Identification and functional characterization of a β-glucosidase from Bacillus tequelensis BD69 expressed in bacterial and yeast heterologous systems

Strains, vectors, media and culture conditions

The strains and vectors used in this study are listed in Table 1. Briefly, B. tequilensis BD69 (GenBank nucleotide accession number: MF767893) is a fiber-degrading strain previously isolated from buffalo dung samples in the Fermentation Technology Group at the National Institute for Biotechnology and Genetic Engineering (NIBGE), Pakistan. Competent E. coli DH10 β (high efficiency) cells were used for construction and sequencing of the recombinant expression vectors. For expression of the β-glucosidase gene in pET-30a(+) (Novagen, Germany) and pPIC9K (Invitrogen, USA), we used E. coli BL21 (DE3) pLysS and P. pastoris GS115 (his4, Mut+) cells, respectively.

Cultures of E. coli BL21 (DE3) pLysS and DH10 β were grown in Luria-Bertani (LB, 0.5% yeast extract, 1% peptone and 1% NaCl) medium (Sambrook, Russell & Russell, 2001), to which ampicillin (100 µg/mL) or kanamycin (50 µg/mL) were added as required. Isopropyl- β-D-thiogalactoside (IPTG) was used to induce expression in BL21 cells, while Pichia transformants were selected for their ability to grow on minimal dextrose (MD) plates (1.34% yeast nitrogen base, 4 × 10 −5% biotin, 2% dextrose, lacking histidine) at 30 °C. Yeast peptone dextrose (YPD; 1% yeast extract, 2% peptone, 2% glucose) agar plates containing geneticin G418 (0.5–2.0 mg/mL) were used for screening of multi-copy Pichia colonies. Buffered glycerol-complex medium (BMGY; 1% yeast extract, 2% peptone, 4 × 10 −5% biotin, 1.34% yeast nitrogen base, 1% glycerol, 100 mmol/L potassium phosphate, pH 6.0), and buffered methanol-complex medium (BMMY; 1% yeast extract, 2% peptone, 4 × 10 −5% biotin, 1.34% yeast nitrogen base, 1% methanol, 100 mmol/L potassium phosphate, pH 6.0) were used for expression in P. pastoris. Restriction enzymes (fast digest EcoRI and NotI), T4 DNA ligase, DreamTaq Green PCR master mix (2X) and PureLink™ Quick Gel Extraction and PCR Purification Combo Kit were purchased from Thermo Fisher Scientific (USA), whereas the QIAprep spin miniprep kit was procured from Qiagen (USA). All other chemicals used were of analytical grade.

DNA extraction, primer design and gene amplification

A single colony of B. tequilensis BD69 was inoculated into sterile LB broth and after overnight incubation at 37 °C, the DNA was extracted from bacterial cells with a conventional phenol-chloroform-isoamyl alcohol method (Sambrook, Russell & Russell, 2001). The presence of DNA was confirmed by 1% agarose gel electrophoresis. NanoDrop was used to quantify the purity and concentration of DNA.

Specific primers were designed to amplify the Bteq βgluc gene based on a multiple sequence alignment to identify conserved sequences among β-glucosidase gene sequences retrieved from the NCBI database (Magrane & UniProt Consortium, 2011). The designed PCR primers were confirmed in silico (San Millán et al., 2013). The forward (GAATTCATGACAAAAGGATTGAAGATTGTAAC) and reverse (GCGGCCGCTTACGCTTCAATTTTGTTGAAAAACTGC) primers containing EcoR1 and Not1 restriction sites (underlined), respectively, were used for amplification of Bteq βgluc gene from B. tequilensis. DreamTaq Green PCR master mix (2X) was used as per manufacturer’s instructions with thermocycling conditions as follows: initial denaturation 95 °C for 5 min, followed by 30 cycles of 95 °C for 1 min, 60 °C for 30 s, 72 °C for 1 min, and final elongation at 72 °C for 10 min. The amplified product was checked for size and purity on a 1% (w/v) agarose gel.

Construction of recombinant plasmids and sequence analysis

The PCR amplicon from the reaction detailed above and expression vectors pET-30a(+) and pPIC9K were restricted by fast digest EcoRI and NotI enzymes, and then ligated by T4 DNA ligase to generate the pET-Bteqβgluc and pPIC-Bteqβgluc expression constructs, respectively. Constructs were chemically transformed into DH10 β E. coli cells according to the manufacturer’s instructions. Recombinants were selected on LB plates containing kanamycin (50 µg/mL) and ampicillin (100 µg/mL) for pET-30a(+) and pPIC9K, respectively. Plasmid extraction was done using the QIAprep spin miniprep kit, and positive recombinant clones were identified by double digestion (Fig. S1) and colony PCR followed by sequencing (Genomics Core, University of Tennessee, Knoxville, USA). Sequencing results were analyzed using DNAMAN 9.0 (http://dnaman.software.informer.com/9.0) and searching the NCBI database using BLAST (McGinnis & Madden, 2004). The β-glucosidase open reading frame (ORF) was predicted by ORF Finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) and used to confirm proper reading frame in recombinants. A BLASTx search was also performed to retrieve homologous proteins, translated amino acid sequences were obtained using the EXPASY translate tool (Gasteiger, 2003). The signal peptide for the β-glucosidase gene was predicted using SignalP 4.0 (http://www.cbs.dtu.dk/services/SingnalP/). Putative N-glycosylation sites were located using the NetNGlyc program 1.0 (http://www.cbs.dtu.dk/services/NetNGlyc/). Physicochemical properties of the recombinant Bteqβgluc were identified using Protparam (http://au.expasy.org/tools/protparam.html). The conserved domains of the Bteqβgluc were identified by InterProScan (http://www.ebi.ac.uk/Tools/InterProScan/).

Transformation and expression

Bacterial expression was performed by transforming E. coli BL21 (DE3) pLysS (Promega) cells with the pET-Bteqβ gluc vector. One mL of an overnight culture of the recombinant bacteria in LB broth containing 50 µg/mL kanamycin was transferred to 50 mL of the same medium and incubated at 37  °C in a shaking incubator until OD600 = 0.4–0.8. The maximum expression conditions of recombinant Bteqβgluc were optimized by performing induction at different temperatures (16−37 °C), IPTG concentrations (0.3–1 mM) and time (2–16 h) of induction. Cells were harvested by centrifugation at 6,000 × g for 10 min, and pellets were re-suspended in 50 mM Tris–HCl (pH 7.5). Collected cells were disrupted by sonication (Fisherbrand™ Q500 Sonicator, 5 s on 5 s off, 7 pulses) and the sample cleared from debris by centrifugation at 12,000 × g for 30 min. Expression and solubilization of Bteqβgluc were confirmed by SDS-10% PAGE according to Laemmli (1970).

Expression of Bteqβgluc in P. pastoris was started by first linearizing pPIC-Bteqβgluc with SalI and then transforming it into P. pastoris GS115 (Invitrogen) by electroporation (1,500 V, 25 µF and 200 Ω). Transformed cells were grown on MD plates at 30 °C for 4 days and then screened on YPD plates containing G418 (0.5–2 mg/ml) for multi-copy insertion of vector. Genomic DNA of P. pastoris was extracted using the Yeast DNA Extraction Kit (Thermo Fisher Scientific, USA) for confirmation of positive transformed colonies by PCR with AOX1 and gene specific primers.

Ten confirmed transformants were grown in 10 mL of BMGY medium at 30 °C for 24–36 h at 250 rpm in order to identify the transformant with the relative maximal expression levels. The selected transformant (based on β-glucosidase activity assay) was grown in 300 mL of BMGY medium for 24 h at 250 rpm and 30 °C. The cells were harvested at 3,000× g in a swinging bucket rotor at 4 °C for 5 min, and then grown to 500 ml of BMMY at 30 °C and 250 rpm for production of enzyme for enzyme assays.

Purification of recombinant proteins

The chromatograms for purification of Bteqβgluc are presented in the supplementary files (Fig. S2). Recombinant Bteqβgluc produced in E. coli was purified by immobilized metal affinity chromatography (IMAC) using a HisTrap HP 5 ml column connected to an AKTA Pure FPLC system (GE Healthcare, Sweden). The mobile phase was 50 mM Tris–HCl (pH 7.5), 300 mM NaCl and 20 mM imidazole with a flow rate of five mL/min. Elution was performed on a gradient of imidazole (20 to 500 mM over 10 column volumes) at a flow rate of 3 mL/min.

Anion exchange chromatography was used for purification of recombinant Bteqβgluc produced in yeast cultures. Protein samples were loaded on a HiTrap Q-FF column (five mL) pre-equilibrated in 50 mM Tris–HCl (pH 7.5), and elution was performed on a linear gradient of NaCl (0–1 M, 10 column volumes) at a flow rate of 3 mL/min.

Purified Bteqβgluc was concentrated and desalted against Milli Q water using Pierce Concentrators (9K MWCO, Thermo Fisher Scientific, USA) at 4 °C. Samples were then analyzed on 10% SDS-PAGE and by Western blotting as detailed below.

Western blotting and electrophoretic analysis

Purified recombinant Bteqβgluc (50 µg) was resolved in SDS-10% PAGE and stained with Protoblue stain (National diagnostics, USA) or electrotransferred at 60 V for 1 h in Towbin’s transfer buffer (192 mM glycine, 25 mM Tris, 0.1% SDS, 20% methanol) to nitrocellulose membranes (0.45 µm, Thermo Fisher Scientific, USA). After blotting, the filter was blocked in 1X PBS, 0.1% Tween-20, 3% BSA for 45 min with continuous shaking, and then incubated for 1 h at room temperature in washing buffer (1X PBS, 0.1% Tween-20, 0.1% BSA) containing anti-6X-His tag monoclonal antibody (1:2,000 dilution). After washing once for 15 min and 5 times for 10 min each with washing buffer, the filter was developed using enhanced chemoluminescence (SuperSignal™ West Pico Chemiluminescent Substrate, Thermo Scientific, USA) and imaged in an Imager 6000 (GE Healthcare).

β-glucosidase assay

Concentration of purified Bteqβgluc was determined using the Qubit Protein Assay Kit (Invitrogen, USA). Activity of β-glucosidase in purified samples was determined by measuring hydrolysis of ρ-nitrophenyl- β-D-glucopyranoside (pNPG) as previously described (Cai, Buswell & Chang, 1998; Shi et al., 2017), with minor modifications. Briefly, reaction mixtures (150 µL) containing 10 mM pNPG in 50 mM acetate buffer (pH 5.0) were mixed with an appropriately diluted enzyme solution (150 µg total protein) and incubated at 50 °C for 30 min as previously optimized. The reactions were stopped by adding 150 µL of 1 M Na₂CO₃, and ρ-nitrophenol produced was detected at 400 nm in a spectrophotometer. One unit of β-glucosidase was defined as the amount of enzyme liberating 1 µmol of ρ-nitrophenol per minute under the assay conditions. Activity of β-glucosidase (amount of substrate converted by the enzyme per unit time) is expressed as units per min (U/min), while specific activity (defined as Bteqβgluc activity per milligram of total enzyme) values are presented as units per milligram (U/mg).

Functional characterization of Bteqβgluc

Stability of purified recombinant Bteqβgluc was tested by incubating the enzyme under different temperature and pH conditions and then assaying the relative activity. The optimal temperature for enzyme activity was determined by testing a temperature range from 30 °C to 80 °C. Whereas, thermal-stability of pET-Bteqβgluc and pPIC-Bteqβgluc was determined by pre-incubating enzyme at different temperature ranging from 50−80 °C. The effect of temperature on the rate of reaction was expressed in terms of temperature quotient (Q ₁₀), which is the factor by which rate increases due to rise in temperature by 10 °C (Reyes, Pendergast & Yamazaki, 2008). The optimum pH for the recombinant enzyme was determined by testing β-glucosidase activity over a pH range from 2–12 in increments of 2 pH units at 50 °C. While, pH stability was determined by pre-incubating Bteqβgluc at different pH values (2–12) at 50 °C for 1 h.

The effect of incubation time on enzyme activity was tested by incubating the enzyme with the substrate (pNPG) for different time intervals (10–100 min) and then determining the amount of product (ρ-nitrophenol) released. Kinetic parameters, including the Michaelis constant (K_m), maximum reaction velocity (V_max), turn over number (k_cat), and catalytic efficiency (k_cat/K_m) were estimated by determining the β-glucosidase activity at 50 °C against different pNPG concentrations ranging from 2–20 mM with a constant enzyme concentration. Both K_m and V_max were calculated from Michaelis–Menten equation using non-linear adjustment and Lineweaver-Burk double reciprocal plots (Lineweaver & Burk, 1934) using the following equation: ${\displaystyle 1∕\mathrm{\upsilon }=\left(1∕V_{\max }\right)+\left(K_{\mathrm{m}}∕V_{\max }\right)1∕\left[S\right]}$ Where υ is reaction velocity (reaction rate), V_max is the maximum reaction velocity, K_m isthe Michaelis–Menten constant, and [S] is the substrate concentration.

Thermodynamic parameters for processing of the pNPG substrate were calculated by rearranging Eyring’s absolute rate equation derived from transition state theory (Eyring & Stearn, 1939) and Arrhenius plot using equations as described by Siddiqui et al. (1997a): ${\displaystyle k_{\mathrm{cat}}=\left(\left(k_{\mathrm{B}}\times T\right)∕h\right)e\wedge \left(\left(\left(-\Delta H‡\right)∕RT\right)\right)e\wedge \left(\left(\Delta S‡∕R\right)\right)}$ Where, k_B is Boltzmann’s constant = 1. 38 × 10⁻²³ J/K, T is absolute temperature (K), h is Plank’s constant = 6.626 ×10⁻³⁴ J s, R is gas constant = 8.314 J/K/mol, ΔH‡ is enthalpy of activation, and ΔS‡ is entropy of activation. The pK_a1 and pK_a2 values of ionizable groups of active site residues were calculated using Dixon plots (Dixon & Webb, 1979). The statistical significance between kinetic parameters and Bteqβgluc expressed in E. coli and P. pastoris was determined by using two-way analysis of variance (ANOVA) using statistical software (IBM SPSS Statistics 25).

Molecular docking

We used homology modeling in the SWISS-MODEL workspace (Arnold et al., 2005) to construct the predicted 3D structure for the recombinant Bteq βgluc. The quality of models was evaluated using Ramachandran’s plot on PDBSum as well as QMEAN scoring function and SAVES Package (PROCHECK, and VERIFY3D). The best model was chosen for docking. The 3D protonation of the selected model was carried out using the Molecular Operating Environment (MOE) software. The protein structure was minimized using MMFF94X ForceField. The 3D structures of ρ-nitrophenyl-linked substrates, ρ-nitrophenyl- β-D-1,4-glucopyranoside (pNPG), 4-methylumbelliferyl- β-D-glucopyranoside (MUG), and saligenin β-D-glucopyranoside (salicin) were downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) and minimized using MOE MMFF94S ForceField. MOE was used for enzyme-substrate docking and subjected to hydrogen bonding analysis. The LigX feature of MOE was used to find interactions among Bteqβ gluc and ligands. Protein ligand interaction fingerprints (PLIF) was used to determine common active residues involved in catalysis.

Cloning and bioinformatics analysis of the Bteqβ gluc gene

The Bteqβgluc gene (1,329 bp) was amplified from the genome of B. tequelensis BD69 by PCR using sequence specific primers and submitted to GenBank with accession number MK774666. Sequence analysis revealed that this β-glucosidase gene encodes a protein of 443 amino acids including a putative signal peptide of 10 amino acids at the N-terminus. The protein, which we named Bteqβgluc, contains a typical glycoside hydrolase (GH) family 4 signature between residues 141 and 172 and was predicted to contain a single potential N-glycosylation site at position Asn171. The protein also includes a sugar binding site (amino acids 96 to 315) and a divalent metal binding site (amino acids 172 to 202), as shown in Fig. 1.

Multiple sequence alignment showed highest homology (98%) between Bteqβgluc and β-glucosidases from Bacillus sp. CMAA 1185 (Fig. 2). The predicted 3D structure for Bteqβgluc showed structural homology with β-glucosidases from B. stearothermophilus (79.91%) and Geobacillus stearothermophilus (79.50%).

Expression and purification of Bteqβ gluc

Expression of Bteqβgluc in E. coli and P. pastoris resulted in recombinant proteins of the predicted molecular weight (54.34 kDa) in SDS-PAGE (Fig. 3A) and Western blotting (Fig. 3B). In the bacterial expression system, Bteqβgluc was produced as inclusion bodies even under different tested culture conditions (temperature 16−37 °C, pH 5-11). Solubilization of inclusion bodies in 8 M urea followed by refolding resulted in comparatively much lower levels of purified soluble Bteqβgluc protein isolated in E. coli compared to P. pastoris. Relative Bteqβgluc yields were 1.75 mg/L of culture for P. pastoris and 0.68 mg/L of culture for E. coli. In addition, the maximum activity (V_max) using same amount (150 µg) of recombinant Bteqβgluc was also higher when the enzyme was purified from P. pastoris (1,462.25 U/mg) compared to E. coli (1,445.09 U/mg).

Kinetics and Functional characterization of recombinant Bteqβgluc

Purified Bteqβgluc proteins from bacterial and yeast cultures showed statistically significant (p < 0.05) difference in specific activities (Table S1). The activity of Bteqβgluc was examined over a range of pH from 2-12 at 50 °C, with optimum activity observed at pH 5 (Fig. 4A). The enzyme was stable at a range of pH from 4–8 for 1 h (Fig. S5A). Enzyme activity increased with increasing temperature from 30 °C and reached a maximum at 50 °C, after which higher temperatures drastically reduced activity suggesting denaturation (Fig. 4B). The enzyme was stable at 50 °C for 180 min and showed degradation above when incubated above 50 °C (Figs. S5B and S5C). Dixon analysis was carried out to evaluate the pK_a of ionizable groups in active amino-acid residues for pNPG hydrolysis at 50 °C. The 0 slope (on the top of bell shaped curve), +1 and −1 slope lines were drawn. The intersection points of +1 and −1 slope lines on the 0 slope line represented the pK_a1 and p K_a2 as 4.8 and 6.4, respectively (Fig. 5). The activation energy (E_a) was observed to be 44.18 and 45.29 kJ/mol for pPIC-Bteqβgluc and pET-Bteqβgluc, respectively (Fig. 6), while temperature quotient for Bteqβgluc was 1.66. Variables of ΔH‡, ΔG‡ and ΔS‡ were calculated as 41.69 kJ/mol, 57.88 kJ/mol and −54.29 kJ/mol/K, respectively for pPIC-Bteqβgluc. While ΔH‡, ΔG‡ and ΔS‡ for pET-Bteqβgluc were calculated as 42.82 kJ/mol, 57.85 kJ.mol and −50.46 kJ/mol/K, respectively. Free energy of transition state binding (ΔG‡_E−T) and free energy of substrate binding (ΔG‡ _E-S) were −10.13 kJ/mol and 4.97 kJ/mol, respectively. The K_m and V_max values for pPIC-Bteqβgluc and pET-Bteqβgluc were determined through Lineweaver-Burk and Michaelis–Menten plots for the hydrolysis of pNPG at 50 °C. Results obtained for pPIC-Bteqβgluc were similar for both methods, with K_m values between 7.43 and 7.81 mM and V_max values between mM and 1,462.25 and 1,505.00 U/mg, from Linewaver-Burk and Michealis-Menten plots, respectively (Fig. 7 and Fig. S6). Similarly, K_m values for pET-Bteqβgluc ranging between 8.43 and 11.43 mM and V_max values between 1,445.09 and 1685.00 U/mg were determined through Lineweaver-Burk and Michaelis–Menten plots, respectively. The k_cat value was found to be 454.05 min⁻¹ for pPIC-Bteqβgluc and 448.72 min⁻¹ for pET-Bteqβgluc, and the efficiency constant (k_cat/ K_m) was estimated to be 61.12 for pPIC-Bteqβgluc and 53.24 for pET-Bteqβgluc, indicating high catalytic power (conversion of bound substrate to product) of the enzyme.

Molecular docking

The QMEAN score and Ramachandran’s plot predicted a good quality model with >90% of residues in the most favored regions (Fig. S3), indicating an appropriately modeled structure. All other validation results of the SAVES Package (PROCHECK, and VERIFY3D) indicated that the predicted model was of good quality and could be further used for docking studies.

The predicted structure with minimized energy (Fig. 8A) was used for docking studies with ρ-nitrophenyl-linked substrates using MOE software. The best docking complex was used to analyze the enzyme-substrate interactions and to determine the conserved amino acid residues in the predicted catalytic cleft/active site. Figure 8B shows hydrogen-bonding interactions of Bteqβgluc with pNPG involving Glu112, Asn150, Phe88 and Tyr16 residues, expected to play an essential role in hydrolysis (2D presentation shown in Fig. S4). Surrounding amino acids Thr113, Phe148, Lys121, Arg89, Thr149, Leu20, Ser316, Tyr315 and Ala313 provide a hydrophobic environment enhancing enzyme activity and thermostability. MUG molecules showed hydrogen-bonding interactions with Bteqβgluc involving Asn150, Thr113 and Gln87 while, Ser14, Phe88, Thr149, Asn235, Arg89, Phe148, Lys121, His202, Tyr256, Gln111, Glu112, Tyr315 and Ala313 provide hydrophobic surrounding environment as showed in Fig. 8C (for 2D presentation). Hydrogen-bonding interactions of salicin involved Asn235 and Tyr16 residues of Bteqβgluc, with Met233, Tyr256, Glu112, Asn150, Thr113, His202, Tyr315, Arg89, Phe148, Ala313 and Val284 as surrounding residues providing stability to the complex as showed in Fig. 8D Fig. S4) for 2D presentation).