Role of the non-conserved amino acid asparagine 285 in the glycone-binding pocket of Neosartorya fischeri β-glucosidase

Abstract

Neosartorya fischeri β-glucosidase (NfBGL595) is distinguished from other BGLs by its high turnover for p-nitrophenyl β-D-glucopyranoside (pNPG) and flavones. The role of non-conserved amino acids in the glycone-binding pocket of NfBGL595 was studied using sequence alignment and homology modeling, followed by site-directed mutagenesis. Nine amino acids (Y223, I224, I225, V283, L284, N285, M359, H361, and V428) were identified as variable residues around the active site residues, E221 and E430, and selected for mutagenesis. Mutation of the residues to Ala resulted in a drastic alteration in the k_cat and K_m values when compared to the wild type NfBGL595. Among these nine residues, mutation of N285 to Ala resulted in a complete loss of activity toward pNPG and flavonoid glucosides. Further mutation, structural, and docking analyses revealed that residue N285 is crucial in maintaining the pK_a and polarity around E221, which is surrounded by non-polar residues. This study suggests the importance of the pK_a and microenvironment around the active site pocket for BGL catalysis.

---

Introduction

Glycoside hydrolases (GHs; EC 3.2.1.-) are a widespread group of enzymes that hydrolyze the glycosidic bond between two or more carbohydrates or between a carbohydrate and a non-carbohydrate moiety.¹ β-Glycosidases (BGLs) constitute a major group of glycoside hydrolases. BGLs can recognize a broad range of glycones, including glucose, galactose, fucose, mannose, xylose, 6-phospho-glucose, and 6-phospho-galactose. Their diversity for aglycones is even higher, including monosaccharides, oligosaccharides, and aryl or alkyl groups.² The physiological function of BGLs varies greatly depending on their origin (plants, fungi, animals, or bacteria) and substrate specificity.³

Based on their substrate specificity, BGLs have been classified as: (1) aryl BGLs, which act on aryl-glucosides; (2) true cellobioses, which hydrolyze cellobiose to release glucose; and (3) broad substrate specificity BGLs, which act on a spectrum of substrates. Some fungal glucosidases classified under GH family 1 (GH1) are flavonoid-hydrolyzing enzymes.⁴ Flavonoids are polyphenolic compounds that are generally not found as free aglycones (e.g., quercetin and kaempferol) but rather as complex conjugates with sugar residues (e.g., glucose or rhamnose). The enzymatic hydrolysis of flavor and aroma glycosides (e.g., monoterpenols, norisoprenoids, and shikimate derivatives) involves the sequential operation of rhamnosidase (EC 3.2.1.40) and arabinosidase (EC 3.2.1.55), followed by BGL (EC 3.2.1.21) reactions; the latter catalyze the release of flavor compounds from the monoglucosides.⁵ BGLs are also used in the preparation of isoflavone aglycones in soybean biotransformation.⁶

GHs are classified by their tertiary structure and are placed into groups designated as superfamilies or clans. Family 1 is included in clan GH-A, which is characterized by an 8-fold (α/β) barrel structure, in which the two active site residues that act as a acid/base catalyst and the nucleophile are located in β strand numbers 4 and 7, respectively.^7–9 All family 1 BGLs retain the anomeric configuration of the glycone in which the β-D-glucoside substrate is converted into the β-D-glucose product. The hydrolysis of the β-glycosidic bond involves two steps (glycosylation and deglycosylation) and the active site glutamic acids occur in the motifs TFNEP and I/VTENG, respectively, in all β-O-glucosidases.¹⁰

The three-dimensional (3D) structures of fungal BGLs were determined from Phanerochaete chrysosporium BGLA (PcBGLA, 2E3Z)¹¹ and Trichoderma reesei BGL2 (3AHY);¹² however, few structural studies have been performed to understand its reaction mechanism,^13,14 substrate recognition, and active-site machinery.^14–16 Despite the lack of these structural details, the properties of BGLs have been investigated using random mutagenesis and site-directed mutagenesis, leading to improvements in the enzymes' performance. Glucose production has reportedly been enhanced by site-directed mutation (G294W and G294Y) of Aspergillus oryzae β-glucosidase, leading to a 1.5- and 1.6-fold higher conversion.¹⁷ Random mutagenesis and a site-directed triple mutant (N317Y/L444F/A433V) of BglC from Thermobifida fusca led to a 5 °C increase in denaturation temperature.¹⁸ In addition, the role of conserved residues in the substrate-binding pocket (SBP) of BGL has been previously described by site-directed mutagenesis.^18–21

In this study, a systematic screening strategy was used to identify the importance of variable residues around the active site of BGLs. Neosartorya fischeri β-glucosidase (NfBGL595) is distinguished from other BGLs by its high turnover for aryl and flavones substrate.²² The role of non-conserved residues of NfBGL595 was analyzed through an initial screening of residues based on sequence alignment, a second screening was conducted using homology modeling, using crystal structure of T. reesei BGL2 (TrBGL2, PDB entry 3AHY) as template and subsequently, site-directed mutagenesis was employed to alter the individual screened residues. Nine residues (Y223, I224, I225, V283, L284, N285, M359, H361, and V428) within 5 Å of the catalytic amino acids were identified as non-conserved among BGLs. Subsequent mutagenesis, docking, and pK_a analysis of an N285 mutant revealed the importance of pK_a and protonation state of this residue in NfBGL595 catalysis.

Material and methods

Materials

Reagents for the polymerase chain reaction (PCR), Taq DNA polymerase and T4 DNA ligase, were purchased from Takara (Takara Corp. Osaka, Japan). pGEM-T Easy was purchased from Promega (Madison, WI, USA).²³ A total RNA extraction kit and nickel-nitrilotriacetic acid (Ni-NTA) superflow column for purification were purchased from Qiagen (Hilden, Germany).²⁴ Restriction enzymes were obtained from New England Biolabs (MA, USA). The pET28a expression vector was purchased from Novagen (Madison, WI, USA).²⁵ A plasmid isolation kit and oligonucleotide primers were obtained from Bioneer (Daejeon, South Korea). Electrophoresis reagents were purchased from Bio-Rad, and all chemicals used for assays were from Sigma-Aldrich (St. Louis, Mo, USA).

Fungal strains and culture conditions

N. fischeri NRRL181 was obtained from Korean Agricultural Culture Collection (KACC, Suwon, Korea). The mycelia of N. fischeri were grown in 100 mL of potato dextrose broth at 27 °C for 4 days. Escherichia coli DH5α²⁶ and BL21 (DE3, codon plus) cells were used as hosts for the transformation of the plasmids and for expression of the proteins, respectively. E. coli DH5α cells were grown in Luria–Bertani (LB) media supplemented with ampicillin (100 μg mL⁻¹) at 37 °C. E. coli BL21 (DE3, codon plus) cells were grown in LB media supplemented with kanamycin (25 μg mL⁻¹) and chloramphenicol (50 μg mL⁻¹) at 37 °C.

Cloning and expression of the Nfbgl595 gene from Neosartorya fischeri

Cloning and expression of the Nfbgl595 gene (XM_001258595) were performed as previously described.²² The wild type pET28a–Nfbgl595 plasmid, harboring the BGL NfBGL595, was expressed in E. coli BL21 (DE3, codon plus) cells, and the expression of recombinant enzyme was induced using 1 mM of isopropyl-β-D-thiogalactopyranoside (IPTG) at 16 °C for 18 h. The induced cells were harvested by centrifugation at 4 °C for 15 min at 10 000 × g, rinsed with phosphate-buffered saline, and stored at −20 °C.

Site-directed mutagenesis of NfBGL595

Site-directed mutagenesis was carried out using the Gene Tailor Site-Directed Mutagenesis System from Invitrogen protocol (http://tools.lifetechnologies.com/content/sfs/manuals/genetailor_man.pdf?q=snpware-96 accessed June 2011). The pET28a–Nfbgl595 plasmid was used as the DNA template. The list of oligonucleotide primers used in the study is provided in ESI Table S1.† The plasmids containing the mutant genes were confirmed through sequencing (Macrogen; Korea) and then transformed into E. coli BL21 (DE3, codon plus) cells. The colonies were selected based on kanamycin and chloramphenicol resistance and then grown for protein expression. The pET28a–Nfbgl595 mutants were expressed according to the same procedure for the wild type enzyme described earlier.

Purification of NfBGL595

To purify the wild type NfBGL595 and the mutants, cell pellets were resuspended in 20 mM sodium phosphate buffer (pH 7.5). The cell suspension was incubated on ice for 30 min in the presence of 1 mg mL⁻¹ lysozyme. Cell disruption was carried out by sonication at 4 °C for 5 min, and the lysate was centrifuged at 14 000 × g for 20 min at 4 °C to remove the cell debris. The resulting crude extract was retained for purification. The cell-free extract was applied to a Ni-NTA superflow column (3.4 × 13.5 cm; Qiagen) previously equilibrated with binding buffer (50 mM NaH₂PO₄, 300 mM NaCl, pH 8.0). Unbound proteins were washed from the column with a washing buffer (50 mM NaH₂PO₄, 300 mM NaCl, 150 mM imidazole, pH 8.0). The NfBGL595 protein was eluted from the column with an elution buffer (50 mM NaH₂PO₄, 300 mM NaCl, 250 mM imidazole, pH 8.0). The enzyme fractions were analyzed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and visualized by staining with Coomassie Brilliant Blue R-250.

Enzyme assay

BGL activity was assayed using p-nitrophenyl β-D-glucopyranoside (pNPG) as the substrate. The enzymatic reaction mixture (1 mL), containing 100 μL of enzyme solution and 10 mM pNPG (final concentration) in 100 mM sodium phosphate buffer (pH 6.0), was incubated for 15 min at 40 °C. The amount of p-nitrophenol released was measured at A₄₁₅ (ε₄₁₅ = 17.0 mM⁻¹ cm⁻¹) after addition of 2 M Na₂CO₃ to the reaction mixture. The kinetic parameters of NfBGL595 were determined by incubation with 1–40 mM pNPG in 100 mM potassium phosphate buffer at pH 6.0 and 40 °C. K_m and V_max were determined using the Michaelis–Menten method with Prism 5 software (GraphPad Software, Inc.; CA, USA). One unit (U) of pNPG-hydrolyzing activity was defined as the amount of enzyme equivalent to the release of 1 μmol of p-nitrophenol per minute under optimum conditions of pH 6, 40 °C, and 10 mM substrate concentration. Measurements were performed in the range where product versus time was linear.²⁷ For assays at different pH values, the reactions were performed at 40 °C for 15 min in the following buffers (100 mM) and pH values: citrate phosphate (pH 2.8–8) and glycine (pH 8–10). The pH dependence of V_max was fitted to a bell-shaped curve described previously,²⁸ and the pK_a values of the catalytic residues of the NfBGL595 variants were obtained from the fitted pH profiles.

Deglycosylation of flavonoids and high-performance liquid chromatography (HPLC) analysis

Apigenine-7-O-β-D-glucopyranoside was dissolved in dimethyl sulfoxide as 2.5 mM stock solutions stored at 4 °C. The kinetic parameters of NfBGL595 were determined by incubation with 0.1–1 mM of the corresponding flavones in 100 mM potassium phosphate buffer at pH 6.0. Reactions were carried out at 50 °C for 20 min and stopped by the addition of 0.9 mL 100% methanol containing 1% trifluoroacetic acid. The samples were filtered using 0.2 μm filter units (Advantec; Dublin, CA, USA), and the reaction products were analyzed using HPLC. HPLC-ultraviolet (UV) analysis was performed using the chromatographic system UltiMate 3000 from Dionex (Germering, Germany). An Agilent Zorbax SB-C18 column (100 × 2.1 mm, 3.5 μm) was used for isocratic separation with a 100% methanol and 10 mM ammonium acetate buffer or water (50 : 50) containing 1 mL of trifluoroacetic acid per liter of solvent mixture. The flow rate was set to 0.95 mL min⁻¹. The injection volumes of the flavone standards (0.1–1 mM) and the samples were 100 μL throughout the run time of 20 min.

Homology modeling and validation

The 3D homology model of NfBGL595 was generated using the Build Homology Models (MODELER) module in Discovery Studio 2.5 (DS 2.5, Accelrys Software Inc.; San Diego, CA, USA). The crystal structure of T. reesei BGL2 (TrBGL2, PDB entry 3AHY) was used as the template.¹² The fitness of the model's sequence in its current 3D environment was evaluated using Profile-3D Score/Verify Protein as implemented in DS 2.5. PROCHECK^29,30 was used to validate the folding integrity of the model. The generated model was then validated using a Ramachandran plot.²⁷

Substrate docking analysis

The resulting model was used for further docking studies. Hydrogen atoms were added to the protein model and minimized. Apigenine-7-glucoside (Api-7-glc) was docked into the active site pocket of the NfBGL595 model using C-DOCKER, a molecular dynamics (MD)-simulated annealing-based algorithm module from DS 2.5. Random substrate conformations were generated using high-temperature MD. Candidate poses were then created using random rigid body rotations followed by simulated annealing. The protein structure was subjected to energy minimization using the CHARMM force field as implemented in DS 2.5.³¹ A final full potential minimization was then used to refine the substrate poses. Based on the CDOCKER energy, the docked conformation of the substrate was retrieved for post-docking analysis. As a result, the substrate orientation that gave the lowest interaction energy was chosen for other round of docking.

Protein database search, sequence analysis, and pK_a calculation

The amino acid sequences deduced from the BGL gene sequences of N. fischeri were compared with those of related enzymes from other sources using the BLAST network at the National Center for Biotechnology Information. A multiple sequence alignment was performed with the ClustalW program. The PDB structures were downloaded from the RCSB data bank. The prepared structures were submitted to the PROPKA web server.^32–35 PROPKA 3.1 was used for all pK_a calculations with its default settings. The protonation state [A⁻]/[HA] of the active site residues was derived from the Henderson–Hasselbach equation at the enzyme pH optimum of 6.

Results and discussion

Sequence alignment, homology modeling, and substrate docking analysis

Using the crystal structure of TrBGL2 (3AHY) as a template, a 3D model of NfBGL595 was constructed. The Nfbgl595 gene encodes a polypeptide of 529 amino acids with a calculated molecular mass of 60.2 kDa (ESI Fig. S1a†). NfBGL595 exhibits the highest activity for pNPG among recombinant BGLs and is resistant to inhibition by ethanol, glucose, and glucono-δ-lactone.²⁶ NfBGL595 has been characterized as the BGL with the highest turnover rate for flavonoid glycosides (ESI Table S2†).

The X-ray crystal structure of T. reesei BGL2 (TrBGL2, PDB entry 3AHY) was used as a template for the homology modeling of NfBGL595. NfBGL595 has 44% sequence identity with TrBGL2. The generated model was then validated using Ramachandran plots.²⁷ The NfBGL595 model had 99.6% of its residues located within the allowed region, of which 90% were in the favourable region. Only 0.4% of the residues were located in outlier regions of the Ramachandran plot. The Profile-3D score of the model was 213 against a 218 maximum expected score. The model was also evaluated by superimposition on the template crystal structure. The root-mean-square deviation between the model and the template was 0.25 Å based on the Cα atoms. NfBGL595 contained an 8-fold (α/β) barrel structure, in which the two glutamate residues, E221 and E430, of the putative active site appeared to be directly involved in catalysis. They were located in β strand numbers 4 and 7, respectively, a common feature of the GH1 family of enzymes.²¹ The NfBGL595 protein sequence showed ∼50% amino acid identity with the fungal GH1 family of enzymes.^{1,3,11,12,36,37} Analysis of the sequence alignment, homology model, and site-directed mutagenesis followed by study of enzyme kinetics showed that E221 and E430 were the active site residues in NfBGL595. When NfBGL595 was compared with other related enzymes, NfBGL595 contained TL/FNEP and I/VTENG motifs (ESI Fig. S2†), which are highly conserved in the GH1 family.^38,39

The pNPG and Api-7-glc substrates were docked into the homology model using DS 3.5. In total, 23 residues, including the two putative active site residues E221 and E430, were found within 5 Å of the SBP (ESI Fig. S3†). Among these, 14 residues were highly conserved throughout the BGL family, and A66 and F219 were conserved residues in almost many GH1 BGLs. The other nine residues were variable or non-conserved (Fig. 1). The roles of the conserved residues in glycone binding and catalysis have been studied in various BGLs.^{9,10,13,14,40} Therefore, the non-conserved residues in the SBP (Fig. 1) were taken for further analysis.

Fig. 1 Non-conserved residues in the substrate-binding pocket of NfBGL595 from N. fischeri. Non-conserved residues, namely Y223, I224, I225, V283, L284, N285, M359, H361, and V428, around the catalytic amino acids E221 and E430 were selected for site-directed mutagenesis.

Alanine substitution of the non-conserved residues in the SBP

The non-conserved residues mentioned above were each mutated to Ala. The recombinant enzymes carrying the Y223A, I224A, I225A, V283A, L284A, N285A, M359A, H361A, and V428A mutations were expressed and purified (ESI Fig. S1b and c†). The kinetic constants using the Api-7-glc substrate were measured and compared with that of wild type NfBGL595 (Table 1). All of the mutants showed a decreased k_cat value compared to the wild type protein, except H361A. The mutant N285A did not have any detectable activity with either the Api-7-glc or pNPG substrates (Tables 1 and 2). For Api-7-glc the NfBGL595 had a k_cat and K_m value of 1350 min⁻¹ and 30 μM, respectively. The Y223A, I224A, I225A, V283A, L284A, M359A, H361A, and V428A mutants showed significant activity. The N285A mutant, which showed complete loss of activity, was taken for further analysis.

Table 1 Kinetic parameters determined for the NfBGL595 and its mutants with Api-7-glc as substrate^a

Protein	Km (μM)	kcat (min^−1)	kcat/Km (mM^−1 s^−1)
a ND – not determined due to loss in enzyme activity.
WT	30 ± 4	1350 ± 220	747
Y223A	70 ± 16	38 ± 7	213
I224A	100 ± 20	293 ± 35	6.4
I225A	99 ± 12	880 ± 42	49.4
V283A	10 ± 1	890 ± 32	1480
L284A	20 ± 2	152 ± 28	130
N285A	ND	ND	ND
M359A	590 ± 119	931 ± 119	26.3
H361A	30 ± 6	1450 ± 180	794
V428A	27 ± 3	421 ± 64	260

---

Table 2 Kinetic parameters determined for the NfBGL595 and its mutants with pNPG as substrate^a

Protein	Km (mM)	kcat (min^−1)	kcat/Km (mM^−1 s^−1)
a ND – not determined due to no activity of the enzyme.
WT	2.8 ± 0.3	1640 ± 60	35 100
Y223A	1.4 ± 0.2	899 ± 101	38 500
I224A	2.2 ± 0.3	409 ± 37	185
I225A	27 ± 3	175 ± 18	388
V283A	1.1 ± 0.2	542 ± 22	29 000
L284A	1.6 ± 0.2	929 ± 127	34 900
N285A	ND	ND	ND
M359A	5.4 ± 2.2	147 ± 22	1630
H361A	2.8 ± 0.3	535 ± 30	11 500
V428A	3.8 ± 0.4	135 ± 8	2140

---

Site-directed mutagenesis and kinetic analysis of the N285 mutants

Among the non-conserved amino acids analyzed, residue N285 was further mutated into non-polar, hydrophobic, acidic, basic, and polar residues. The N285 was replaced with the amino acids His, Asp, Glu, Gln, Ser, Arg, Lys, Ile, and Phe, and were designated N285H, N285D, N285E, N285Q, N285S, N285R, N285K, N285I, and N285F mutants, respectively. The kinetic parameters were determined for the purified NfBGL595 and mutant NfBGL595 enzymes using Api-7-glc as the substrate (Table 3). When N285 was replaced with Ile, Phe, Lys, and Arg, NfBGL595 completely lost its catalytic activity. The mutant N285H had k_cat and K_m values of 1920 min⁻¹ and 34 μM, respectively, a 1.4-fold higher k_cat value than the wild type NfBGL595. The N285D and N285E mutants retained their catalytic activity, with k_cat values of 165 and 380 min⁻¹ and K_m values of 5.1 and 4.1 μM, respectively. The N285S and N285Q mutants had k_cat values of 790 and 2710 min⁻¹ and K_m values of 25 and 114 μM, respectively.

Table 3 Kinetic parameters determined for the NfBGL595 and its N285 mutants with Api-7-glc as substrate^a

Protein	Km (μM)	kcat (min^−1)	kcat/Km (mM^−1 s^−1)
a ND – not determined due to loss in enzyme activity.
WT	30 ± 2	1350 ± 180	747
N285A	ND	ND	ND
N285I	ND	ND	ND
N285F	ND	ND	ND
N285R	ND	ND	ND
N285K	ND	ND	ND
N285H	34 ± 4	1920 ± 130	934
N285D	5.1 ± 0.7	165 ± 22	541
N285E	4.1 ± 0.5	380 ± 20	1530
N285S	25 ± 2	790 ± 101	507
N285Q	114 ± 14	2710 ± 240	394

---

The mutagenesis and subsequent docking analysis revealed that the nucleophile E430 was surrounded by polar and charged residues (R128, N360, Y363, and T429), allowing it to retain a low pK_a of 5.5, remain unprotonated, and act as a nucleophile at a BGL's optimal pH of 6 (Fig. 2). In contrast, E221 was surrounded by non-polar residues (I224, I225, P222, V283, and L284), shifting its pK_a up to 9.4 and allowing it to remain protonated in the pH range that NfBGL595 is active (pH 4–8). When the ionization environment in the fungal BGLs from Phanerochaete chrysosporium (2E3Z) and T. reesei (3AHY) was analyzed (ESI Fig. S4†), a similar arrangement was seen around the active sites, supporting the hypothesis that catalysis depends on the ionization state and electrostatics around the active site.^41,42

Fig. 2 Distribution of the polar and non-polar residues in the substrate-binding pocket of NfBGL595. The catalytic nucleophile residue E430 is surrounded by the polar residues R128, N360, Y363, and T429 (violet). The general acid/base residue E221 is surrounded by the non-polar residues I224, I225, P222, V283, and L284 (magenta). N220 and N285 are the two polar residues (cyan). The catalytic residues are colored and labelled red.

Docking analysis and pK_a values of the NfBGL595 mutants

The refined models of NfBGL595 and the mutants were docked with the substrate Api-7-glc. We compared the interactions of the N285 residue of NfBGL595 with the N285A inactive mutant. For the wild type protein, N285 had an electrostatic interaction with the active site E221 residue, whereas it did not show any interaction with E221 in the N285A mutant. The pK_a values of NfBGL595 and its mutants (N285A, N285H, N285D, N285E, and N285R) with the substrate Api-7-glc docked were calculated using the online pK_a prediction program PROPKA 3.1 (Table 4). The pK_a values of the active site residues E221 and E430 were compared between the wild type and mutant proteins. For NfBGL595, the nucleophile E430 had a low pK_a value of 5.50, indicating a deprotonated state consistent with its role as a nucleophile. The acid/base E221 had a pK_a of 9.44, indicating a protonated state. For the N285A, N285H, N285E, and N285D mutants, the pK_a value of the nucleophile E430 was lower than the acid/base E221. The pK_a values of the mutants N285R and N285K were drastically changed; the nucleophile E430 had pK_a values of 9.41 and 8.10 and the acid/base E221 had pK_a values of 4.41 and 5.73, respectively.

Table 4 pK_a values for NfBGL595 and its mutants with Api-7-glc

Protein	pKa values^a		[A^−]/[HA]^a		pKa values^b
	E221	E430	E221	E430	E221	E430
a PROPKA3.1 predicted values.b Experimentally determined values. Each value represents the mean of triplicate measurements and varied from the mean by not more than 20%.
WT	9.44	5.50	4.0 × 10^−4	3.2	9.7	5.4
N285A	10.64	6.23	2.3 × 10^−5	0.6	11.2	5.7
N285H	9.02	5.92	1.0 × 10^−3	1.2	9.2	5.6
N285E	10.87	6.87	1.4 × 10^−5	0.2	10.1	6.2
N285D	9.78	6.37	2.0 × 10^−4	0.4	9.4	6.3
N285R	4.41	9.41	38.9	3.9 × 10^−4	4.1	9.7
N285K	5.73	8.10	1.9	7.9 × 10^−3	5.3	8.7

---

The protonation state of NfBGL595 and the N285A, N285H, N285E, and N285D mutants were calculated for the active site residues E221 and E430 at the enzyme's optimum pH of 6 (Table 4). The ratio [A⁻]/[HA] gives a useful measure of the protonation state of the active site. The NfBGL595 E430 had an [A⁻]/[HA] ratio of 3.2, which indicates its role as a nucleophile, and E221 had an [A⁻]/[HA] ratio of 4.0 × 10⁻⁴. A similar trend of [A⁻]/[HA] values (higher for the nucleophile and lower for the acid/base) was seen for mutants N285A, N285H, N285E, and N285D, whereas the mutants N285R and N285K had an [A⁻]/[HA] ratio of 3.9 × 10⁻⁴ and 7.9 × 10⁻³ for the nucleophile E430, and 38.9 and 1.9 for the acid/base E221, respectively. The predicted pK_a values of the N285 mutants were verified by carrying out experiments as described in Materials and methods. The obtained values (Table 4) were in agreement with the predicted values indicating that the PROPKA3.1 predicted pK_a values were reasonably accurate.

pK_a is a useful indicator of the ionization state and microenvironment of amino acids.⁴³ The ionization properties of the active-site residues in enzyme are important in the study of the catalytic mechanism of enzymes. Knowledge of ionization constant (pK_a values) helps in identifying the proton donor and the catalytic nucleophile in the reaction mechanism of enzymes. It has been shown that pK_a values are accurate enough for identifying the proton donor in an enzyme active site by considering in detail only the active-site residues and their immediate electrostatic interaction partners.⁴⁴ The electrostatic interactions between charged groups are among the strongest and most long-range interactions in biology. Hence the ionizable groups can have significant influence on the structure, function, stability, and dynamics of proteins. The properties of ionizable groups are influenced significantly by their surroundings. The pK_a values are very sensitive to changes in protein structure. The calculation of pK_a values of ionizable groups in proteins is one of the best ways to test our understanding of the structural basis of the energetics and function of proteins.⁴⁵

The PROPKA-predicted pK_a values of NfBGL595 and the mutants (N285A, N285H, N285E, and N285D) revealed that the nucleophile E430 of mutants had a low pK_a value close to the pK_a of NfBGL595 E430 (5.50) and the acid/base E221 of mutants had a high pK_a value close to the pK_a of NfBGL595 E221 (9.44). The protonation state calculated for the active site of NfBGL595 further supported the role of the nucleophile E430, with a [A⁻]/[HA] value of 3.2 indicating its deprotonated state. Similarly, the acid/base E221 had a [A⁻]/[HA] value of 4.0 × 10⁻⁴, indicating its protonated state and allowing it to donate a proton. However, the N285R and N285K mutants had an altered pK_a that led to a drastic increase in the protonation state of E430 and prevented the nucleophilic attack, as well as the acid/base catalysis of E221. The pK_a value predicted for the P. chrysosporium crystal structure complex with gluconolactone (PDB ID: 2 × 10⁴⁰) gave similar values, with a low pK_a value of 5.11 for the nucleophile and a high pK_a value of 9.29 for the acid/base residue (Table 4). The Bacillus polymyxa BGLA and BGLB crystal structures are available for the free enzyme, covalent intermediate stage, and glycone-bound complex state, which aided in elucidating the role of the pK_a shift during catalysis. In the glycone-bound enzyme complexes of BGLA and BGLB, the nucleophile had low pK_a values of 5.24 and 5.40 and the acid/base had high pK_a values of 9.15 and 7.49, respectively. The pK_a during the covalent intermediate stage showed an elevated value of 12.53 and 12.54 for the nucleophile residues and 6.14 and 5.36 for the acid residues of BGLA and BGLB, respectively (Table 5). Thus, the lowered pK_a value of the acid/base catalyst during the covalent intermediate stage enables the enzyme to perform base catalysis and aids in the uptake of the proton from the water molecule.

Table 5 PROPKA3.1 predicted pK_a values for the GH1 bacterial and fungal BGL crystal structures

Organism and isoform	PDB ID	Ligand	pKa of acid/base	pKa of nucleophile
B. polymyxa BGLA	1BGA	Free enzyme	9.15	5.24
	1E4I	Deoxy-fluoro-β-D-glucose	6.14	12.53
	1BGG	Gluconate	7.66	2.95
B. polymyxa BGLB	2O9P	Free enzyme	7.49	5.40
	2JIE	Deoxy-fluoro-β-D-glucose	5.36	12.54
	2O9T	Glucose	7.14	2.78
P. chrysosporium BGL1A	2E3Z	Free enzyme	9.44	6.14
	2E40	Gluconolactone	9.29	5.11

---

The docking of NfBGL595 with the Api-7-glc substrate revealed that N285 had an inter-atomic electrostatic interaction with the catalytic acid/base E221 and the neighboring residue Y310. However, the N285A mutant docked with the substrate did not show an interaction between A285 and E221 or Y310 (Fig. 3). The mutation of N285 to non-polar amino acids (Ile and Ala) and a hydrophobic amino acid (Phe) resulted in a loss of enzyme activity, indicating the importance of electrostatic interaction. Replacement with non-polar and hydrophobic amino acids resulted in an increase in hydrophobicity at the acid–base E221, which is not favorable for the water molecule during the deglycosylation step of catalysis. When N285 was replaced with the polar, charged amino acid His, the polar contact and inter-atomic interaction were maintained. Among the active mutants (N285H, N285E, and N285D) that had charged side chains, the His mutant, whose side chain pK_a is close to the optimal pH of NfBGL595, had similar K_m (34 μM) and k_cat (1920 min⁻¹) values to those of the wild type.

Fig. 3 Homology model of the NfBGL595 active site with bound Apigenin-7-glucose substrate. Apigenin-7-glucose was docked into the substrate-binding pocket of wild type NfBGL595 (a) and N285A mutant (b). The interaction between the N285 residue and the catalytic acid/base E221 and Y310 is visible.

The pK_a of the histidine side chain (pK_a 6) also falls near the optimal pH, and thus, the ionization state of this residue does not affect the protein environment. The N285E and N285D mutants, both with acidic side chains, altered the pK_a and ionization state around the active sites. The [A⁻]/[HA] ratio for the acid/base E221 in the N285E and N285D mutants was lower than that in NfBGL595, which resulted in an increase in the protonated form of E221 available for the acid/base catalysis. The [A⁻]/[HA] ratio for the nucleophile E430 was reduced, indicating that a decrease in the deprotonated state of E430 led to the lower k_cat of the mutants. Thus, the ionization environment and electrostatic interaction around the two active site residues are maintained by the neighboring residues to aid BGL catalysis. N285 is a crucial residue for NfBGL595 activity because it contributes to the appropriate ionization environment and electrostatic interaction and assists the acid–base E221 catalyst during hydrolysis. Recent studies by Granum et al. 2014 have shown that the solution pH and pK_a values of ionizable residues are critical factors known to influence enzyme catalysis, structural stability, and dynamical fluctuations. The study demonstrated the complex and critical role of coupled ionizable residues to the proper functioning of cellobiohydrolase Cel7B, functionally related glycosyl hydrolases, and enzymes in general. The predicted pK_a values support the role of E212 as the catalytic nucleophile and E217 as the acid–base residue for cellobiohydrolase Cel7B from Melanocarpus albomyces. The simulations also support the use of prediction of residue pK_a values and to evaluate the impact of pH on protein structure and charge dynamics.⁴⁶ In a study with E. coli enzyme dihydrofolate reductase revealed that electrostatic interactions play an important role in enzyme catalysis by guiding ligand binding and facilitating chemical reactions.⁴⁷

Conclusions

The present work in NfBGL595 reveals the importance of ionization environment and electrostatics around the active site amino acid of BGL in aiding catalysis. A systematic analysis was performed on NfBGL595 using sequence, structure models and mutagenesis to study the importance of non-conserved amino acids in the substrate binding pocket. The alanine screening and subsequent mutagenesis experiment revealed the importance of the residue N285 in catalysis. The residue N285 was found to be crucial in maintaining the pK_a and polarity around E221, which is surrounded by non-polar residues.

Acknowledgements

This work was supported by the Energy Efficiency & Resources Core Technology Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), granted financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea (20153030091450). This research was also supported by a grant from the Intelligent Synthetic Biology Center of Global Frontier Project (2011-0031955) funded by the Ministry of Science, ICT and Future Planning, Republic of Korea. This research was supported by the 2014 KU Brain Pool of Konkuk University. This work was supported by WTU Joint Research Grants of Konkuk University.

References

1. S. Tribolo, J.-G. Berrin, P. A. Kroon, M. Czjzek and N. Juge, J. Mol. Biol., 2007, 370, 964–975.
2. S. R. Marana, IUBMB Life, 2006, 58, 63–73.
3. W. Chuenchor, S. Pengthaisong, R. C. Robinson, J. Yuvaniyama, W. Oonanant, D. R. Bevan, A. Esen, C.-J. Chen, R. Opassiri, J. Svasti and J. R. K. Cairns, J. Mol. Biol., 2008, 377, 1200–1215.
4. M. Pandey and S. Mishra, Gene, 1997, 190, 45–51.
5. R. N. Barbagallo, G. Spagna, R. Palmeri, C. Restuccia and P. Giudici, Enzyme Microb. Technol., 2004, 35, 58–66.
6. M. Kaya, J. Ito, A. Kotaka, K. Matsumura, H. Bando, H. Sahara, C. Ogino, S. Shibasaki, K. Kuroda, M. Ueda, A. Kondo and Y. Hata, Appl. Microbiol. Biotechnol., 2008, 79, 51–60.
7. S. S. Jagtap, S. S. Dhiman, T.-S. Kim, I.-W. Kim and J.-K. Lee, Appl. Microbiol. Biotechnol., 2014, 98, 661–669.
8. B. Henrissat, I. Callebaut, S. Fabrega, P. Lehn, J. P. Mornon and G. Davies, Proc. Natl. Acad. Sci. U. S. A., 1995, 92, 7090–7094.
9. J. Sanz-Aparicio, J. A. Hermoso, M. Martínez-Ripoll, J. L. Lequerica and J. Polaina, J. Mol. Biol., 1998, 275, 491–502.
10. M. Czjzek, M. Cicek, V. Zamboni, D. R. Bevan, B. Henrissat and A. Esen, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 13555–13560.
11. Y. Nijikken, T. Tsukada, K. Igarashi, M. Samejima, T. Wakagi, H. Shoun and S. Fushinobu, FEBS Lett., 2007, 581, 1514–1520.
12. W.-Y. Jeng, N.-C. Wang, M.-H. Lin, C.-T. Lin, Y.-C. Liaw, W.-J. Chang, C.-I. Liu, P.-H. Liang and A. H. J. Wang, J. Struct. Biol., 2011, 173, 46–56.
13. W. P. Burmeister, S. Cottaz, H. Driguez, R. Iori, S. Palmieri and B. Henrissat, Structure, 1997, 5, 663–676.
14. J. V. Jan Zouhar, J. Marek, J. Damborský, X.-D. Su and B. Brzobohatý, Plant Physiol., 2001, 127, 973–985.
15. Q. Wang, D. Trimbur, R. Graham, R. A. J. Warren and S. G. Withers, Biochemistry, 1995, 34, 14554–14562.
16. S. G. Withers, R. A. J. Warren, I. P. Street, K. Rupitz, J. B. Kempton and R. Aebersold, J. Am. Chem. Soc., 1990, 112, 5887–5889.
17. T. Fukuda, M. Kato-Murai, T. Kadonosono, H. Sahara, Y. Hata, S.-I. Suye and M. Ueda, Appl. Microbiol. Biotechnol., 2007, 76, 1027–1033.
18. T. Jeoh, J. O. Baker, M. K. Ali, M. E. Himmel and W. S. Adney, Anal. Biochem., 2005, 347, 244–253.
19. M. K. Tiwari, K.-M. Lee, D. Kalyani, R. K. Singh, H. Kim, J.-K. Lee and P. Ramachandran, Process Biochem., 2012, 47, 2365–2372.
20. M. Sue, K. Yamazaki, S. Yajima, T. Nomura, T. Matsukawa, H. Iwamura and T. Miyamoto, Plant Physiol., 2006, 141, 1237–1247.
21. T. Kaper, J. H. G. Lebbink, J. Pouwels, J. Kopp, G. E. Schulz, J. van der Oost and W. M. de Vos, Biochemistry, 2000, 39, 4963–4970.
22. P. Ramachandran, M. K. Tiwari, R. K. Singh, J.-R. Haw, M. Jeya and J.-K. Lee, Process Biochem., 2012, 47, 99–105.
23. S.-M. Yang, G. Shim, B.-G. Kim and J.-H. Ahn, Microb. Cell Fact., 2015, 14, 1–12.
24. S. Han, T.-V. Pham, J.-H. Kim, Y.-R. Lim, H.-G. Park, G.-S. Cha, C.-H. Yun, Y.-J. Chun, L.-W. Kang and D. Kim, Arch. Biochem. Biophys., 2015, 575, 1–7.
25. T.-S. Kim, H.-M. Jung, S.-Y. Kim, L. Zhang, J. Li, S. Sigdel, J.-H. Park, J.-R. Haw and J.-K. Lee, J. Microbiol. Biotechnol., 2015, 25, 1093–1100.
26. S. K. S. Patel, C. Selvaraj, P. Mardina, J.-H. Jeong, V. C. Kalia, Y. C. Kang and J.-K. Lee, Appl. Energy, 2016, 171, 383–391.
27. S. K. S. Patel, S. H. Choi, Y. C. Kang and J.-K. Lee, Nanoscale, 2016, 8, 6728–6738.
28. W. W. Cleland, Steady-state kinetics, in The enzymes, ed. P. D. Boyer, Academic Press, Inc., New York, N.Y, 3rd edn, 1970, vol. 2.
29. R. A. Laskowski, J. Appl. Crystallogr., 1993, 26, 283–291.
30. J.-A. Yoon, B.-G. Kim, W. J. Lee, Y. Lim, Y. Chong and J.-H. Ahn, Appl. Environ. Microbiol., 2012, 78, 4256–4262.
31. B. R. Brooks, R. E. Bruccoleri, B. D. Olafson, D. J. States, S. Swaminathan and M. Karplus, J. Comp. Chem., 1983, 4, 187–217.
32. C. R. Søndergaard, M. H. M. Olsson, M. Rostkowski and J. H. Jensen, J. Chem. Theory Comput., 2011, 7, 2284–2295.
33. D. Bas, D. Rogers and J. Jensen, Proteins, 2008, 73, 765–783.
34. H. Li, A. Robertson and J. Jensen, Proteins, 2005, 61, 704–721.
35. M. H. M. Olsson, C. R. Søndergaard, M. Rostkowski and J. H. Jensen, J. Chem. Theory Comput., 2011, 7, 525–537.
36. T. Akiba, I. Nishio M Fau-Matsui, K. Matsui I Fau-Harata and K. Harata, Proteins, 2004, 57, 422–431.
37. P. Isorna, J. Polaina, L. Latorre-García, F. J. Cañada, B. González and J. Sanz-Aparicio, J. Mol. Biol., 2007, 371, 1204–1218.
38. H. D. Ly and S. G. Withers, Annu. Rev. Biochem., 1999, 68, 487–522.
39. M. Machida, I. Ohtsuki, S. Fukui and I. Yamashita, Appl. Environ. Microbiol., 1988, 54, 3147–3155.
40. C. Wiesmann, W. Hengstenberg and G. E. Schulz, J. Mol. Biol., 1997, 269, 851–860.
41. D. Kalyani, K.-M. Lee, M. Tiwari, P. Ramachandran, H. Kim, I.-W. Kim, M. Jeya and J.-K. Lee, Appl. Microbiol. Biotechnol., 2012, 94, 413–423.
42. M. Schubert, D. K. Y. Poon, J. Wicki, C. A. Tarling, E. M. Kwan, J. E. Nielsen, S. G. Withers and L. P. McIntosh, Biochemistry, 2007, 46, 7383–7395.
43. T. K. Harris and G. J. Turner, IUBMB Life, 2002, 53, 85–98.
44. J. E. Nielsen and A. McCammon, Protein Sci., 2003, 12, 1894–1901.
45. J. Nielsen, M. R. Gunner and M. García, Proteins, 2011, 79, 3249–3259.
46. D. M. Granum, S. Vyas, S. V. Sambasivarao and C. M. Maupin, J. Phys. Chem. B, 2014, 118, 434–448.
47. C. T. Liu, J. P. Layfield, R. J. Stewart, J. B. French, P. Hanoian, J. B. Asbury, S. Hammes-Schiffer and S. J. Benkovic, J. Am. Chem. Soc., 2014, 136, 10349–10360.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra28017f

‡ These authors equally contributed to this paper.

---