1 min at 72°C. The amplified PCR products were separated by electrophoresis on a 1.2% agarose gel, followed by purifying the desired gene fragments using a NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Düren, Germany). The obtained gene fragments (975-bp) were then ligated into a pGEM-T easy vector (Promega, Madison, WI, United States). After transformation of the constructed pGEM-T easy/gluLR vectors into Escherichia coli DH5α competent cells, they were purified from the recombinant cells cultured in 100 mL of ampicillin (100 mg/L)-containing Luria-Bertani (LB) broth (BD Difco, Franklin Lakes, NJ, United States) for 14 h at 37°C. Subsequently, the recombinant plasmids were digested with restriction enzymes NdeI and HindIII and the resulting gluL fragments with the corresponding sticky ends were purified employing a NucleoSpin Gel and PCR Clean-up (Macherey-Nagel). For the recombinant production of GluL proteins, the obtained gene fragments were inserted into a pET-28a(+) expression vector (Novagen, Darmstadt, Germany) with the same sticky ends, after which the constructed pET-28a(+)/gluL vectors were transformed into E. coli BL21.

To produce N-terminal (His)₆-tagged rGluL, the cultivation of recombinant E. coli BL21 cells containing pET-28a(+)/gluL was carried out in a 5-L baffled flask, which included LB broth (1 L) and kanamycin (25 mg/L), in a rotary shaker (150 rpm) for 15 h at 30°C. The overexpression of gluL was induced by adding 1.0 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) after the optical density of the bacterial culture at 600 nm reached around 0.45. The rGluL-expressing cells were recovered from the culture broth by centrifugation (8,000 × g) for 20 min at 4°C and then stored at –20°C for 3 h. For the isolation of active rGluL proteins, the frozen recombinant E. coli BL21 cells were thoroughly suspended in a binding buffer (pH 7.4) consisting of 20 mM sodium phosphate, 0.5 M NaCl, and 20 mM imidazole, after which they were disrupted by sonication. The soluble fraction including rGluL proteins displaying endo-β-1,4-glucanase activity toward CMC was carefully collected by centrifugation (15,000 × g) for 20 min at 4°C. In this study, the simple purification of rGluL proteins was achieved by affinity chromatography using a HisTrap HP (Cytiva, Uppsala, Sweden) (5.0 mL) column connected to a fast protein liquid chromatography system (Amersham Pharmacia Biotech, Uppsala, Sweden), according to the manufacturer’s protocol. The N-terminal (His)₆-tagged rGluL proteins were collected from the column by a linear gradient elution method with imidazole (20–500 mM) at a flow rate of 2.0 mL/min. The fractions exhibiting high endo-β-1,4-glucanase activity were then collected, combined, and desalted with a HiPrep 26/10 desalting column (GE Healthcare) using 50 mM sodium phosphate buffer (pH 6.0) as the mobile phase. The desalted active fractions were recovered, pooled, and applied for further analysis.

The relative molecular mass of the denatured rGluL proteins was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in a 12.0% gel. After the electrophoresis, the gel was stained for 3 h with a 0.05% Coomassie Brilliant Blue R-250 solution (Bio-Rad Laboratories, Inc., Seoul, South Korea) to visualize the proteins separated by SDS-PAGE. Quantitative determination of the protein concentrations was performed using Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories, Inc., Seoul, South Korea) using bovine serum albumin as a standard.

The quantitative assay of endo-β-1,4-glucanase activity was conducted by measuring the amount of reducing sugars produced after the enzymatic degradation of CMC employing 3,5-dinitrosalicylic acid (DNS) reagent. In this case, a standard curve for the target reducing sugar by plotting the mean absorbance against the D-glucose concentration was constructed and used for the quantification of reducing sugars. The standard assay mixture (0.5 mL) was comprised of 1.0% CMC and rGluL solution (0.05 mL) diluted in 50 mM sodium citrate buffer (pH 4.0). The enzyme reactions were routinely conducted at 45°C for 10 min. After termination of the biocatalytic reaction, 0.75 mL of the DNS reagent was added to the assay mixture and then boiled for 5 min to develop the red-brown color. The absorbance of the red-brown color developed was measured at 540 nm. One unit (U) of endo-β-1,4-glucanase activity for CMC or barley β-glucan was defined as the amount of rGluL required to release 1 μmol of reducing sugar per min under standard reaction conditions.

The effect of pH on the endo-β-1,4-glucanase activity of rGluL toward CMC was investigated at different pH values (3.0–8.5) at 45°C for 10 min using the following buffer systems at 50 mM: sodium citrate (pH 3.0–5.5), sodium phosphate (pH 5.5–7.5), and Tris-HCl (pH 7.5–8.5). However, the pH stability of rGluL in the aforementioned pH buffers was evaluated by measuring its residual endo-β-1,4-glucanase activity after stopping the enzyme reaction that was performed at 45°C for 10 min. In this case, the biocatalytic reaction was initiated by adding 1.0% CMC to the reaction mixture after pre-incubation of 1 h at 3°C in the absence of substrate. The optimum temperature of rGluL to degrade CMC was examined by reacting it with substrate at 3, 10, 18, 25, 30, 35, 40, 45, 50, 55, 60, and 65°C for 10 min in 50 mM sodium citrate buffer (pH 4.0). The thermal stability of the enzyme at 3, 10, 18, 25, 30, 35, 40, 45, 50, 55, 60, and 65°C was estimated by ascertaining the residual endo-β-1,4-glucanase activity after completing the hydrolytic reaction that was conducted at pH 4.0 for 10 min. In this case, the enzyme reaction was started by adding 1.0% CMC to the assay mixture after pre-incubation of 1 h at the corresponding reaction temperature in the absence of substrate. The stimulatory or inhibitory effects of metal ions (each 1 mM) and chemical substances (each 5 mM or 0.5%) on the endo-β-1,4-glucanase activity of rGluL were assessed after pre-incubation of the enzyme at 3°C for 10 min in the reaction mixture that included the chemical of concern in the presence of 1.0% CMC. The salt tolerance of rGluL was assessed by reacting the enzyme with CMC for 10 min in the standard assay mixture containing NaCl at concentrations of 0, 0.1, 0.5, 1.0, 2.0, and 4.0 M.

The enzymatic hydrolysis of D-cellooligosaccharides (C₂–C₆, each 1 mg) and CMC (2 mg) was accomplished by reacting rGluL (10 μg) with the cellulosic substrates in 50 mM sodium citrate buffer (pH 4.0) for 3 h at 45°C, during which the enzyme retained more than 75% of its original endo-β-1,4-glucanase activity. Next, the enzyme reaction was stopped by heating the reaction mixtures at 100°C for 5 min, after which the resulting products formed by the degradation of the cellulosic substrates were identified by liquid chromatography/tandem mass spectrometry (LC-MS/MS) using D-glucose (C₁) and D-cellooligosaccharides (C₂–C₆) as standards. High performance liquid chromatography (HPLC) analysis was conducted employing a Finnigan Surveyor Modular HPLC system (Thermo Electron Co., Waltham, MA, United States) connected with an Asahipak NH2P-50 2D column (5 μm, 2.0 × 150 mm, Showa Denko K.K., Tokyo, Japan). Mobile phase A consisted of water and 0.05% formic acid, whereas mobile phase B contained acetonitrile and methanol at a ratio of 7:3 together with 0.05% formic acid. Degradation products were eluted from the column using the following conditions: a linear gradient of 90–85% B from 0 to 20 min and isocratic elution with 50% B from 20 to 25 min at a flow rate of 0.25 mL/min. Additionally, LC-MS was accomplished by a Finnigan LCQ Advantage MAX ion trap mass spectrometer (Thermo Electron Co.) equipped with an electrospray ionization (ESI) source, as described elsewhere (Kim et al., 2010).

The 1044-bp GluL gene (GenBank accession number: ON016586) identified by an in silico analysis of the complete genome sequence of L. cladoniae PAMC 26568 was predicted to encode an extracellular GH8 endo-β-1,4-glucanase comprised of 347 amino acids (Figure 1). Accordingly, the deduced molecular mass and calculated isoelectric point (pI) of the premature GluL were assessed to be 38,630 Da and 7.80, respectively, as analyzed using the Compute pI/MW tool.² In addition, the cleavage site for its signal peptide was predicted to be located between Ala27 and Asp28 in the N-terminal region, as analyzed by the SignalP 6.0 server.³ Compared to the premature GluL, the mature form lacking a signal sequence was estimated to be a protein with a deduced molecular mass of 35,576 Da and a calculated pI of 6.73. As displayed in Figure 1, protein BLAST and Pfam analyses revealed that similar to other characterized cold-adapted GH8 functional homologs (Bhat et al., 2013; Chen et al., 2020; de Melo et al., 2021), the premature GluL might be a non-modular endo-β-1,4-glucanase consisting of a single catalytic GH8 domain (from Gln32 to Ile339) without other additional substrate-binding domains such as carbohydrate-binding module and ricin-type β-trefoil lectin domain-like domain.

The phylogenetic tree exhibited in Figure 2 clearly revealed that the primary structure of GluL shared a close evolutionary relationship with that of its functional homologs within GH family 8, which constituted inverting endo-β-1,4-glucanase (EC 3.3.1.4), chitosanase (3.2.1.132), licheninase (EC 3.2.1.73), endo-β-1,4-xylanase (EC 3.2.1.8), reducing-end-xylose releasing exo-oligoxylanase (EC 3.2.1.156), and endo-β-1,3-1,4-glucanase/lichenase-laminarinase (EC 3.2.1.6).⁴ Moreover, multiple sequence alignment indicated that the catalytic GH8 domain of GluL was most similar to that of Gluconacetobacter takamatsuzukensis glycoside hydrolase family 8 (GenBank accession number: WP_182950054) with a sequence identity of 55%, which had not been functionally characterized but identified only through a genome survey (Figure 1). A protein BLAST survey also revealed that the catalytic GH8 domain of GluL shared a relatively low sequence identity of <53% with that of most uncharacterized structural homologs deposited in the National Center for Biotechnology Information (NCBI) database. For example, the catalytic GH8 domain of GluL was assessed to be 51, 50, and 48% identical to that of Novacetimonas hansenii endoglucanase (QOF96800), Komagataeibacter medellinensis NBRC 3288 endo-β-1,4-glucanase (BAK84916), and Acetobacter orientalis endo-β-1,4-glucanase (GAN65284), respectively, which have not yet been biochemically characterized (Figure 1). In GluL, the two conserved residues related to catalysis, Glu59 (proton donor) and Asp246 (nucleophile/base), were identified in its active site, as observed in other GH8 endo-β-1,4-glucanases (Bhat et al., 2013; Cano-Ramírez et al., 2016; Huang et al., 2021). However, it has also been reported that in the case of a cold-adapted endo-β-1,4-glucanase from Burkholderia pyrrocinia JK-SH007, the two Glu83 and Glu271 residues in the active site play a key role in catalysis as the proton donor and the acceptor, respectively (Chen et al., 2020).

It has been reported that when recombinantly overexpressed in E. coli BL21, some carbohydrolases are often produced as inactive inclusion bodies (Kim et al., 2009; Bai et al., 2021a,b), which can be converted to their native state by an on-column refolding protocol (Singh et al., 2015). Of such biocatalysts, a bi-modular GH6 endo-β-1,4-glucanase with a carbohydrate-binding module 2 domain from Cellulosimicrobium funkei HY-13 (Kim et al., 2016) and a non-modular GH5 endo-β-1,4-glucanase from a rhizosphere metagenomic library (Wierzbicka-Woś et al., 2019) are examples of enzymes that were normally produced as protein aggregates in an inactive form. However, the rGluL proteins with a (His)₆-tag at the N-terminus region were intracellularly produced in an active form, similar to other cold-adapted GH8 endo-β-1,4-glucanases (Chen et al., 2020; de Melo et al., 2021). Therefore, based on their solubility, the active (His)₆-tagged rGluL proteins were simply isolated to electrophoretic homogeneity by basic Ni-NTA affinity chromatography.

The relative molecular mass of purified rGluL was estimated to be approximately 38.0 kDa, as analyzed by SDS-PAGE (Figure 3). This value corresponded well to the deduced molecular mass (37,870 Da) of the enzyme that was determined using the Compute pI/MW tool (see text footnote 2). As listed in Table 1, the molecular size (38.0 kDa) of rGluL was assessed to be relatively similar to that of some GH8 endo-β-1,4-glucanases with relative molecular masses between 36.0 and 40.0 kDa. However, its molecular size was found to be much smaller than that of B. pyrrocinia JK-SH007 cold-adapted endo-β-1,4-glucanase (60.0 kDa) (Chen et al., 2020) and that of Bacillus subtilis B111 endo-β-1,4-glucanase (53.0 kDa) (Huang et al., 2021), which were composed of a single catalytic GH8 domain.

Most GH8 endo-β-1,4-glucanases exhibited the highest biocatalytic activity toward CMC or barley β-glucan at a pH value between 6.0 and 7.0 (Table 1). However, rGluL was found to maximally degrade CMC at pH 4.0 and 45°C (Figures 4A,B), indicating that it was an acidic endo-β-1,4-glucanase. The optimum pH of rGluL to decompose CMC was lower than that of a GH8 endo-β-1,4-glucanase (pH 5.0) from Halomonas sp. S66-4 (Huang et al., 2010) and that of a cold-active GH8 endo-β-1,4-glucanase (pH 4.5) from cold desert soil in Ladakh (Bhat et al., 2013) identified by functional metagenomics. In contrast to the aforementioned acidic GH8 endo-β-1,4-glucanases, Bacillus circulans KSM-N257 GH8 endo-β-1,4-glucanase was an alkaline enzyme that maximally catalyzed the degradation reaction of CMC at pH 8.5 (Hakamada et al., 2002). It is also interesting to note that rGluL displayed more than 80% of its maximum endo-β-1,4-glucanase activity even at pH 3.0, although a gradual decrease in the biocatalytic activity was observed as the pH of the reaction mixture increased in the range of 4.0–8.5. As shown in Figure 4C, rGluL was fairly stable in a pH range from 4.0 to 8.0 because the enzyme maintained over 95% of its residual endo-β-1,4-glucanase activity after pre-incubation of 1 h at those pH values. Conversely, the pH stability of rGluL was significantly reduced by <55% when exposed to pH 3.0 for 1 h in the absence of CMC. The optimum temperature (45°C) of rGluL was higher than that of a cold-active GH8 endo-β-1,4-glucanase (28°C) from Ladakh soil (Bhat et al., 2013), that of a cold-adapted GH8 endo-β-1,4-glucanase (35°C) from B. pyrrocinia JK-SH007 (Chen et al., 2020), and that of a cold-adapted GH8 endo-β-1,4-glucanase (40°C) from X. citri subsp. citri (de Melo et al., 2021). However, similar to the aforementioned cold-adapted GH8 enzymes, rGluL also showed over 60% of its maximum endo-β-1,4-glucanase activity when reacted with CMC at 25°C (Figure 4B). Furthermore, rGluL was capable of easily deconstructing CMC to some extent (approximately 23% of its maximum hydrolytic activity in a cold environment of 3°C), which suggested that the enzyme was a cold-adapted endo-β-1,4-glucanase. It is assumed that the cold adaptation of rGluL might be due to an increase of its structural flexibility for efficient biocatalytic reactions at low temperatures (Parvizpour et al., 2015). rGluL is likely to be relatively thermostable at temperatures below 50°C because it retained over 90% of its residual biocatalytic activity after pre-incubation of 1 h (Figure 4D). In contrast, the enzyme was drastically inactivated when exposed to temperatures exceeding 50°C for the same pre-incubation time in the absence of CMC. The thermal stability of cold-adapted GH8 rGluL was comparable to the thermal instability of a cold-active GH8 endo-β-1,4-glucanase from Ladakh soil (Bhat et al., 2013) and a cold-adapted GH8 endo-β-1,4-glucanase from B. pyrrocinia JK-SH007 (Chen et al., 2020), which were observed to lose over 55% of their residual biocatalytic activity at temperatures exceeding 40°C in a pre-incubation period of 1 h.

The stimulatory or inhibitory effect of various divalent cations and chemical reagents on the endo-β-1,4-glucanase activity of rGluL is displayed in Figure 5. When pre-incubated in the presence of 1 mM Ni²⁺, Ba²⁺, or Mn²⁺, the endo-β-1,4-glucanase activity of rGluL was noticeably stimulated by >1.25-fold. In particular, the substantial activation (1.34-fold) of rGluL by 1 mM Mn²⁺ was comparable to the partial inhibition (<80%) of a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021) by the same metal ions. Previously, the stimulatory or inhibitory effect of 1 mM Mn²⁺ on the biocatalytic activity of a GH8 endo-β-1,4-glucanase from Bacillus subtilis B111 (Huang et al., 2021) and a GH8 endo-β-1,4-glucanase from Halomonas sp. S66-4 (Huang et al., 2010) was also shown to be marginal. It is noteworthy that the CMC-decomposing activity of rGluL was partially downregulated by >18% in the presence of 1 mM Mg²⁺, Cu²⁺, Sn²⁺, or N-ethylmaleimide, although no enzyme inhibition by 1 mM Zn²⁺ or EDTA was observed. On the other hand, Halomonas sp. S66-4 GH8 endo-β-1,4-glucanase was reported to be insensitive to 1 mM Mg²⁺ but partially suppressed by approximately 22% in the presence of 1 mM EDTA (Huang et al., 2010). Moreover, compared to rGluL, S. proteamaculans CDBB-1961 GH8 endo-β-1,4-glucanase was completely inactivated when pre-incubated in the presence of 1 mM Zn²⁺, although the stimulatory or inhibitory effect of 1 mM EDTA on the enzyme activity was negligible (Cano-Ramírez et al., 2016). As shown in diverse cellulolytic and hemicellulolytic enzymes (Yin et al., 2010; Kim et al., 2011, 2018; Liu et al., 2021), rGluL was also found to be highly sensitive to tryptophan (Trp) residue-specific modifiers, such as Hg²⁺ and N-bromosuccinimide, oxidizing the indole ring of conserved Trp residues in the active site of endo-type GH enzymes that play a key role in the binding of enzyme to substrate (Zolotnitsky et al., 2004). Specifically, it appeared that the enzyme almost completely lost its endo-β-1,4-glucanase activity toward CMC when pre-incubated at 3°C for 10 min in the presence of 1 mM Hg²⁺ or 5 mM N-bromosuccinimide without the substrate. Similar observations were also made by a GH8 endo-β-1,4-glucanase from Halomonas sp. S66-4 (Huang et al., 2010) and a cold-active GH8 endo-β-1,4-glucanase from Ladakh soil (Bhat et al., 2013) treated with 1 mM Hg²⁺ and/or N-bromosuccinimide. However, compared to rGluL, GH8 endo-β-1,4-glucanases from S. proteamaculans CDBB-1961 (Cano-Ramírez et al., 2016) and Bursaphelenchus xylophilus (Zhang et al., 2013) exhibited approximately 64 and 36% of their maximum CMC-degrading activity in the presence of 5 and 10 mM Hg²⁺ ions, respectively. It is interesting to note that the strong inhibition (>90%) of rGluL by 1 mM Fe²⁺ was very comparable to the powerful stimulation (1.58-fold) of Halomonas sp. S66-4 GH8 endo-β-1,4-glucanase (Huang et al., 2010) exerted by the same metal ions. In addition, in the case of a GH8 endo-β-1,4-glucanase from S. proteamaculans CDBB-1961 (Cano-Ramírez et al., 2016) and a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021), no significant inhibition or stimulation was seen by pre-incubation with 1 mM Fe²⁺. Meanwhile, rGluL was proven to be almost completely inactivated when pre-incubated in the presence of 0.5% SDS, although it was slightly upregulated by approximately 1.1-fold in the presence of a non-ionic detergent (0.5% Tween 80 or 0.5% Triton X-100). The results suggested that SDS was highly toxic to rGluL, as previously exhibited by other characterized GH8 endo-β-1,4-glucanases (Bhat et al., 2013; Zhang et al., 2013). Conversely, the inhibition of B. subtilis B111 GH8 endo-β-1,4-glucanase by SDS resulted in an approximately 50% reduction in its original CMC-hydrolyzing activity (Huang et al., 2021). It should also be noted that the endo-β-1,4-glucanase activity of rGluL toward CMC could be remarkably enhanced by >1.3-fold in the presence of NaCl between 0.1 and 0.5 M (Figure 6). Moreover, the suppression of its endo-β-1,4-glucanase activity exerted by 2.0 M NaCl was only observed to be approximately 5%, implying that it was a relatively salt-tolerant enzyme. These findings suggested that the salt tolerance of rGluL was superior to that of Paenibacillus xylanivorans A59 GH8 endo-β-1,4-glucanase (Ghio et al., 2020), which showed biocatalytic activity below 50% in the presence of NaCl between 0.5 and 2.0 M.

Different types of true endo-β-1,4-glucanases lacking other additional carbohydrolase activities in addition to bi- and multi-functional endo-β-1,4-glucanases, which belong to GH family 8, have been identified and biochemically characterized to date (Table 1). Bacterial GH8 enzymes from Enterobacter sp. R1 (Ontañon et al., 2019), X. citri subsp. citri (de Melo et al., 2021), Halomonas sp. S66-4 (Huang et al., 2010), B. circulans KSM-N257 (Hakamada et al., 2002), and B. pyrrocinia JK-SH007 (Chen et al., 2020) are examples of true endo-β-1,4-glucanases that preferentially cleaved only β-1,4-glycosidic bond in cellulosic polysaccharides. In contrast, a GH8 endo-β-1,4-glucanase from B. subtilis B111 has been recently reported to decompose diverse cellulosic substrates as well as chitosan (Huang et al., 2021). Similarly, bi-functional GH8 endo-β-1,4-glucanases, which were specific to both cellulosic and xylosic polysaccharides, have also been identified from S. proteamaculans CDBB-1961 (Cano-Ramírez et al., 2016) and B. xylophilus (Zhang et al., 2013). Compared to the aforementioned GH8 endo-β-1,4-glucanases, a cold-active endo-β-1,4-glucanase from Ladakh soil (Bhat et al., 2013) and an endo-β-1,4-glucanase from P. xylanivorans A59 (Ghio et al., 2020) were reported to be multi-functional GH8 enzymes capable of deconstructing the polysaccharides with different microstructures, such as CMC, xylan, and chitosan, to some extent. Therefore, the substrate specificity of cold-adapted GH8 rGluL in this study was evaluated using structurally different sugar-based polymeric materials together with pNP-sugar derivatives (Table 2). Of the investigated polysaccharides, rGluL was able to preferentially deconstruct amorphous CMC (β-1,4-D-glucan) and barley β-glucan (β-1,3-1,4-D-glucan) to low molecular weight derivatives and with specific activities toward CMC and barley β-glucan estimated at 8.3 and 15.1 U mg^–1, respectively. However, the enzyme did not exhibit any detectable hydrolytic activity toward Avicel PH-101 (crystalline β-1,4-D-glucan), β-1,3-D-glucan, β-1,4-D-xyloglucan, β-1,4-D-chitosan, β-1,4-D-mannan, β-1,4-D-glucomannan, β-1,4-D-xylan, pNP-glucopyranoside, and pNP-cellobioside. Based on these findings, it was proposed that like some GH8 endo-β-1,4-glucanases (Huang et al., 2010; Ontañon et al., 2019; Chen et al., 2020; de Melo et al., 2021), the cold-adapted rGluL enzyme is a new type of true GH8 endo-β-1,4-glucanase without other additional glycoside hydrolase activities. It is worth noting that the specific activity (15.1 U mg^–1) of rGluL for barley β-glucan was approximately 7.9- and 1.2-fold higher than that (1.9 U mg^–1) of Halomonas sp. S66-4 GH8 endo-β-1,4-glucanase (Huang et al., 2010) and that (12.8 U mg^–1) of B. subtilis B111 GH8 endo-β-1,4-glucanase (Huang et al., 2021), respectively. However, the barley β-glucan-degrading activity (15.1 U mg^–1) of rGluL was assessed to be slightly lower than that (16.6 U mg^–1) of a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021). Likewise, rGluL (8.3 U mg^–1) was more active on CMC compared to a GH8 endo-β-1,4-glucanase (0.8 U mg^–1) from S. proteamaculans CDBB-1961 (Cano-Ramírez et al., 2016), a GH8 endo-β-1,4-glucanase (4.1 U mg^–1) from Enterobacter sp. R1 (Ontañon et al., 2019), and a cold-active GH8 endo-β-1,4-glucanase (4.7 U mg^–1) from Ladakh soil (Bhat et al., 2013; Table 1). Conversely, the biocatalytic activity (8.3 U mg^–1) of rGluL to decompose CMC was estimated to be relatively lower than that (12.7 U mg^–1) of a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021) and that (11.2 U mg^–1) of a cold-adapted GH8 endo-β-1,4-glucanase from B. pyrrocinia JK-SH007 (Chen et al., 2020).

The kinetic parameters of rGluL for CMC and barley β-glucan in concentration range 0.2–1.2%, which was determined by non-linear regression of the Michaelis-Menten equation, are shown in Table 3. Under the optimal pH (4.0) and temperature (45°C) conditions, rGluL exhibited a V_max value of 27.48 U mg^–1, a K_m value of 2.73 mg mL^–1, and a k_cat value of 17.34 s^–1 toward barley β-glucan. In this case, the catalytic efficiency (k_cat/K_m: 6.35 mg^–1 s^–1 mL) of rGluL toward barley β-glucan was approximately 1.23- and 1.60-fold higher than that (3.98 mg^–1 s^–1 mL) of a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021) and that (5.16 mg^–1 s^–1 mL) of a thermostable GH8 endo-β-1,4-glucanase from Enterobacter sp. R1 (Ontañon et al., 2019), respectively, for the same polysaccharide. However, the apparent k_cat/K_m value of rGluL toward CMC was calculated to be approximately 2.83 mg^–1 s^–1 mL, which was lower than its k_cat/K_m value (6.35 mg^–1 s^–1 mL) toward barley β-glucan (Table 3). The results indicated that the kinetic efficiency (2.83 mg^–1 s^–1 mL) of rGluL for CMC was slightly higher than that (2.50 mg^–1 s^–1 mL) of a cold-active GH8 endo-β-1,4-glucanase from Ladakh soil (Bhat et al., 2013) for the same substrate but was relatively lower than that (6.05 mg^–1 s^–1 mL) of a cold-adapted GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021).

Despite the inability of rGluL to cleave C₂, the enzyme could hydrolyze CMC as well as D-cellooligosaccharides with a degree of polymerization in the range of 3–6, although C₃ seemed to be only slightly susceptible to the enzyme (Figure 7 and Table 4). Specifically, after the enzymatic hydrolysis of C₃ for 3 h at pH 4.0 and 45°C, the resultant products in the reaction mixture were identified as C₃ (93.6%), which was assumed to be unhydrolyzed by rGluL, and a small amount of C₂ (6.4%). These results implied that the hydrolytic degradation of C₃ by rGluL was proceeded very slowly without the production of longer D-cellooligosaccharides under the reaction conditions. In contrast to rGluL, five GH8 endo-β-1,4-glucanases from Enterobacter sp. R1 (Ontañon et al., 2019), X. citri subsp. citri (de Melo et al., 2021), B. pyrrocinia JK-SH007 (Chen et al., 2020), P. xylanivorans A59 (Ghio et al., 2020), and B. subtilis B111 (Huang et al., 2021) are known to exhibit no cleavage activity for C₃, indicative of their lack of exoenzyme activity. Of the evaluated cellulosic substrates, the GluL-mediated biocatalytic degradation of C₅ resulted in the formation of a mixture consisting of C₂ (58.2%) and C₃ (41.8%) without C₁ and D-cellooligosaccharides with a degree of polymerization of ≥4, suggesting that the enzyme did not have transglycosylation activity. However, the results of HPLC analysis clearly indicated that compared to the inability of X. citri subsp. citri GH8 endo-β-1,4-glucanase (de Melo et al., 2021) and B. circulans KSM-N257 GH8 endo-β-1,4-glucanase (Hakamada et al., 2002) to hydrolyze C₄, the biocatalytic degradation of C₄ by rGluL yielded a mixture of C₁ (3.8%), C₃ (31.1%), and C₄ (2.3%) together with C₂ (62.8%) as the dominant end product. The prominent formation of C₂ was also identified in a mixture of the degradation products that were formed by the enzymatic hydrolysis of C₅ and C₆. The degradation patterns of D-cellooligosaccharides (C₄-C₆) were very comparable to those of the same substrates by three GH8 endo-β-1,4-glucanases from B. pyrrocinia JK-SH007 (Chen et al., 2020), X. citri subsp. citri (de Melo et al., 2021), and Enterobacter sp. R1 (Ontañon et al., 2019), which resulted in the formation of primarily C₃ as the end product. Based on these results, rGluL was believed to be a true GH8 endo-β-1,4-glucanase because it could easily degrade CMC, which was generally employed for the assay of endo-β-1,4-glucanase activity in other known GH8 functional homologs (Table 1), to a mixture composed of C₁ (6.6%), C₂ (33.9%), C₃ (30.1%), C₄ (24.4%), and C₅ (5.0%) (Figure 7G and Table 4). Conversely, unlike rGluL, a GH8 endo-β-1,4-glucanase from Halomonas sp. S66-4 was previously identified as a C₃ and C₄-releasing enzyme specific to CMC (Huang et al., 2010). Moreover, a cold-adapted GH8 endo-β-1,4-glucanase from B. pyrrocinia JK-SH007 was a biocatalyst releasing exclusively C₁ from CMC, although it could not hydrolyze C₂ and C₃ but catalyzed the hydrolytic degradation of C₄ and C₅ that yielded a mixture of C₁-C₄ and a mixture of C₁-C₅, respectively, as the final degradation products (Chen et al., 2020). Furthermore, no release of C₁ as the end product was observed after the biocatalytic reaction of CMC with a cold-active GH8 endo-β-1,4-glucanase from Ladakh soil (Bhat et al., 2013), a thermostable GH8 endo-β-1,4-glucanase from Enterobacter sp. R1 (Ontañon et al., 2019), and a GH8 endo-β-1,4-glucanase from X. citri subsp. citri (de Melo et al., 2021). Taken together, the substrate specificity of rGluL and its mode of action on cellulosic substrates elucidated in this study strongly indicated that it was a novel GH8 endo-type β-1,4-D-glucan-degrading enzyme distinct from other characterized GH8 functional homologs (Table 1).