Altering the substrate specificity of glutamate dehydrogenase from Bacillus subtilis by site-directed mutagenesis

Article abstract of DOI:10.1271/bbb.69.1802

The Lys80, Gly82 and Met101 residues of glutamate dehydrogenase from Bacillus subtilis were mutated into a series of single mutants. The wild-type enzyme was highly specific for 2-oxoglutarate, whereas G82K and M101S dramatically switched to increased specificity for oxaloacetate with k _cat values 3.45 and 5.68s^-1, which were 265-fold and 473-fold higher respectively than those for 2-oxoglutarate.

Full text of DOI:10.1271/bbb.69.1802

Biosci. Biotechnol. Biochem., 69 (9), 1802–1805, 2005
Note
Altering the Substrate Specificity of Glutamate Dehydrogenase
from Bacillus subtilis by Site-Directed Mutagenesis
Md. Iqbal Hassan KHAN,1 Hyeung KIM,1 Hiroyuki ASHIDA,2
Takahiro ISHIKAWA,1 Hitoshi SHIBATA,1 and Yoshihiro SAWA
y
1;
1Department of Life Science and Biotechnology, Faculty of Life and Environmental Science,
Shimane University, 1060 Nishikawatsu, Matsue, Shimane 690-8504, Japan
2Department of Molecular and Functional Genomics, Center for Integrated Research in Science,
Shimane University, 1060 Nishikawatsu, Matsue, Shimane 690-8504, Japan
Received April 18, 2005; Accepted May 31, 2005
The Lys80, Gly82 and Met101 residues of glutamate
dehydrogenase from Bacillus subtilis were mutated into
a series of single mutants. The wild-type enzyme was
highly specific for 2-oxoglutarate, whereas G82K and
M101S dramatically switched to increased specificity for
oxaloacetate with kcat values 3.45 and 5.68 sÀ1, which
were 265-fold and 473-fold higher respectively than
those for 2-oxoglutarate.
(Bs-GluDH), the "-amino group of Lys80 (Cs-GluDH
Lys89) and side chain hydroxyl group of Ser358 (Cs-
GluDH Ser380) form salt bridge and hydrogen bond
interactions respectively with the ꢀ-carboxylic group of
L-glutamate (Fig. 1). The three-dimensional structure
model information on Bs-GluDH suggests that apart
from those residues, Gly82 and Met101 residues lie
close to the binding site for 2-OG in Bs-GluDH (Fig. 1).
Residue Gly82 is conserved in the GluDH family
whereas the equivalent residue in Cs-GluDH is Gln110
for Met101. In this study, which aimed at altering
the substrate specificity of Bs-GluDH from 2-OG to
oxaloacetate (OAA), we chose K80, G82, and M101
residues and constructed a series of single mutant
enzymes that form electrostatic interactions or hydrogen
bonds with oxygen atoms of the ꢁ-carboxylic group of
OAA. Here we report that among the constructed single
mutant enzymes, G82K and M101S dramatically shifted
their specificity and showed preferentially good activity
for OAA, the probable substrate for L-aspartate produc-
tion in the reductive amination reaction.
The rocG gene encoding NADþ-dependent Bs-
GluDH has been cloned and expressed at considerable
magnitude in Escherichia coli.8) Site-directed muta-
genesis was introduced into the rocG gene with a Mutan
Super Express kit (Takara, Ohtsu, Japan).9) The dideoxy-
nucleotide chain-termination method was employed for
sequencing using an ABI PRISM 3100-Avant Genetic
Analyzer (Boston, MA, U.S.A.) and sequence data were
analyzed using GENETYX-SV/RC 7.0 software (Soft-
ware Development, Tokyo, Japan). The mutant enzymes
and synthetic primers used were K80R, 50-CGGTCC-
GACCGTGGGGGGCGT-30, G82K, 50-CAAAGGGG-
AAAGTTCGCTT-30; G82R, 50-CAAAGGGGCGTGT-
TCGCTTCC-30, M101K, 50-CCATTTGGAAAACGCT-
CAA-30, and M101S 50-CCATTTGGAGTACGCTCA-
A-30. The transformants were grown at 37 ꢀC for 20 h in
Key words: glutamate dehydrogenase; Bacillus subtilis;
substrate specificity
Glutamate dehydrogenase (GluDH) is one of the
enzymes that offer information concerning enzymolog-
ical properties and relationships between structure and
function. The extremely small equilibrium constant of
GluDH allows it to act as a useful catalyst in the analysis
of amino acids, 2-oxo acids, and ammonia, which are
important tools in clinical chemistry, bioprocess control,
and nutrition studies.1) One of the major goals of our
studies has been to understand the structural basis of
substrate specificity in GluDH from Bacillus subtilis
(Bs-GluDH), and to apply that knowledge to the
engineering of novel substrate specificities. Among the
amino acid dehydrogenase super family, the amino acid
specificity of GluDH, which catalyzes the reversible
oxidative deamination of L-glutamate to 2-oxoglutarate
(2-OG) and ammonia, is relatively strict.2,3)
Analysis of enzyme homology with other GluDHs has
shown that the residues conserved among the GluDH
families, regardless of their quaternary structure, are
clustered in the three-dimensional structure, emphasiz-
ing the remarkably similar layout of the glutamate
binding site and the active site pocket.4,5) In the context
of crystal structure information on the Clostridium
symbiosum GluDH–glutamate complex (Cs-GluDH),6,7)
in the equivalent position of GluDH from B. subtilis
y
To whom correspondence should be addressed. Fax: +81-852-32-6092; E-mail: ysawa@life.shimane-u.ac.jp
Abbreviations: GluDH, glutamate dehydrogenase; 2-OG, 2-oxoglutarate; IPTG, isopropyl thio-ꢁ-D-galactoside; OAA, oxaloacetate

Altering Substrate Specificity of Bs-GluDH
1803
Fig. 1. The Active Site of Cs-GluDH in Complex with L-Glutamate (PDB ID: 1BGV).
The equivalent positions of Bs-GluDH shown in parentheses were determined from the 3D alignment of modeled BsGluDH with that of
CsGluDH. Dotted lines indicate electrostatic and hydrogen bonding interactions.
Table 1. Substrate Specificity of Single Mutant Enzymes in Reductive Amination Reaction
2-Oxo acids
2-OG
Wild typea
G82Ka
G82Ra
K80Ra
M101Ka
M101Sa
100
(440b)
0.23
0.11
6.00
0.05
2.70
0.10
NDc
0.05
100
(0.026b)
28,000
2,270
1,080
731
100
(0.019b)
NDc
91
100
(0.014b)
208
100
(0.012b)
117
126
69
100
(0.019b)
49,500
1,530
2,050
132
OAA
Pyruvate
216
188
ꢁ-Phenylpyruvate
2-Oxo-3-methylvalerate
2-Oxo-iso-valerate
2-Oxo-iso-caproate
2-Ketohexonoate
p-Hydroxyphenylpyruvate
145
129
225
408
215
96
26
73
356
213
289
236
191
279
54
1,850
3,430
502
329
129
31
432
121
182
aRelative activity: One Hundred percent corresponds to the specific activity for 2-OG of each enzyme.
bSpecific activity: unitsꢂmg proteinꢃ1
.
cND, not detectable.
LB medium containing 0.005% ampicillin and 1 mM
IPTG. The cells were suspended in potassium phosphate
(KP) buffer, pH 7.3, and subjected to French press
(Ohtake, Tokyo, Japan) two times at a pressure of
1500 kg/cm2. After ultracentrifugation (181;000 ꢁ g,
1 h), the recombinant enzyme was purified from the
supernatant according to a procedure used for the native
enzyme.8) In each step of purification, expression and
purity were checked by SDS–PAGE (12%) under the
conditions described by Laemmli.10) All mutant en-
zymes were purified to homogeneity, and exhibited a
single band with subunit Mr similar to that of wild-type
enzyme (Mr 46,000) on SDS–PAGE (data not shown).
Homology models for the 3D structures of the wild-type
and mutant enzymes were built using the same method.8)
Table 1 shows the substrate specificity of the wild-
type and mutant enzymes on a number of 2-oxo acids as
their relative activities as well as the specific activities
for 2-OG. The wild-type enzyme exhibited negligible
activity on OAA as well as on pyruvate in relation to its
considerably higher activity on 2-OG. On the other
hand, among the five single mutant enzymes, G82K and
M101S had 280-fold and 495-fold higher preferences
respectively for OAA over 2-OG, while those enzymes
showed very low specific activities for 2-OG compared
with that of the wild-type enzyme. Those mutant
enzymes took pyruvate and aromatic 2-oxoacids as the
substrates of second-order preference. G82K and M101S
showed 23-fold and 15-fold higher reactivity respec-
tively for pyruvate, and 34-fold and 21-fold higher
reactivity respectively for aromatic 2-oxoacid than their
activities for 2-OG. Thus the mutation at those ascribed
positions somewhat broadened the specificity of G82K
and M101S. But, regardless of the context of relative
activity, the specific activities of mutant enzymes
compared to the wild-type on the various 2-oxo acids
listed in Table 1 were not drastically altered as to those
of their magnitude of activities for OAA from 2-OG.
Thus, the preferences for OAA of both mutant enzymes,
G82K and M101S were distinctly superior to those for
other 2-oxo acids.
The reactivities of three other single mutant enzymes,
G82R, K80R, and M101K, were not found to be
significant compared to those of G82K and M101S on
the 2-oxo acids listed in Table 1. Because only G82K
and M101S showed substantially high reactivity on

1804
Md. I. H. KHAN et al.
Table 2. Steady-State Kinetics Parameters of Mutant Enzymes in Reductive Amination Reaction
ꢃ1
)
Km (mM)
NH4Cl
kcat (sꢃ1
)
kcat=Km (sꢃ1ꢂmM
2-OG
Enzyme
2-OG
OAA
NADH
2-OG
OAA
OAA
Wild type
G82K
M101S
0.65
>100
>100
NDa
4.16
2.27
55.1
125
135
0.07
0.095
0.091
344
0.013
0.012
NDa
3.45
5.68
529
<10ꢃ5
<10ꢃ5
NDa
0.829
2.50
aND, Not detectable.
OAA, we determined the kinetic parameters of those
two mutants. The steady-state kinetic parameters of the
wild-type and the mutant enzymes are shown in Table 2.
G82K showed a kcat value of 3.45 sꢃ1 for OAA
compared to its kcat value of 0.013 sꢃ1 for 2-OG. The
former value was 265 fold higher than the latter one, and
was 1.2% of that of the wild-type enzyme for 2-OG. The
Km values of G82K for OAA and NADH were 4.16 and
0.095 mM respectively, whereas for 2-OG the Km was
greater than 100 mM indicating that the G82K mutant
showed binding affinity good enough to endorse dehy-
drogenation catalysis of OAA but not of 2-OG.
Similarly, the kcat value of M101S for OAA was
5.68 sꢃ1, which is 1.7% of that of wild-type enzyme
for 2-OG, whereas the kcat value for 2-OG was
0.012 sꢃ1, showing that about a 473-fold higher catalytic
shift had been achieved for OAA. The Km values of
M101S for OAA, NADH, and 2-OG were 2.27, 0.091,
and greater than 100 mM respectively. Therefore, the
estimated catalytic efficiency, kcat=Km values of G82K
that Met101 is close to the ꢁ-carboxyl group of OAA
and that the hydroxyl group of Ser101 therefore makes a
hydrogen bond with that of one ꢁ-carboxylate oxygen of
OAA, and thus facilitates binding and catalysis. The
most evident means of achieving the changed specificity
of the substrate is to change the substrate binding
residues without disturbing the catalytic residues ex-
clusively involved in catalytic chemistry. G82K and
M101S, however, still display low kcat values for OAA
in relation to the figure for the wild-type enzyme with its
natural substrate, 2-OG. It might be more fruitful to
combine molecular evolutionary engineering with spe-
cific selection to achieve further subtle improvements in
catalytic capability. In addition, we are making progress
in resolving the crystal structure of these single mutant
enzymes and hope to identify the details of the structural
basis of their substrate specificities to OAA.
Acknowledgments
ꢃ1
and M101S for OAA were 0.829 and 2.50 sꢃ1ꢂmM
This work was supported in part by a Grant-in-Aid
for Scientific Research (C) (no. 15580079) from the
Ministry of Education, Culture, Sports, Science and
Technology of Japan.
respectively. Although these kcat=Km values were 0.16
and 0.47% of that of wild-type enzyme for 2-OG
(529 sꢃ1ꢂmMꢃ1), the use of a large amount of mutant
enzymes gave comparable activities of aspartate dehy-
drogenase (AspDH). Furthermore, in the oxidative
deamination reactions, the single mutants G82K and
M101S showed specific activity on L-aspartate 0.39 and
0.21 unitꢂmgꢃ1 respectively. Hence the NADPþ-type
AspDH from archaea, the only one found in nature, has
been reported to show low Kcat 4.9 sꢃ1 even at temper-
atures as high as 70 ꢀC.11)
References
1) Ohshima, T., and Soda, K., Biochemistry and biotech-
nology of amino acid dehydrogenases. Adv. Biochem.
Eng. Biotechnol., 42, 187–189 (1990).
2) Olson, J. A., and Anfinsen, C. B., Kinetic and equilib-
rium studies on crystalline L-glutamic acid dehydrogen-
ase. J. Biol. Chem., 202, 841–856 (1953).
The differences in substrate specificity can be ascrib-
ed to the amino acid residues that line in the active site
pocket; very small changes that orient active site
residues with that of its native substrate can play a
large role in substrate binding and catalysis. The 3D
structures of the G82K-OAA, G82K-2-OG, M101S-
OAA, and M101S-2OG complexes were determined by
using the MOE 2004.03 docking program. Because
OAA is a smaller molecule than 2-OG, the G82 residue,
which forms the active site pocket, is assumed to be
closer to the ꢁ-carboxylic group of OAA than to the ꢀ-
carboxylic group of 2-OG. Conversion by lysine residue
might ensure binding of the "-amino group of Lys82
with one of the ꢁ-carboxylate oxygens of OAA
effectively through a salt bridge. The possibility of
M101S forming complex with OAA, similarly, suggests
3) Struck, J., and Sizer, I. W., The substrate specificity of
glutamate dehydrogenase. Arch. Biochem. Biophys., 86,
260–266 (1960).
4) Britton, K. L., Baker, P. J., Engel, P. C., Rice, D. W., and
Stillman, T. J., Evolution of substrate diversity in the
superfamily of amino acid dehydrogenases: Prospects
for rational chiral synthesis. J. Mol. Biol., 234, 938–945
(1993).
5) Britton, K. L., Baker, P. J., Rice, D. W., and Stillman,
T. J., Structural relationship between the hexameric and
tetrameric family of glutamate dehydrogenases. Eur. J.
Biochem., 209, 851–859 (1992).
6) Stillman, T. J., Baker, P. J., Britton, K. L., and Rice,
D. W., Conformational flexibility in glutamate dehydro-
genase: role of water in substrate recognition and
catalysis. J. Mol. Biol., 234, 1131–1139 (1993).
7) Baker, P. J., Britton, K. L., Engel, P. C., Farrants, G. W.,

Altering Substrate Specificity of Bs-GluDH
1805
Lilley, K. S., Rice, D. W., and Stillman, T. J., Subunit
assembly and active site location in the structure of
glutamate dehydrogenase. Proteins, 12, 75–86 (1992).
8) Khan, M. I. H., Ito, K., Kim, H., Ashida, H., Ishikawa,
T., Shibata, H., and Sawa, Y., Molecular properties and
enhancement of thermostability by random mutagenesis
of glutamate dehydrogenase from Bacillus subtilis.
Biosci. Biotechnol. Biochem., in press.
dual amber method for site-directed mutagenesis. Gene,
152, 271–275 (1995).
10) Laemmli, U. K., Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature,
227, 680–685 (1970).
11) Yang, Z., Savchenko, A., Yakunin, A., Zhang, R.,
Edwards, A., Arrowsmith, C., and Tong, L., Aspartate
dehydrogenase, a novel enzyme identified from struc-
tural and functional studies of TM1643. J. Biol. Chem.,
278, 8804–8808 (2003).
9) Hashimoto-Gotoh, T., Mizuno, T., Ogasahara, Y., and
Nakagawa, M., An oligodeoxyribonucleotide-directed