Determinants of dual substrate specificity revealed by the crystal structure of homoisocitrate dehydrogenase from Thermus thermophilus in complex with homoisocitrate·Mg2+·NADH

Article abstract of DOI:10.1016/j.bbrc.2016.09.004

HICDH (Homoisocitrate dehydrogenase) is a member of the β-decarboxylating dehydrogenase family that catalyzes the conversion of homoisocitrate to α-ketoadipate using NAD⁺ as a coenzyme, which is the fourth reaction involved in lysine biosynthesis through the α-aminoadipate pathway. Although typical HICDHs from fungi and yeast exhibit strict substrate specificities toward homoisocitrate (HIC), HICDH from a thermophilic bacterium Thermus thermophilus (TtHICDH) catalyzes the reactions using both HIC and isocitrate (IC) as substrates at similar efficiencies. We herein determined the crystal structure of the quaternary complex of TtHICDH with HIC, NADH, and Mg²⁺ ion at a resolution of 2.5??. The structure revealed that the distal carboxyl group of HIC was recognized by the side chains of Ser72 and Arg85 from one subunit, and Asn173 from another subunit of a dimer unit. Model structures were constructed for TtHICDH in complex with IC and also for HICDH from Saccharomyces cerevisiae (ScHICDH) in complex with HIC. TtHICDH recognized the distal carboxyl group of IC by Arg85 in the model. In ScHICDH, the distal carboxyl group of HIC was recognized by the side chains of Ser98 and Ser108 from one subunit and Asn208 from another subunit of a dimer unit. By contrast, in ScHICDH, which lacks an Arg residue at the position corresponding to Arg85 in TtHICDH, these residues may not interact with the distal carboxyl group of shorter IC. These results provide a molecular basis for the differences in substrate specificities between TtHICDH and ScHICDH.

Full text of DOI:10.1016/j.bbrc.2016.09.004

Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications
journal homepage: www.elsevier.com/locate/ybbrc
Determinants of dual substrate specificity revealed by the crystal
structure of homoisocitrate dehydrogenase from Thermus
2
þ
thermophilus in complex with homoisocitrate$Mg $NADH
*
Kento Takahashi, Takeo Tomita, Tomohisa Kuzuyama, Makoto Nishiyama
Biotechnology Research Center, The University of Tokyo, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
HICDH (Homoisocitrate dehydrogenase) is a member of the
b
-decarboxylating dehydrogenase family
Received 30 August 2016
Accepted 1 September 2016
Available online 4 September 2016
þ
that catalyzes the conversion of homoisocitrate to
a-ketoadipate using NAD as a coenzyme, which is the
fourth reaction involved in lysine biosynthesis through the
a
-aminoadipate pathway. Although typical
HICDHs from fungi and yeast exhibit strict substrate specificities toward homoisocitrate (HIC), HICDH
from a thermophilic bacterium Thermus thermophilus (TtHICDH) catalyzes the reactions using both HIC
and isocitrate (IC) as substrates at similar efficiencies. We herein determined the crystal structure of the
Keywords:
b-decarboxylating dehydrogenase
2þ
quaternary complex of TtHICDH with HIC, NADH, and Mg ion at a resolution of 2.5 Å. The structure
Homoisocitrate dehydrogenase
Crystal structure
Thermus thermophilus
revealed that the distal carboxyl group of HIC was recognized by the side chains of Ser72 and Arg85 from
one subunit, and Asn173 from another subunit of a dimer unit. Model structures were constructed for
TtHICDH in complex with IC and also for HICDH from Saccharomyces cerevisiae (ScHICDH) in complex
with HIC. TtHICDH recognized the distal carboxyl group of IC by Arg85 in the model. In ScHICDH, the
distal carboxyl group of HIC was recognized by the side chains of Ser98 and Ser108 from one subunit and
Asn208 from another subunit of a dimer unit. By contrast, in ScHICDH, which lacks an Arg residue at the
position corresponding to Arg85 in TtHICDH, these residues may not interact with the distal carboxyl
group of shorter IC. These results provide a molecular basis for the differences in substrate specificities
between TtHICDH and ScHICDH.
©
2016 Elsevier Inc. All rights reserved.
1. Introduction
lysine biosynthesis via a-aminoadipate [6]. Since homologous en-
zymes are considered to have evolved from a common ancestral
enzyme into specialized enzymes [7e11], comparisons of recogni-
tion mechanisms between each enzyme provide useful information
on how these homologous enzymes acquired the separate func-
tions in the course of evolution.
Homoisocitrate dehydrogenase (HICDH) is an enzyme involved
in the lysine biosynthetic pathway ( -aminoadipate pathway) of
yeast, fungi, and some kinds of bacteria and archaea [1e3]. HICDH
a
catalyzes the decarboxylating dehydrogenation of homoisocitrate
(
HIC) into
a-ketoadipate (Fig. 1). HICDH is a member of the
b
-
Lysine is known to be synthesized via the diaminopimelate
decarboxylating dehydrogenase family, which consists of 3-
isopropylmalate dehydrogenase [IPMDH; EC 1.1.1.85] for leucine
biosynthesis [4], isocitrate dehydrogenase [ICDH; EC 1.1.1.42] for
the tricarboxylic acid (TCA) cycle [5], and HICDH [EC 1.1.1.87] for
(DAP) pathway in bacteria, and via the a-aminoadipate (AAA)
pathway in fungi and yeast. We previously found that Thermus
thermophilus, which is an extremely thermophilic bacterium,
synthesized lysine via AAA [12], cloned the biosynthetic gene
cluster, and characterized most of the enzymes involved in the
pathway [13e19]. Some of these enzymes catalyze not only the
reaction in lysine biosynthesis but also those of related pathways
such as branched-chain amino acid biosynthesis, the TCA cycle,
Abbreviations: IPMDH, 3-isopropylmalate dehydrogenase; ICDH, isocitrate de-
hydrogenase; HICDH, homoisocitrate dehydrogenase; KEGG, Kyoto Encyclopedia of
Genes and Genomes; IPM, 3-isopropylmalate; IC, isocitrate; HIC, homoisocitrate;
EDTA, Ethylenediaminetetraacetic acid.
and/or arginine
biosynthesis
[13e15,17].
HICDH
from
T. thermophilus (TtHICDH) is an enzyme that recognizes HIC and IC
as substrates [16] (Fig. 1). Since ancestral enzymes are considered
to have been promiscuous and capable of catalyzing reactions for
*
Corresponding author. Biotechnology Research Center, The University of Tokyo,
1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan.
E-mail address: umanis@mail.ecc.u-tokyo.ac.jp (M. Nishiyama).
http://dx.doi.org/10.1016/j.bbrc.2016.09.004
006-291X/© 2016 Elsevier Inc. All rights reserved.
0

K. Takahashi et al. / Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
1689
multiple related reactions, TtHICDH may retain the characteristic
2.2. Enzyme assay
features of its ancestral enzyme. By contrast, HICDHs from yeast,
Saccharomyces cerevisiae (ScHICDH) and Schizosaccharomyces
pombe (SpHICDH), distinguish HIC from IC and exhibit strict
In order to examine the substrate specificity of TtHICDH, we
performed an enzyme assay using HIC, and IC as substrate candi-
þ
specificity toward HIC [20,21]. The substrate specificities of
decarboxylating dehydrogenase family members appeared to be
determined by a few amino acid residues in the loop 3- 4 and
the N-terminal region of the following 4 helix [21]. Substrate
b
-
dates, with NAD as a coenzyme. The activities of the enzymes
ꢂ
were measured at 60 C by monitoring increases in absorbance at
ꢁ1
ꢁ1
b
a
340 nm (ε340 ¼ 6.22 mM cm ) of NADH. The enzyme reaction
ꢁ
ml of enzyme solution (0.5 mg ml ) to
1
a
was initiated by adding 10
990 l of reaction mixture (50 mM HEPES-NaOH, pH 8.0, 200 mM
KCl, and 5 mM MgCl
specificity was converted by amino acid replacement in the cor-
responding regions of IPMDH and ICDH. We also showed that
m
2
) that contained 5 mM substrate and 2 mM
ꢂ
Arg85 on the
a
4 helix of TtHICDH is a crucial determinant for
coenzyme and was pre-incubated at 60 C for 5 min. One unit of
substrate specificity and plays a key role in the dual functions of
TtHICDH as HICDH and ICDH in a site-directed mutagenesis study
enzyme activity was defined as the amount of the enzyme that
ꢂ
produced 1
mmol of NADH per min at 60 C in the reaction.
[
16]. Although determinants of substrate specificity have been
In order to determine the kinetic constants of TtHICDH, the
reactions for HIC and IC were preformed in the presence of NAD .
The concentrations of substrates ranged between 10 and 500 mM
þ
proposed for TtHICDH and SpHICDH based on crystallographic
analyses coupled with site-directed mutagenesis, the molecular
mechanisms underlying substrate recognition by HICDHs has not
yet been elucidated due to the lack of substrate-bound structures
in most cases [22e24]. In the present study, we determined the
crystal structure of a quaternary complex of TtHICDH binding HIC,
for HIC and IC in the reaction mixture with a fixed concentration of
þ
NAD (2 mM). Kinetic constants were calculated using Cleland's
initial velocity program Hyper [25].
NADH, and Mg2 , which revealed the structural feature of
TtHICDH exhibiting dual substrate specificity, and also provided
structural basis that defined the strict substrate preference of
ScHICDH for HIC using its modeled structure.
þ
2.3. Crystallization
Crystallization conditions were screened by the hanging drop
vapor diffusion method using the screening kits, Crystal Screen
(
Hampton Research, USA) and Wizard (Emerald BioSystems, USA).
ꢁ1
One microliter of 10 mg ml TtHICDH solution supplemented with
5 mM MgSO , 10 mM HIC, and 10 mM NADH was added to an equal
2
. Experimental procedures
4
volume of the reservoir solutions. The mixtures were equilibrated
ꢂ
2.1. Protein preparation
against 500
m
l of reservoir solutions at 20 C. Crystals of the
2
þ
TtHICDH$HIC$Mg $NADH complex were obtained with Crystal
We constructed an expression system for TtHICDH as follows.
Screen II (No.42) containing 0.1 M Tris-HCl, pH8.5, 1.5 M (NH
and 12% glycerol.
4 2 4
) SO ,
The plasmid pET-tHICDH101 for the expression of TtHICDH which
was constructed in our previous study [16] was introduced into
Escherichia coli BL21(DE3)CodonPlus-RIL cells. E. coli cells harboring
pET-tHICDH101 were grown in 2 ꢀ YT medium (1.6% tryptone, 1%
yeast extract, and 0.5% NaCl) supplemented with kanamycin
2.4. Data collection, structural determination, and refinement
Prior to data collection, crystals were briefly soaked in a series of
reservoir solutions finally supplemented with 26% glycerol (v/v) as
a cryoprotectant by increasing the concentrations of the cryopro-
ꢁ ꢁ1
1
ꢂ
(
50
tical density of approximately 0.6 at 600 nm. Gene expression was
induced by adding isopropyl -thiogalactopyranoside (IPTG) at a
mg ml ) and chloramphenicol (30 mg ml ) at 37 C to an op-
b
-D
2
tectant stepwise by 4%. Crystals were then flash-frozen in a N gas
final concentration of 1.0 mM and the culture was continued at
stream at 95 K. Diffraction data sets were collected at the AR-NW12
station of the Photon Factory, High Energy Accelerator Research
Organization (KEK) (Tsukuba, Japan). Diffraction images were
indexed, integrated, and scaled using the HKL-2000 program suite
[26]. Data collected at a wavelength of 1.0 Å were used for subse-
quent molecular replacement and crystallographic refinement.
Molecular replacement was performed with Phaser [27] in the
CCP4 program suite [28] using the model of a monomer of the apo
form of TtHICDH, which is deposited in the RCSB PDB with acces-
sion number 1X0L, as a search model. Model correction in the
electron density map was performed with Coot [29]. Refinement
was performed with Refmac 5.2 [30]. The final model contained
four polypeptide chains in an asymmetric unit. Data collection,
refinement statistics and the results of Ramachandran plots pro-
duced by the program PROCHECK [31] are summarized in Table 1.
Figures were prepared using PyMOL [32]. The atomic coordinates
have been deposited in the RCSB PDB with accession number 4YB4.
ꢂ
37 C for an additional 12 h. Cells were harvested by centrifugation
at 4000 ꢀ g for 10 min, washed with buffer A (20 mM Tris-HCl pH
8
.0, 0.5 mM EDTA), and centrifuged in the same manner. Cells were
suspended in buffer A supplemented with 150 mM NaCl, disrupted
by sonication, and centrifuged at 35,000 ꢀ g for 20 min. The su-
ꢂ
pernatant was heat-treated at 85 C for 20 min to denature the
proteins from E. coli. The precipitate was removed by centrifugation
at 35,000 ꢀ g for 20 min and the supernatant was applied to an
anion exchange column, DE52 (Whatman, England), which had
been equilibrated with buffer A. Since most TtHICDH was not
adsorbed to the column, the passing fractions were pooled.
Ammonium sulfate was added to the pooled fraction at a final
saturation of 45%, and precipitated proteins collected by centrifu-
gation at 35,000 ꢀ g for 30 min were dissolved in buffer A. The
protein solution was applied to a HiLoad 16/10 Phenyl Sepharose
column (GE Healthcare, USA) equilibrated with buffer A supple-
mented with 1.5 M ammonium sulfate. The adsorbed proteins were
eluted with a linear gradient of 1.5e0 M ammonium sulfate.
TtHICDH-containing fractions were concentrated by the ultrafil-
tration spin column, Vivaspin 10,000 MWCO (Sartorius Stedim
Biotech GmbH, Germany), and applied to the gel filtration chro-
matography column, HiLoad 26/60 Superdex 200 (GE Healthcare,
USA), equilibrated with buffer A supplemented with 150 mM NaCl.
The active fractions that showed a single band on SDS-PAGE were
collected and used for enzyme assays and crystallization.
2.5. Modeling of TtHICDH in complex with IC and ScHICDH in
complex with HIC
A model of TtHICDH in complex with IC was constructed using
the structure of the quaternary complex of TtHICDH as a template.
IC was docked manually to superpose its malate moiety to that of
HIC. A standard molecular dynamics cascade was subsequently
performed by Discovery studio 4.0 (Accelrys) using CHARMm as a

1690
K. Takahashi et al. / Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
Table 1
COOH
COOH
COOH
X-ray data collection and refinement statistics.
TtHICDH$HIC$Mg2 $NADH
COOH
HOOC
þ
HOOC
OH
Homoisocitrate
(HIC)
OH
Isocitrate
(IC)
Data collection
Beamline
Wavelength (Å)
Space group
Cell dimensions
a (Å)
AR-NW12
1.0
P4 2 2
3 1
+
+
NAD
NAD
159.7
159.7
148.5
2.50 (2.50e2.56)
485,173
B (Å)
c (Å)
Resolution (Å)a
CO2
NADH
CO2
NADH
Total reflections
Unique reflections
60,576
COOH
HOOC
COOH
a
Completeness (%)
95.1 (96.3)
18.7 (3.9)
11.1 (45.3)
HOOC
a
Average I/ s(I)
a,b
O
O
Rmerge (%)
α-Ketoadipate
α-Ketoglutarate
Refinement
Resolution (Å)
112.90e2.50
15.3/20.5
10,076
12
Fig. 1. Reaction of TtHICDH. The HICDH and ICDH reactions are shown on the left and
right sides, respectively.
R-factor/Rfree (%)
No. of protein atoms
No. of ligands
No. of water molecules
Average B-factor (Å )
318
2
crystal structure of TtHICDH was also determined as a form binding
a designed inhibitor, (2S, 3S)-thiahomoisocitrate [23]. This
inhibitor-bound structure was similar to the apo-TtHICDH struc-
ture, with the former structure being in a slightly more closed state
than the latter. The inhibitor-bound structure exhibited RMSDs for
Protein
HIC
32.5
27.5
29.1
37.3
32.8
2
þ
Mg
SO
2-
4
Waters
C
a
of 0.7
Å
and 1.6
Å
for the apo-TtHICDH and
R.m.s.d from ideal values
Bond length (Å)
Bond angles (deg.)
Ramachandran plot (%)
Favored
2þ
0.009
1.5
TtHICDH$HIC$Mg $NADH structures, respectively. These results
indicate that TtHICDH took a fully closed conformation that was
more closed than the previously published structure of the binary
complex TtHICDH-(2S, 3S)-thiahomoisocitrate, due to a marked
96.2
3.3
0.5
Allowed
Disallowed
2
þ
domain closure induced upon the binding of HIC$Mg and the
2þ
a
coenzyme. The electron densities of HIC, Mg , and NADH were
clearly observed in the cleft of each subunit (Fig. 2C). The distance
between the C4N atom of the nicotinamide ring of NADH and the
C2 atom of HIC was 3.5 Å, which was suitable for the transfer of
hydride between them. Thus, the quaternary complex mimicked
the enzymatically competent structure. The arrangement of these
ligands was very similar to that of the quaternary complex of the
IPM$Mg $NADH complex of IPMDH from T. thermophilus
(TtIPMDH) [34]. In TtIPMDH, Glu270 binds a K ion at the active site
and plays an essential role in stimulating activity by the K ion [35].
Values in parentheses are for the highest resolution shell.
b
R
merge ¼ S jI ꢁ〈I〉j/ S〈I〉.
i
force field [33]. A model of ScHICDH in complex with HIC was also
constructed using the structure of the quaternary complex of
TtHICDH as a template. Homology modeling of ScHICDH was per-
formed by Discovery studio 4.0 using CHARMm as a force field. HIC
was docked in the active site of the model of ScHICDH in order to
achieve a similar conformation to HIC in TtHICDH. A standard
molecular dynamics cascade was then performed in a similar
manner to that of the IC complex of TtHICDH.
2
þ
þ
þ
In TtHICDH, Glu257 equivalent to Glu270 in TtIPMDH was found in
the same position in the active site; however, a water molecule, but
þ
not a K ion was found at this position, which may have been due to
þ
the lack of K ions in the crystallization buffer.
3
. Results and discussion
NAD(H) recognition of TtHICDH e NADH was added to the protein
solution for crystallization. NADH was found in the cleft between
domain 1 and domain 2 (Fig. 2B), as observed in the quaternary
complex of TtIPMDH. NADH formed specific contacts with TtHICDH
2
þ
Crystal structure of TtHICDH in complex with HIC, NADH, and Mg
2
þ
e The crystal structure of TtHICDH complexed with HIC, Mg , and
2
þ
NADH (TtHICDH$HIC$Mg $NADH) was determined by molecular
replacement using the apo-form of TtHICDH (PDB ID 1X0L) at a
resolution of 2.5 Å. There were four monomers in the crystallo-
graphic asymmetric unit (Fig. 2A). Models were built from Ala2 to
Leu334 for each chain. The structure of the four chains were very
similar to each other with a root mean square deviation (RMSD) of
0 0
Fig. 3A). The 2 - and 3 - OHs of the adenine ribose of NADH formed
(
hydrogen bonds with Asp265 conserved among HICDHs, which
may serve as a determinant for the preference of HICDH family
þ
þ
members for NAD to NADP .
HIC recognition of TtHICDH e In the quaternary complex of
TtHICDH, the basic residues, Arg88, Arg96, Arg118, Tyr125, and
Lys171*(asterisks indicate residues from another subunit in a
dimer), recognized the malate moiety of HIC. Asp204*, Asp228,
approximately 0.2 Å for C
a between the chains. Similar to apo-
TtHICDH, TtHICDH formed a tetrameric structure composed of
two dimers interacting with each other head-to-head. A monomer
was composed of two domains: domain 1 (Ala2-Asp92 and Gly240-
Leu334) and domain 2 (Leu93-Gly239). The active site was located
in the cleft between two domains (Fig. 2B). RMSD for Ca was 1.9 Å
between TtHICDH$HIC$Mg $NADH and apo-TtHICDH; however,
when the RMSD was calculated for each domain, small RMSDs
1.0 Å for domain 1 and 0.7 Å for domain 2) were obtained. This was
attributed to the rigid-body rotation by 20 of domain 1 from
domain 2 upon the binding of HIC$Mg and NADH (Fig. 2B). The
2þ
Asp232, and water molecules bound a Mg ion in an octahedral
coordination manner similar to those of other substrate-bound
2
þ
structures such as the TtIPMDH$IPM$Mg $NADH complex (PDB
2
þ
ID 4F7I) (Fig. 3B). These amino acid residues were conserved in all
the enzymes belonging to
members, suggesting that
b
b
-decarboxylating dehydrogenase family
-decarboxylating dehydrogenase family
(
ꢂ
enzymes including TtHICDH share a common catalytic mechanism.
The amino acid residues responsible for the recognition of the
2þ
g
-

K. Takahashi et al. / Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
1691
Fig. 2. Structure of TtHICDH in complex with HIC, Mg2 , and NADH (A) The TtHICDH homo-tetramer with A, B, C, and D chains are shown in green, yellow, cyan, and magenta,
þ
respectively. (B) The TtHICDH monomer with the two structural domains shown with dotted lines. Ribbon models of the apo-form of TtHICDH (PDB ID 1X0L) and the quaternary
complex of TtHICDH are shown in purple and green, respectively. The rotation of domain 1 relative to domain 2 is shown by an arrow. (C) An omit difference (F
o
-F
level for the ligand bound to TtHICDH$HIC$Mg $NADH. The distance between the C2 atom of HIC and the C4N atom of the bound NADH is indicated.
D) Water molecules found at the catalytic site in the structure of IPMDH from T. thermophilus. The water molecule (Wat1) and Glu257, which play an essential role in stimulating
c
) electron density
2þ
map contoured at the 3.3-
s
(
þ
the activity of the K ion are underlined. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 3. Active site of TtHICDH. (A) NADH-binding site. (B) HIC-binding site. Hydrogen bonds and electrostatic interactions are shown as dotted lines. The distances of the bonds are
shown.

1692
K. Takahashi et al. / Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
moiety of substrates were mainly located on the
helix. The carboxyl group at the terminal of the
b
C-
g
a
4 loop and
-moiety (car-
a
4
been recognized by introduced Val85 which provided a hydro-
phobic environment to accept the -isopropyl group of IPM in this
g
boxyethyl group) of HIC was recognized by Arg85 with ionic bonds
and the aliphatic ethyl side chain of HIC was stabilized by a hy-
drophobic interaction with the Cd carbon of Arg85 (Figs. 3B and 4A).
pocket. These residues are conserved in HICDH from Deinococcus
radiodurans, which exhibits activity towards a broad range of sub-
strates, namely HIC, IC, and IPM [36]. We speculated that the side
chain conformation of Arg85 may be slightly altered in order to not
prevent IPM from binding in DrHICDH; however, this need to be
verified by a crystal structure analysis of DrHICDH.
In addition, Ser72 and Asn173* recognized the distal carboxyl group
via hydrogen bonds. TtHICDH exhibited similar kinetic parameters
for HIC and IC (Table 2). In order to elucidate the mechanism
responsible for the recognition of IC, we constructed a model of the
ScHICDH exhibits a strict preference for HIC [20,21].
2þ
IC complex of TtHICDH using the HIC$Mg $NADH complex of
TtHICDH as a model (Fig. 4B). In this model, the -moiety (meth-
In order to elucidate the mechanisms responsible for the strict
substrate specificities of HICDHs from yeast, we constructed a
model structure of the HIC complex of ScHICDH using the
g
ylcarboxyl group) of IC was oriented toward Arg85 and formed
bidentate ionic bonds with Arg85, while the methylcarboxyl group
was too short to be recognized by Ser72 or Asn173*. We previously
identified Arg85 as a determinant for the substrate specificity of
TtHICDH for IC based on the finding that TtHICDH with the
Arg85Val replacement lost its activity for IC, but retained that for
2þ
TtHICDH$HIC$Mg $NADH complex as the template. The con-
struction of a model around the 3- 4 loop of ScHICDH was not
b
a
difficult because the amino acid sequence around this region was
aligned well between TtHICDH and ScHICDH without a gap
sequence (Fig. 4). In the model of the HIC complex of ScHICDH, the
carboxyethyl group of HIC formed hydrogen bonds with Ser98,
Ser108, and Asn208*. Ser98 and Asn208* corresponded to Ser72
and Asn173* of TtHICDH, respectively, suggesting that these resi-
dues have the same functions as in TtHICDH. In ScHICDH, Arg85 in
2
þ
HIC [16]. The crystal structure of the TtHICDH$HIC$Mg $NADH
complex and its model binding IC definitively demonstrated that
Arg85 is a key residue for recognizing both of HIC and IC. These
results shows that Arg85, Ser72, and Asn173* are involved in the
recognition of HIC, while the model structure suggests that only
Arg85 is involved in the recognition of IC. It is important to note
that the TtHICDH variant with Arg85Val mutation exhibited activ-
ities for HIC and IPM. The present crystal structure provides an
insight into the mechanism underlying the dual substrate speci-
ficity of the HICDH-R85V variant. In the HICDH-R85V variant, the
distal carboxyl group of HIC was still recognized by Ser72 and
Table 2
Kinetic constants of TtHICDH.
Enzyme
HIC
IC
k
cat
Km
M]
k
[s
cat
ꢁ1
/Km
k
[s
cat
ꢁ1
Km
M]
k
[s
cat
ꢁ1
/Km
m
[sꢁ1]
[
m
m
M
ꢁ1
]
]
[
m
ꢁ1
M ]
Asn173* without Arg85, the
g-isopropyl group of IPM may have
TtHICDH 33 ± 1 210 ± 10 1.6 ꢀ 10ꢁ1 76 ± 1 290 ± 10 2.6 ꢀ 10ꢁ1
Fig. 4. Comparison of substrate-binding sites of TtHICDH. (A) The HIC-binding site of TtHICDH. (B) The IC-binding site of the model of the TtHICDH/IC complex. (C) The HIC-binding
site of the model of the ScHICDH/HIC complex. (D) Amino acid sequence alignment of TtHICDH and related enzymes. The amino acid sequences of TtHICDH, DrHICDH, SpHICDH, and
ScHICDH are aligned. The residues directly involved in the recognition of substrates in TtHICDH and ScHICDH are indicated as green and yellow circles, respectively. The secondary
structure of TtHICDH is shown at the upper side of the alignment. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article).

K. Takahashi et al. / Biochemical and Biophysical Research Communications 478 (2016) 1688e1693
1693
TtHICDH was replaced with Val111 (Fig. 4C). In the absence of the
Arg residue at the corresponding position, ScHICDH had a Ser
residue at 108 in order to stabilize HIC binding. However, all the
residues with the potential to bind to the distal carboxyl group of
HIC were located at distance that were too far to make contact with
the carboxyl group of the distal end of IC. This is the most important
difference between TtHICDH and ScHICDH. Val111 in ScHICDH,
which corresponded to Arg85 in TtHICDH in the alignment of
amino acid sequences, interacted with the hydrophobic part of the
aliphatic carboxyethyl chain of HIC and may play a role in pre-
venting the carboxymethyl group of IC from binding.
bifunctional hyperthermostable carboxypeptidase/aminoacylase from Pyro-
coccus horikoshii OT3, Appl. Environ. Microbiol. 67 (2001) 673e679.
11] K. Miyazaki, Bifunctional isocitrate-homoisocitrate dehydrogenase: a missing
link in the evolution of b-decarboxylating dehydrogenase, Biochem. Biophys.
Res. Commun. 331 (2005) 341e346.
12] N. Kobashi, M. Nishiyama, M. Tanokura, Aspartate kinase-independent lysine
synthesis in an extremely thermophilic bacterium, Thermus thermophilus:
lysine is synthesized via alpha-aminoadipic acid not via diaminopimelic acid,
J. Bacteriol. 181 (1999) 1713e1718.
13] J. Miyazaki, N. Kobashi, M. Nishiyama, H. Yamane, Functional and evolutionary
relationship between arginine biosynthesis and prokaryotic lysine biosyn-
thesis through alpha-aminoadipate, J. Bacteriol. 183 (2001) 5067e5073.
[14] J. Miyazaki, N. Kobashi, T. Fujii, M. Nishiyama, H. Yamane, Characterization of
a lysK gene as an argE homolog in Thermus thermophilus HB27, FEBS Lett.
[
[
[
512 (2002) 269e274.
In conclusion, the quaternary complex of TtHICDH unambigu-
ously revealed the molecular mechanism responsible for the dual
substrate specificity of TtHICDH and also explained the difference
in substrate specificities between bacterial type- and yeast type-
HICDHs. These results provide an insight into the evolutionary
scenario responsible for alterations of substrate specificities of
HICDH family members.
[
15] A.P. Wulandari, J. Miyazaki, N. Kobashi, M. Nishiyama, T. Hoshino, H. Yamane,
Characterization of bacterial homocitrate synthase involved in lysine
biosynthesis, FEBS Lett. 522 (2002) 35e40.
[
16] J. Miyazaki, N. Kobashi, M. Nishiyama, H. Yamane, Characterization of
homoisocitrate dehydrogenase involved in lysine biosynthesis of an
extremely thermophilic bacterium, Thermus thermophilus HB27, and evolu-
tionary implication of beta-decarboxylating dehydrogenase, J. Biol. Chem. 278
(2003) 1864e1871.
[
17] T. Miyazaki, J. Miyazaki, H. Yamane, M. Nishiyama, alpha-Aminoadipate
aminotransferase from an extremely thermophilic bacterium, Thermus ther-
mophilus, Microbiology 150 (2004) 2327e2334.
Author contributions
[
18] Y. Jia, T. Tomita, K. Yamauchi, M. Nishiyama, D.R. Palmer, Kinetics and product
analysis of the reaction catalysed by recombinant homoaconitase from
Thermus thermophilus, Biochem. J. 396 (2006) 479e485.
K. T. performed expression, purification, and biochemical ana-
lyses as well as crystallization experiments. K. T. and T. T. performed
the crystallographic analysis. K. T., T. T., and M.N. analyzed the 3D
structure. K. T., T. T., T. K., and M.N. planned the experiments, and
wrote the manuscript.
[19] A. Horie, T. Tomita, A. Saiki, H. Kono, H. Taka, R. Mineki, T. Fujimura,
C. Nishiyama, T. Kuzuyama, M. Nishiyama, Discovery of proteinaceous N-
modification in lysine biosynthesis of Thermus thermophilus, Nat. Chem. Biol.
5
(2009) 673e679.
[20] Y. Lin, S.S. Alguindigue, J. Volkman, K.M. Nicholas, A.H. West, P.F. Cook,
Complete kinetic mechanism of homoisocitrate dehydrogenase from
Saccharomyces cerevisiae, Biochemistry 46 (2007) 890e898.
Acknowledgments
[21] R. Chen, S.S. Jeong, Functional prediction: identification of protein orthologs
and paralogs, Protein Sci. 9 (2000) 2344e2353.
[
22] J. Miyazaki, K. Asada, S. Fushinobu, T. Kuzuyama, M. Nishiyama, Crystal
structure of tetrameric homoisocitrate dehydrogenase from an extreme
thermophile, Thermus thermophilus: involvement of hydrophobic dimer-
dimer interaction in extremely high thermotolerance, J. Bacteriol. 187
(2005) 6779e6788.
23] E. Nango, T. Yamamoto, T. Kumasaka, T. Eguchi, Structure of Thermus ther-
mophilus homoisocitrate dehydrogenase in complex with a designed inhibi-
tor, J. Biochem. 150 (2011) 607e614.
This work was supported in part by JSPS KAKENHI (Grant No.
4228001 to M.N). We are grateful to the staff of the Photon Factory
for their assistance with data collection, which was approved by the
Photon Factory Program Advisory Committee (Proposal no.
011G524, 2013G618).
2
[
2
Transparency document
[24] S.L. Bulfer, J.M. Hendershot, R.C. Trievel, Crystal structure of homoisocitrate
dehydrogenase from Schizosaccharomyces pombe, Proteins 80 (2012)
6
61e666.
Transparency document related to this article can be found
online at http://dx.doi.org/10.1016/j.bbrc.2016.09.004.
[
[
[
25] W.W. Cleland, Statistical analysis of enzyme kinetic data, Methods Enzymol.
63 (1979) 103e138.
26] Z. Otwinowski, W. Minor, Processing of X-ray diffraction data collected in
oscillation mode, Methods Enzymol. 276 (1997) 307e326.
27] A.J. McCoy, R.W. Grosse-Kunstleve, P.D. Adams, M.D. Winn, L.C. Storoni,
R.J. Read, Phaser crystallographic software, J. Appl. Crystallogr. 40 (2007)
658e674.
[28] N. Collaborative Computational Project, The CCP4 suite: programs for protein
crystallography, Acta Crystallogr. D. 50 (1994) 760e763.
[29] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta
Crystallogr. D. 60 (2004) 2126e2132.
[30] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of macromolecular
structures by the maximum-likelihood method, Acta Crystallogr. D. 53 (1997)
240e255.
[31] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, PROCHECK: a
program to check the stereochemical quality of protein structures, J. Appl.
Crystallogr. 26 (1993) 283e291.
[32] W.L. DeLano, The PyMol Molecular Graphics System on World Wide Web,
2002. http://www.pymol.org.
[33] F.A. Momany, R. Rone, Validation of the general purpose QUANTA 3.2/
CHARMm force field, J. Comput. Chem. 13 (1992) 888e900.
[34] A. Pall oꢀ , J. Ol ꢀa h, E. Gr aꢀ czer, A. Merli, P. Z aꢀ vodszky, M.S. Weiss, M. Vas, Struc-
References
[
1] H. Nishida, M. Nishiyama, N. Kobashi, T. Kosuge, T. Hoshino, H. Yamane,
A prokaryotic gene cluster involved in synthesis of lysine through the amino
adipate pathway: a key to the evolution of amino acid biosynthesis, Genome
Res. 9 (1999) 1175e1183.
[
[
[
[
[
[
2] T.M. Zabriskie, M.D. Jackson, Lysine biosynthesis and metabolism in fungi, Nat.
Prod. Rep. 17 (2000) 85e97.
3] H. Xu, B. Andi, J. Qian, A.H. West, P.F. Cook, The alpha-aminoadipate pathway
for lysine biosynthesis in fungi, Cell Biochem. Biophys. 46 (2006) 43e64.
4] S.J. Parsons, R.O. Burns, Purification and properties of b-isopropylmalate de-
hydrogenase, J. Biol. Chem. 244 (1969) 996e1003.
5] P.D. Weitzman, Patterns of diversity of citric acid cycle enzymes, Biochem. Soc.
Symp. 54 (1987) 33e43.
6] M. Strassman, L.N. Ceci, Enzymatic formation of a-ketoadipic acid from
homoisocitric acid, J. Biol. Chem. 240 (1965) 4357e4361.
7] M.W. Bauer, E.J. Bylina, R.V. Swanson, R.M. Kelly, Comparison of a b-glucosi-
dase and a b-mannosidase from the hyperthermophilic archaeon Pyrococcus
furiosus. Purification, characterization, gene cloning, and sequence analysis,
J. Biol. Chem. 271 (1996) 23749e23755.
tural and energetic basis of isopropylmalate dehydrogenase enzyme catalysis,
FEBS J. 281 (2014) 5063e5076.
[
8] V. Durbecq, C. Legrain, M. Roovers, A. Pierard, N. Glansdorff, The carbamate
kinase-like carbamoyl phosphate synthetase of the hyperthermophilic
archaeon Pyrococcus furiosus, a missing link in the evolution of carbamoyl
phosphate biosynthesis, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 12803e12808.
9] I. Matsui, E. Matsui, Y. Sakai, H. Kikuchi, Y. Kawarabayasi, H. Ura, S. Kawaguchi,
S. Kuramitsu, K. Harata, The molecular structure of hyperthermostable aro-
matic aminotransferase with novel substrate specificity from Pyrococcus
horikoshii, J. Biol. Chem. 275 (2000) 4871e4879.
[35] Eꢀ . Gr aꢀ czer, A. Pall oꢀ , J. Ol aꢀ h, T. Szimler, P.V. Konarev, D.I. Svergun, A. Merli,
P. Z aꢀ vodszky, M.S. Weiss, M. Vas, Glutamate 270 plays an essential role in
K(þ)-activation and domain closure of Thermus thermophilus iso-
propylmalate dehydrogenase, FEBS Lett. 589 (2015) 240e245.
[36] K. Miyazaki, Identification of a novel trifunctional homoisocitrate dehydro-
genase and modulation of the broad substrate specificity through site-
directed mutagenesis, Biochem. Biophys. Res. Commun. 336 (2005)
596e602.
[
[10] K. Ishikawa, H. Ishida, I. Matsui, Y. Kawarabayasi, H. Kikuchi, Novel