This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.
Agricola
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, June 2005, p. 3285-3293, Vol. 71, No. 6
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.6.3285-3293.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Cooperative Effect of Two Surface Amino Acid Mutations (Q252L and E170K) in Glucose Dehydrogenase from Bacillus megaterium IWG3 on Stabilization of Its Oligomeric State

Sang-Ho Baik,1,3* Fabrice Michel,2 Nushin Aghajari,2 Richard Haser,2 and Shigeaki Harayama1,3

Marine Biotechnology Institute, 3-75-1 Heita Kamaishi, Iwate 026-0001, Japan,1 Institut de Biologie et Chimie des Protéines, UMR5086-CNRS, Laboratoire de BioCristallographie, IFR128 BioSciences Lyon-Gerland, 7 Passage du Vercors, 69367 Lyon cedex 07, France,2 Biotechnology Development Center, National Institute of Technology and Evaluation, 2-5-8 Kazusa-Kamatari, Kisarazu-shi, Chiba 292-0818, Japan3

Received 11 August 2004/ Accepted 21 December 2004


arrow
ABSTRACT
 
A thermostable glucose dehydrogenase (GlcDH) mutant of Bacillus megaterium IWG3 harboring the Q252L substitution (Y. Makino, S. Negoro, I. Urabe, and H. Okada, J. Biol. Chem. 264:6381-6385, 1989) is stable at pH values above 9, but only in the presence of 2 M NaCl. Another GlcDH mutant exhibiting increased stability at an alkaline pH in the absence of NaCl has been isolated previously (S.-H. Baik, T. Ide, H. Yoshida, O. Kagami, and S. Harayama, Appl. Microbiol. Biotechnol. 61:329-335, 2003). This mutant had two amino acid substitutions, Q252L and E170K. In the present study, we characterized three GlcDH mutants harboring the substitutions Q252L, E170K, and Q252L/E170K under low-salt conditions. The GlcDH mutant harboring two substitutions, Q252L/E170K, was stable, but mutants harboring a single substitution, either Q252L or E170K, were unstable at an alkaline pH. Gel filtration chromatography analyses demonstrated that the oligomeric state of the Q252/E170K enzyme was stable, while the tetramers of the enzymes harboring a single substitution (Q252L or E170K) dissociated into dimers at an alkaline pH. These results indicated that the Q252L and E170K substitutions synergistically strengthened the interaction at the dimer-dimer interface. The crystal structure of the E170K/Q252L mutant, determined at 2.0-Å resolution, showed that residues 170 and 252 are located in a hydrophobic cavity at the subunit-subunit interface. We concluded that these residues in the wild-type enzyme have thermodynamically unfavorable effects, while the Q252L and E170K substitutions increase the subunit-subunit interactions by stabilizing the hydrophobic cavity.


arrow
INTRODUCTION
 
Glucose dehydrogenase (GlcDH) (E.C. 1.1.1.47) from Bacillus megaterium IWG3 is an oligomeric tetramer composed of four identical subunits with a molecular mass of 28 kDa, and it is a member of the short-chain alcohol dehydrogenase family (9, 8, 14). Wild-type GlcDH is inactivated at pH values above 9 in the absence of a high concentration (2 M) of NaCl. At such pH values, the quaternary structure of GlcDH is significantly perturbed, as the tetrameric enzyme dissociates into inactive monomers due to weak subunit-subunit interactions. When the pH is brought to neutral, the enzyme is reversibly renatured by reconstruction of the oligomeric tetramer (16, 17, 22). Although the dissociation and association of the tetrameric oligomer of wild-type GlcDH accompanying changes in pH or the urea concentration have been well characterized (18), the detailed mechanisms of the subunit-subunit interactions of GlcDH must still be elucidated.

Several amino acid residues that affect the stability of this enzyme have been identified by isolating GlcDH mutants exhibiting improved thermostability (1, 15, 18, 19). One of the mutations giving rise to improved thermostability is Q252L. However, the Q252L mutant showed dramatically decreased stability at alkaline pHs in the absence of a high concentration of NaCl (15, 19). Recently, we isolated a GlcDH mutant with two substitutions, E170K and Q252L. The stability of the Q252L/E170K enzyme at alkaline pHs was high even in the absence of NaCl.

In this study, we examined the effect of the E170K and Q252L mutations on the secondary and quaternary structures of GlcDH. Moreover, the crystal structure of the Q252L/E170K enzyme was determined and compared with the structure of wild-type GlcDH which has been determined previously (30). The results obtained in this study indicated that the stability of the Q252L/E170K GlcDH is driven by stabilization of the quaternary structure and that a hydrophobic cavity is formed at the subunit-subunit interfaces and is stabilized by the E170K and Q252L mutations.


arrow
MATERIALS AND METHODS
 
Site-specific mutagenesis.
The L252Q substitution (wild-type allele) was introduced into the Q252L/E170K mutant gene by PCR-based site-specific mutagenesis to obtain the E170K mutant gene. The template was pTDN-46 carrying the Q252L/E170K mutant gene (20), and the PCR amplification was carried out with two primers, primers A (5'-ATGCAGCAAGTAAAGGCG3') and B (5'-AAGTCGACTTATTCGCGTTCTGCTTGGAATGATGGGTACTGTGTCATACCG-3'). The residue of primer B that leads to the L252Q mutation is underlined, while the SalI site in primer B is indicated by boldface. The PCR-amplified DNA fragment was restricted by the SalI restriction endonuclease (New England Biolabs), and the resultant 322-bp fragment was isolated from an agarose gel with a gel purification kit (QIAGEN). The isolated DNA fragment was ligated to SalI-restricted pTDN-46 to produce the gene encoding a mutant GlcDH carrying E170K. The mutation was confirmed by dye termination cycle sequencing using a 377 DNA sequencer (Applied Biosystems).

Purification of GlcDHs.
The GlcDHs carrying the Q252L, E170K, and Q252L/E170K mutations were overexpressed in Escherichia coli JM109 and purified by anion-exchange chromatography using a high-pressure liquid chromatography system (Tosoh). The active fraction was heated in 10 mM phosphate buffer (pH 6.5) at 60°C to remove heat-labile proteins of E. coli, and soluble proteins were purified further by gel filtration as described previously (1, 19). For purification of the Q252L and E170K enzymes, heat treatment at 60°C was carried out in 10 mM phosphate buffer (pH 6.5) containing 2 M NaCl to prevent inactivation of the enzymes at this temperature. The soluble proteins after the heat treatment were dialyzed overnight at 4°C against 10 mM phosphate buffer (pH 6.5) to remove the excess NaCl before gel filtration. The purity of the purified enzymes was checked by sodium dodecyl sulfate-12.5% polyacrylamide gel electrophoresis at a constant current of 15 mA, and the enzymes were visualized by Coomassie blue or silver staining (11).

Enzyme assay.
Under the standard assay conditions, the GlcDH activity was determined spectrophotometrically at 25°C by measuring the rate of NADH formation at 340 nm in 100 mM Tris-HCl (pH 8.0) containing 0.1 M D-glucose and 0.5 mM NAD+ as described previously (1, 30). One unit of GlcDH was defined as the amount of the enzyme that produced 1 µmol of NADH per min under the standard assay conditions. The protein concentration was determined by the bicinchoninic acid method (28) with bovine serum albumin as a standard. The analyses of kinetic constants and substrate specificity were carried out as described previously (1).

Enzyme inactivation at alkaline pH.
Enzyme inactivation was determined at 25°C by incubating the enzyme (10 µg ml–1) in 10 mM phosphate buffers at pHs ranging from 8 to 10.5 and measuring the residual activity in 100 mM Tris-HCl (pH 8.0).

CD analysis.
Circular dichroism (CD) spectra were obtained with a J-725 spectropolarimeter (Jasco) with a cell having a 1-cm path length at 25°C. The protein samples were continuously stirred to keep the temperature uniform. To monitor the alkali-induced denaturation of the secondary structures of GlcDHs, changes in the CD spectra around 222 nm were recorded in 10 mM phosphate buffer (pH 6.5) before and after incubation of the enzymes at alkaline pHs as described above.

Quaternary structure analysis by gel filtration chromatography.
Purified enzymes (300 µg) were incubated in 10 mM phosphate buffer (pH 6.5) at pH values between 8 and 10.5 for 1 h at 25°C and loaded onto a Superdex R 200HR 10/30 column (Amersham Biosciences) which had been equilibrated with 10 mM potassium phosphate buffer at pH values between 8 and 10.5 and fitted to a Tosoh high-pressure liquid chromatography system (Tosoh). The proteins were eluted at a flow rate of 0.4 ml/min. For equilibration to a specific pH, the column was washed with buffer at the pH overnight at a flow rate of 0.4 ml/min. The protein elution profile was monitored at 280 nm at room temperature.

Crystallization of the Q252L/E170K mutant.
Crystallization was performed using the hanging-drop vapor diffusion method with the Q252L/E170K enzyme (10 mg/ml) in the presence of NAD+ (4 mM) in 50 mM sodium phosphate buffer (pH 6.3) as described for the native enzyme by Yamamoto et al. (30). Crystals were grown from 0.1 M morpholineethanesulfonic acid (MES) (pH 6.5) and 30% (wt/vol) polyethylene glycol 6000 at 18°C and appeared within 3 to 4 weeks.

Data collection and model refinement.
Crystals were flash frozen at 100 K in a nitrogen stream before data collection using 50% ethylene glycol as a cryoprotectant. Diffraction data were collected at 2.0-Å resolution with a MarCCD detector at European Synchrotron Radiation Facility beamline BM30A ({lambda} = 0.98 Å). Diffracted intensities were integrated with the program MOSFLM (13), as implemented in the CCP4 software package and scaled with SCALA (4). Refinement of the model was done with the simulated annealing protocol as implemented in the software package CNS (3) using all reflections between 15 and 2.0 Å and then alternating cycles of positional refinement and manual building using TURBO-FRODO (23). Water molecules present at similar positions in the Q252L/E170K mutant and wild-type structures have the same numbering. Free and conventional R factors were monitored (2) to avoid overrefinement. Model qualities were examined with PROCHECK (12) and WHATCHECK (6).

Protein structure accession number.
Atomic coordinates and structure factors have been deposited in the Protein Data Bank (PDB) under accession code 1RWB.


arrow
RESULTS
 
Enzyme inactivation at alkaline pHs and changes in the secondary structures of GlcDHs.
In this study, we constructed a gene encoding GlcDH possessing the E170K mutation. The mutant enzyme was expressed in E. coli and purified. The kinetic constant of the E170K enzyme was determined and is shown in Table 1 together with those of the Q252L and Q252L/E170K enzymes. We investigated the stabilities of these three mutant enzymes at pH values ranging from 8 to 10.5. The enzymes were incubated in 10 mM phosphate buffer at each pH for 1 h, and the residual activity was measured under the standard assay conditions. As reported previously, the Q252L enzyme was inactivated at pH values above 9 (19). The newly isolated mutant enzyme carrying E170K was shown to be relatively resistant to alkaline pHs, as shown in Fig. 1. Approximately 26% of the initial enzyme activity remained even after incubation at pH 10 or 10.5. The Q252L/E170K enzyme was more stable than the E170K enzyme and retained more than 95% of the initial enzyme activity after treatment at pH values between 8 and 10.5. The results indicated that there was a synergetic effect of the E170K and Q252L mutations that maintained the structural integrity of GlcDH at alkaline pHs. On the other hand, the stabilities of the Q252L and E170K enzymes at alkaline pHs were significantly increased when 2 M NaCl was added, and they were indistinguishable from the stability of the Q252L/E170K enzyme.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Apparent kinetic constants of glucose dehydrogenases harboring E170K, Q252L, and E170K/Q252L substitutionsa



View larger version (56K):
[in this window]
[in a new window]
  		FIG. 1. Stability of glucose dehydrogenase at different pHs. The activities of the wild-type enzyme (A) and three mutant enzymes, the Q252L (B), E170K (C), and Q252L/E170K (D) enzymes, were measured after incubation of the enzymes for 1 h at various pHs in the absence (open bars) or presence (dotted bars) of 2 M NaCl. The error bars indicate standard errors of the means (n = 4).

Although the results shown in Fig. 1 clearly demonstrated the difference in the stabilities of the Q252L, E170K, and Q252L/E170K enzymes, it should be mentioned that the assays probably overestimated the stabilities of these enzymes at alkaline pHs. It is known that wild-type GlcDH shows reversible dissociation-association between inactive monomers and active tetramers when the pH is shifted between 9 and 7 (16, 17). Thus, a fraction of an inactivated enzyme might be reactivated during enzyme assays conducted at pH 8.0.

The enzymes were analyzed by CD spectroscopy. A local negative maximum in ellipticity was observed around 226 nm in the spectra of all enzymes before denaturation. As shown in Fig. 2, the alpha-helix content of each enzyme (monitored at 222 nm) did not change significantly at pH 9.5, at which the Q252L enzyme was completely inactivated and the E170K enzyme was partially inactivated, indicating that the inactivation of this enzyme at pH 9.5 was not due to the loss of secondary structures. At pH 10.5, the alpha-helix contents of the mutant Q252L and E170K enzymes slightly decreased, as shown in Fig. 2, suggesting that prolonged incubation of these enzymes at pH 10.5 resulted in partial unfolding of the secondary structures. In contrast, the secondary structures of the double-mutant Q252L/E170K enzyme were not affected even at pH 10.5, illustrating that the Q252L/E170K double mutant was more resistant to alkaline pHs than the E170K and Q252L single mutants.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 2. Far-UV circular dichroism spectra of the Q252L, E170K, and Q252L/E170K enzymes. Far-UV circular dichroism spectra were obtained with the enzymes incubated at pH 8.0 (thick lines), pH 9.5 (thin lines), and pH 10.5 (dashed lines) for 1 h. Protein concentrations were adjusted to 0.1 mg/ml in the same buffers (10 mM), and the spectra obtained were processed using the smoothing function installed in a standard Jasco analysis program. (A) Q252L enzyme; (B) E170K enzyme; (C) Q252L/E170K enzyme.

Gel filtration chromatography.
Since inactivation of the wild-type GlcDH at an alkaline pH was proven to be strongly correlated with destruction of the quaternary structure (tetramer) of the enzyme (16, 17, 22), gel filtration chromatography was carried out to analyze the enzyme carrying either the Q252L, E170K, or Q252L/E170K mutation(s) before and after alkali-induced inactivation. The native forms of these three enzymes at neutral pH were eluted as a single peak around 110 kDa, a size corresponding to a tetramer (Fig. 3). The Q252L/E170K enzyme, which was stable at alkaline pHs, showed absolutely identical chromatographic profiles before and after the alkali treatment, as shown in Fig. 3. On the other hand, two peaks at 110 and 60 kDa, corresponding to the tetrameric and dimeric oligomers, respectively, were detected with the Q252L enzyme after treatment at pH 8.5 in 10 mM potassium phosphate buffer, conditions under which only 50% of the enzyme activity was retained. Moreover, when the pH was shifted to 9.5, the tetramer peak of the Q252L enzyme disappeared completely, and there was a single peak at 60 kDa corresponding to the dimeric state (Fig. 3). The intensity of the dimer peak after treatment at pH 10.5 was significantly lower than that at pH 9.5, indicating partial unfolding of the secondary structures or aggregation. In the case of the E170K enzyme, no change in the elution pattern was observed after treatment at pH 8.5. At pH 9.5, two peaks of the E170K enzyme corresponding to the tetrameric and dimeric oligomers appeared, while at pH 10.5, a single peak corresponding to a dimer was observed. No trace of monomers was present under these conditions.



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 3. Gel filtration chromatography analyses of glucose dehydrogenase mutants after incubation at various pHs. (A) Q252L/E170K enzyme; (B) E170K enzyme; (C) Q252L enzyme. The standard molecular mass markers used to establish the calibration curve were bovine thyroglobulin (670 kDa), glutamate dehydrogenase (290 kDa), lactate dehydrogenase (142 kDa), enolase (67 kDa), chicken ovalubumin (44 kDa), myokinase (32 kDa), horse myoglobin (17 kDa), vitamin B12 (13.5 kDa), and cytochrome c (12.4 kDa; Amersham Biosciences).

Three-dimensional structure determination.
Crystals of the Q252L/E170K enzyme belonged to monoclinic space group C2, and their unit cell parameters were quite similar to those of the wild-type GlcDH structure. Diffraction data were collected to 2.0 Å, and the statistics are summarized in Table 2. Since the crystals of the Q252L/E170K enzyme were isomorphous with those of wild-type GlcDH, the three-dimensional structure of the mutant was determined by using the phases of the wild-type enzyme structure as solved at 1.7-Å resolution (30), in which NAD+ as well as water molecules had been removed in order to avoid bias. The refined model (Rcryst = 22.4%, Rfree = 26.6%) contained 1,044 amino acid residues, 549 water molecules, and four NAD+ molecules. The details of the refinement results are given in Table 2.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Data collection and refinement statistics

As expected, the Q252K/E170K enzyme, like the wild-type GlcDH, consisted of four identical subunits, each displaying a seven-stranded parallel ß-sheet flanked by two arrays of three {alpha}-helices (Fig. 4A). The sites of substitutions within the Q252L and E170K enzymes were surrounded by several hydrophobic amino acid residues, which formed a hydrophobic cavity-like interface of oligomeric structures, as shown in Fig. 4B. The hydrophobic cavity around residue 252 of monomer A [252 (A)] is lined up by residues Y253 (A), P214 (A), M215 (A), P254 (A), F256 (A), I150 (C), P151 (C), W152 (C), A173 (D), and L174 (D). In both wild-type and mutant enzymes, there is an interaction between Y253 (A) and T251 (C) at one edge of the hydrophobic cavity. In the wild-type enzyme, Q252 (A) interacts with G241 (D) to make intermolecular contacts across the P axis, as defined in Yamamoto et al. (30). This interaction was not observed in the Q252L/E170K structure. Instead, in the mutant enzyme, G241 (D) forms intramolecular contacts with K170 (D) bordering another edge of the hydrophobic cavity. Moreover, in the latter mutant enzyme, two water molecules (Wat269 and Wat320 according to the nomenclature of the wild-type structure) observed around residue 252 in the wild-type structure were lacking (Fig. 4B).




View larger version (109K):
[in this window]
[in a new window]
  		FIG. 4. (A) Crystal structure of the Q252L/E170K enzyme with a close-up of the region harboring Leu252 and Lys170. Monomers B, C, and D are grey, while monomer A is different colors. Leu252, Lys170, and NAD+ are represented by ball and stick models and are color coded by atom type (yellow, carbon; red, oxygen; blue, nitrogen). Leu252 and Lys170 on monomer A are indicated by larger spheres. The P-axis interface is between molecules A and D, whereas the Q-axis interface is between molecules A and B. Helices in monomer A are different colors, ranging from blue at the N terminus and then green followed by yellow and finally red at the C terminus. (B) Superimposition of the Q252L/E170K (black lines) and wild-type (grey lines) glucose dehydrogenases. Direct interactions between amino acids are indicated by thick dashed lines, and interactions mediated by water molecules are indicated by thin dashed lines. The figures were produced with the program MOLSCRIPT (10).


arrow
DISCUSSION
 
In a search for a more stable GlcDH for determination of the blood glucose level in clinical diagnosis and for coenzyme regeneration in various biotechnological applications, we previously isolated a thermostable GlcDH double mutant with the Q252L and E170K substitutions. Since the mutated amino acid residues are located at the monomer-monomer or dimer-dimer interfaces, as shown in Fig. 4, we expected these mutations to be involved in stabilization of the oligomeric structure of the enzyme (30). In the present study, we showed by using gel filtration chromatography and circular dichroism analyses that inactivation of the enzyme carrying either the E170K or Q252L mutation occurred with dissociation of the oligomeric structure from tetramer to dimers without a significant change in the CD spectra. In contrast to an enzyme carrying a single mutation (either E170K or Q252L), the enzyme carrying both mutations in a single molecule was stable in terms of activity, as well as quaternary structure, under all conditions examined. At pH 10.5 in the absence of 2 M NaCl, more than 90% of the E170K enzyme was dissociated into dimers (Fig. 3), while the residual activity of the E170K enzyme incubated under the same conditions was 26% (Fig. 1). We thought that the inconsistency between the tetramer content and the enzyme activity of the E170K enzyme might have been due to reactivation of inactivated enzymes during the enzyme assay at pH 8.0. To test this hypothesis, we carried out an experiment in which the E170K enzyme was incubated at either pH 8.0 or 10.5 for 1 h, and the residual activity was assayed at pH 10.5. The E170K enzyme incubated at pH 8.0 showed activity, and the activity decreased gradually at pH 10.5, indicating that enzyme inactivation took place during the measurement of activity at pH 10.5. The E170K enzyme incubated at pH 10.5, in contrast, showed very low activity at pH 10.5. Although it was difficult to measure accurately the enzyme activity at pH 10.5, the residual activity of the E170K enzyme incubated at pH 10.5 was estimated to be less than 10% of the activity of the enzyme incubated at pH 8.0 (unpublished data),

Thus, a striking difference in stability between the E170K, Q252L, and Q252L/E170K mutants was observed at alkaline pHs in the absence of a high concentration of NaCl. In these conditions, the stabilities of the GlcDHs were as follows: Q252L/E170K enzyme > E170K enzyme > Q252L enzyme {approx} wild-type enzyme. Although the stability of the Q252L enzyme was almost identical to that of the wild-type enzyme at alkaline pHs and in the absence of NaCl, its thermostability was improved at neutral pHs (15, 19). Since residue 252 is located at the interface involved in the association between subunits A and D (Fig. 4B), increased stability of the Q252L enzyme at high temperatures was expected to be related to the reinforcement of the subunit association around the P axis due to the Q252L mutation. On the other hand, the E170K mutation, located near the interface between subunits A and B (Q axis), improved the stability at alkaline pHs, suggesting that the Q-axis association is stabilized by the E170K mutation. In order to understand the details of the stabilization mechanisms induced by these two mutations, we determined the crystal structure of the E170K/Q252L enzyme at 2.0-Å resolution (Fig. 4A) and compared it to the wild-type three-dimensional structure.

On the basis of the crystal structure, we propose that L252 of the mutant enzyme stabilizes the tetramer structure by changing the environment in the vicinity of Y253. The latter residue forms intermolecular contacts, especially with T251 [namely, Y253 (A)-T251 (C)] in both the wild-type and mutant enzymes. Y253 is in a conformation with dihedral angles ({phi}, {psi}) characteristic of a left-handed helix, meaning that the carbonyl groups of both residues 252 and 253 point in the same direction. The side chain carbonyl carbon of Q252 in the wild-type enzyme is in a hydrophobic environment; thus, the side chain of L252 in the mutant enzyme may strengthen the hydrophobic nature of this region. One edge of this region is closed by the interaction between Y253 (A) and T251 (C) in the wild-type enzyme, while another H-bond interaction bordering the cavity is found in the mutant enzyme between K170 (D) and G241 (D).

Sequence alignments of enzymes that belong to the short-chain alcohol dehydrogenase family and whose three-dimensional structures are known show that only 2 of 18 sequences contain the glutamine mentioned above at position 252, whereas 12 sequences contain a hydrophobic residue at this position (Fig. 5). Therefore, it seems that hydrophobic residues are preferable at this position in order to stabilize the oligomeric structure.



View larger version (91K):
[in this window]
[in a new window]
  		FIG. 5. Protein sequence alignment based on the three-dimensional structures of enzymes belonging to the short-chain alcohol dehydrogenase family. The secondary structure of the B. megaterium glucose dehydrogenase (PDB accession code 1GCO) is shown above the sequence alignment. Strongly conserved residues are enclosed in boxes, in which white letters with a red background indicate absolutely conserved residues and residues with similar biochemical properties are indicated by red letters. Residues equivalent to residue 170 or 252 in the 1GCO sequence are indicated by stars. Positively charged residues at positions equivalent to E170 in the 1GCO structure are indicated by a yellow background, while hydrophobic residues at positions equivalent to Q252 are indicated by a green background. The triad of catalytically important residues (S, Y, and K) in the short-chain alcohol dehydrogenase family are indicated by triangles, and the N-terminal coenzyme-binding motif (G-X3-G-X-G) is indicated by solid circles. The enzymes corresponding to PDB codes are as follows: 1GEG, Klebsiella pneumoniae acetoin reductase; 1EDO, Brassica napus ß-keto acyl carrier protein reductase; 1DOH, Magnaporthe grisea trihydroxynaphthalene reductase; 1GCO, B. megaterium glucose dehydrogenase; 1HXH, Comamonas testosteroni 3ß/17ß-hydroxysteroid dehydrogenase; 1HDC, Streptomyces exfoliatus 3{alpha},20ß-hydroxysteroid dehydrogenase; 1AHH, E. coli 7{alpha}-hydroxysteroid dehydrogenase; 1AE1, Datura stramonium tropinone reductase I; 2AE1, Datura stramonium tropinone reductase II; 1E3S, Rattus norvegicus short-chain 3-hydroxyacyl-coenzyme A dehydrogenase; 1CYD, Mus musculus carbonyl reductase; 1H5Q, Agaricus bisporus NADP-dependent mannitol dehydrogenase; 1BDB, Pseudomonas sp. cis-biphenyl-2,3-dihydrodiol-2,3-dehydrogenase; 1FJH, Comamonas testosteroni 3{alpha}-hydroxysteroid dehydrogenase/carbonyl reductase; 1A27, Homo sapiens 17-ß-hydroxysteroid dehydrogenase; 1A4U, Drosophila alcohol dehydrogenase; 1NAS, Mus musculus sepiapterin reductase; and 1DHR, Rattus norvegicus dihydropteridine reductase. The relative accessibility of each residue (blue, accessible; cyan, intermediate; white, buried) and the hydropathic character (pink, hydrophobic; grey, intermediate; cyan, hydrophilic) of the 1GCO sequence are shown below each block. The sequence alignment was generated by CLUSTALW (26) and displayed with ESPript (5).

In our previous study (1), we assumed that anion-anion repulsion between subunits is a main reason for the instability of the wild-type and Q252L enzymes under alkaline conditions and in the absence of a high concentration of NaCl. Since the stability of GlcDH under alkaline conditions was significantly improved by the E170K mutation, residue 170 seems to be crucial for the stability at alkaline pHs. This residue is located on the {alpha}E helix, which belongs to the Q-axis-related interface. In the wild-type enzyme, E170 interacts with K149 across the Q axis between the A and B subunits and between the C and D subunits via water-mediated hydrogen bonds to Wat41 and Wat56. In the structure of the E170K/Q252L mutant, the water molecules mentioned above are displaced, and the water-mediated interactions between two subunits across the Q axis are absent (Fig. 4B). Moreover, the side chain of K170 (D subunit) forms direct hydrogen bonds with the carbonyl oxygen of G241 (D subunit), resulting in a new intramolecular interaction. Thus, the E170K mutation probably increases the hydrophobicity of this region in relation to the longer side chain of a lysine compared to a glutamic side chain. Therefore, the Lys residue may help prevent access of water molecules to the hydrophobic core due to steric hindrance and to direct hydrogen bonding to the carbonyl oxygen of G241.

The profile of the pH-dependent inactivation of GlcDH shown in Fig. 1 indicated that an ionizable species with a pKa of 8 or more is crucial for the subunit-subunit interaction. We assumed that E170 has an anomalously high pKa due to a hydrophobic environment that stabilizes the protonated form of this residue, as has been shown for other proteins (21, 27). We further assumed that the negative charge of E170 contributes to disruption of the dimer-dimer association. This hypothesis explains why K170, under any conditions, renders GlcDH resistant to alkaline pHs. It also explains the salt-dependent stability of the wild-type and Q252L enzymes under alkaline conditions, suggesting that in the presence of a high salt concentration, coulombic interactions are essentially screened out.

Sequence comparison of enzymes belonging to the short-chain alcohol dehydrogenase family (Fig. 5) showed that the majority of the amino acid residues at position 170 were basic (R or K) and that the E residue was found only in GlcDH (PDB accession code 1GCO) and in 17-ß-hydroxylsteroid dehydrogenase (PDB accession code 1A27).

In conclusion, the stabilization of the quaternary structure of GlcDH by the E170K/Q252L mutations is interpreted to be the result of the cooperative effects of these two mutations. Residue 252 is in a relatively unstable hydrophobic cavity surrounded by each of the monomers. This cavity appears to be easily accessible by water molecules, and their presence in such a hydrophobic core is certainly not thermodynamically favorable for stabilization of the surrounding region. As this hydrophobic region interlinks the four monomers, its instability would lead to instability of the active enzyme. On the other hand, the Q252L mutation may contribute to stabilization of the hydrophobic core in two ways: by removing a carbonyl group which is a potential hydrogen bond acceptor candidate and by displacing two water molecules from this region. The E170K mutation may also help to stabilize the enzyme structure by increasing the hydrophobicity around the Q-axis interface with the displacement of two water molecules in the vicinity of K170 and by preventing the development of a negative charge at residue 170 under alkaline conditions.

Since many enzymes consist of oligomeric structures for maintaining their activities, it is important to elucidate the factors involved in stabilization of their oligomeric states. Unfortunately, our knowledge concerning these aspects is still far from complete. Generally, the oligomeric structure of proteins is formed by various subunit-subunit interactions, including disulfide bonding, hydrogen bonding, and hydrophobic and electrostatic interactions. To understand how to increase the stability of these oligomeric organizations, conditions which increase or decrease the stability of an oligomeric structure have been investigated (7, 24, 25, 29). In this report, we propose an efficient stabilization mechanism for GlcDH which involves the shielding of a hydrophobic core by preventing its interactions with water molecules, a strategy which may be useful for engineering other proteins.


arrow
ACKNOWLEDGMENTS
 
We thank Hideko Yoshida for technical assistance.

This work was supported by the New Energy and Industrial Technology Development Organization (NEDO).


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Biotechnology Development Center, National Institute of Technology and Evaluation, 2-5-8 Kazusa-Kamatari, Kisarazu-shi, Chiba 292-0818, Japan. Phone: 81-438-20-5908. Fax: 81-438-20-5766. E-mail: baik-sangho{at}nite.go.jp. Back


arrow
REFERENCES
 
    1
  1. Baik, S.-H., T. Ide, H. Yoshida, O. Kagami, and S. Harayama. 2003. Significantly enhanced stability of glucose dehydrogenase IWG3 from Bacillus megaterium by directed evolution. Appl. Microbiol. Biotechnol. 61:329-335.[Medline]
    2. 2
  2. Brünger, A. T. 1992. Free R value: a novel statistical quantity for assessing the accuracy of crystal structures. Nature 355:472-475.[CrossRef][Medline]
    3. 3
  3. Brünger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. Rice, T. Simonson, and G. L. Warren. 1998. Crystallography and NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. Sect. D 54:905-921.[CrossRef][Medline]
    4. 4
  4. CCP4 Collaborative Computational Project. 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. Sect. D 50:760-763.[CrossRef][Medline]
    5. 5
  5. Gouet, P., E. Courcelle, D. I. Stuart, and F. Metoz. 1999. ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15:305-308.[Abstract/Free Full Text]
    6. 6
  6. Hooft, R. W., G. Vriend, C. Sander, and E. E. Abola. 1996. Errors in protein structures. Nature 381:272.[Medline]
    7. 7
  7. Jaenicke, R., and H. Lilie. 2000. Folding and association of oligomeric and multimeric proteins. Adv. Protein Chem. 53:329-401.[Medline]
    8. 8
  8. Jany, K. D., W. Ulmer, M. Fröschle, and G. Phleiderer. 1984. Complete amino acid sequence of glucose dehydrogenase from Bacillus megaterium. FEBS Lett. 165:6-10.[CrossRef][Medline]
    9. 9
  9. Jörnvall, H., B. Persson, M. Krook, S. Atrian, R. Gonzàlez-Duarte, J. Jeffery, and D. Ghosh. 1995. Short-chain dehydrogenase/reductases (SDR). Biochemistry 34:6003-6013.[CrossRef][Medline]
    10. 10
  10. Kraulis, P. J. 1991. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24:946-950.[CrossRef]
    11. 11
  11. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685.[CrossRef][Medline]
    12. 12
  12. Laskowski, R. A., M. W. MacArthur, and D. S. Moss. 1993. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26:283-291.[CrossRef]
    13. 13
  13. Leslie, A. G. W. 1991. Molecular data processing, p. 50-61. In D. Moras, A. D. Podjarny, and J. C. Thierry (ed.), Crystallography computing. Oxford University Press, Oxford, United Kingdom.
    14. 14
  14. Makino, Y., J. Y. Ding, S. Negoro, I. Urabe, and H. Okada. 1989. Purification and characterization of new glucose dehydrogenase from vegetative cells of Bacillus megaterium. J. Ferment. Bioeng. 67:372-379.
    15. 15
  15. Makino, Y., S. Negoro, I. Urabe, and H. Okada. 1989. Stability-increasing mutants of glucose dehydrogenase from Bacillus megaterium IWG3. J. Biol. Chem. 264:6381-6385.[Abstract/Free Full Text]
    16. 16
  16. Maurer, E., and G. Pfleiderer. 1985. Reversible pH-induced dissociation of glucose dehydrogenase from Bacillus megaterium. I. Conformational and functional changes. Biochim. Biophys. Acta 827:381-388.
    17. 17
  17. Maurer, E., and G. Pfleiderer. 1987. Reversible pH-induced dissociation of glucose dehydrogenase from Bacillus megaterium. II. Kinetics and mechanism. Z. Naturforsch. Sect. C 42:905-915.
    18. 18
  18. Mendoza-Hernandez, G., F. Minauro, and J. L. Rendon. 2000. Aggregation, dissociation and unfolding of glucose dehydrogenase during urea denaturation. Biochim. Biophys. Acta 1478:221-231.[CrossRef][Medline]
    19. 19
  19. Nagao, T., Y. Makino, K. Yamamoto, I. Urabe, and H. Okada. 1989. Stability-increasing mutants of glucose dehydrogenase. FEBS Lett. 253:113-116.[CrossRef][Medline]
    20. 20
  20. Nelson, R. M., and G. L. Long. 1989. A general method of site-specific mutagenesis using a modification of the Thermus aquaticus polymerase chain reaction. Anal. Biochem. 180:147-151.[CrossRef][Medline]
    21. 21
  21. Paddock, M. L., S. H. Rongey, G. Feher, and M. Y. Okamura. 1989. Pathway of proton transfer in bacterial reaction centers: replacement of glutamic acid 212 in the L subunit by glutamine inhibits quinone (secondary acceptor) turnover. Proc. Natl. Acad. Sci. USA 86:6602-6606.[Abstract/Free Full Text]
    22. 22
  22. Pauly, H. E., and G. Pfleiderer. 1977. Conformation and functional aspects of the reversible dissociation and denaturation of glucose dehydrogenase. Biochemistry 16:4599-4604.[CrossRef][Medline]
    23. 23
  23. Roussel, A., and C. Cambillau. 1992. TURBO-FRODO. Biographics AFMB, Marseille, France.
    24. 24
  24. Russell, R. J., J. M. Ferguson, D. W. Hough, M. J. Danson, and G. L. Taylor. 1997. The crystal structure of citrate synthase from the hyperthermophilic archaeon Pyrococcus furiosus at 1.9 Å resolution. Biochemistry 36:9983-9994.[CrossRef][Medline]
    25. 25
  25. Schliebs, W., N. Thanki, R. Jaenicke, and R. K. Wierenga. 1997. A double mutation at the tip of the dimer interface loop of triosephosphate isomerase generates active monomers with reduced stability. Biochemistry 36:9655-9662.[CrossRef][Medline]
    26. 26
  26. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.[Abstract/Free Full Text]
    27. 27
  27. Tozawa, K., H. Ohbuchi, H. Yagi, T. Amano, T. Matsui, M. Yoshida, and H. Akutsu. 1996. Unusual pKa of the carboxylate at the putative catalytic position of the thermophilic F1-ATPase subunit determined by 13C-NMR. FEBS Lett. 397:122-126.[CrossRef][Medline]
    28. 28
  28. Wiechelman, K., R. Braun, and J. Fitzpatrick. 1988. Investigation of the bicinchoninic acid protein assay: identification of the groups responsible for color formation. Anal. Biochem. 175:231-237.[CrossRef][Medline]
    29. 29
  29. Xu, D., C. J. Tsai, and R. Nussinov. 1998. Mechanism and evolution of protein dimerization. Protein Sci. 7:533-544.[Medline]
    30. 30
  30. Yamamoto, K., G. Kurisu, M. Kusunoki, S. Tabata, I. Urabe, and S. Osaki. 2001. Crystal structure of glucose dehydrogenase from Bacillus megaterium IWG3 at 1.7 Å resolution. J. Biochem. (Tokyo) 129:303-312.[Abstract/Free Full Text]

Applied and Environmental Microbiology, June 2005, p. 3285-3293, Vol. 71, No. 6
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.6.3285-3293.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Vazquez-Figueroa, E., Yeh, V., Broering, J.M., Chaparro-Riggers, J.F., Bommarius, A.S. (2008). Thermostable variants constructed via the structure-guided consensus method also show increased stability in salts solutions and homogeneous aqueous-organic media. Protein Eng Des Sel 21: 673-680 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.
Agricola
Right arrow Articles by Baik, S.-H.
Right arrow Articles by Harayama, S.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»