Structural and enzymatic analysis of soybean beta-amylase mutants with increased pH optimum.

Comparison of the architecture around the active site of soybean beta-amylase and Bacillus cereus beta-amylase showed that the hydrogen bond networks (Glu380-(Lys295-Met51) and Glu380-Asn340-Glu178) in soybean beta-amylase around the base catalytic residue, Glu380, seem to contribute to the lower pH optimum of soybean beta-amylase. To convert the pH optimum of soybean beta-amylase (pH 5.4) to that of the bacterial type enzyme (pH 6.7), three mutants of soybean beta-amylase, M51T, E178Y, and N340T, were constructed such that the hydrogen bond networks were removed by site-directed mutagenesis. The kinetic analysis showed that the pH optimum of all mutants shifted dramatically to a neutral pH (range, from 5.4 to 6.0-6.6). The Km values of the mutants were almost the same as that of soybean beta-amylase except in the case of M51T, while the Vmax values of all mutants were low compared with that of soybean beta-amylase. The crystal structure analysis of the wild type-maltose and mutant-maltose complexes showed that the direct hydrogen bond between Glu380 and Asn340 was completely disrupted in the mutants M51T, E178Y, and N340T. In the case of M51T, the hydrogen bond between Glu380 and Lys295 was also disrupted. These results indicated that the reduced pKa value of Glu380 is stabilized by the hydrogen bond network and is responsible for the lower pH optimum of soybean beta-amylase compared with that of the bacterial beta-amylase.

crystallography and examined the mechanism underlying their different pH optima.

EXPERIMENTAL PROCEDURES
Site-directed Mutagenesis-Mutant SBA genes were constructed by a double PCR method. The PCR was performed in a reaction volume totaling 50 l and containing 2.5 units of TAKARA EX Taq polymerase, 10ϫ buffer, 200 M dNTP, two synthesized mutagenesis primers at 100 ng/ml, and 5 ng/l pKSBA (pKK233-2 vector including the SBA gene), which had been constructed previously (6). The following primers, with the desired changes indicated in bold, were used in the mutagenesis procedures (F and R denote the upstream and the downstream primers, respectively): M51T-F, 5Ј-GGTTACCGTTGATGTGTGGTGGGGG-3Ј; M51T-R, 5Ј-CC-GTCAACACCTGCAGCTCGAAGCTG-3Ј; E178Y-F, 5Ј-GTTGGGCTTG-GCCCTGCAGGAGAGC-3Ј; E178Y-R, 5Ј-GTAAATGTCTATAATAAGT-CCAGAT-3Ј; N340T-F, 5Ј-GCCATTTTTGACCTTTACATGTCTTG-3Ј; N340T-R, 5Ј-ATGATGTCTAGACAGCATTCTTGCA-3Ј. The mutagenesis primers were extended using a thermal cycler (model 2400, PerkinElmer Life Sciences) for the PCR (denaturation for 30 s at 95°C, annealing for 20 s at 55°C, and elongation for 5 min at 72°C for 18 cycles). The PCR products were separated by electrophoresis on a 1.2% (w/v) agarose gel, and the products were purified using glass powder (Bio-Rad). The resulting (mutant) 1.5-kilobase pair fragments were blunted by a blunting kit, phosphorylated by T4 polynucleotide kinase at the 5Ј-segment, and selfligated using a ligation kit, version 2 (Takara). The constructed mutant vectors were transformed into Escherichia coli strain JM105 using heat shock. The mutant DNA sequences were confirmed by a DNA sequencer.
Preparation of Mutant Proteins-Protein expression was induced by the addition of isopropyl-␤-D-thiogalactopyranoside to reach a final concentration of 1.0 mM in E. coli strain JM105. The purification and activity assay of the mutant SBA were performed by a method described previously (6).
Enzyme Kinetics and pH Dependence of ␤-Amylase Activity-The values of k cat and K m for both SBA and the mutants were measured using potato amylopectin as a substrate (30). Concentrations of the substrate were varied from 1 ⁄10 to 5 times the K m value. Calculations of V max and K m were performed using the Michaelis-Menten equation with KaleidaGraph Software (Synergy Software). The pH-dependent activities of SBA and the mutants were determined using potato amylopectin at 37°C in 0.05 M Britton-Robinson buffer at pH 3.0 -9.0. The ionic strength of the buffer solution at each pH was adjusted to be 0.2 with NaCl. The apparent pK 1 and pK 2 values were calculated with the KaleidaGraph nonlinear curve-fitting program using Equation 1 (31), where v and V are specific activity (units/mg) at each pH and pHindependent activity, respectively.
[H] is the concentration of hydrogen ions, and K 1 and K 2 are dissociation constants of catalytic groups of the enzyme.

Estimation of Dissociation Constants of the Mutant-Maltose
Complexes-The dissociation constants (K d ) of SBA and the mutants respectively complexed with maltose were estimated by the titration of emission spectra based on tryptophan residues (excitation ϭ 295 nm, emission ϭ 330 nm) in 0.1 M acetate buffer at pH 5.4 and 25°C; observations were made with a Hitachi F-3000 fluorescence spectrometer.
Crystallization and Data Collection-Based on the crystallized condition of the enzyme in a previous report (6), crystallization of wild-type and mutant SBA was performed at 4°C by the hanging drop vapor diffusion method, i.e. mixing 5 l of a 24 mg/ml protein solution in 0.1 M sodium acetate buffer (pH 5.4) with 5 l of reservoir liquid of 48% ammonium sulfate containing 0.1 M sodium acetate buffer, 1 mM EDTA, and 18 mM 2-mercaptoethanol at pH 5.4. Crystals grew for 1 month. The wild-type and mutant crystals used for data collection were ϳ1 ϫ 0.5 ϫ 0.5 mm. These crystals were all trigonal and belonged to P3 1 21; the cell dimensions are shown in Table I. The wild-type and mutant crystals were soaked in the mother liquor in the presence of 200 mM maltose for 30 min at 20°C, and then the samples were mounted in thin walled glass capillaries filled with the mother liquor and sealed with wax. We avoided crystal freezing to prevent unexpected changes, such as a shift in the pH, from occurring. X-ray data of the wild-type crystal was collected at up to 1.86-Å resolution at room temperature using CuK ␣ radiation with a Rigaku RAXIS IIC detector coupled to a Rigaku rotating anode generator, and the data were processed with Rigaku software (Rigaku). X-ray data of the respective mutant SBA crystals were collected at up to 1.95-2.1-Å resolution at room temperature using CuK ␣ radiation with a Bruker Hi-Star area detector coupled to a MAC Science M18XHF rotating anode generator, and the results were processed with the SADIE and SAINT software packages (Bruker).
Phasing and Refinement-The initial model utilized the refined coordinates of SBA complexed with maltose (Protein Data Bank accession code 1BYC). The refinement program CNS (32) and the graphic program TURBO-FRODO (Architecture et Fonction des Macromolécules Biologiques-CNRS, Marseille, France) were used to refine and rebuild the SBA and the mutant models. The 2F o Ϫ F c and F o Ϫ F c maps were used to locate the correct models. Several rounds of minimization and B-factor refinement followed by manual model building were carried out to improve the model. The structures were refined against all reflections from 15.0 Å to the highest resolution available without any (F) cut-off (see Table I). The R free values of these mutants were calculated for randomly separated 10% data, respectively. The number of water molecules incorporated and the final refinement parameters are also indicated in Table I. The stereoquality of the model was assessed using the program PROCHECK (33). The molecular models of SBA, the mutants, and BCB were superimposed using a fitting program implemented in TURBO-FRODO.

RESULTS AND DISCUSSION
Effect of pH on Mutant ␤-Amylase Activity-In the preset study, the three mutants, M51T, E178Y, and N340T, were all shown to have an increased pH optimum as indicated in Fig. 2. The specific activity for the hydrolysis of potato amylopectin at 37°C in each optimum pH of the respective ␤-amylase mutants is given in Table II. M51T, E178Y, and N340T had decreased specific activities of 11, 43, and 32%, respectively, of the wildtype SBA. The relative pH activity profiles were bell-shaped except in the case of N340T (Fig. 2). Although we did not examine the pH dependence of the V max and K m values separately, the relative pH activity profiles appeared to be mainly due to the V max value; notably the K m value of the wild type was reported previously to be constant from pH 3.0 to 7.0 (34). Thus, the observed specific activities were fitted to Equation 1, assuming two pK a values. Based on this curve-fitting, the op- timum pH and postulated pK a values of the catalytic residues were calculated (Table II). The optimum pH values of the three mutants were shifted from pH 5.4 (wild type) to 6.5 (M51T), 6.0 (E178Y), and 6.6 (N340T), all of which are close to that of BCB (pH 6.7). The relative pH activity profiles were explained by the two pK a values, pK 1 and pK 2 , of the assumed basic and acidic catalysts, respectively. These values for the wild type were in agreement with values reported previously for ␤-amylases from soybean (34,35), sweet potato (36), and barley (37). The pK 1 values of the mutants increased about 0.8 -0.9 pH units from that of the wild-type enzyme, while the pK 2 values were almost the same as that of the wild-type enzyme. These results indicated that the mutations affected the environment of the base catalyst and that due to these mutations the changes in the base catalyst were more significant than those in the acid catalyst, resulting in the increased pK 1 characteristic of the mutant enzymes.
As shown in Table III, the apparent K m values of E178Y and N340T were approximately the same as that of the wild type, whereas that of M51T was over 20 times larger than that of the wild-type enzyme. On the other hand, the V max values of E178Y and N340T were approximately 2 and 3 times lower, respectively, than that of the wild-type enzyme. In contrast to the K m values, the K d value of M51T for maltose was 10-fold lower than that of the wild type, whereas the K d values of the two mutants were both approximately equal to that of the wild-type enzyme. These results suggest that conformational changes occurred at the active site of M51T and consequently reduced    Quality of the Refined Models-To investigate the structural changes in the mutant SBA, we determined the crystal structures of the wild-type and mutant SBA complexed with maltose at 1.86 -2.1-Å resolutions. The x-ray data collection and refinement statistics are summarized in Table I. From the Luzzati plots (38), the mean absolute positional errors were estimated to be 0.21-0.22 Å. Ramachandran plots of the main-chain conformation angles (39) showed that 88.8 -88.9% of the residues lie within the core region, and 99.7-99.8% lie within the allowed region. With the exception of a small displacement of the main-chain region (residues 341-344) over the active site cleft, no significant changes occurred in the backbone structure of the mutants. The root mean square deviations between the wild-type and the mutant structures, calculated for all C␣ ␤-Amylase Mutants with Increased pH Optimum atoms, were less than 0.2 Å. The flexible loop (residues 96 -103), which was reported to play an important role in the catalytic step (16,17), was in the closed form (Table I). As shown in Fig. 3, the 2F o Ϫ F c maps of the active site and the bound maltose, except in the case of subsite ϩ3, were sufficiently clear to provide a reliable interpretation of the structural changes in the mutants.
The Structure of Wild-type SBA-Maltose Complex-In a previous report concerned with the SBA-maltose complex (17), the electron density map of two tandem maltose molecules at subsites Ϫ2 to ϩ2 was interpreted to indicate that one maltotetraose and two maltose molecules (␤-anomer at subsites Ϫ2 to Ϫ1 and subsites ϩ1 to ϩ2, respectively) bind simultaneously. A better interpretation was considered to be that both 4 C 1 ␣and distorted ␤-anomeric maltoses can bind at subsite Ϫ2 to Ϫ1 without forming maltotetraose from the excess maltose. It is also assumed that due to steric hindrance, no glucose residue can bind at subsite ϩ1 when an ␣-anomeric glucose residue binds at subsite Ϫ1. For the previous data, the ratio of the two anomers at subsite Ϫ1 was estimated to be 0.2 and 0.8 for ␣ and ␤, respectively (40). However, the present data set of SBAmaltose complex could be refined by two maltoses at subsites Ϫ2 to Ϫ1 and ϩ1 to ϩ2, respectively, with a distorted sugar ring almost assuming the boat form with unknown anomer type at subsite Ϫ1 (Fig. 3A). It was suggested that the ratio of the distorted glucose and 4 C 1 ␣-glucose residue at subsite Ϫ1 depended on the slight change in pH.
The Structure of M51T-Maltose Complex- Fig. 3B shows the 2F o Ϫ F c map and the residues around Glu 380 of M51T with the bound maltose determined at pH 5.4. In the structure of M51T, the substitution of Thr 51 created a space that had been occupied by S␦ and C⑀ in the original Met 51 . Instead of being occupied by the side chain of Met 51 , this space was occupied by one water molecule, which was hydrogen-bonded to O⑀-2 of Glu 380 (2.8 Å) and N of Lys 295 (3.2 Å). The loss of the side chain of Met 51 resulted in the disruption of the hydrogen bond between N of Lys 295 and O⑀-2 of Glu 380 (from 2.9 to 3.8 Å), which may have allowed for the splitting of the side chain of Glu 380 into two alternate positions. Both of the positions were shifted by 0.6 and 2.1 Å from that of the Glu 380 O⑀-2 in the wild-type SBA. The disposition of Glu 380 O⑀-2 also disrupted the hydrogen bond between O⑀-2 of Glu 380 and N␦-2 of Asn 340 . The altered side-chain position of Glu 380 was so close to O-1 of the distorted glucose at subsite Ϫ1 that it prevented the binding of ␤-anomeric glucose and the localization of a catalytic water molecule, which would be in accord with the low activity of M51T.
In contrast to the electron density map of the wild-type enzyme, that of maltose in M51T was interpreted to show a mixture of three maltose molecules bound at subsites Ϫ2 to Ϫ1

␤-Amylase Mutants with Increased pH Optimum
(a distorted glucose and an ␣-4 C 1 glucose at subsite Ϫ1), at subsites ϩ1 to ϩ2, and at subsites ϩ2 to ϩ3. It is assumed that a maltose molecule can bind at subsites ϩ1 to ϩ2 only when a distorted glucose residue binds at subsite Ϫ1. If an ␣-anomeric glucose residue binds at subsite Ϫ1, the second maltose will bind at subsites ϩ2 to ϩ3 due to a collision between O-1 of the glucose at subsite Ϫ1 and O-4 of the glucose at subsite ϩ1. The occupancy refinement of the two alternate positions of maltose coupled with the alternate position of the side chain of Glu 380 showed that their respective occupancies were both about 0.5.
The Structure of E178Y-Maltose Complex- Fig. 3C shows the 2F o Ϫ F c map and the residues around Glu 380 of E178Y with the bound maltose determined at pH 5.4. The substituted side chain of Y178 created a novel hydrogen bond with the side chain of Asn 340 (O Tyr 178 -O␦-1 Asn 340 , 2.8 Å). The side chain of Asn 340 was found to flip toward O of Tyr 178 by rotating the 2 torsion angle about 53°, resulting in the disruption of the hydrogen bond between O⑀-2 of Glu 380 and O␦-1 of Asn 340 . Two maltose molecules were found, one at subsite Ϫ2 to Ϫ1 and one at subsite ϩ2 to ϩ3. The anomer of the glucose residue at subsite Ϫ1 was clearly identified as having the ␣-configuration, suggesting that the collision of this O-1 with O-4 of the glucose residue at subsite ϩ1 prevented the second maltose from binding at subsite ϩ1. Instead of a glucose residue at subsite ϩ1, there were four water molecules. One water molecule was assigned as the catalytic water molecule, and it was found to be in almost the same position as that of O-1 of the distorted glucose residue at subsite Ϫ1 in the wild-type SBA-maltose complex. The alternative position of the ␣-anomeric glucose residue at subsite Ϫ1 may be ascribed to the increased pK a value of Glu 380 , which was induced by the disruption of the hydrogen bond with Asn 340 . It is thought that both distorted maltose and normal maltose bind primarily at subsite Ϫ2 to Ϫ1, depending on the protonated and deprotonated states of Glu 380 .
To examine this hypothesis, the structure of the E178Ymaltose complex was determined at pH 7.1 as shown in Fig. 3D. The orientations of the side chains of Glu 380 , Asn 340 , and Tyr 178 were the same as those found at pH 5.4. In contrast to the results obtained at pH 5.4, both a distorted maltose and a normal ␣-anomeric maltose bind at subsites Ϫ2 to Ϫ1, and the second maltose binds to subsites ϩ1 to ϩ2 or subsites ϩ2 to ϩ3 as found in the case of M51T-maltose complex at pH 5.4 (Fig.  3B). It is suggested that only the deprotonated form of Glu 380 can enable the binding of the distorted maltose to subsites Ϫ2 to Ϫ1.
The Structure of N340T-Maltose Complex- Fig. 3E shows the 2F o Ϫ F c map and the residues around Glu 380 of N340T with the bound maltose determined at pH 5.4. O␥-1 of the substituted Thr 340 was found to face the side chain of Glu 178 , creating a weak hydrogen bond with O⑀-1 of Glu 178 at a distance of 3.4 Å and resulting in disruption of the original hydrogen bond between O⑀-2 of Glu 380 and O␦-1 of Asn 340 . Two separated maltose molecules were found, one at subsites Ϫ2 to Ϫ1 (␣-anomer) and one at subsites ϩ2 to ϩ3 as was the case for E178Y at pH 5.4. The distance between O⑀-1 of Glu 380 and oxygen of the catalytic water molecule was found to be 2.9 Å, which is slightly longer than that found in E178Y. The temperature factor of this catalytic water molecule was also slightly higher in N340T (35 Å 2 ) than in E178Y (31 Å 2 ). These parameters of the catalytic water molecule may account for the decreased specific activity in these mutants and for the fraction of bound distorted maltose at subsites Ϫ2 to Ϫ1 around their respective pH optima.
Structural Changes Required for Shifts in pH Optimum- Fig. 4 shows the superimposition of the residues around Glu 380 in the mutants and the wild-type SBA onto those of BCB. In the mutant M51T, the substituted side chain of Thr 51 formed a hydrogen bond with the side chain of Gln 87 (O␥-1 of Thr 51 -O⑀-2 of Gln 87 , 2.8 Å) instead of the disrupted two interactions between S␦ of Met 51 and N of Lys 295 and N of Lys 295 and O⑀-2 of Glu 380 . The space was filled with a water molecule that formed a hydrogen bond with O⑀-2 of Glu 380 and N of Lys 295 . These changes led to the instability of the side chain of Glu 380 , resulting in the two disordered alternate positions. The disruption of the hydrogen bond between O⑀-2 of Glu 380 and N␦-2 of Asn 340 and the disordered side chain of Glu 380 were in accord with the increased pH optimum and the most decreased activity of the mutant (Tables II and III). The situation in the case of BCB was similar in that two water molecules occupied the vacant space between the substituted Thr 51 and the catalytic base residue. But in the case of BCB, the other water molecule was hydrogen-bonded to O␥ of Thr 47 (3.2 Å), O⑀-2 of Glu 367 (3.2 Å), and N of Lys 287 (2.8 Å), indicating that these water-mediated hydrogen bonds stabilized the position of Glu 367 . The results for M51T suggest that the side chain of Met 51 or the water mediated-hydrogen bonds is important for fixing the position of the base catalyst.
In contrast to M51T, E178Y and N340T have 42 and 32% of the activity of the wild-type enzyme with increased pH optimum, respectively. Their conformation changes are restricted only around the mutated side chains, providing precise information about the shift in optimum pH. In the case of E178Y, despite the bulkiness of the side chain of Tyr 178 , the side chain of Tyr 178 was found to occupy almost the same orientation as that of Glu 178 in the wild-type SBA. Met 51 also held the same ␤-Amylase Mutants with Increased pH Optimum position as it did in the wild-type SBA. However, the side chain of Asn 340 was turned to face the side chain of Tyr 178 by a 53°r otation of the 2 torsion angle, thus forming a hydrogen bond (O of Tyr 178 -O␦-1 of Asn 340 , 2.8 Å), and this resulted in the disruption of the hydrogen bond between O⑀-2 of Glu 380 and O␦-1 of Asn 340 . In the case of N340T, the mutation did not affect the side-chain conformation of Met 51 or Glu 178 . The mutated side chain of Thr 340 created a hydrogen bond with the side chain of Glu 178 as was the case with Tyr 164 and Thr 328 in BCB. As the interactions between N of Lys 295 and O⑀-2 of Glu 380 and between S␦ of Met 51 and N of Lys 295 were still maintained in E178Y and N340T, the disruption of the hydrogen bond between O⑀-2 of Glu 380 and O␦-1 of Asn 340 was suf-ficient to increase the optimum pH from 5.4 to 6.0 -6.6 (Table II).
In the structures of E178Y and N340T determined at pH 5.4, the binding mode of maltose was altered from subsites Ϫ2 to Ϫ1 and ϩ1 to ϩ2 (wild type) to Ϫ2 to Ϫ1 and ϩ2 to ϩ3, respectively, due to the predominant binding of an ␣-anomeric glucose residue at subsite Ϫ1. In every case in which a distorted glucose residue binds at subsite Ϫ1, the sugar rings are distorted to form conformations that resemble boat or halfchair conformations. Alteration of the binding mode occurred due to changes in pH, as shown in Fig. 3D, in the case of E178Y, suggesting that the binding of a distorted glucose residue at subsite Ϫ1 required the deprotonation of Glu 380 . The strong nucleophile of the ionized side chain of Glu 380 may have enabled the distortion of the sugar ring at subsite Ϫ1. The position of O-1 in the distorted glucose residue at subsite Ϫ1 was very close to that of a catalytic water molecule found in the binding of ␣-glucose at the subsite, suggesting that the ionized side chain of Glu 380 can activate the catalytic water molecule to attack C-1 of the glucose residue, provided that a true substrate with an ␣-configuration binds at subsite Ϫ1. Thus, the catalytic activity is thought to be very sensitive to the nucleophilic nature of Glu 380 in terms of the distance between O⑀-1 of Glu 380 and the catalytic water molecule; this sensitivity may account for the decreased activity of E178Y and N340T relative to that of the wild-type enzyme.
Hydrogen Bond Networks Altering the pH Optimum in Other Glycoside Hydrolases-Alteration of the pH optimum by changing the pK a of the acid/base catalyst in other members of the glycoside hydrolase family has been observed previously (41)(42)(43). In a study of Aspergillus awamori glucoamylase (GA) and Bacillus circulans xylanase (BCX), hydrogen bonds between the acid/base catalyst and residues at the active site were shown to play significant roles in regulation of the pH optimum (41, 43). To provide a general concept to describe the control of the pH optimum of enzymes, we compared the active site of SBA with those of GA and BCX. Fig. 5 shows the conformations of active site residues of SBA-maltose, GA-acarbose, and BCX-2FXb (2-deoxy-2-fluoro-xylobiose), which were arranged such that they had the same substrate orientation; in each case, the two catalytic residues were positioned above and below the substrate, respectively. Like SBA, GA (glycoside hydrolase family 15) is an inverting exoglycosidase (1, 2) that produces ␤-D-glucose by hydrolyzing ␣-1,4and ␣-1,6-glucosidic linkage from the non-reducing ends of starch and related oligo-and polysaccharides (44,45). Two carboxyl groups are known to be involved in the catalytic mechanism of GA in which Glu 179 and Glu 400 are the acid/base catalyst (A. awamori and Aspergillus niger numbering system) and correspond to Glu 186 and Glu 380 in SBA, respectively (46 -49). Glu 400 in GA also forms hydrogen bond networks (Tyr 48 -Glu 400 -Ser 411 and Glu 400 -Gln 401 -Asn 315 including catalytic water) (Fig. 5B). Frandsen et al. (48) conducted a mutational analysis of Y48W, and suggested that Tyr 48 is functionally linked to Glu 400 and is important for maintaining the active site geometry and for the stabilization of an oxocarbonium ion intermediate. The k cat value of Y48W in GA was reduced 80 -100-fold, while K m was increased 2-3-fold. Y48W had unusually high activity at pH values below 4.0. In contrast, Fang et al. (41) reported that disruption of the hydrogen bond between O␥ of Ser 411 and O⑀-1 of Glu 400 by sitedirected mutagenesis of Ser 411 to either Ala (S411A) or Cys (S411C) reduced the catalytic efficiencies by only 46 and 26% that of the wild-type enzyme, respectively, but the mutant S411A and S411C increased the pH optima when maltose or maltoheptaose was used as the substrate by 0.8 -0.9 pH units; this latter effect was mainly due to the increase in the pK 1 of Glu 400 . Although they did not confirm the loss of the hydrogen bond between Ser 411 and Glu 400 by x-ray crystallographic analysis, the increase in the pK 1 values of S411A or S411C in GA was about the same as that found in SBA mutants (shown in Table II), suggesting that the effect of the disruption of the hydrogen bond between Ser 411 and Glu 400 in S411A or S411C of GA corresponds to that of the disruption of the hydrogen bond between Asn 340 and Glu 380 in both E178Y and N340T in SBA (Fig. 3, C and E).
BCX (glycoside hydrolase family 11) hydrolyzes xylosidic substrates with a net retention of an anomeric configuration (1,2). This proceeds via a double displacement mechanism in which a covalent intermediate is formed in the glycosylation step, and this intermediate is subsequently hydrolyzed in the deglycosylation step. Previous studies have determined that Glu 78 serves as a nucleophile and Glu 172 functions as a general acid/base residue (50 -52). Glu 172 in BCX forms hydrogen bond networks (Asn 35 -Glu 172 -Tyr 80 , Wat-Glu 172 -CW-Tyr 80 where Wat is water and CW is the catalytic water molecule) as shown in Fig. 5C. The pH optima of family 11 xylanases are well correlated with the nature of the residue adjacent to the acid/ base catalyst. In xylanases that function optimally under acidic conditions, this residue is Asp 35 , whereas under more alkaline conditions, it is Asn 35 . Joshi et al. (43) reported that substitution of Asn 35 with Asp decreased the pH optimum of BCX from 5.7 to 4.6. The crystal structure analysis of N35D-2FXb showed the formation of a hydrogen bond between Asp 35 and Glu 172 by 2.7 Å, which was stronger than that in the wild-type BCX-2FXb by 3.3 Å. They proposed a mechanism to explain the dependence of the pH optimum of BCX on the distance of the hydrogen bond between Asn/Asp 35 and Glu 172 . They also showed that the pH optimum of Y80F in BCX was shifted from 5.7 to 6.3, mainly due to an increase of ϳ1 unit in the apparent pK a value of Glu 172 (53). Although the mutant showed that k cat decreased by ϳ2 orders of magnitude, the crystal structure of Y80W showed a loss of the hydrogen bond between Glu 172 and Tyr 80 in the wild-type BCX. These studies suggest that SBA, GA, and BCX share the same features responsible for controlling the pH optimum. It is concluded that hydrogen bonds, as well as networks between the acid/base catalyst and the residues near the active site, control the pH optimum despite the different reaction mechanisms used by inverting and retaining enzymes.
In this study, we demonstrated that the pH optimum of SBA mutants could be shifted toward the alkaline region by removing hydrogen bonds with the side chain of Glu 380 , the catalytic base, possibly as a result of increasing its pK a . The present study also suggested that the formation or deformation of hydrogen bonds between the catalytic residue and residues near the active site can control the pH optimum of enzymes; such changes in hydrogen bonding can lead to shifts in the pH optimum toward either the acidic or alkaline region.