Identification of catalytically important amino acid residues for enzymatic reduction of glyoxylate in plants

Article abstract of DOI:10.1016/j.bbapap.2013.09.013

NADPH-dependent glyoxylate reductases from Arabidopsis thaliana (AtGLYR) convert both glyoxylate and succinic semialdehyde into their corresponding hydroxyacid equivalents. The primary sequence of cytosolic AtGLYR1 reveals several sequence elements that are consistent with the β-HAD (β-hydroxyacid dehydrogenase) protein family, whose members include 3-hydroxyisobutyrate dehydrogenase, tartronate semialdehyde reductase and 6-phosphogluconate dehydrogenase. Here, site-directed mutagenesis was utilized to identify catalytically important amino acid residues for glyoxylate reduction in AtGLYR1. Kinetic studies and binding assays established that Lys170 is essential for catalysis, Phe231, Asp239, Ser121 and Thr95 are more important in substrate binding than in catalysis, and Asn174 is more important in catalysis. The low activity of the mutant enzymes precluded kinetic studies with succinic semialdehyde. The crystal structure of AtGLYR1 in the absence of substrate was solved to 2.1 A? by molecular replacement using a previously unrecognized member of the β-HAD family, cytokine-like nuclear factor, thereby enabling the 3-D structure of the protein to be modeled with substrate and co-factor. Structural alignment of AtGLYR1 with β-HAD family members provided support for the essentiality of Lys170, Phe173, Asp239, Ser121, Asn174 and Thr95 in the active site and preliminary support for an acid/base catalytic mechanism involving Lys170 as the general acid and a conserved active-site water molecule. This information established that AtGLYR1 is a member of the β-HAD protein family. Sequence and activity comparisons indicated that AtGLYR1 and the plastidial AtGLYR2 possess structural features that are absent in Arabidopsis hydroxypyruvate reductases and probably account for their stronger preference for glyoxylate over hydroxypyruvate.

Full text of DOI:10.1016/j.bbapap.2013.09.013

Biochimica et Biophysica Acta 1834 (2013) 2663–2671
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta
journal homepage: www.elsevier.com/locate/bbapap
Identification of catalytically important amino acid residues for
enzymatic reduction of glyoxylate in plants
Gordon J. Hoover a,1, René Jørgensen b,1,2, Amanda Rochon a,1, Vikramjit S. Bajwa a,1
,
A. Rod Merrill b, Barry J. Shelp a,
⁎
a
Department of Plant Agriculture, University of Guelph, Guelph, ON N1G 2W1, Canada
Department of Molecular and Cellular Biology, University of Guelph, Guelph, ON N1G 2W1, Canada
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 10 May 2013
Received in revised form 15 September 2013
Accepted 18 September 2013
Available online 27 September 2013
NADPH-dependent glyoxylate reductases from Arabidopsis thaliana (AtGLYR) convert both glyoxylate and
succinic semialdehyde into their corresponding hydroxyacid equivalents. The primary sequence of cytosolic
AtGLYR1 reveals several sequence elements that are consistent with the β-HAD (β-hydroxyacid dehydrogenase)
protein family, whose members include 3-hydroxyisobutyrate dehydrogenase, tartronate semialdehyde reduc-
tase and 6-phosphogluconate dehydrogenase. Here, site-directed mutagenesis was utilized to identify catalyti-
cally important amino acid residues for glyoxylate reduction in AtGLYR1. Kinetic studies and binding assays
established that Lys170 is essential for catalysis, Phe231, Asp239, Ser121 and Thr95 are more important in
substrate binding than in catalysis, and Asn174 is more important in catalysis. The low activity of the mutant
enzymes precluded kinetic studies with succinic semialdehyde. The crystal structure of AtGLYR1 in the absence
of substrate was solved to 2.1 Å by molecular replacement using a previously unrecognized member of the
β-HAD family, cytokine-like nuclear factor, thereby enabling the 3-D structure of the protein to be modeled
with substrate and co-factor. Structural alignment of AtGLYR1 with β-HAD family members provided support
for the essentiality of Lys170, Phe173, Asp239, Ser121, Asn174 and Thr95 in the active site and preliminary sup-
port for an acid/base catalytic mechanism involving Lys170 as the general acid and a conserved active-site water
molecule. This information established that AtGLYR1 is a member of the β-HAD protein family. Sequence and ac-
tivity comparisons indicated that AtGLYR1 and the plastidial AtGLYR2 possess structural features that are absent
in Arabidopsis hydroxypyruvate reductases and probably account for their stronger preference for glyoxylate over
hydroxypyruvate.
Keywords:
Arabidopsis thaliana
Glyoxylate reductase
β-hydroxyacid dehydrogenase protein family
Hydroxypyruvate reductase
Photorespiration
Succinic semialdehyde reductase
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
with overlapping specificities for glyoxylate versus hydroxypyruvate
and NADPH versus NADH, and for distinct cytosolic and plastidial
In 1953, Zelitch [1] provided the first report of NAD(P)H-dependent
glyoxylate reductase (GLYR, EC 1.2.1.79) activity in crude or partially
purified extracts from spinach leaves. Subsequently, evidence from
highly purified enzyme preparations of several plant species, as well as or-
ganelle fractionation, was obtained for the existence of multiple enzymes
NADPH-dependent glyoxylate reductases that have only low activity
for hydroxypyruvate and/or NADH [2–5]. More recently, Arabidopsis
genes for the two isoforms (cytosolic AtGLYR1, GenBank accession no.
AY044183, and plastidial AtGLYR2, GenBank accession no. AAP42747)
were identified using a combination of strategies, including yeast com-
plementation and in silico analysis of gene/protein sequence, in vivo
targeting analysis of green fluorescent protein fusions in tobacco BY-2
suspension cells and Arabidopsis plants, and production and kinetic
characterization of recombinant proteins [6–10]. Both recombinant
AtGLYRs prefer NADPH over NADH and convert glyoxylate to glycolate,
and the AtGLYR1 has negligible hydroxypyruvate-dependent activity
[7,9]. Notably, the two isoforms also convert SSA (succinic semialdehyde)
to gamma-hydroxybutyrate, albeit with much lower catalytic efficiency
than for glyoxylate [7,9], and the use of single atglyr1 and atglyr2 knock-
outs indicates that they can function in γ-aminobutyrate metabolism
during stress [11,12].
Abbreviations: AtGLYR, Arabidopsis thaliana glyoxylate reductase; β-HAD, β-
hydroxyacid dehydrogenase; 3-HIBADH, 3-hydroxyisobutyrate dehydrogenase; HPR,
hydroxypyruvate reductase; 6-PGDH, 6-phosphogluconate dehydrogenase; SSA, succinic
semialdehyde; TA, tartaric acid; TSAR, tartronate semialdehyde reductase
⁎
Corresponding author at: Department of Plant Agriculture, Bovey Bldg., Rm 4237,
University of Guelph, 50 Stone Rd. E., Guelph, ON N1G 2W1, Canada. Tel.: +1 519 824
4120x53089; fax: +1 519 767 0755.
E-mail address: bshelp@uoguelph.ca (B.J. Shelp).
These authors contributed equally to the paper.
Present address: Department of Microbiology and Infection Control, Statens Serum
1
2
Institut, 5 Artillerivej, Copenhagen S, DK-2300, Denmark.
1570-9639/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.bbapap.2013.09.013

2664
G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
Preliminary analysis of the primary sequence of AtGLYR1 revealed
2.2. Site-directed mutagenesis of AtGLYR1 active site
several sequence elements that are consistent with members of the
β-HAD (β-hydroxyacid dehydrogenase) protein family [6]. Identification
of the β-HAD family was originally based on similarities in hydroxyacid
substrate specificity, oxidative reaction mechanism and conserved
glycine-rich sequence elements [13]. The crystal structure of NADP-
dependent sheep 6-PGDH (6-phosphogluconate dehydrogenase) in the
presence of its substrate 6-phosphogluconate revealed 14 strictly
conserved amino acid residues with the potential to be catalytically
important [14] (PBD ID 1PGP), and sequence alignments between this
protein and NAD-dependent rat 3-HIBADH (3-hydroxyisobutyrate dehy-
drogenase) enabled identification of a conserved glycine-rich consensus
element that includes two residues from the 6-PGDH active site [13].
Directed mutagenesis demonstrated K173 as catalytically crucial and
N177 as catalytically important residues in 3-HIBADH. Subsequent
mutagenesis and modeling studies enabled identification of additional
amino acids as catalytically important residues at the 6-PGDH active
site [15–19]. Recently, the crystal structures for Thermus thermophilus
HB8 3-HIBADH [20] (PDB ID 2CVZ), as well as Salmonella typhimurium
LT2 TSAR (tartronate semialdehyde reductase) [21] (PDB ID 1VPD) and
NADPH-dependent Eubacterium barkeri 2-(hydroxymethyl)glutarate
dehydrogenase [22] (PDB ID 3CKY) became available. This information
indicated that TSAR and 2-(hydroxymethyl)glutarate dehydrogenase
are also members of the β-HAD family.
In the present paper, information about the β-HAD protein family
allowed us to identify catalytically important amino acid residues
for glyoxylate reduction in AtGLYR1 by characterization of the kinetic
and binding properties of mutant enzymes. Initially, we had hoped to
gain insights into the mechanistic importance of the residues in both
glyoxylate and SSA reduction; however, kinetic studies of the mutant
enzymes were not possible with SSA since the activities were too low.
We were also able to solve the crystal structure of AtGLYR1 in the
absence of substrate (apoprotein; PDB ID 3DOJ) to 2.1 Å by molecular
replacement using a previously unrecognized member of the β-HAD
family, cytokine-like nuclear factor. This allowed us to model the 3-D
structure of the protein with substrate and co-factor. Together, these
studies provided support for the importance of these residues and
established the basis for including AtGLYR1 in the β-HAD protein family.
The QuickChange site-directed mutagenesis kit (Stratagene) was
used to introduce S121A, K170A, K170R, K170E, K170H, N174A,
F231A, D239A, T95A mutations into AtGLYR1 expression plasmid
pET15b, using the appropriate primers (see Table S1 in the supplemen-
tal material). Mutated plasmids were sent to Genologics (University of
Guelph, Laboratory Services Branch, Guelph, ON, Canada) for sequenc-
ing (ABI PRISM Sequencer Model 377, PerkinElmer Life Sciences, Foster
City, CA, USA).
2.3. Analysis of recombinant protein stability
Purified recombinant proteins were separated by electrophoresis on
a 10% SDS-PAGE gel. The separated proteins were detected by immuno-
blot analysis on nitrocellulose membrane with anti-His monoclonal
antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:500 dilution.
Blots were developed using alkaline phosphatase-linked secondary
antibody at 1:10,000 dilution (goat anti-mouse AP-linked, Sigma, St.
Louis, MO) and the alkaline phosphate conjugate substrate kit (Biorad,
Hercules, CA).
The purified recombinant enzymes were analyzed by steady-state
Trp fluorescence emission scans using PTI QuantaMaster C-61 steady-
state fluorimeter (Photon Technology International, London, Ontario,
Canada). The temperature of the cuvette was maintained at 22 °C
with a circulating water bath. For emission scans, the excitation
wavelength was 290 nm and the emission was scanned from 305 to
400 nm with a scan rate of 1.0 nm/s. The excitation and emission wave-
lengths were set to 2 and 4 nm, respectively. The final concentrations of
the native enzyme, denatured native enzyme (using 5 M guanidine-
HCl) and the mutant proteins were 60 μg/mL. Each protein spectrum
was corrected for the buffer only spectrum and the corrected spectra
were normalized. The normalized corrected spectra of the native and
mutant enzymes were superimposed to assess differences in shape
and emission maxima.
2.4. Enzyme assays
Enzymatic activity of the purified native and mutant AtGLYR1
enzymes was measured continuously as the oxidation of NADPH using
the protocol described previously [7]. Briefly, the reaction mixture
consisted of 50 mM 4-(2-hydroxyethyl)-piperazine-1-ethanesulfonic
acid, pH 7.8, 10% sorbitol, NADPH, glyoxylate and purified recombinant
enzyme. Kinetics with glyoxylate as the varied substrate were
conducted at a fixed concentration of NADPH (50 μM); for each enzyme
the concentration of glyoxylate was varied to give five to six data points
both above and below the Km. When NADPH was varied, the fixed con-
centration of glyoxylate was typically 8–10 times the Km for glyoxylate;
the only exception was for the F231A mutant where the maximum
concentration of glyoxylate achievable was 2.5 times the Km. For each
enzyme, the concentration of enzyme used was determined to be within
the linear range (0.75–2500 nM). One unit of activity equals 1 μmol/min,
and all kinetic data were obtained from measurements of initial rate.
Measurements were obtained in triplicate at each concentration and typ-
ically three to four biological replicates were used. Kinetic data were fit to
the Michaelis–Menten equation using non-linear least-squares analysis
(SigmaPlot2000, version 6.1; Enzyme Kinetics Module, version 1.0; Systat
Software Inc., Point Richmond, CA). A fit was deemed acceptable if it ex-
hibited random residuals, passed the runs test and reached minimum
error in all fitting parameters. Kinetics of the native AtGLYR1 enzyme
with lithium β-hydroxypyruvate (Sigma, St. Louis, MO) as the varied sub-
strate were performed essentially as described above for glyoxylate with
fixed concentrations of NADPH (50 μM) and enzyme (25 nM). For native
AtGLYR2, kinetics using glyoxylate as a varied substrate were essentially
the same as for AtGLYR1 with the exception that enzyme concentration
was 25 nM.
2. Materials and methods
2.1. Expression and purification of recombinant AtGLYR1
Recombinant AtGLYR2 was produced by amplifying a truncated
AtGLYR2 (Gene ID 838342) cDNA sequence lacking the N-terminal 58
amino acids using primers 5′-GCATCATATGTCTACCAGAGATGAACTTG
GAAC-3′ and 5′-GCATGGATCCCTAAGCTTCTCGGGATTTTGC-3′, thereby
providing a NdeI restriction site on the forward primer and a BamHI site
on the reverse primer. The amplified PCR product was cloned into the
pET-15b expression vector using NdeI/BamHI (Novagen, EMD Biosciences
Inc., Madison, WI, USA) and sequenced (University of Guelph, Laboratory
Services Branch, Guelph, ON, Canada).
The recombinant AtGLYR1 and AtGLYR2 were expressed from
the isopropyl-β-D-thiogalactopyranoside-inducible pET15b plasmid in
Escherichia coli BL21 pLysS and purified as described previously [7].
Briefly, post induction (2 h) the BL21 cells were collected by centrifuga-
tion, frozen overnight at −20 °C, lysed with lysozyme (1 mg/mL) in
buffer (100 mM Tris, pH 8.0, 0.5 mM phenylmethylsulfonyl-fluoride,
1 μg/mL pepstatin A, 2 μg/mL leupeptin) and precipitated with 10% poly-
ethylene glycol 8000 overnight. The precipitate was resuspended in
equilibration buffer (50 mM Tris, pH 8.0, 0.5 M NaCl, 20 mM imidazole,
100 μM phenylmethylsulfonyl-fluoride) and loaded on a 4 mL HIS-
Select® Nickel Affinity Gel (Sigma) to purify the His6-tagged protein (as
per manufacturer's protocol). The purity of the eluted protein was veri-
fied by SDS-PAGE (coomassie stained and western blot) and its concen-
tration was determined with the Bradford assay [7].

G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
2665
Glyoxylate-dependent quenching of intrinsic tryptophan fluores-
3. Results and discussion
cence was used to determine the binding constants (Kd). Triplicate
reactions were performed in the above described reaction buffer over
a concentration range of 0–50 μM glyoxylate in the presence of 1 μM
enzyme. Fluorescence measurements were obtained using a Cary
Eclipse fluorescence spectrophotometer (Santa Clara, CA) according to
the method reported elsewhere [23].
3.1. Sequence alignment of AtGLYR1 and β-HAD family members
Sequence alignment reveals that the amino acid identity among
AtGLYR1, two known members of the β-HAD family (3-HIBADH and
TSAR), and N-PAC ranges from 23 to 48% (Fig. 1). However, the identity
between these four proteins and 6-PGDH ranges from only 10 to 19%,
and the subunit size is ~30 kD and ~50 kD, respectively. Furthermore,
3-HIBADH, TSAR and AtGLYR1 all catalyze low energy electron transfer
oxidations and reductions [13,14,34], whereas 6-PGDH catalyzes a
three-step, high energy oxidative decarboxylation [35]. Also evident
from the sequence alignment is that all five proteins share 10 strictly
conserved glycines within their primary sequence, contributing to a
common structural framework. They contain a rather large conserved
primary structure consensus element at their N-terminus that was
originally identified by Hawes et al. [13] and later confirmed as the
3-HIBADH signature sequence by Prosite ([LIVMFY](2)-G-L-G-x-[MQ]-
G-x(2)-[MA]-[SAV]-x-[SNHR]) (Prosite (http://expasy.org/prosite/) ac-
cession number: PS00895). This element is forgiving in amino acid
identity but not in character; therefore, small deviations can be seen
in the case of 6-PGDH and N-PAC.
2.5. Crystallization of recombinant AtGLYR1
Purified AtGLYR1 was desalted and concentrated to 10 mg/mL
in 50 mM Tris, pH 7.8, using Microcon-10 centrifugal concentrators
(Amicon). Crystallization trials were set up using the sitting-drop
vapor-diffusion method at 20 °C. The crystallization conditions were
identified from a preformulated commercial screening kit (Hampton
Research, Aliso Viejo, CA). The best crystals were obtained by equilibrat-
ing a mixture of 2 μL of protein solution and an equivalent volume of
reservoir solution (0.2 M calcium acetate hydrate, 20% polyethylene
glycol 3350, pH 6.5). Needlelike crystals formed within a few days and
grew to diffracting quality size single crystals after approximately six
weeks.
2.6. Data collection, structure solution and refinement
3.2. Kinetic characterization of putative active-site mutants of AtGLYR1
Before flash freezing in liquid N2 the crystals were transferred to
paratone-N (Hampton research) for cryo-protection. A native 2.35 Å
dataset was collected with our in-house Enraf-Nonius FR571 diffrac-
tometer equipped with a rotating Cu anode and a proteum pt135 ccd
detector (Bruker AXS). The diffraction images were processed using
Proteum2 software package (Bruker AXS). The space group was deter-
mined to be body-centered tetragonal with a Matthews coefficient indi-
cating only one molecule per asymmetric unit with a solvent content of
66%. The structure of AtGLYR1 was solved by molecular replacement
using Molrep [24] and the cytokine-like nuclear factor (N-PAC) (PDB
entry, 2UYY) as search model. The solution from molecular replacement
was used as input to warpNtrace [25], which was able to trace ~70% of
the structure. The model was rebuilt in Coot [26] and refined in Refmac5
[27] using TLS at 2.35 Å resolution. Finally, a 2.1 Å data set was collected
on the same crystal at beamline 08-ID at the Canadian Light Source in
Saskatoon, SK, Saskatchewan. After processing the new data using the
HKL2000 software package [28], the model was re-refined at 2.1 Å in
Phenix [29] using the new reflection file after extension and transfer
of the test set. The buried surface area was calculated by PISA [30].
Based on earlier studies of 3-HIBADH, 6-PGDH and TSAR [34,36]; see
also references above), five putative active-site residues and a charac-
teristically conserved T95 residue were identified in AtGLYR1 (Fig. 1)
and replaced with Ala (S121A, K170A, N174A, F231A, D239A, T95A).
Lys170 was also replaced with Arg, Glu or His. The mutant enzymes,
like the native AtGLYR1 [7], could be highly expressed in E. coli and
then extracted and purified to homogeneity (see Fig. S1A in the
supplemental material). Trp emission scans revealed that the emission
maximum and the shape of the fluorescence scan were similar for
mutant and native enzymes (see Fig. S1B in the supplemental material),
indicating that stability and probably folding were unaffected by the
introduced mutations.
We had planned to investigate the importance of the various resi-
dues in the reduction of SSA, as well as glyoxylate; however, the activi-
ties of the mutant enzymes with SSA were generally too low for kinetic
studies. Thus, we focused on the glyoxylate-dependent reaction; the
kinetic and binding parameters are summarized in Tables 1 and 2.
The native enzyme displayed values for catalytic efficiency (kcat/Km)
and affinity (Km) for glyoxylate and NADPH of 3407 s−1 mM−1, 18 μM
and 3.4 μM, respectively, as determined with the use of a microplate
reader. These values are in agreement with previous data obtained
with a double beam spectrophotometer, which demonstrated a much
2.7. Alignments
higher catalytic efficiency and affinity for glyoxylate (2870 s−1 mM−1
,
The multiple sequence alignment was carried out using ClustalW
[31] (http://www.ebi.ac.uk) and the output was loaded into the
ESPRIPT utility (http://espript.ibcp.fr) to incorporate secondary struc-
ture assignment from the AtGLYR1 crystal structure. The AtGLYR1
structure was superimposed on the structures of 3-HIBADH-NADPH,
6-PGDH-NAPDH and TSAR-TA using the Brute force alignment com-
mand in LSQMAN [32]. Structural figures were prepared in PyMOL
[33].
4.5 μM) than for SSA (11.6 s−1 mM−1, 870 μM), as well as a compara-
ble affinity for NADPH (2.2 μM) [7]. Here, even less favorable values
were obtained for these kinetic parameters using hydroxypyruvate
(kcat/Km = 0.232 s−1 mM−1 and Km = 6.1 mM), indicating that
hydroxypyruvate is a very poor substrate for AtGLYR1. These bio-
chemical properties are consistent with those reported in early stud-
ies of NADPH-dependent glyoxylate reductase activity from tobacco
and spinach [2,3,5].
There was a four- to 70-fold range in the Km and kcat/Km values for
NADPH across the enzymes tested (Table 1), whereas the values for
glyoxylate varied by 800- to 18,000-fold (Table 2), indicating that
interactions with glyoxylate were dramatically more sensitive to the
putative active-site mutations. For the K170A mutant, glyoxylate turn-
over could not be detected even at elevated enzyme and substrate
concentrations, and the Kd value was comparable to that in the native
enzyme (Table 2), indicating that this Lys is crucial for catalysis. The
kcat values for glyoxylate in the K170R and K170E mutants were about
three and four orders of magnitude, respectively, lower than that for
2.8. Statistical analyses
All statistical analyses for the kinetic data were done using SAS 9.1 at
the α = 0.05 level (SAS Institute, Cary, NC). Data were first checked for
homogeneity of variance (Brown and Forsythe's test for homogeneity
of variance) and log transformed in cases where equal variance was
not observed before being analyzed using Duncan's multiple range
(Proc GLM). Unless otherwise indicated, values represent means of
three to four biological replicates
SE.

2666
G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
Fig. 1. Primary sequence and secondary structure alignments of native AtGLYR1 (PDB ID 3DOJ) and N-PAC (PDB ID 2UYY) with the original members of the β-HAD family: 3-HIBADH (PDB
ID 2CVZ), 6-PGDH (PDB ID 1PGO), and TSAR (PDB ID 1VPD). The black bar denotes the position of the N-terminal Prosite 3-HIBADH signature sequence [LIVMFY](2)GLGx[MQ]Gx(2)[MA]-
[SAV]x[SNHR]. The blue stars indicate the positions of the AtGLYR1 active-site consensus sequence Sx(48)Kx(3)Nx(56)Fx(7)D. The red stars indicate the positions of the 6-PGDH active-
site consensus sequence Nx(25)Sx(54)Kx(2)HNx(2)EYx(68)KxTx(24)Rx(158)Rx(2)Fx(2)H [7]. The numbering for 6-PGDH in the above alignment is one amino acid higher than the
numbering related to the 1PGO entry, wherein the sequence starts at the Ala, which corresponds to the second amino acid in the alignment above.

G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
2667
Table 1
(α6 and α7). Domain II (residues 195–287) consists of only helices
(α8–α13) from the C-terminal segment of the protein. The two do-
mains are connected by a long α-helix, α8 (residues 166–194). Like
the 3-HIBADH [23], TSAR (PDB ID 1VPD) and N-PAC (PDB code 2UYY)
structures, the AtGLYR1 exists as an apparent tetramer in the crystal.
Even though the four enzymes were crystallized in different space
groups they all contain a globally identical tetrameric structure
(Fig. 2B), which lends further support that these enzymes function as
a homotetramer. As carefully described for the 3-HIBADH structure [20]
each subunit of the tetramer interacts primarily with two neighboring
subunits. The AB subunits show a major buried surface area of
2344.5 Å2 corresponding to 17.4% of the total protomer surface. The
AC and AD interface show minor areas of 676.4 Å2 and 294.3 Å2, respec-
tively. Furthermore, the AB interface is essentially mediated by hydro-
phobic interactions, whereas several hydrogen bonds mediate the AC
and AD interface. The tetrameric structure of the AtGLYR1 oligomer in
solution was further confirmed using gel exclusion chromatography
(data not shown).
Kinetic parameters for recombinant native and mutant AtGLYR1s with NADPH. Values
represent the mean of two–four biological replicates
column indicate no significant difference between samples as determined by a Duncan's
multiple range test (when necessary, data were log transformed in order to ensure
equal variance).
SE. Shared letters within each
AtGLYR1
kcat/Km (s−1 μM−1
)
Km (μM)
kcat (s−1
)
Native
S121A
N174A
K170A
F231A
D239A
T95A
24.3
51.7
10.9
NDa
1.6 b
6.6 a
4.0 c
3.44
1.80
0.902
ND
2.04
2.73
64.8
0.26 b
0.07 cd
0.108 d
84.1
93.7
9.09
ND
10.4 a
14.1 a
2.38 b
4.34
10.4
0.779
0.55 c
2.3 c
0.186 d
0.65 cd
0.67 bc
3.1 a
9.56
25.4
51.1
4.26 b
2.8 b
14.5ab
a
Not detectable at a minimum limit of 120 nM s−1
.
the native enzyme; however, there was much less change, if any, in the
Km and Kd values. Furthermore, it was not possible to saturate the reac-
tion for the K170H mutant, even up to a glyoxylate concentration of
200 mM, indicating extremely low interaction with the substrate
(data not shown). These findings are in agreement with similar studies
on the K183 that is identically positioned within 6-PGDH [14,17].
The S121A, N174A, F231A, D239A and T95A mutants had lower
kcat/Km values for glyoxylate than the native enzyme, but enzymatic
activity was not eliminated (Table 2), indicating that these residues
are not crucial participants in the catalytic event. Of these mutants,
N174A had the lowest kcat (turnover), but the Kd was unaffected, indi-
cating that N174 is more important in catalysis than in binding. F231A
and D239A had lower kcat values, and much higher (weaker affinity)
or undeterminable Kd values due to their extremely low binding,
indicating that these residues are more important in binding than in
catalysis. The kcat for S121A was unaffected, whereas the Kd was slightly
compromised, suggesting that S121 is more important in binding than
catalysis. Similarly, the kcat for T95A was unaffected, but the Kd was
undeterminable, indicating that T95 is very important in binding.
3.4. Structural alignment and NADPH binding in AtGLYR1 and β-HAD
family members
The solved crystal structure of AtGLYR1 without substrate superim-
poses nicely onto the structures of 3-HIBADH-NADPH, 6-PGDH-NADPH
and TSAR-tartaric acid (TSAR-TA) with r.m.s.d. values of 1.5 Å, 1.9 Å
and 1.2 Å, respectively, for the Cα atoms (Fig. 3A). This supports
the structural relationship of AtGLYR1 to the β-HAD family. Many
NAD(H)/NADP(H)-dependent dehydrogenases contain the Rossmann
fold for binding the dinucleotide cofactor [37]; the pyrophosphate
group interacts with the GxGxx(G/A) sequence fingerprint motif found
in the Rossmann fold. This characteristic glycine-rich fingerprint motif
is present in the N-terminal domain (G7–G12, as numbered in
AtGLYR1). We were not successful in crystallizing AtGLYR1 in the pres-
ence of substrate or infusing glyoxylate into the AtGLYR1 crystal, but the
presence of NADPH (from 3-HIBADH-NADPH) and TA (from TSAR-TA)
enables identification of residues that are most likely to contribute to
substrate binding in AtGLYR1. Modeling of each substrate into the
crystal structure of AtGLYR1 illustrates a good fit within the proposed
binding site (Fig. 3B and C, respectively). Like the 3-HIBADH-NADPH
structure, NADPH is most likely bound to the interdomain cleft of the
AtGLYR1 structure. The ADP portion of the cofactor is exposed to the
solvent, whereas the nicotinamide ring as well as the TA substrate are
shielded from the solvent.
3.3. Structure of AtGLYR1 protomer and oligomer
Diffraction data for the AtGLYR1 crystal without substrate were
collected to a maximum resolution of 2.1 Å and the structure was solved
by molecular replacement using the phase coordinates from N-PAC,
which has a sequence identity of 49%, resulting in high quality electron
density maps (see Fig. S2 in the supplemental material). The data reduc-
tion and refinement statistics for the model are summarized in Table S2
in the supplemental material. The overall structure of AtGLYR1 consists
of two domains and contains eight β-strands (named β1 through β8)
placed in domain I and 13 α-helices (α1 through α13) distributed
between the two domains (Fig. 2A). Domain I, with the dinucleotide-
binding region, comprises residues 1–165 in the N-terminus. This
typical Rossmann fold domain contains two α/β units: a six-stranded
parallel β-sheet (β1-β6a) covered by four helices (α1-α5) and followed
by a mixed three-stranded β-sheet (β6b-β8) covered by two helices
The residues surrounding the NADPH and TA substrates (M11, N30,
R31, L64, S121, K170, N174,D239, F231, and T95 as numbered in
AtGLYR1) in the superposition of the four structures are highly con-
served and have almost perfectly overlapping positions (Fig. 3D). The
potential hydrogen bonds in the apo structure of AtGLYR1 are listed in
Table S3 in the supplemental material and compared to the correspond-
ing bonds in 3-HIBADH-NADPH and TSAR-TA structures. Residues K170,
N174 and D239 are positioned within hydrogen bonding distance of
Table 2
Kinetic parameters for recombinant native and mutant AtGLYR1s with glyoxylate. Values for native and mutant enzymes represent the mean of nine and three–four biological
replicates SE, respectively. Shared letters within each column indicate no significant difference between samples as determined by a Duncan's multiple range test (when necessary,
data were log transformed in order to ensure equal variance). ND, not detectable.
AtGLYR1
kcat/Km (s−1 mM−1
)
Km (mM)
kcat (s−1
)
Kd (μM)
Native
S121A
N174A
K170A
K170R
K170E
F231A
D239A
T95A
3407
480
983 a
17 b
12.9 c
0.018
0.181
0.088
ND
0.061
0.033
12.4
3.00
0.002 e
0.015 c
0.011 d
54.6
86.4
6.06
ND
0.051
0.0052
11.0
22.0
67.8
13.2 a
6.7 a
0.32 c
1.47
15.13
1.92
3.89
1.41
1.77
ND
0.42 d
4.02 b
0.32 cd
1.61 c
0.28 d
0.18 cd
72.8
NDa
0.86
0.19
0.87
7.45
14.6
0.05 e
0.03 f
0.13 e
0.53 d
2.8 d
0.008 d
0.014 e
1.70 a
0.18 b
0.16 b
0.004 d
0.0012 e
2.6 c
0.4 b
10.4 a
183.9
ND
14.49 a
4.6
a
Not detectable at a minimum limit of 120 nM s−1
.

2668
G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
Fig. 2. Ribbon representation of native AtGLYR1 in the 2.1 Å crystal structure. (A) Protomer with domain and secondary structure assignments; α-helices are shown in green, loops in red
and β-strands in light brown. (B) Tetramer as packed in the crystal structure, with subunit A colored in blue, B in orange, C in green and D in purple.
each other, with D239 always hydrogen bonded to a conserved active-
site water molecule, as in the 3-HIBADH-NADPH and TSAR-TA struc-
tures. These hydrogen-bond interactions could act as a proton relay
from the active-site water to K170 via D239 and N174 residues, and
then to glyoxylate. Furthermore, the highly conserved water, as seen
in the HIBADH holo structure (PDB code 1WP4), could move closer to
the catalytically crucial Lys and becomes highly coordinated, giving it
the potential to also function as a general acid/base catalyst.
K170R mutant would be capable of accepting a proton from the water
molecule, and therefore could function as a general acid by donating a
proton to the developing product. However, the catalytic activity with
Arg would be lower because it likely has a higher pK in the enzyme
environment than does Lys and is therefore a weaker acid [17]. This is
not unexpected as the pK of the guanidinium group of Arg is about
12.5 in solution compared to 10.5 for the primary amine group of Lys.
With the K170H mutant, there would be a dramatically lower binding
affinity for glyoxylate, likely due to the result of steric interference by
the imidazole group of the His side chain. We attempted to determine
the pK value of the K170 in AtGLYR1 using nuclear magnetic resonance;
however, it was unsuccessful due to the large molecular mass of
the AtGLYR1 tetramer (~120 KDa) and the sorbitol required for enhanc-
ing protein solubility/stability in the buffer used. Alternatively, we
used the computational pKa feature in the modeling suite, Molecular
Operating Environment 2011.10 (Chemical Computing Group, INC) to
calculate the pKa for K170 in AtGLYR1 and it was found to be 8.09. How-
ever, it must be considered that the AtGLYR1 (3DOJ) structure does not
have bound cofactor or substrate, which will undoubtedly shift the pKa
for K170.
The D239A or N174A substitution in AtGLYR1 would affect the pro-
ton relay between active-site water and active-site Lys and decrease cat-
alytic efficiency and turnover in these mutant enzymes, as is evident
from the observed kinetic data (Table 2). For N174, the hydrogen bond-
ing also supports its involvement in orienting K170 during catalysis.
The D239A mutant of AtGLYR1 also had a much weaker binding affinity
for glyoxylate than the native enzyme (Table 2), and the binding affinity
for the T95A and F231A mutants was below the detection limit
(Table 2). However, the kcat values for these three mutants were quite
comparable, indicating that T95, F231 and D239 serve a more important
role in substrate orientation and docking than in catalysis. All β-HAD
family members (except 6-PGDH) have a Phe in their primary sequence,
which is identical to the position of F231 in the AtGLYR1 model. The
TSAR model suggests that the steric presence of F231 positions TA
towards the cofactor position [21]; a similar relationship most likely ex-
ists in AtGLYR1. The T95 residue of AtGLYR1 is also characteristically
conserved as Ser, Thr or Asn among all β-HAD family members, indicat-
ing that a small polar residue is probably required at this position in
order to form polar interactions with the substrate molecule. The Kd of
the S121A mutant with glyoxylate was slightly compromised; however,
the catalytic efficiency was not significantly affected (Table 2), suggest-
ing that the S121 hydroxyl is likely important for recognizing and
orienting the substrate.
3.5. Comparison of glyoxylate and hydroxypyruvate reductases and
potential role during abiotic stress
Photorespiratory flux in plants is typically a function of the ambient
levels of CO2 and O2 (Fig. 4, [38]). Abiotic stress conditions such as
drought and salinity can cause stomatal closure, thereby reducing
the internal CO2/O2 ratio and resulting in corresponding increases in
the oxygenation of ribulose 1,5 bisphosphate, glycolate-2-phosphate,
glycolate, and glyoxylate production [39]. In contrast, under hypoxia
the photorespiratory pathway would not be expected to generate
glyoxylate due to limiting glycolate oxidase activity [40]. All three
stresses are likely associated with an elevated NAD(P)H/NAD(P)
ratio [11,12], which could limit the mitochondrial conversion of glycine
into serine via glycine decarboxylase and where appropriate, cause
glyoxylate to further accumulate. The generation of glyoxylate from
hydroxypyruvate via the non-photorespiratory serine pathway might
also be restricted by unfavorable redox conditions during abiotic stress.
Notably, the activities of NAD+-dependent succinic semialdehyde
dehydrogenase and glutamate decarboxylase and NADPH-dependent
Overall, these data support the assignment of K170 as a general acid
in the acid/base catalytic mechanism of AtGLYR1. With this in mind, we
can speculate on the impact of the K170 mutations on AtGLYR1 activity
(Table 2). The K170A mutation would eliminate the side chain of the Lys
residue and therefore any possibility of acid-base catalysis. The side
chain of Glu in the K170E mutant would be less likely to accept a proton
from the active-site water, probably due to its lower pK and shorter side
chain compared to Lys, even though it could form hydrogen-bonding
interactions with glyoxylate. The guanidinium group of Arg in the

G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
2669
Fig. 3. Structural comparison of native AtGLYR1 with the β-HAD family. (A) Superposition of the AtGLYR1 structure (green) with the structures of 3-HIBADH-NADPH (yellow) (PDB ID
2CVZ), 6-PGDH-NADPH (red) (PDB ID 1PGO) and TSAR-TA (blue) (PDB ID 1VPD). Overall ribbon representation of the four structures with the NADPH bound to the 3-HIBADH is
shown in yellow sticks, the NADPH bound to the 6-PGDH is shown in red sticks, and the TA bound to the TSAR is shown in blue sticks. (B) Surface representation of AtGLYR1 with
NADPH is shown in gray ball and stick. The surface is colored according to conservation based on the sequence alignment shown in Fig. 2. (C) ~90° rotation of the view as seen in B showing
the other side of the binding pocket with the bound TA in black ball and stick. (D) Ball and stick representation of conserved residue side chains important for the catalytic activity of the
four structures (colored as in A). NADPH from the HIBADH-NADPH structure is shown in gray and TA from the TSAR-TA structure is shown in black. For clarity the Cα atoms are presented
in slightly larger spheres.
SSA reductase activities would be decreased and increased, respectively,
resulting in increases in γ-aminobutyrate and γ-hydroxybutyrate pro-
duction [41,42].
NADH and glyoxylate over hydroxypyvruvate (HPR3) [43,44]. Based
on substrate specificity, it seems plausible that the NADPH-dependent
cytosolic and plastidial AtGLYRs and AtHPRs could support the optimum
reduction of glyoxylate, as well as hydroxypyruate [12,44]. However,
purified recombinant preparations of both AtGLYR1 and AtGLYR2
strongly favored the reduction of glyoxylate over hydroxypyuvate,
whereas AtHPR2 and AtHPR3 had similar maximal activities with both
glyoxylate and hydroxypyruvate (Table 3). Comparison of primary
sequences indicated that AtHPR2 and AtHPR3 were 45% identical to
each other at the amino acid level, but only 19–25% identical to
AtHPR1, the NADH-dependent form, and 8–9% identical to the AtGLYRs
Hydroxypyruvate reductase (HPR) activity is important in the
recycling of metabolites derived during photorespiration (Fig. 4).
Recent evidence indicated that there are distinct HPR isoforms located
in different cellular compartments and with dual substrate specificity:
HPR1 is a peroxisomal protein that prefers NADH over NADPH and
hydroxypyruvate over glyoxylate; HPR2 is a cytosolic protein that pre-
fers NADPH over NADH but utilizes hydroxypyruvate and glyoxylate
similarly; and HPR3 is a plastidial protein that prefers NADPH over

2670
G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
Fig. 4. Enzymatic reduction of glyoxylate via GLYR and HPR activities and potential relationships between photorespiration and γ-aminobutyrate metabolism. Pathway intermediate
abbreviations are: Ala, alanine; GABA, γ-aminobutyrate; GHB, γ-hydroxybutyrate; 2-OG, 2-oxoglutarate; Pyr, pyruvate. Enzyme abbreviations are indicated in italics as: AGT, alanine/
glyoxylate transaminase; GABA-T, GABA transaminase; GAD, glutamate decarboxylase; GDC, glycine decarboxylase; GGT, glutamate:glyoxylate transaminase; GK, glycerate kinase;
GLYR/SSAR, glyoxylate/succinic semialdehyde reductase; GO, glycolate oxidase; GP, glycerate-3-P phosphatase; HPR, hydroxypyruvate reductase; PGP, phosphoglycolate phosphatase;
SGT, serine:glyoxylate transaminase; SHMT, serine hydroxymethyl transferase; SSADH; succinic semialdehyde dehydrogenase; TCA, tricarboxylic acid. The primary metabolites and
enzymes under consideration are indicated in black/white balloons and squares, respectively. The thick arrow represents reactions of glycolysis, and the question mark represents uncer-
tainty about the fate of GHB [38,39].
(see Fig. S3 in the supplemental material). Furthermore, none of the
AtHPRs contained the active-site residues conserved in AtGLYR1 and
AtGLYR2, indicating that the sites responsible for reducing glyoxylate
differ greatly between the AtGLYRs and AtHPRs.
characterization of AtGLYR1 should contribute to our understanding of
the structure-function relations in the small β-HAD protein family and
to identification of new family members.
A general acid/base kinetic mechanism is a common feature of
oxidoreductase enzymes [45] and previous studies of 6-PGDH have pro-
posed that the oxidation of 6-phosphogluconate to ribulose 5-phosphate
utilizes K183 as a general base for deprotonation and stabilization of
transition state intermediates [14,17]. In this paper, we proposed that
the active site of AtGLYR1 is an excellent candidate for general acid/
base catalysis and addressed the relative kinetic importance in catalysis
and binding of five potentially active-site residues (i.e., F231, S121,
K170, N174, D239) and a conserved hydroxyl in a Thr residue.
Sequence and activity comparisons indicated that AtGLYR1 and
AtGLYR2 possess structural features that are absent in the AtHPRs and
that these features could account for their stronger preference for
glyoxylate over hydroxypyruvate. On this basis, it is possible that the
AtGLYRs serve a more important role than the AtHPRs in glyoxylate
detoxification and the recycling of reducing equivalents during the
4. Conclusion
Despite low sequence identity, site-directed mutagenesis and struc-
tural information revealed that AtGLYR1 shares common secondary
structure, 3-D architecture and functionally important amino acid resi-
dues with the β-HAD family, which includes 3-HIBADH, TSAR, 6-PGDH
and 2-(hydroxymethyl)glutarate dehydrogenase. The data reported
here establishes AtGLYR1 as the fifth member of the β-HAD family of
proteins with known function and as the first family member reported
to catalyze the production of either a β-hydroxy (i.e., glycolate from
glyoxylate) or a γ-hydroxyacid (i.e., γ-hydroxybutyrate from SSA) [4].
Unfortunately, it was not possible in this paper to assess the mechanistic
role of the important amino acid residues in SSA reduction. Further
Table 3
Substrate specificity for NADPH-dependent activities of known recombinant native AtGLYRs and AtHPRs. Kinetic parameters for AtGLYR1 with glyoxylate as the substrate were
recalculated from Table 2. Kinetic parameters for AtGLYR1 with hydroxypyruvate as the substrate and for AtGLYR2 with glyoxylate as the substrate were determined using one typical
biological replicate. Values for maximal activity of the AtHPRs for glyoxylate or hydroxypyruvate were taken from published literature [43,44] and represent the mean of three biological
replicates.
Enzyme (subcellular location)
Substrate
Glyoxylate
Hydroxypyruvate
Vmax or maximal activity
(μmol min−1 mg−1 prot)
Km
(μM)
Vmax or maximal activity
(μmol min−1 mg−1 prot)
Km
(mM)
AtGLYR1 (cytosol)
AtGLYR2 (plastid)
AtHPR1 (peroxisome)
AtHPR2 (cytosol)
AtHPR3 (plastid)
109
25
0.26
4.6
18
16
2.8
–
6.1
–
a
–
12
5.5
1.1
–
–
–
–
2.3
–
a
Data not available.

G.J. Hoover et al. / Biochimica et Biophysica Acta 1834 (2013) 2663–2671
2671
[16] E. Tetaud, S. Hanau, J.M. Wells, R.W. Le Page, M.J. Adams, S. Arkison, M.P. Barrett,
6-Phosphogluconate dehydrogenase from Lactococcus lactis: a role for arginine
residues in binding substrate and coenzyme, Biochem. J. 338 (1999) 55–60.
[17] L. Zhang, L. Chooback, P.F. Cook, Lysine 183 is the general base in the
6-phosphogluconate dehydrogenase-catalyzed reaction, Biochemistry 38 (1999)
11231–11238.
[18] L. Li, F.S. Dworkowski, P.F. Cook, Importance in catalysis of the 6-phosphate-binding
site of 6-phosphogluconate in sheep liver 6-phosphogluconate dehydrogenase,
J. Biol. Chem. 281 (2006) 25568–25576.
response to abiotic stress. Recent research demonstrated that the dele-
tion of the NADPH-dependent AtHPRs influences the photorespiratory
phenotype [43,44], but further work is required to assess the involve-
ment of the AtGLYRs in photorespiration and during abiotic stress.
These findings have potential implications for engineering stress resis-
tance in important crop plants.
[19] R. Sundaramoorthy, J. Iulek, M.P. Barrett, O. Bidet, G.F. Ruda, I.H. Gilbert, W.N.
Hunter, Crystal structures of a bacterial 6-phosphogluconate dehydrogenase reveal
aspects of specificity, mechanism and mode of inhibition by analogues of high-
energy reaction intermediates, FEBS J. 274 (2007) 275–286.
[20] N.K. Lokanath, N. Ohshima, K. Takio, I. Shiromizu, C. Kuroishi, N. Okazaki, S.
Kuramitsu, S. Yokoyama, M. Miyano, N. Kunishima, Crystal structure of novel
NADP-dependent 3-hydroxyisobutyrate dehydrogenase from Thermus thermophilus
HB8, J. Mol. Biol. 352 (2005) 905–917.
[21] J. Osipiuk, M. Zhou, S. Moy, F. Collart, A. Joachimiak, X-ray crystal structure of
GarR-tartronate semialdehyde reductase from Salmonella typhimurium, J. Struct.
Funct. Genomics 10 (2009) 249–253.
[22] S. Reitz, A. Alhapel, L.-O. Essen, A.J. Pierik, Structural and kinetic properties of a
beta-hydroxyacid dehydrogenase involved in nicotinate fermentation, J. Mol. Biol.
382 (2008) 802–811.
Acknowledgments
This research was supported by funds from the Natural Sciences and
Engineering Research Council of Canada to B.J.S. [grant number 42718-
2009] and A.R.M. [grant number 105440-2008], the Canadian Institutes
of Health Research to R.J. [fellowship number MFE-78132], and the
Ontario Ministry of Agriculture Food and Rural Affairs to B.J.S. [grant
number 200131].
Appendix A. Supplementary data
[23] B.K. Beattie, G.A. Prentice, A.R. Merrill, Investigation into the catalytic role for
the tryptophan residues within domain III of Pseudomonas aeruginosa exotoxin A,
Biochemistry 35 (1996) 15134–15142.
[24] A. Vagin, A. Teplyakov, MOLREP: an automated program for molecular replacement,
J. Appl. Crystallogr. 30 (1997) 1022–1025.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.bbapap.2013.09.013.
[25] A. Perrakis, R. Morris, V.S. Lamzin, Automated protein model building combined
with iterative structure refinement, Nat. Struct. Biol. 6 (1999) 458–463.
[26] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta
Crystallogr. D Biol. Crystallogr. 60 (2004) 2126–2132.
References
[1] I. Zelitch, Oxidation and reduction of glycolic and glyoxylic acids in plants. II.
Glyoxylic acid reductases, J. Biol. Chem. 201 (1953) 719–726.
[2] I. Zelitch, A.M. Gotto, Properties of a new glyoxylate reductase from leaves, Biochem.
J. 84 (1962) 541–546.
[3] L.A. Kleczkowski, D.D. Randall, D.G. Blevins, Purification and characterization of a novel
NADPH (NADH)-dependent glyoxylate reductase from spinach leaves. Comparison on
immunological properties of leaf glyoxylate reductase and hydroxypryruvate reduc-
tases, Biochem. J. 239 (1986) 653–659.
[4] C.V. Givan, L.A. Kleczkowski, The enzymic reduction of glyoxylate and hydroxypyruvate
in leaves of higher plants, Plant Physiol. 100 (1992) 552–556.
[5] L.A. Kleczkowski, Kinetics and regulation of the NAD(P)H-dependent glyoxylate-
specific reductase from spinach leaves, Z. Naturforsch. 50c (1995) 21–28.
[6] K.E. Breitkreuz, W.L. Allan, O.R. Van Cauwenberghe, C. Jakobs, D. Talibi, B. André, B.J.
Shelp, A novel γ-hydroxybutyrate dehydrogenase: identification and expression of
an Arabidopsis cDNA and potential role under oxygen deficiency, J. Biol. Chem. 278
(2003) 41552–41556.
[27] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of macromolecular structures
by the maximum-likelihood method, Acta Crystallogr. D Biol. Crystallogr. 53 (1997)
240–255.
[28] Z. Otwinowski, W. Minor, Processing of x-ray diffraction data collected in oscillation
mode, Methods Enzymol. 276 (1997) 307–326.
[29] P.D. Adams, R.W. Grosse-Kunstleve, L.W. Hung, T.R. Ioerger, A.J. Mc-Coy, N.W.
Moriarty, R.J. Read, J.C. Sacchettini, N.K. Sauter, T.C. Terwilliger, PHENIX: building
new software for automated crystallographic structure determination, Acta Crystallogr.
D Biol. Crystallogr. 58 (2002) (2002) 1948–1954.
[30] E. Krissinel, K. Henrick, Inference of macromolecular assemblies from crystalline
state, J. Mol. Biol. 372 (2007) 774–797.
[31] M.A. Larkin, G. Blackshields, N.P. Brown, R. Chenna, P.A. McGettigan, H. McWilliam,
F. Valentin, I.M. Wallace, A. Wilm, R. Lopez, J.D. Thompson, T.J. Gibson, D.G. Higgins,
Clustal W and Clustal X version 2.0, Bioinformatics 23 (2007) 2947–2948.
[32] G.J. Kleywegt, Use of non-crystallographic symmetry in protein structure refine-
ment, Acta Crystallogr. D Biol. Crystallogr. 52 (1996) 842–857.
[7] G.J. Hoover, O.R. Van Cauwenberghe, K.E. Breitkreuz, S.M. Clark, A.R. Merrill, B.J.
Shelp, Characteristics of an Arabidopsis glyoxylate reductase: general biochemical
properties and substrate specificity for the recombinant protein, and developmental
expression and implications for glyoxylate and succinic semialdehyde metabolism
in planta, Can. J. Bot. 85 (2007) 883–895.
[33] W.L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, Palo Alto,
California, USA, 2002.
[34] R.K. Njau, C.A. Herndon, J.W. Hawes, Novel beta-hydroxyacid dehydrogenases
in Escherichia coli and Haemophilus influenzae, J. Biol. Chem. 275 (2000)
38780–38786.
[8] G.J. Hoover, G.A. Prentice, A.R. Merrill, B.J. Shelp, Kinetic mechanism of an Arabidopsis
glyoxylate reductase: studies of initial velocity, dead-end inhibition and product
inhibition, Can. J. Bot. 85 (2007) 896–902.
[35] S. Hanau, K. Montin, C. Cervellati, M. Magnani, F. Dallacchio, 6-Phosphogluconate
dehydrogenase mechanism: evidence for allosteric modulation by substrate, J. Biol.
Chem. 285 (2010) 21366–21371.
[9] J.P. Simpson, R. DiLeo, P. Dhanoa, W.L. Allan, A. Makhmoudova, S.M. Clark, G.J. Hoover,
[36] R.K. Njau, C.A. Herndon, J.W. Hawes, New developments in our understanding of the
beta-hydroxyacid dehydrogenases, Chem. Biol. Interact. 130 (2001) 785–791.
[37] Collaborative computational project, number 4, Acta Crystallogr. D Biol. Crystallogr.
50 (1994) 760–763.
[38] H. Bauwe, M. Hagemann, A.R. Fernie, Photorespiration: players, partners and origin,
Trends Plant Sci. 15 (2010) 330–336.
R.T. Mullen, B.J. Shelp, Identification and characterization of
a plastid-localized
Arabidopsis glyoxylate reductase isoform: comparison with a cytosolic isoform and
implications for cellular redox homeostasis and aldehyde detoxification, J. Exp. Bot.
58 (2008) 2545–2554.
[10] S.L.K. Ching, S.K. Gidda, A. Rochon, O.R. Van Cauwenberghe, B.J. Shelp, R.T. Mullen,
Glyoxylate reductase isoform 1 is localized in the cytosol and not peroxisomes in
plant cells, J. Integr. Plant Biol. 54 (2012) 152–168.
[39] W.L. Allan, S.M. Clark, G.J. Hoover, B.J. Shelp, Role of plant glyoxylate reductases
during stress: a hypothesis, Biochem. J. 423 (2009) 15–22.
[11] W.L. Alan, J.P. Simpson, S.M. Clark, B.J. Shelp, γ-Hydroxybutyrate accumulation in
Arabidopsis and tobacco plants is a general response to abiotic stress: putative reg-
ulation by redox balance and glyoxylate reductase isoforms, J. Exp. Bot. 59 (2008)
2555–2564.
[12] W.L. Allan, K.E. Breitkreuz, J.C. Waller, J.P. Simpson, G.J. Hoover, A. Rochon, D.J. Wolyn, D.
Rentsch, W.A. Snedden, B.J. Shelp, Detoxification of succinate semialdehyde in
Arabidopsis glyoxylate reductase and NAD kinase mutants subjected to submergence
stress, Botany 90 (2012) 51–61.
[40] R. Narsai, K.A. Howell, A. Carroll, A. Ivanova, A.H. Millar, J. Whelan, Defining core
metabolic and transcriptomic responses to oxygen availability in rice embryos and
young seedlings, Plant Physiol. 151 (2009) 306–322.
[41] B.J. Shelp, G.G. Bozzo, C.P. Trobacher, G. Chiu, V.S. Bajwa, Strategies and tools for
studying the metabolism and function of γ-aminobutyrate in plants. I. Pathway
structure, Botany 90 (2012) 651–668.
[42] B.J. Shelp, G.G. Bozzo, A. Zarei, J.P. Simpson, C.P. Trobacher, W.L. Allan, Strategies
and tools for studying the metabolism and function of γ-aminobutyrate in plants.
II. Integrated analysis, Botany 90 (2012) 781–793.
[13] J.W. Hawes, E.T. Harper, D.W. Crabb, R.A. Harris, Structural and mechanistic similarities
of 6-phosphogluconate and 3-hydroxyisobutyrate dehydrogenases reveal
a
new
[43] S. Timm, A. Nunes-Nesi, T. P rnik, K. Morgenthal, S. Wienkoop, O. Keerberg, W.
Weckwert, L.A. Kleczkowski, A.R. Fernie, H. Bauwe, A cytosolic pathway for the
conversion of hydroxypyruvate to glycerate during photorespiration in Arabidopsis,
Plant Cell 20 (2008) 2848–2859.
[44] S. Timm, A. Florian, K. Jahnke, A. Nunes-Nesi, A.R. Fernie, H. Bauwe, The
hydroxypyruvate-reducing system in Arabidopsis: multiple enzymes for the same
end, Plant Physiol. 155 (2011) 694–705.
enzyme family, the 3-hydroxyacid dehydrogenases, FEBS Lett. 389 (1996) 263–267.
[14] M.J. Adams, G.H. Ellis, S. Gover, C.E. Naylor, C. Phillips, Crystallographic study of
coenzyme, coenzyme analogue and substrate binding in 6-phosphogluconate
dehydrogenase: implications for NADP specificity and the enzyme mechanism,
Structure 2 (1994) 651–668.
[15] W.E. Karsten, L. Chooback, P.F. Cook, Glutamate 190 is a general acid in the
6-phosphogluconate dehydrogenase-catalyzed reaction, Biochemistry 37 (1998)
15691–15697.
[45] C.P. Nicholas, S. Lewis, Fundamentals of Enzymology: The Cell and Molecular
Biology of Catalytic Proteins, third ed. Oxford University Press, Oxford, U.K., 1999.