A gene fragment of xynSL3 (318 bp) was amplified using the CODEHOP primers X10-F and X10-R (Wang et al., 2010). DNA fragments amplified by TAIL-PCR were assembled with the core region and an ORF (starts with ATG and terminates with TAA) of 4056 bp was identified. The full-length gene (xynSL3, GenBank accession number: KX611103) encoded a polypeptide, which consists of 1,351 amino acids with a calculated mass of 149.4 kDa and theoretical isoelectric point of 4.82. Sequence analysis suggested that XynSL3 is a multi-domain protein with a putative signal peptide predicted from Met¹ to Ala³¹, followed by three domains of CBM family 22 (Gln⁴⁴-Glu²⁰³, Val²⁵¹-Pro⁴⁰⁹, and Ala⁴¹⁵-Asp⁵⁷⁴), a catalytic domain belonging to family 10 from Leu⁵⁷⁵ to Asp⁹⁰⁷ and a family 9 Carbohydrate-binding domain from Val⁹²² to Gln¹¹⁰⁶ (Figure 1).

The protein sequence of xynSL3 shared the highest identity of 66% with a hypothetical xylanase from Alkalibacterium sp. AK22 (WP_051503745). However, XynSL3 has low identities (41–46%) with other putative xylanases from Amphibacillus jilinensis (WP_017470487), Gracilibacillus sp. Awa-1 (WP_058309001), Bacillus akibai (WP_052012994), Amphibacillus xylanus (WP_015010454), and Gorillibacterium massiliense (WP_040950848). Furthermore, XynSL3 showed 33–45% identities with the functionally characterized xylanase of Paenibacillus sp. JDR-2 (CAI79477, Stjohn et al., 2006), the xylanase 10A from Paenibacillus curdlanolyticus (ABZ80916, Waeonukul et al., 2009), and xylanase A from an uncultured microorganism (AEI17090, Zhao et al., 2013).

Protein sequence of the catalytic domain of XynSL3, its closest homologs and alkaline or alkali-tolerant xylanases were extracted to build a phylogenetic tree (Figure 2). As can be seen from the tree, high bootstrap values separated these xylanases into three major groups (Figure 2). XynSL3 was closely related to the putative xylanases from Alkalibacterium, but distantly related to alkaline or alkali-tolerant xylanases that have been reported to date.

Based on the multiple sequence alignment of the catalytic domain of XynSL3 and 8 GH10 xylanases, two putative catalytic residues Glu⁷⁰⁴ and Glu⁸³⁰ were identified in XynSL3. Modeled XynSL3 had a structure typical of GH10 hydrolases, consisting of eight α-helices on the outside and eight β-strands on the inside that alternate along the peptide backbone to form a typical eightfold TIM-barrel structure of family 10 xylanases (Figure 3).

The gene fragment encoding the mature protein was expressed in E. coli BL21 (DE3). After induction with 1 mM IPTG at 30°C for 12 h, xylanase activity was detected in the cell lysate. No xylanase activity was detected in the cultures or cell lysate of the uninduced transformant. The crude enzyme was purified to electrophoretic homogeneity by ammonium sulfate precipitation and Ni-affinity chromatography (Figure 4). The purified rXynSL3 migrated as a single band at around 150 kDa on Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), which was identical to the calculated value (149.4 kDa). Three internal peptides obtained from LC–ESI-MS/MS, AVVQLPSGDYPVVGGADFVPGEEFTIEGTR, QVADENGWDDMVLYYNDYNDHVQGK, and VNYENEASF NPGR, matched the protein sequence of XynSL3, confirming that the purified enzyme was indeed XynSL3.

At pH 9.0 and 55°C, the rXynSL3 had high specific activities toward beechwood xylan (180.4 U mg^-1) and corncob xylan (151.5 U mg^-1), but no activity against CMC-Na, lichenan, pullulan and barley-glucan.

When assayed at 37°C, the purified rXynSL3 showed apparently optimal xylanase activity at pH 9.0, and retained greater than 55% of the maximum activity in the range of pH 7.0 and 11.0 and around 45% at pH 12.0 (Figure 5A). The thermal activity of purified rXynSL3 was apparently optimal at 55°C when assayed at pH 9.0, and retained at least 50% of the maximum activity when assayed at 40–55°C (Figure 5B). Without substrate, the purified rXynSL3 exhibited more than 90% of the initial activity after incubation in buffers ranging from pH 7 to 12 at 37°C for 1 h (Figure 5C). Without substrate, the enzyme was stable at 45 and 50°C for more than 60 min, whereas at 55 and 60°C, the half-life of the enzyme was approximately 18 and 4 min (Figure 5D), respectively.

Using beechwood xylan as the substrate, the K_m, V_max, and k_cat values were 1.40 ± 0.08 mg mL^-1, 285.71 ± 4.53 μmol mg^-1⋅min^-1 and 709.52 ± 7.59 s^-1, respectively, based on a Lineweaver-Burk plot.

The effect of numerous chemicals on rXynSL3 activity was also investigated (Table 1). Ca²⁺ at 5 mM enhanced the activity by 1.14 fold. Li⁺, Na⁺, K⁺, Mg²⁺, Ni²⁺, Zn²⁺, Co²⁺, Ag⁺, and Pb²⁺ had little or no effect on rXynSL3 activity. Other chemicals strongly inhibited the activity of rXynSL3, and 5 mM of Hg²⁺ and EDTA resulted in an almost complete loss of activity. The activity of rXynSL3 was enhanced by 1 mM β-mercaptoethanol, but was inhibited completely at a higher concentration (5 mM). In addition, it was highly resistant to SDS, retaining more than 53 and 37% of the initial activity at concentrations of 1 and 5%, respectively.

Purified rXynSL3 retained greater than 60% xylanase activity in the presence of 0.25–3.0 M NaCl, and even 26.3% at 4.5 M NaCl (Figure 6A). Moreover, purified rXynSL3 showed strong tolerance to high concentrations of NaCl, retaining more than 87% xylanase activity after 1-h incubation with 3 or 4 M NaCl at 37°C and pH 9.0 (Figure 6B).

Purified rXynSL3 retained greater than 40% xylanase activity in the presence of 0.1–5% ethanol, and 16.3% at 10% ethanol (Figure 7A). Moreover, purified rXynSL3 showed strong tolerance to ethanol, retaining more than 83% xylanase activity after 1-h incubation with 5 or 10% ethanol at 37°C and pH 9.0 (Figure 7B). At 20% ethanol, more than 50% xylanase activity was retained after 1-h incubation at the same concentrations.

The hydrolysis products of corncob xylan by purified rXynSL3 were determined by thin-layer chromatography (TLC). The major hydrolysis products from corncob xylan were xylobiose and xylotetraose, plus a small portion of xylose and xylotriose (Figure 8).

The bacterial strain Alkalibacterium sp. SL3 used for obtaining the xylanase gene was isolated from a soda lake during a previous study (Wang et al., 2016). E. coli Top 10 and the pMD 18-T vector (TaKaRa, Otsu, Japan) were used for gene cloning. Vector pET-22b(+; Novagen, San Diego, CA, USA) and E. coli BL21 (DE3; TaKaRa, Otsu, Japan) were used for gene expression. Kits for genomic DNA isolation, DNA purification, and plasmid isolation were purchased from Omega (Norcross, GA, USA). The Genome Walking kit, restriction endonucleases, T4 DNA ligase, DNA polymerase, dNTPs, and Isopropyl-β-D-1-thiogalactopyranoside (IPTG) were purchased from TaKaRa (Otsu, Japan). Nickel-NTA agarose (Qiagen, Valencia, CA, USA) was used to purify the His6-tagged protein. The substrate beechwood xylan was purchased from Sigma (St. Louis, MO, USA). Corncob xylan was purchased from Yuanye (Shanghai, China) All other chemicals were of analytical grade and commercially available.

Genomic DNA of Alkalibacterium sp. SL3 was extracted using the Omega genomic DNA isolation kit following the manufacturer’s instructions. A previously designed degenerate primer set specific for GH10 xylanases (Wang et al., 2010) was used for PCR amplification to obtain the xylanase gene fragment, using the purified genomic DNA as template. The PCR products were purified and ligated into vector pMD 18-T vector, transformed into E. coli DH5α, and sequenced by Invitrogen (Carlsbad, CA, USA). The flanking regions of the gene fragment were obtained using the thermal asymmetric interlaced-PCR (TAIL-PCR) procedure (Liu and Whittier, 1995), with a Genome Walking kit (TaKaRa, Otsu, Japan) following the manufacturer’s instructions and with three nested specific primers for both 5′ and 3′ ends (Table 4). PCR products of the expected size appearing between the second and third rounds of amplification were purified, cloned into the pMD 18-T vectors, sequenced, and then assembled with the known fragment sequence.

Identification of the open reading frame (ORF) was performed using the program Vector NTI 10.3 (InforMax, Gaithersburg, MD, USA). The full-length xylanase gene was designated as xynSL3. The signal peptide sequence was predicted using SignalP 4.1². The DNA and protein sequence similarities were assessed using the BLASTn and BLASTp programs³. Multiple sequence alignments were performed with Clustal W⁴. A phylogenetic tree including XynSL3, its closest homologs, alkaline or alkali-tolerant xylanases was constructed using the neighbor-joining (NJ) algorithm in MEGA 4.0 (Tamura et al., 2007). Catalytic domains of these xylanases were predicted by using CD search programs of NCBI and protein sequences of these catalytic domains were trimmed manually. Confidence for the tree topology was estimated using the bootstrap values based on 1,000 replicates.

We used the following procedure to model the three dimensional (3D) structure of the catalytic domain of XynSL3. First, the target protein sequence was searched against the Protein Data Bank (PDB) using the BLAST program. Two homologous proteins (PDB entries: 4E4P and 3RDK) were identified with 50% sequence identity. The 3RDK protein was selected as the template because it had better resolution than 4E4P. Second, the target protein was directly submitted to the SwissModel server with 3RDK as the assigned template and a reliable model was obtained. The predicted model and its surface electrostatic potential were visualized using Pymol program.

The coding sequence of mature XynSL3 without the predicted signal peptide was amplified by PCR using primers xynSL3-m-F and xynSL3-m-R (Table 4), and cloned into the SalI-XhoI site of pET-22b(+). The recombinant plasmid, pET-xynSL3, was transformed into E. coli BL21 (DE3) competent cells. Transformants harboring the recombinant plasmid (pET-xynSL3) were identified by PCR and further confirmed by DNA sequencing. The cells were grown in LB medium containing 100 μg mL^-1 of ampicillin at 37°C to an A₆₀₀ of 0.5 and induced by addition of IPTG at a final concentration of 1 mM at 30°C for 12 h. Xylanase activities of the cell pellet and culture supernatant were assayed as described below.

To purify the His-tagged recombinant protein (rXynSL3), the cells were harvested by centrifugation (12,000 × g, 4°C for 10 min) and washed with sterile distilled water. The cells were then resuspended in sterilized ice-cold buffer (20 mM Tris-HCl, 0.5 M NaCl, pH 7.6) and disrupted by sonication (6 s, 160 W) on ice. The crude enzyme was collected (12,000 × g for 10 min at 4°C) and loaded onto a Ni2+-NTA agarose gel column. The purified enzyme was eluted with a linear imidazole gradient of 20–300 mM in Tris-HCl buffer (20 mM Tris-HCl, 500 mM NaCl, pH 7.6).

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12.0% running gel) was used to determine the purity and apparent molecular mass of rXynSL3. The protein concentration was determined by the Bradford method (Bradford, 1976), using bovine serum albumin as a standard. To verify that the purified protein was rXynSL3, gel band was cut and sent to ProTech (Suzhou, China) which provides analysis of liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS).

Xylanase activity was determined by measuring the release of reducing sugar from substrates using the 3,5-dinitrosalicylic acid (DNS) method (Miller et al., 1960). The standard reaction contained 0.1 mL of appropriately diluted enzyme and 0.9 mL of Tris-HCl buffer (pH 9.0) containing 1% (w/v) beechwood xylan as substrate. After incubation at 55°C for 10 min, the reaction was stopped with 1.5 mL of DNS reagent and boiled for 5 min. The absorption at 540 nm was measured when the aforementioned mixture cooled to room temperature. Using a standard curve generated with D-xylose, the absorbance was converted into moles of reducing sugars produced. One unit (U) of xylanase activity was defined as the amount of enzyme that released 1 μmol of reducing sugar equivalent to xylose per minute. The enzyme activity was assayed by following this standard procedure unless otherwise noted. All reactions were performed in triplicate.

For identifying the substrate specificity of purified rXynSL3, assays were performed by incubating the enzyme solution with 1% (w/v) of beechwood xylan, corncob xylan, carboxymethyl cellulose sodium salt (CMC-Na), lichenan, pullulan or barley-glucan under standard conditions (pH 9.0, 55°C, 10 min).

The optimal pH for xylanase activity of the purified rXynSL3 was determined at 37°C in buffers with pH ranging from 4.0 to 12.0. The buffers used were McIlvaine buffer (0.2 M Na₂HPO₄/0.1 M citric acid) for pH 4.0–7.0, 0.1 M Tris-HCl for pH 7.0–9.0, and 0.1 M glycine-NaOH for pH 9.0–12.0. The stability of purified rXynSL3 at different pH values was estimated by incubating the enzyme solution at 37°C in various buffers for 1 h without substrate. The remaining activity was measured in Tris-HCl buffer (pH 9.0) at 55°C for 10 min. The initial activity of rXynSL3 was set as 100%.

The optimal temperature for purified rXynSL3 activity was determined over the range of 10–65°C in Tris-HCl buffer (pH 9.0). Thermostability of rXynSL3 was determined by measuring the residual activities after pre-incubation of the enzyme in Tris-HCl buffer (pH 9.0) at 45, 50, 55, and 60°C without substrate for various periods.

To investigate the effects of different metal ions and chemical reagents on the purified rXynSL3 activity, activities were measured at 37°C in Tris-HCl buffer (pH 9.0) containing 5 mM (final concentration) of LiCl, NaCl, KCl, MgSO₄, CaCl₂, MnSO₄, ZnSO₄, FeSO₄, CoCl₂, NiSO₄, CuSO₄, CrCl₃, FeCl₃, Pb(CH₃COO)₂, AgCl, HgCl₂, EDTA, CrCl₃, SDS, or β-mercaptoethanol.

The K_m, V_max, and k_cat values for rXynSL3 were determined in Tris-HCl buffer (pH 9.0) containing 1–10 mg mL^-1 beechwood xylan and 0.1 mL of enzyme solution (2.16 U mL^-1) at 55°C. K_m and V_max were determined from a Lineweaver-Burk plot using the non-linear regression computer program GraFit (Erithacus, Horley, Surrey, UK).

The effect of NaCl on the purified rXynSL3 was determined at 55°C in Tris-HCl buffer (pH 9.0) containing 0.25–4.5 M NaCl. To examine its resistance to salt, rXynSL3 was incubated with 3 or 4 M of NaCl at 37°C for 1 h, and the residual enzyme activities were measured.

The effect of ethanol on the purified rXynSL3 was determined at 55°C in Tris-HCl buffer (pH 9.0) containing 0.5–10% ethanol. To examine its resistance to ethanol, rXynSL3 was incubated with 5, 10, and 20% ethanol at 37°C for 1 h, and the residual enzyme activities were measured.

The hydrolysis of 1% (w/v) corncob xylan was carried out with rXynSL3 for 12 h at 50°C and pH 9.0. The hydrolysis products were analyzed by TLC on silica gel G-60 using chloroform/acetic acid/water (3:6:1 by volume) as a mobile phase system. Hydrolysis products were detected by spraying a mixture of sulphuric acid and ethanol (5:95) on the plate.

The nucleotide sequences of the GH10 xylanase gene (xynSL3) was submitted to the GenBank database, and the corresponding data is publicly available using the accession number KX611103.