Rabbit 3-hydroxyhexobarbital dehydrogenase is a NADPH-preferring reductase with broad substrate specificity for ketosteroids, prostaglandin D2, and other endogenous and xenobiotic carbonyl compounds

Article abstract of DOI:10.1016/j.bcp.2013.08.024

3-Hydroxyhexobarbital dehydrogenase (3HBD) catalyzes NAD(P) ⁺-linked oxidation of 3-hydroxyhexobarbital into 3-oxohexobarbital. The enzyme has been thought to act as a dehydrogenase for xenobiotic alcohols and some hydroxysteroids, but its physiological function remains unknown. We have purified rabbit 3HBD, isolated its cDNA, and examined its specificity for coenzymes and substrates, reaction directionality and tissue distribution. 3HBD is a member (AKR1C29) of the aldo-keto reductase (AKR) superfamily, and exhibited high preference for NADP(H) over NAD(H) at a physiological pH of 7.4. In the NADPH-linked reduction, 3HBD showed broad substrate specificity for a variety of quinones, ketones and aldehydes, including 3-, 17- and 20-ketosteroids and prostaglandin D₂, which were converted to 3α-, 17β- and 20α-hydroxysteroids and 9α,11β- prostaglandin F₂, respectively. Especially, α-diketones (such as isatin and diacetyl) and lipid peroxidation-derived aldehydes (such as 4-oxo- and 4-hydroxy-2-nonenals) were excellent substrates showing low K_m values (0.1-5.9 μM). In 3HBD-overexpressed cells, 3-oxohexobarbital and 5β-androstan-3α-ol-17-one were metabolized into 3-hydroxyhexobarbital and 5β-androstane-3α,17β-diol, respectively, but the reverse reactions did not proceed. The overexpression of the enzyme in the cells decreased the cytotoxicity of 4-oxo-2-nonenal. The mRNA for 3HBD was ubiquitously expressed in rabbit tissues. The results suggest that 3HBD is an NADPH-preferring reductase, and plays roles in the metabolisms of steroids, prostaglandin D₂, carbohydrates and xenobiotics, as well as a defense system, protecting against reactive carbonyl compounds.

Full text of DOI:10.1016/j.bcp.2013.08.024

Biochemical Pharmacology 86 (2013) 1366–1375
Contents lists available at ScienceDirect
Biochemical Pharmacology
journal homepage: www.elsevier.com/locate/biochempharm
Rabbit 3-hydroxyhexobarbital dehydrogenase is a NADPH-preferring
reductase with broad substrate specificity for ketosteroids,
prostaglandin D2, and other endogenous and xenobiotic carbonyl
compounds
a
a
a
Satoshi Endo a, , Toshiyuki Matsunaga , Atsuko Matsumoto , Yuki Arai ,
Satoshi Ohno b, Ossama El-Kabbani c, Kazuo Tajima d, Yasuo Bunai e,
Shigeru Yamano f, Akira Hara b,e, Yukio Kitade b
*
a Gifu Pharmaceutical University, Gifu 502-1196, Japan
b Faculty of Engineering, Gifu University, Gifu 501-1193, Japan
c Victorian College of Pharmacy, Monash University, Parkville, Victoria 3052, Australia
d Faculty of Pharmaceutical Sciences, Hokuriku University, Kanazawa 920-1181, Japan
e Graduate School of Medicine, Gifu University, Gifu 501-1193, Japan
f Faculty of Pharmaceutical Sciences, Fukuoka University, Fukuoka 814-0180, Japan
A R T I C L E I N F O
A B S T R A C T
3-Hydroxyhexobarbital dehydrogenase (3HBD) catalyzes NAD(P)+-linked oxidation of 3-hydroxyhex-
obarbital into 3-oxohexobarbital. The enzyme has been thought to act as a dehydrogenase for xenobiotic
alcohols and some hydroxysteroids, but its physiological function remains unknown. We have purified
rabbit 3HBD, isolated its cDNA, and examined its specificity for coenzymes and substrates, reaction
directionality and tissue distribution. 3HBD is a member (AKR1C29) of the aldo-keto reductase (AKR)
superfamily, and exhibited high preference for NADP(H) over NAD(H) at a physiological pH of 7.4. In the
NADPH-linked reduction, 3HBD showed broad substrate specificity for a variety of quinones, ketones
Article history:
Received 27 June 2013
Accepted 14 August 2013
Available online 27 August 2013
Keywords:
Aldo-keto reductase
Ketosteroid reductase
3-Hydroxyhexobarbital dehydrogenase
4-Oxo-2-nonenal
and aldehydes, including 3-, 17- and 20-ketosteroids and prostaglandin D2, which were converted to 3
, 17 - and 20 -hydroxysteroids and 9 ,11 -prostaglandin F2, respectively. Especially, -diketones
(such as isatin and diacetyl) and lipid peroxidation-derived aldehydes (such as 4-oxo- and 4-hydroxy-2-
nonenals) were excellent substrates showing low Km values (0.1–5.9 M). In 3HBD-overexpressed cells,
3-oxohexobarbital and 5 -androstan-3 -ol-17-one were metabolized into 3-hydroxyhexobarbital and
-androstane-3 ,17 -diol, respectively, but the reverse reactions did not proceed. The overexpression
a-
b
a
a
b
a
Prostaglandin D2
m
b
a
5
b
a
b
of the enzyme in the cells decreased the cytotoxicity of 4-oxo-2-nonenal. The mRNA for 3HBD was
ubiquitously expressed in rabbit tissues. The results suggest that 3HBD is an NADPH-preferring
reductase, and plays roles in the metabolisms of steroids, prostaglandin D2, carbohydrates and
xenobiotics, as well as a defense system, protecting against reactive carbonyl compounds.
ß 2013 Elsevier Inc. All rights reserved.
to 3-oxohexobarbital (3OB) by 3-hydroxyhexobarbital dehydroge-
nase (3HBD) (Fig. 1) [1,2]. While 3HB is excreted after conjugation
with glucuronic acid, electrophilic 3OB reacts with glutathione to
produce 1,5-dimethylbarbituric acid and cyclohexenone-glutathi-
one adduct [3,4]. Because 3HBD is involved in the formation of the
reactive 3OB as well as the hexobarbital metabolism, it has been
purified and characterized from liver cytosols of rabbits [5–7],
guinea-pigs [8,9], mice [10] and golden hamsters [11]. The
enzymes are monomeric proteins with molecular weights of
31–42 kDa, and oxidize several other alicyclic alcohols including
1. Introduction
Hexobarbital, a short time-acting hypnotic, is metabolized to
3-hydroxyhexobarbital (3HB) by cytochrome P450, and then
Abbreviations: 3HB, 3-hydroxyhexobarbital; 3OB, 3-oxohexobarbital; 3HBD, 3-
hydroxyhexobarbital dehydrogenase; AKR, aldo-keto reductase; HSD, hydroxys-
teroid dehydrogenase; PG, prostaglandin; TBE, 6-tert-butyl-2,3-epoxy-5-cyclohex-
ene-1,4-dione; 4R-TBEH, 6-tert-butyl-2,3-epoxy-4(R)-hydroxy-5-cyclohexen-1-
one; RT, reverse transcription; BAEC, bovine aortic endothelial cell; LC–ESI-MS,
liquid chromatography–electrospray ionization-mass spectrometry.
some 3a- and/or 17b-hydroxysteroids in the presence of either
NAD+ or NADP+ as the coenzyme. However, they differ from one
another in their reactivity toward the substrates. In the reactivity
*
Corresponding author at: Laboratory of Biochemistry, Gifu Pharmaceutical
University, Gifu 501-1196, Japan. Tel.: +81 58 230 8100; fax: +81 58 230 8105.
toward the
a- and b-isomers of 3HB (a-3HB and b-3HB,
E-mail address: sendo@gifu-pu.ac.jp (S. Endo).
0006-2952/$ – see front matter ß 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.bcp.2013.08.024

S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
1367
Fig. 1. Metabolism of hexobarbital into 3HB and 3OB. Hexobarbital is oxidized by cytochrome P450 (CYP) into
a- and
b-isomers of 3HB, which are converted into 3OB by
3HBD. The structures of (+)-hexobaribital and its metabolites are shown.
respectively, Fig. 1), all the enzymes oxidize
preferentially than -3HB, but the activities of the mouse and
hamster enzymes toward -3HB are much lower compared to
those of the other animal enzymes. The mouse and hamster
enzymes also oxidize trans-benzene dihydrodiol that is
a
-3HB more
the metabolism of the substrates in the enzyme-overexpressed
cells. Our data show that AKR1C29 acts as a NADPH-preferring
reductase with a wide range of both aromatic and aliphatic
carbonyl compounds including ketosteroids, prostaglandin (PG) D2
and lipid peroxidation-derived aldehydes.
b
b
a
representative substrate of dihydrodiol dehydrogenase [10,11].
Guinea-pig 3HBD is distinguishable from the other animal
enzymes in its low activity for other xenobiotic alicyclic alcohols
2. Materials and methods
(such as 1-indanol and 1-tetralol) and high activity for 17
hydroxy-C19-steroids, thereby suggesting its identity with testos-
terone 17 -dehydrogenase (EC. 1.1.1.64) [8,9]. The mouse enzyme
also exhibits 17 -hydroxysteroid dehydrogenase (HSD) activity
for 17 -hydroxy-C19-steroids and 17 -estradiol [10] that is not
the substrate for the guinea-pig enzyme [8]. The rabbit enzyme
oxidizes limited 17 -hydroxysteroids such as 5 -androstane-
,17 -diol [7]. The hamster enzyme exhibits 3 -HSD activity
toward bile acids, in addition to the 17 -HSD activity for 5
androstane-3 ,17 -diol [11]. The enzymatic properties of mam-
b
-
2.1. Chemicals
b
Steroids were obtained from Sigma Chemicals (Perth, WA) and
Steraloids (Newport, RI), and PGs, 4-oxo-2-nonenal and 4-hydroxy-
2-nonenal were from Cayman Chemicals (Ann Arbor, MI). 3OB [5],
trans-benzene dihydrodiol [17], 4-oxo-2-nonenol [18] geranylger-
anial [19], 6a/b-naloxols [20,21], 6-tert-butyl-2,3-epoxy-5-cyclo-
hexene-1,4-dione (TBE) [22], 6-tert-butyl-2,3-epoxy-4(R)-hydroxy-
5-cyclohexen-1-one (4R-TBEH) and its 4S-isomer (4S-TBEH) [23]
b
b
b
b
b
a
3a
b
b
b-
a
b
were synthesized as described previously. a- and b-3HBs were
malian 3HBDs were studied under optimal pH conditions of 8.0–
10.5, where their NAD+-linked dehydrogenase activities are higher
than the NADP+-linked activities. Therefore, 3HBD has been
reported to act as a dehydrogenase for the above xenobiotic and
steroidal alcohols, but its physiological role has not yet been
clarified. Among the mammalian 3HBDs, only hamster 3HBD is so
far molecularly cloned and shown to be a member of the aldo-keto
reductase (AKR) superfamily [11]. This suggests a possibility that
the mammalian enzymes reduce carbonyl compounds, but their
substrate specificities in the reduction reaction have not yet been
thoroughly characterized.
kindly denoted by Dr R. Takenoshita (Fukuoka University, Japan). 3-
Deoxyglucosone and befunolol were gifts from Nippon Zoki
Pharmaceutical Co. (Osaka, Japan) and Kaken Pharmaceutical Co.
(Tokyo, Japan), respectively. All other chemicals were of the highest
grade that could be obtained commercially.
2.2. cDNA isolation and RT-PCR analysis
A cDNA for AKR1C29 was isolated from a total RNA sample of
small intestine of a Japanese white rabbit by reverse transcription
(RT)-PCR. The preparation of total RNA, RT, and DNA techniques
followed the standard procedures described by Sambrook et al.
[24]. The PCR was performed using Pfu DNA polymerase (Agilent
Technologies, Santa Clara, CA) and a pair of sense and antisense
primers, 50-ttttcatatgatggatcccaagcatcagcgttatgggcatt-30 and 50-
ttttgtcgactcaatattcatcagaaaatgggtaattt-30, which contain under-
lined NdeI and SalI sites, respectively. PCR amplification consisted
of an initial denaturation step at 94 8C for 5 min, 29 cycles of
denaturation (94 8C/30 s), annealing (60 8C/30 s) and extension
(72 8C/2 min), and a final incubation step at 72 8C for 5 min. The
PCR products were purified, digested with the two restriction
enzymes (Life Technologies, Carlsbad, CA), and ligated into the
pCold I vectors (Takara Bio Inc., Otsu, Japan) that had been digested
with the two restriction enzymes [25]. The sequence of the coding
region of the cloned cDNA was analyzed using a Beckman Coulter
CEQ8000XL DNA sequencer. The 972-base pair sequence of the
In rabbits, hepatic 3HBD was reported to be identical to indanol
dehydrogenase [6]. In contrast, six multiple forms of indanol
dehydrogenase associated with 17
the liver cytosol [12]. In addition, cytosolic 17
dihydrodiol dehydrogenase with 17 -HSD activity [15] exist in
b
-HSD activity are isolated from
b
-HSD [13,14] and
b
multiple forms in rabbit liver. Furthermore, the recent rabbit
genomic analysis has predicted several genes that encode HSD-like
proteins belonging to the AKR1C subfamily, three of which have
recently been identified as NAD+-preferring
3a/17b-HSDs
(AKR1C26-AKR1C28) [16]. In this study, we isolated the cDNA
for rabbit 3HBD that has been assigned AKR1C29 in the AKR
superfamily. To elucidate the physiological role of AKR1C29 and its
functional relationship with the enzymes that metabolize steroids
and xenobiotic alcohols, we characterized the properties of the
recombinant AKR1C29 at a physiological pH of 7.4, and examined

1368
S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
cDNA was deposited in DDBJ database with the accession number
AB821295.
In the RT-PCR analysis of the expression of mRNA for AKR1C29
in rabbit tissues, the preparation of total RNA samples, RT and PCR
0.1-M potassium phosphate, pH 7.4. The substrate and products
were extracted into 4-mL ethyl acetate 30 min after the reaction
was started at 37 8C. The products of oxidoreduction of steroids
[25] and reduction of PGD2 [28], farnesal [29] and 4-oxo-2-
nonenal [18] were analyzed by TLC, as described. The reduced
products of TBE were identified by the HPLC methods [23]. The
were carried out as described above. The cDNA for rabbit
b-actin
was also amplified as an internal control with a pair of sense and
antisense primers, 50-ccggcttcgcgggcgacg-30 and 50-tcccggccagc-
caggtcc-30, respectively. The PCR products were separated by
agarose gel electrophoresis, and revealed with ethidium bromide.
products of 3HB oxidation, 3OB reduction,
5b-androstane-
3
a
,17 -diol oxidation and 5 -androstan-3 -ol-17-one reduc-
b
b
a
tion were analyzed by the liquid chromatography–electrospray
ionization-mass spectrometry (LC–ESI-MS) using a Hewlett-
Packard HP 1100 Series LC/MSD system attached with a diode
2.3. Enzyme purification
array detector and
a column (Mightysil RP-18 GP 5 mm,
The expression construct was transfected into Escherichia coli
BL21 (DE3) pLysS, and the recombinant AKR1C29 fused to the N-
terminal 6-His tag was expressed and purified to homogeneity
from the cell extracts using the nickel-charged Sepharose 6FF resin
(GE Healthcare, Chalfont St Giles, UK) as described previously [25].
3HBD was purified to homogeneity from the rabbit liver (50 g) by
the methods of Takenoshita and Toki [5], except that Q-Sepharose
(GE Healthcare) was used instead of TEAE-cellulose. The protein
concentrations of the enzymes were determined with Bradford
reagent using bovine serum albumin as the standard.
4.6 mm ꢀ 250 mm, Kanto Chemical Co., Tokyo, Japan). Separa-
tions were carried out at a flow rate of 0.5 mL/min and 40 8C
using the following mobile phases: 25% acetonitrile aqueous
solution containing 0.1% formic acid for 3OB and
80% acetonitrile aqueous solution containing 0.1% formic acid
for the two steroids. 3OB, -3HB, -3HB, 5 -androstan-3 -ol-
17-one and -androstane-3 ,17 -diol were detected by
a/b-3HBs, and
a
b
b
a
5
b
a
b
monitoring their total ions (m/z 249.1, 251.1, 251.1, 289.4
and 291.4, respectively) in the negative ESI mode, and eluted at
the retention times of 20.1, 17.6, 16.8, 14.9 and 12.7 min,
respectively. The detection limits of 3OB,
a/b-3HBs and the two
2.4. Mass spectrometry
steroids were 0.1, 0.1 and 1 nmol, respectively.
The identity of the purified hepatic 3HBD with AKR1C29 was
examined by the mass spectrometry method. The purified protein
(0.1 mg) was digested by lysylendopeptidase, and then the digest
was analyzed using a Bruker-Franzen Mass Spectrometry System
2.7. Cell culture experiments
Bovine aortic endothelial cells (BAECs) were generous gift from
Taisho Pharmaceutical Co. (Saitama, Japan), and cultured in
Dulbecco’s modified Eagle medium supplemented with 10% fetal
bovine serum, penicillin (100 units/mL) and streptomycin (0.1 mg/
mL) at 37 8C in a humidified incubator containing 5% CO2. The
construction and transfection of the pGW1 plasmids harboring the
cDNA for AKR1C29 into BAECs were performed as described
previously [30]. Briefly, the cDNA was amplified by PCR using the
primers. The sense primer, 50-ccccgaattcGCCACCatggatcccaagcat-
cagcg-30, contains an EcoRI site, a Kozak sequence and a start
codon, which are shown in underlined, capital and italic letters,
respectively. The antisense primer was the same as that used for
the cDNA cloning. After the sequences of the PCR products were
verified, they were subcloned at the EcoRI and SalI sites of the
pGW1 expression vector. Using a Lipofectamine 2000 reagent (Life
Technologies), the cells were transfected with the expression
plasmids. The transfected cells were maintained in the medium
(2 mL) containing 2% fetal bovine serum for 24 h, and then
Ultraflex TOF/TOF with a saturated
a-cyano-4-hydroxycinnamic
acid matrix as described previously [26].
2.5. Assay of enzyme activity
The dehydrogenase activities for the enzymes were assayed by
measuring the rate of change in fluorescence (at 455 nm with an
excitation wavelength of 340 nm) or absorbance (at 340 nm) of
NAD(P)H. The corresponding standard reaction mixture consisted
of 0.1-M potassium phosphate buffer, pH 7.4, 0.25-mM NADP+ or
2.0-mM NAD+, substrate and enzyme, in a total volume of 2.0 mL.
The reductase activities, except for all-trans-retinal reductase
activity, were determined by measuring the rate of change in
NAD(P)H absorbance in the phosphate buffer, pH 7.4, containing
0.1-mM NADH or NADPH and an appropriate amount of carbonyl
substrate. The assay of all-trans-retinal reductase activity was
´
`
performed according to the method of Pares and Julia [27], except
that 0.1-M potassium phosphate, pH 7.4, containing 0.01% Tween
80 was added as the buffer. One unit of enzyme activity was
defined as the amount of enzyme that catalyzes the formation or
incubated with
a
-3HB, 3OB (each 200
m
M), 5
b
-androstan-3
a-ol-
17-one or 5 -androstane-3
b
a
,17 -diol (each 100 m
b
M) at different
times. The culture medium was collected by centrifugation, and
the lipidic fraction of the medium was extracted twice by 4-mL
ethyl acetate. The extract was evaporated to dryness, and the
oxidation of 1
mmol NAD(P)H per min at 25 8C.
The apparent Km and kcat values for coenzymes and substrates
were determined by fitting the initial velocities to the Michaelis–
Menten equation. The IC50 values for inhibitors were determined
with 2-mM b-ionol as the substrate in the above standard reaction
mixture. The inhibitor constant, Ki, was estimated from the Dixon
residue was dissolved in methanol (0.1 mL). The portion (10 mL) of
the sample was analyzed by the LC–ESI-MS method as described
above. The amounts of the eluted substrates and metabolites were
quantified from their peak areas using the standard curves
constructed with the authentic samples. The effects of the
AKR1C29 overexpression on the cytotoxicity of 4-oxo-2-nonenal
were examined using the above BAECs as described previously
[25].
plot and/or Cornish–Bowden plot of the velocities obtained in the
b
-ionol range (0.5–20 ꢀ Km) with three concentrations of the
inhibitor. The kinetic constants and IC50 values are expressed as the
means of at least three determinations. The standard deviations of
the determinations were less than 15%, unless otherwise noted.
3. Results
2.6. Product identification
3.1. Amino acid sequence and cDNA cloning of 3HBD
The reaction was conducted at 37 8C in a 2.0-mL reaction
mixture, containing coenzyme (1-mM NADP+ or 0.1-mM
NADPH), substrate (0.05–0.1 mM), enzyme (0.1–0.3 mg), and
3HBD was purified from rabbit liver and its 13 lysylendopepti-
dase-digested peptides were subjected to amino acid sequence
analysis by the mass spectrometry (Fig. 2). The sequences

S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
1369
Fig. 2. Amino acid sequence of AKR1C29. The underlined residues are identical to those determined by peptide sequencing of rabbit liver 3HBD. The NAD(P)(H)- and
substrate-binding residues of the structurally evaluated AKR1C subfamily enzymes [33,39–42] are shown by open and closed arrowheads, respectively.
(composed of a total 213 residues) coincided with that of an
Oryctolagus cuniculus PGE2 9-reductase like protein (NCBI acces-
sion no. XP_002721727.1) among the AKR1C subfamily proteins
predicted from the rabbit genomic analysis. This protein has been
assigned as AKR1C29 in the AKR superfamily. The cDNA for
AKR1C29 was cloned by RT-PCR, and its nucleotide sequence was
identical to that predicted from the genomic analysis, with the
exception of two nucleotide replacements (847A ! C and 855C ! T)
that cause one amino acid replacement (Phe283 ! Val) in the
predicted sequence.
values determined in the phosphate buffer. The difference in kcat/
Km values for (S)-1-tetralol between the NAD+- and NADP+-linked
activities was low compared to that for the coenzymes.
In order to confirm the preferences for NADP(H) over NAD(H),
we compared the inhibitory effects of NAD+ and NADP+ on the
NADH- and NADPH-linked reductase activities in the phosphate
buffer (Table 2). The NADH-linked activity was potently inhibited
by NADP+ in contrast to low inhibition by NAD+, whereas the
NADPH-linked activity was weakly inhibited by NADP+ and hardly
affected by the addition of NAD+. In the 0.1-mM NAD(P)+-linked
The purified recombinant AKR1C29 showed both NAD+- and
reverse reaction with a-3HB as the substrate, NADPH inhibited the
NADP+-linked dehydrogenase activities toward
a
-3HB and (S)-1-
NAD+-linked activity more potently than the NADP+-linked
activity, while 0.1-mM NADH showed less than 24% inhibition
toward the NAD(P)+-linked activity. The inhibition pattern of
NADPH was competitive with respect to NAD+, showing the Ki
tetralol, and the pH optima of the NAD+- and NADP+-linked
activities were 9.5 and 10.5, respectively, as described previously
with hepatic 3HBD [5]. The Km and kcat values for
a-3HB
determined in 0.1-M glycine–NaOH buffer, pH 9.5, containing
1.0-mM NAD+ were 0.19 mM and 528 minꢁ1, respectively, which
are also comparable to those reported with hepatic 3HBD [5].
AKR1C29 exhibited NADH- and NADPH-linked reductase activities
toward pyridine-3-aldehyde, which showed similar pH optima at
6.0. Therefore, the enzymatic properties of AKR1C29 were
examined at a physiological pH of 7.4, in order to compare the
kinetic constants for the coenzymes and substrates between the
oxidation and reduction reactions.
value of 0.10 mM. The potent inhibition of NAD(H)-linked activity
3.2. Coenzyme specificity
When the (S)-1-tetralol dehydrogenase and pyridine-3-alde-
hyde reductase activities of AKR1C29 were assayed in 0.1-M Tris–
HCl buffer (pH 7.4), the NAD+- and NADH-linked activities were
much lower than those determined in 0.1-M potassium phosphate
buffer (pH 7.4), but the NADP+- and NADPH-linked activities were
independent of the buffer component (Fig. 3). The NAD(H)-linked
activities in the Tris–HCl buffer were enhanced in a dose-
dependent manner (from 0 to 80 mM) by the addition of the
phosphate buffer, and reached the activity level determined in 0.1-
M phosphate buffer. As observed with guinea-pig 3HBD [9],
phosphate ion also acted as an activator of the NAD(H)-linked
activities of AKR1C29. The kinetic constants for the coenzymes and
(S)-1-tetralol determined with the phosphate and Tris–HCl buffers
are summarized in Table 1. In the phosphate buffer, both Km and
kcat values for NAD(H) were higher than those for NADP(H), but the
kcat/Km values indicated approximately 22- and 5-fold preferences
for NADP(H) over NAD(H) in the oxidation and reduction reactions,
respectively. In the Tris–HCl buffer, the preferences for NADP(H)
over NAD(H) were much greater mostly due to the 3-fold decrease
in kcat values of the NAD(H)-linked activities, compared to the
Fig. 3. Effect of phosphate on the NAD(H)- and NADP(H)-linked activities of
AKR1C29. The enzyme activities were determined in 0.1-M Tris–HCl buffer, pH 7.4,
in the absence or presence of indicated concentrations of potassium phosphate
buffer, pH 7.4. The substrate and coenzyme for the dehydrogenase activity (solid
line) were 0.1-mM (S)-1-tetralol and 2.0-mM NAD+ (~) or 0.25-mM NADP+ (~),
respectively and those for the reductase activity (dotted line) were 2.0-mM
pyridine-3-aldehyde and 0.1-mM NADH (*) or NADPH (*). The activities are
expressed as relative activities (means of two determinations), in which the NAD+-
linked dehydrogenase and NADH-linked reductase activities in the absence of the
phosphate buffer are taken as 100%.

1370
S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
Table 1
Kinetic constants for coenzymes and substrates in the oxidation and reduction reactions by AKR1C29.
Varied substrate
Fixed substrate
Phosphate buffer
Km M)
Tris–HCl buffer
Km M)
(
m
kcat (minꢁ1
)
kcat/Km (minꢁ1 mMꢁ1
)
(
m
kcat (minꢁ1
)
kcat/Km (minꢁ1 mMꢁ1
)
Oxidation
NAD+
1-mM (S)-1-Tetralol
0.1-mM (S)-1-Tetralol
2-mM NAD+
0.25-mM NADP+
471
1.6 (0.003)
12
1.0 (0.08)
111
0.24
454
1.2 (0.003)
29
0.9 (0.03)
34
0.075
6.6 (88)
1.3
NADP+
8.4 (0.08)
101
8.5 (0.08)
5.3 (22)
8.4
7.9 (0.2)
39
(S)-1-Tetralol
(S)-1-Tetralol
8.5 (1)
10 (0.3)
11 (8)
Reduction
NADH
2-mM P3Aa
2-mM P3Aa
25
1.5 (0.06)
17
5.0 (0.3)
0.68
13
1.0 (0.08)
2.6
0.20
NADPH
3.3 (5)
5.2 (2)
5.2 (26)
The activity was assayed in 0.1-M potassium phosphate buffer or Tris–HCl buffer (pH 7.4). The value in parenthesis represents the ratio of the NADP(H)-linked activity to the
NAD(H)-linked activity.
a
Pyridine-3-aldehyde.
Table 2
Inhibitory effects of coenzymes on NAD(P)H-linked reductase and NAD(P)+-linked dehydrogenase activities of AKR1C29.
Pyridine-3-aldehyde reductase activity
IC50
(
mM)
a-3HB dehydrogenase activity
IC50 (mM)
NADP+
NAD+
NADPH
NADH
NADH-linked
NADPH-linked
0.3
310
NAD+-linked
NADP+-linked
0.4
47
(24%)
(20%)
212
(12%)
The activities were determined with 0.1 mM coenzyme and 2-mM pyridine-3-aldehyde or 0.5-mM
1 mM NAD+ and 0.1 mM NADH.
a-3HB. The values in parentheses represent the inhibition percentages by
by NADP(H) further indicates that AKR1C29, i.e., 3HBD, is a
NADP(H)-preferring enzyme. Therefore, the substrate specificity of
the enzyme was examined with NADP(H) as the coenzyme in 0.1-
M phosphate buffer, pH 7.4.
3.3.2. Substrates in the NADPH-linked reduction
AKR1C29 showed extremely broad substrate specificity for a
variety of ketosteroids (Table 4), quinones, ketones (Table 5),
dicarbonyl compounds (Table 6), aldehydes and several monosac-
charides (Table 7). The enzyme reduced 3-, 17- and 20-ketosteroids
3.3. Substrate specificity
Table 3
Kinetic constants for substrates in the NADP+-linked oxidation by AKR1C29.
3.3.1. Substrates in NADP+-linked oxidation
AKR1C29 oxidized
a
-3HB more efficiently than its isomer
b
-
Substrate
Km
(
m
M) kcat (minꢁ1
)
kcat/Km
(minꢁ1 m
ꢁ1
)
M
3HB (Table 3), as previously determined at the optimal pH of 9.5
with hepatic 3HBD [5]. The enzyme showed high activity toward
the known 3HBD substrates such as 1-tetralol and 1-indanol
[5,7], in which high preferences for their (S)-isomers over (R)-
isomers were observed. The enzyme showed low activity toward
Alicyclic alcohols
(S)-1-Tetralol
1.0
8.2
5.0
9.8
7.5
18
7.5
1-Acenaphthenol
2.2
a
-3HB
(S)-1-Indanol
-3HB
trans-Benzene dihydrodiol
,11 -PGF2
11
2.2
12
1.2
other alicyclic alcohols including endogenous 9
did not oxidize 0.1 mM PGF2 and 9 ,11 -PGF2, suggesting its
-hydroxy group of PGF2. AKR1C29 also
oxidized several aliphatic alcohols including the most efficient
substrate -ionol showing the highest kcat/Km value of
12 minꢁ1 m ꢁ1, although no activity was observed with sugar
alcohols (glycerol, -threitol and xylitol). Furthermore, AKR1C29
oxidized not only 17 -hydroxysteroids, but also some 3 - and
20 -hydroxysteroids. The oxidized products of 5 -andros-
tane-3 ,17 -diols were identified as their 17-keto metabolites,
and the oxidation of 5 -pregnane-3 ,20 -diols yielded 5
-pregnan-3 -ol-20-ones. AKR1C29 did not show significant
activity toward other 3 -hydroxysteroids (5 -androstan-3 -ol-
17-one, 5 -pregnan-3 -ol-20-ones, 4-pregnen-3
and lithocholic acid) and 20 -hydroxysteroids (5
20 -ol-3-ones and 4-pregnen-20
showed low activity (0.003 unit/mg) toward 50
stan-3 -ol-17-one, but did not oxidize other 3
oids (5 -androstan-3 -ol-17-one, -pregnan-3
ones and isolithocholic acid). In addition, no activity was
observed with the following substrates: 17 -hydroxysteroids
(5 -androstan-17 -ol-3-one and 17 -estradiol), 20 -hydro-
xysteroids (5 -pregnan-20 -ol-3-ones and 4-pregnen-20
a,11b-PGF2, but
b
16
8.4
4.0
0.45
1.6
1.1
0.53
0.33
0.014
0.003
0.002
b
a
12
33
a
9a
b
specific oxidation of 11
b
(R)-1-Indanol
(R)-1-Tetralol
512
488
b
M
Aliphatic alcohols
b-Ionol
0.2
2.5
12
L
Geraniol
5.4
2.0
4.2
3.8
1.9
0.35
b
a
Farnesol
0.16
0.33
0.23
1.0
0.080
0.079
0.061
0.002
a
a/b
1-Decanol
Geranylgeraniol
1-Nonanol
a
b
537
a
/
b
a
a
a/
b
a
3
a
-Hydroxysteroids
4-Androsten-3 -ol-17-one
-Androstan-3 -ol-17-one
a
a
1.2
2.4
0.31
0.15
0.26
a
b
a
5
a
0.063
a
/
b
a
a
-ol-20-one
-pregnan-
a
a
/
b
17b-Hydroxysteroids
5
b
a
b
a
b
a
-Androstane-3
a
,17
b
-ol-3-one
-ol-3-one
b
b
-diol
-diol
7.2
1.8
4.8
1.8
7.0
2.8
0.35
0.27
0.27
0.26
0.16
0.056
0.039
a
a
-ol-3-one). The enzyme
M 5 -andro-
-hydroxyster-
-ol-20-
5
5
5
5
5
-Androstane-3
a,17
0.49
1.3
m
b
a
-Androstan-17
-Androstan-17
b
b
b
0.47
1.1
b
b
5
a
/
b
b
-Androstane-3
-Androstane-3
b,17
-diol
-diol
b,17
b
12
10
0.67
0.39
Testosterone
a
a
/
b
a
a
b
20a-Hydroxysteroids
4-Pregnene-17
a
a
a
,20
a
a
a
-diol-3-one
2.9
35
8.3
0.26
0.60
0.13
0.090
0.017
0.016
a
/
b
b
b-
5a-Pregnane-3
-Pregnane-3
,20
,20
-diol
-diol
ol-3-one), and 21-hydroxysteroids (5
diones and 11-deoxycorticosterone).
a/b-pregnane-21-ol-3,20-
5b

S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
1371
Table 4
Kinetic constants for ketosteroids in the NADPH-linked reduction by AKR1C29.
Table 6
Kinetic constants for dicarbonyl compounds in the NADPH-linked reduction.
Substrate
Km
(
m
M) kcat (minꢁ1
)
kcat/Km
(minꢁ1 mMꢁ1
Substrate
Km
(m
M)
kcat (minꢁ1
)
kcat/Km
(minꢁ1 mMꢁ1
)
)
3-Ketosteroids
a
-Diketones
5
5
5
5
b
a
b
a
-Pregnan-20
a
a
-ol-3-one
-ol-3-one
1.0
1.9
2.3
3.7
0.70a
1.2a
0.93a
0.83a
0.89
0.70
0.63
0.40
0.22
0.056
9,10-Phenanthrenequinone
1-Phenyl-1,2-propanedione
BFAMa
0.1
0.1
0.1
0.3
0.4
1.6
2.5
7.4
5.3
2.5
6.3
8.8
3.9
4.8
74
53
25
21
22
2.4
1.9
-Pregnan-20
-Androstan-17
-Androstan-17
b
b
-ol-3-one
-ol-3-one
Isatin
4-Pregnen-20
a-ol-3-one
16
2,3-Heptanedione
3,4-Hexanedione
Diacetyl
17-Ketosteroids
5
5
5
5
b
b
a
a
-Androstan-3
a
b
a
b
-ol-17-one
0.9
0.9
2.4
1.5
3.4
2.0
2.7
-Androstan-3
-Androstan-3
-Androstan-3
-ol-17-one
-ol-17-one
-ol-17-one
1.7
Other a-dicarbonyls
14
11
0.24
0.18
Methylglyoxal
Glyoxal
9.0
3.8
0.42
33
0.89
0.96
0.027
0.026
3-Deoxyglucosone
37
3,17-Diketosteroids
5
5
b
a
-Androstane-3,17-dione
-Androstane-3,17-dione
1.2
3.6
13
0.86
0.93
1.2
0.72
0.26
0.092
b
- and g-Diketones
2,5-Hexanedione
2,4-Pentanedione
33
2.4
3.4
0.073
0.017
4-Androstene-3,17-dione
205
a
20-Ketosteroids
Benzoylformic acid methyl ester.
5
5
b
-Pregnane-3
a
a
,21-diol-20-one
,21-diol-20-one
0.5
0.9
2.1
0.7
1.1
0.59a
0.83
1.3
0.34a
0.39
1.2
a-Pregnane-3
0.92
0.62
0.49
0.35
4-Pregnen-3a-ol-20-one
Table 7
5
5
a
-Pregnan-3
a
a
-ol-20-one
-ol-20-one
Kinetic constants for aldehydes and monosaccharides in the NADPH-linked
reduction.
b-Pregnan-3
3,20-Diketosteroids
Substrate
Km
(
m
M)
kcat (minꢁ1
)
kcat/Km
5
5
5
a-Pregnane-3,20-dione
b-Pregnane-21-ol-3,20-dione
b-Pregnan-3,20-dione
0.7
3.0
1.4
1.1
1.7
4.5
0.72a
2.1
1.0
(minꢁ1 mMꢁ1
)
0.70
0.60
0.51
0.47
0.24
0.84
0.56a
0.80
1.1a
Aromatic aldehydes
4-Nitrobenzaldehyde
Benzaldehyde
Progesterone
0.5
0.7
5.4
4.1
2.6
6.5
3.9
8.2
4-Pregnene-21-ol-3,11,20-trione
3.7
5a-Pregnane-21-ol-3,20-dione
Pyridine-3-aldehyde
Furfural
1.2
44
0.089
a
Substrate inhibition was observed at concentrations of >5 ꢀ Km
.
Alkanals
1-Decanal
1-Nonanal
1-Heptanal
0.6
0.5
0.6
5.0
2.8
3.2
8.3
5.6
5.3
Table 5
Kinetic constants for quinones and ketones in the NADPH-linked reduction.
Substrate
Km
(
mM)
kcat (minꢁ1
)
kcat/Km
(minꢁ1 mMꢁ1
Alkenals
)
4-Oxo-2-nonenal
Geranylgeranial
trans-2-Decenal
trans-2-Nonenal
4-Hydroxy-2-nonenal
Farnesal
0.3
0.4
0.8
0.5
0.3
1.0
1.0
9.8
5.2
5.4
6.1
2.7
5.2
2.5
1.3
3.9
0.94
3.7
1.4
0.86
3.4
20
6.8
p-Quinones
TBE
6.5
0.7
1.5
0.5
6.7
4.5
1.3
9.1
3.0
2.4
5.0
1,4-Naphthoquinone
Menadione
4.3
3.9
Drug ketones
3OB
4-Hydroxy-2-hexenal
trans-2-Hexenal
Crotonaldehyde
all-trans-Retinal
Acrolein
0.94
0.38
0.27
0.16
0.11
3.8
2.1
0.65
0.94
15
0.55
Naloxone
16
0.043
0.039
0.035
0.024
0.012
0.008
Naltrexone
Befunolol
24
424
67
30
Daunorubicin
Haloperidol
Loxoprofen
1.6
Monosaccharides
55
0.67
2.1
L
-Threose
10
49
1.7
3.6
1.4
2.3
1.4
2.9
0.17
272
L
-Glyceraldehyde
0.073
0.048
0.025
0.022
0.021
Other ketones
Dihydroxyacetone
29
4-Nitroacetophenone
2.9
4.1
2.6
1.4
0.94
3.4
2.9
2.2
1.2
1.4
D
-Glyceraldehyde
-Threose
93
a
-Tetralone
2.8
2.4
4.1
0.93
0.58
0.23
0.20
0.088
0.039
0.034
D
64
Cyclohexanone
PGD2
L
-Erythrulose
138
2-Cyclohexen-1-one
Acetoin
17
33
56
35
ketosteroids,5b-dihydro-C19-steroidsweremoreefficientlyreduced
4-Benzoylpyridine
Acetol
than their 5 -isomers. This tendency was observed in the reduction
a
of androstane-3,17-diones, and 4-androstene-3,17-dione was re-
duced with the lowest kcat/Km value. The reduction of the 3,17-
diketosteroids yielded only their17b-ol metabolites, suggesting that
with Km values in the micromolar or submicromolar range (Table 4).
In the 3-ketosteroids, both 5 - and 5 -dihydro-C19/C21-steroids
were similarly reduced, although substrate inhibition was observed
AKR1C29 exclusively reduces the 17-keto group of the 3,17-
diketosteroids. In contrast to the specificity in the reduction of 3-
and 17-ketosteroids, D4-unsaturated, 5
a-dihydro and 5b-dihydro
a
b
at concentrations higher than 5 ꢀ Km. In contrast, the enzyme
forms of 20-keto-C21-steroids were reduced with similar kcat/Km
values, although the values for the steroids with 21-hydroxy group
were higher. The reduced product of progesterone was identified as
showed a high Km value for 4-pregnen-20a-ol-3-one, and did not
reduce testosterone and 5
acid and dehydrolithocholic acid). The reduced products of 5
pregnan-20 -ol-3-ones and -pregnan-20 -ol-3-ones were
identified as the corresponding 3 -ol metabolites. In the 17-
b-dihydro-C27-steroids (dehydrocholic
a/b-
4-pregnen-20
p-Quinones were efficient substrates, showing high kcat/Km
values comparable to those for (S)-1-tetralol and -3HB (Table 5).
a-ol-3-one.
a
5a/b
a
a
a

1372
S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
Almost equal amounts of 4(S)- and 4(R)-TBEHs were formed in the
reduction of the favorable substrate, TBE. AKR1C29 reduced 3OB
and several drug ketones, although no significant activity toward
ketotifen, metyrapone and acetohexamide was observed. In the
substrates than their
20-mM pentoses ( - and
glucose, -fructose and
D
-epimers. No activity could be detected with
-xyloses and -arabinose) and hexoses (
-glucuronic acid) as the substrates.
L
D
L
D-
D
D
reduction of 3OB both
products, and both 6
a
- and
b
-3HBs were detected as the
3.4. Metabolism of 3OB, 17-ketosteroid and 4-oxo-2-nonenal in cells
overexpressing AKR1C29
a
- and 6
b-naloxols were formed in the
reduction of naloxone. The enzyme also reduced other aromatic,
alicyclic and aliphatic ketones including PGD2, acetoin (3-hydroxy-
2-butanone) and acetol (1-hydroxyacetone) with low Km values of
The substrate specificity for alcohols (Table 3) and carbonyl
compounds (Tables 4, 5 and 7) showed that AKR1C29 catalyzes the
less than 56
m
M. The reduced product of PGD2 was identified as
reversible oxidation of several substrates such as 3HB and 5
androstane-3 ,17 -diol, although many nonsteroidal carbonyl
compounds were irreversibly reduced. To elucidate the reaction
directionality of the enzyme in the intact cells, the oxidation of
3HB and reduction of 3OB in the AKR1C29-overexpressed BAECs
were examined. The expression of cytosolic AKR1C29 in the cells
transfected with the vector harboring its cDNA was confirmed by
assay of (S)-1-tetralol dehydrogenase activity: The NADP+-linked
activity in the cell extract was 0.038 unit/mg, in contrast to much
lower activity (0.003 unit/mg) in the extract of the control BAECs,
which were transfected with the vector alone. While the formation
of 3OB was not observed in the incubation of the AKR1C29-
b-
9
a
,11 -PGF2. It should be noted that no significant reductase
b
a
b
activity toward 100-
PGE2 was observed.
AKR1C29 efficiently reduced
nous isatin and diacetyl, of which those with aromatic and alicyclic
rings were the best substrates, showing high kcat/Km values of more
than 20 minꢁ1 mMꢁ1 (Table 6). AKR1C29 also reduced other
m
M PGA2, 13,14-dihydro-15-keto-PGD2 and
-diketones including endoge-
a-
a
a-
dicarbonyls with an aldehyde and keto or aldehyde groups, in
which methylglyoxal showed high catalytic efficiency compared
with 3-deoxyglucosone and glyoxal. Furthermore, 2,5-hexane-
dione and 2,4-pentanedione, which do not have the conjugated
dicarbonyl structure, were reduced with moderate kcat values (2.4
and 3.4 minꢁ1, respectively) by the enzyme.
Aromatic and lipid-derived aliphatic aldehydes were excellent
substrates, as evidenced by Km values in the micromolar or
submicromolar range (Table 7). Within a series of linear alkanals
and alkenals with one double bond, the catalytic efficiency tended
to increase with increasing chain lengths. In the 9-carbon alkenals,
the kcat value for 4-oxo-2-nonenal was higher than those for 4-
hydroxy-2-nonenal and 2-nonenal, suggesting that the presence of
a keto group at position 4 affects the catalytic activity of AKR1C29.
The enzyme efficiently reduced geranylgeranial and farnesal with
more than three double bonds and branched methyl groups, but
showed low reactivity to all-trans-retinal with a trimethylcyclo-
hexene ring. The reduced product of farnesal was identified as
farnesol. Furthermore, the enzyme reduced short-chain aldoses
overexpressed cells with
a-3HB for 12 h, a-3HB and b-3HB were
formedintheincubationwith3OBandtheiramountswereinversely
correlated with the decrease in 3OB (Fig. 4A). In addition, the
AKR1C29-overexpressed cells proceeded the reduction of 5
androstan-3 -ol-17-one into -androstane-3 ,17 -diol
(Fig. 4B), but not the reverse oxidation of 5 -androstane-3 ,17
diol into 5 -androstan-3 -ol-17-one during 12 h-incubation with
the substrate. No reduced products of both 3OB and 5 -androstan-
-ol-17-one were formed in the control cells (data not shown).
b-
a
5b
a
b
b
a
b-
b
a
b
3a
Furthermore, we examined the protective effect of AKR1C29 against
cytotoxicity of 4-oxo-2-nonenal toward BAECs, because 4-oxo-2-
nonenal, the excellent substrate for the enzyme, is highly cytotoxic
toward cultured cells [31]. The control cells were killed by 4-oxo-2-
nonenal in a dose-dependent manner (Fig. 4C). The toxicity against
thecells was attenuatedbytheoverexpressionofAKR1C29at4-oxo-
and ketoses, of which
L-threose and
L-glyceraldehyde were better
2-nonenal concentrations of higher than 15 mM.
Fig. 4. Metabolism of 3OB, 5
b-androstan-3
a
-ol-17-one and 4-oxo-2-nonenal in AKR1C29-overexpressed BAECs. (A) Reduction of 3OB. The cells were incubated with 0.2-mM
-3HB (~) and -3HB (&) by LC-ESI-MS. The
-3HB and their sum (*) are expressed as the metabolic percentages, relative to the sum of 3OB and
the two metabolites. Values are the means of duplicate experiments. (B) Reduction of 0.1-mM 5 -androstan-3 -ol-17-one. The medium was taken at indicated times for
determination of 5 -androstan-3 -ol-17-one (*) and its reduced metabolite, 5 -androstane-3 ,17 -diol (*). The amounts of the substrate and metabolite are expressed
3OB, and a portion of the medium was taken at indicated times for determination of 3OB (*) and its reduced metabolites,
amount of 3OB represents the remaining percentage, and those of -3HB,
a
b
a
b
b
a
b
a
b
a
b
as the remaining and metabolic percentages, respectively, as described in (A). Values are the means of duplicate experiments. (C) Protective effect of AKR1C29-overexpression
against 4-oxo-2-nonenal-induced cytotoxicity. BAECs were transfected for 48 h with the expression vector alone (control, *) and the vector carrying AKR1C29 cDNA
(overexpressing, *), and then treated for 24 h with various concentrations of 4-oxo-2-nonenal. The viability is shown as the mean ꢂ SD, evaluated by using the unpaired
Student’s t-test and ANOVA followed by Fisher’s test, n = 3. *Significant difference (p < 0.05) from the control cells.

S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
1373
Fig. 5. RT-PCR analysis for expression of mRNA for AKR1C29 in tissues of male and female rabbits. Tissues: brain (B), lung (Lu), heart (H), stomach (S), liver (Li), adrenal gland
(A), small intestine (I), colon (C), testis (T) and ovary (O). The kidney of the male rabbit was divided into the renal cortex (Rc) and renal medulla (Rm), while whole kidney (K) of
female rabbit was used. The expression of mRNA for
b-actin is shown as the control.
3.5. Inhibitor sensitivity
high. In addition to the high cellular ratio of NADPH/NADP+ [35,36],
the high affinity for NADPH suggests that AKR1C29 primarily acts
as a reductase. The directional preferences of HSDs in the cells are
reported to be governed by relative affinities for NAD(P)(H) and
existing coenzyme gradients [36]. The data using the AKR1C29-
overexpressed cells demonstrated that the intracellular enzyme
The
by phenolphthalein (IC50 = 3.4
M phenolphthalin (an open form of the phthalein ring), 4,5,6,7-
b
-ionol dehydrogenase activity of AKR1C29 was inhibited
m
M), but it was not affected by 50-
m
tetrabromophthalein or 30,300,50,500-tetrabromophthalein. The en-
zyme activity was also inhibited by nonsteroidal anti-inflamma-
tory drugs such as tolfenamic acid, mefenamic acid and flufenamic
catalyzes the reduction of both 3OB and 5b-androstan-3a-ol-17-
one, but not the reverse reactions. Therefore, we concluded that
AKR1C29 acts as an NADPH-dependent reductase. This raises the
question of what enzyme(s) are responsible for the metabolism of
3HB into 3OB in the rabbit. With respect to the cellular coenzyme
gradients, NAD+-dependent dehydrogenases may act as the 3HBD.
NAD+-dependent AKRs of rats (AKR1C24) [37] and mice (AKR1C19)
acid (IC50 values of 8.8, 22 and 39
estrogens such as hexestrol, diethylstilbestrol and dienstrol (IC50
values of 27, 31 and 38 M, respectively), and flavonoids such as
kaempferol, 7-hydroxyflavone and quercetin (IC50 values of 12, 17
and 22 M, respectively). The inhibition patterns of phenolphtha-
lein and tolfenamic acid were competitive with respect to -ionol,
and their Ki values were 0.82 and 1.2 M, respectively. No
significant inhibition (less than 25%) was observed with 10-
medroxyprogesterone acetate, 50- M steroidal anti-inflammatory
drugs (dexamethasone, betamethasone and 6 -methylpredniso-
mM, respectively), synthetic
m
m
b
[38] have been reported to oxidize
identified three rabbit orthologs of AKR1C24 (AKR1C26, AKR1C27
and AKR1C28) [16], of which AKR1C26 and AKR1C28 oxidized
3HB more efficiently than -3HB, showing Km values of 7.6 and
7.0
M, respectively, and kcat values of 1.2 and 7.2 minꢁ1
a-3HB and b-3HB. We
m
m
M
a-
m
b
a
m
,
lone), bile acids (isolithocholic acid and lithocholic acid), glycyr-
rhetinic acid and 1-mM 1,10-phenanthroline, which inhibit HSDs
[32,33] of the AKR1C subfamily. The enzyme was also not inhibited
by known inhibitors of aldehyde reductase (1-mM valproic acid,
respectively, at pH 7.4 and 25 8C. The two rabbit NAD+-dependent
dehydrogenases may function in the metabolism of 3HB into 3OB.
AKR1C29 shows 74% overall amino acid sequence identity with
hamster 3HBD [11], but between the two enzymes there is no
difference in the coenzyme-binding residues (Fig. 2) that are
shown by crystallographic studies of other AKRs [33,39–42].
Hamster 3HBD shows higher Vmax/Km value for NADP+ than NAD+
even at its optimal pH of 8.5 [11], and might also act as a reductase
using NADPH as the coenzyme under physiological conditions.
The newly found alcohol substrates of AKR1C29 include trans-
benzene dihydrodiol, which is a representative substrate of
dihydrodiol dehydrogenase [15]. Monomeric AKR1C29 differs
from the tetrameric and dimeric forms of rabbit dihydrodiol
dehydrogenase purified by Klein et al. [15]. Another monomeric
rabbit enzyme with trans-benzene dihydrodiol dehydrogenase
activity is indanol dehydrogenase, which exists in seven multiple
barbital and diphenic acid) [34] and
a rabbit dihydrodiol
dehydrogenase isoform, CM-2 (1-mM phenobarbital) [15].
3.6. Tissue distribution
The tissue distribution of AKR1C29 was assessed by RT-PCR
(Fig. 5). The expression level of the mRNA for the enzyme was high
in the kidney, adrenal gland, intestine and colon of male rabbits,
and low in the other tissues. The mRNA was expressed more highly
in the cortex than the medulla of rabbit kidney. Similar expression
pattern of the transcript in female rabbit tissues was observed,
except that the expression levels in the stomach and liver were
higher compared to male rabbits.
forms associated with 3a-HSD and/or 17b-HSD activities [12].
Among the seven forms (A1–A7), A6 is the most similar to
AKR1C29 in its pH optima of the NAD+- and NADP+-linked
activities, low Km values for 1-acenaphthenol and 1-indanol, and
sensitivity to the three synthetic estrogens and flufenamic acid.
Other endogenous alcohol substrates of AKR1C29 found in this
study are some 3a- and 20a-hydroxysteroids, in addition to 17b-
hydroxysteroids that were previously reported [7]. AKR1C29 is
different from hamster 3HBD [11] in its inability to oxidize bile
4. Discussion
This report describes the cloning and characterization of rabbit
3HBD that has been thought to act as a dehydrogenase [5–7]. The
enzyme is a member (AKR1C29) of the AKR superfamily, and
efficiently reduced carbonyls derived from a broad range of
structural classes, including both aromatic and aliphatic aldehydes
and ketones, using NADPH as the preferred coenzyme. The
preference for NADP(H) is evidenced by the high kcat/Km values
for NADP(H) compared with those for NAD(H) and the potent
inhibition of the NAD(H)-linked activities by NADP(H), in which
the affinity for NADPH (indicated by the Ki value) was significantly
acids and exhibiting 20a-HSD activity, which may result from
differences in the substrate-binding residues (positions 54, 128,
306, 310 and 311, Fig. 2) between the two enzymes. In contrast to
its low kcat/Km values for the hydroxysteroids, AKR1C29 efficiently
reduced various 3-, 17- and 20-ketosteroids in the reverse reaction.

1374
S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
The product identification of the ketosteroid reduction indicates
that AKR1C29 is a reductive 3 /17 /20 -HSD. Because it did not
oxidize 17 -estradiol, AKR1C29 may be distinct from multiple
forms of NADP+-dependent 17
-HSD purified from female rabbit
liver [13]. The broad steroid specificity of AKR1C29 and its
sensitivity to hexestrol resemble the properties of rabbit 3 /3
17 /20 -HSD [14], although the 3 /3 /17 /20 -HSD substrate,
-androstan-3 -ol-17-one, was slightly oxidized by AKR1C29.
AKR1C29 shows the highest sequence identity with rabbit 20
HSD (AKR1C5) [43] among the known AKRs. AKR1C5 is specifically
expressed in ovary and uterus of female rabbits [43,44] and
reduces progesterone, estrone, dehydroepiandrosterone [40,45]
and PGE2 [44], an indication of its major role in the termination of
pregnancy by both inactivation of progesterone and synthesis of
(AKR1B2), its similar enzyme (AKR1B19) [25] and AKR1C5 [45], in
addition to AKR1C29. Among the rabbit enzymes, AKR1C29 shows
the highest kcat/Km values for most alkenals including highly
reactive 4-oxo-2-nonenal and 4-hydroxy-2-nonenal. This efficien-
cy in the cells is demonstrated by the protective effect of AKR1C29
overexpression against the cytotoxicity of 4-oxo-2-nonenal. The
kcat/Km value for methylglyoxal of AKR1C29 is comparable to that
a
b
a
b
b
a
b/
b
a
a
a
b
b
a
5
b
of AKR1B2, and superior to those of AKR1B19 and AKR1C5. The kcat
Km value for glyoxal is higher than those of AKR1B2 and AKR1B19
(0.009 and 0.002 minꢁ1 m ꢁ1, respectively), although the value
/
a
-
M
for 3-deoxyglucosone is lower than that of AKR1B2. AKR1C29
reduced another reactive dicarbonyl compound, diacetyl, more
efficiently than the other rabbit AKRs, and also reduced acetoin, the
reduced product of diacetyl. AKR1C29 appears to function as a
defense system against these naturally occurring reactive carbonyl
compounds in many rabbit tissue cells. (3) Metabolism of isatin
and monosaccharides. Isatin is an endogenous monoamine oxidase
inhibitor [56] and antagonist of atrial natriuretic peptide function
and nitric oxide signaling [57]. It is one of the best endogenous
substrates of AKR1C29, showing much higher kcat/Km value than
those of the above other AKRs in rabbits. AKR1C29 may be involved
in controlling the intracellular concentration of the biologically
active isatin. In addition, AKR1C29 reduced short-chain aldoses, of
which trioses and tetroses generate superoxide anions in glycation
17b-estradiol and PGF2 . AKR1C29 differs from AKR1C5 in its
a
inability to reduce estrone, dehydroepiandrosterone and PGE2, and
its low sensitivity to the inhibition by phenolphthalein, its
derivatives and flavonoids compared with that of AKR1C5 [45],
in addition to its ubiquitous expression in tissues of both male and
female rabbits. Therefore, AKR1C29 may play a pivotal role in
controlling the concentrations of cellular androgen (5
17 -ol-3-one), progesterone and neuroactive steroids (5
nan-3 -ol-20-one and 5 -pregnane-3 ,21-diol-20-one) in rabbit
a-androstan-
b
a
-preg-
a
a
a
tissues other than the ovary and uterus, in addition to hepatic
metabolism of ketosteroids.
Several cytosolic, monomeric (28–35 kDa) and NADPH-depen-
dent drug ketone reductases with HSD activity were purified from
rabbit livers. They are loxoprofen reductase with 3-ketosteroid
reactions [58] and
ascorbic acid, is suggested to be a potent glycation agent [59].
Although no rabbit reductase for -threose has been reported, the
Km value for -threose of AKR1C29 is much lower than those (350–
750 M) of human aldose reductase (AKR1B1) [60], rat aldose
L-threose, one of the oxidation products of
L
L
reductase activity [46], befunolol reductase with 3
a
-HSD activity
-HSD
m
[47], and two naloxone reductases with 3 -HSD or 17
a
b
reductase-like (AKR1B18) [61] and rat AKR1C15 [62]. This ability to
reduce the short-chain aldoses may also contribute to prevention
of the progress of glycation in rabbits.
activity [48]. AKR1C29 reduced the three drugs, but its sequence is
clearly different from the befunolol reductase [49] and naloxone
reductases [48]. Additionally, its broad specificity for 3-, 17- and
20-ketosteroids is distinct from the specificity of loxoprofen
reductase confined to 3-ketosteroids [46]. In the purification of the
reductases for loxoprofen [46] and naloxone [48] other multiple
enzyme forms were found, and more than five multiple forms of
rabbit reductase for naloxone, naltrexone and/or daunorubicin
have been separated [50]. AKR1C29 probably corresponds to one of
these uncharacterized multiple forms of the drug ketone
reductase(s), and may act as an efficient reductase toward
loxoprofen, naloxone, daunorubicin and befunolol because of its
lower Km values compared to those of the above known drug
ketone reductases.
Conflict of interest
None declared.
Acknowledgements
This work was partly founded by grant-in-aid for Young
Scientists (B) from the Japan Society for the Promotion of Science,
and by a Sasakawa Scientific Research Grant from Japan Science
Society.
AKR1C29 represents a unique member of the AKR superfamily
with respect to its extremely broad substrate specificity and low
Km values, and is ubiquitously distributed in rabbit tissues. Based
on the broad substrate specificity of the enzyme and tissue
distribution, the following physiological roles in metabolism of
endogenous carbonyl compounds could be considered. (1) Reduc-
tion of PGD2: A 66-kDa PGD2 11-ketoreductase was purified from
rabbit liver cytosol [51], while the 36-kDa enzymes belonging to
the AKR superfamily were purified or cloned from bovine lung [28]
References
[1] Knodell RG, Dubey RK, Wilkinson GR, Guengerich FP. Oxidative metabolism of
hexobarbital in human liver: relationship to polymorphic S-mephenytoin 4-
hydroxylation. J Pharmacol Exp Ther 1988;245:845–9.
[2] Takenoshita R, Toki S. New aspects of hexobarbital metabolism: stereoselec-
tive metabolism, new metabolic pathway via GSH conjugation, and 3-hydro-
xyhexobarbital dehydrogenases. Yakugaku Zasshi 2004;124:857–71.
[3] Miyano K, Ota T, Toki S. Stereoselective formation of glucuronides in metabo-
lism of hexobarbital enantiomers in vivo: isolation and quantitation of glu-
curonides in rabbit urine. Drug Metab Dispos 1981;9:60–4.
[4] Takenoshita R, Nakamura T, Toki S. Hexobarbital metabolism: a new metabolic
pathway to produce 1,5-dimethylbarbituric acid and cyclohexenone-glutathi-
one adduct via 30-oxohexobarbital. Xenobiotica 1993;23:925–34.
[5] Takenoshita R, Toki S. Rabbit liver 3-hydroxyhexobarbital dehydrogenase.
Purification and properties. J Biol Chem 1974;249:5428–9.
[6] Takenoshita R, Toki S. Dehydrogenation of indanol by rabbit liver 3-hydro-
xyhexobarbital dehydrogenase. Xenobiotica 1977;7:383–91.
[7] Takenoshita R, Toki S. 5b-Androstane-3a,17b-diol: an endogenous substrate
for rabbit liver 3-hydroxyhexobarbital dehydrogenase. Biochem Pharmacol
1978;27:989–93.
[8] Kageura E, Toki S. Guinea pig liver 3-hydroxyhexobarbital dehydrogenase as a
and liver [52], and human tissues [53]. The Km value (4
AKR1C29 is comparable to those (3.4–10 M) of the bovine and
human enzymes, but is much lower than 200 M of the 66-kDa
mM) of
m
m
rabbit enzyme [51]. AKR1C29 may be involved in the metabolism
of PGD2 in many rabbit tissues, although further comparative
studies of the two rabbit enzymes are needed with respect to their
tissue expression and ability to catalyze the conversion of PGH2
into PGF2 that is associated in the human and bovine enzymes. (2)
a
Detoxification of reactive endogenous carbonyl compounds, which
17b-hydroxysteroid dehydrogenase. Biochim Biophys Acta 1974;341:172–7.
[9] Kageura E, Toki S. Guinea pig liver 3-hydroxyhexobarbital dehydrogenase.
Purification and properties. J Biol Chem 1975;250:5015–9.
[10] Takenoshita R, Naitoh K, Toki S. Mouse liver 3-hydroxyhexobarbital dehydro-
genase: purification and substrate specificity. Prog Clin Biol Res 1989;290:
279–91.
include the lipid-derived long-chain aldehydes [18,31,54] and
a-
dicarbonyl compounds such as methylglyoxal, glyoxal and 3-
deoxyglucosone that are related to glycation [55]. These reactive
carbonyl compounds are substrates of rabbit aldose reductase

S. Endo et al. / Biochemical Pharmacology 86 (2013) 1366–1375
1375
[11] Takenoshita R, Nomura Y, Toki S. Cloning and expression of cDNA encoding
hamster liver 3-hydroxyhexobarbital/17 (3 )-hydroxysteroid dehydroge-
dehydrogenase distinct from the type
1053–63.
5
isozyme.
J
Biochem 2006;139:
b
a
nase 1. Chem Biol Interact 2001;130–132:863–70.
[12] Hara A, Kariya K, Nakamura M, Nakayama T, Sawada H. Isolation of multiple
[38] Ishikura S, Horie K, Sanai M, Matsumoto K, Hara A. Enzymatic properties of a
member (AKR1C19) of the aldo-keto reductase family. Biol Pharm Bull
2005;28:1075–8.
forms of indanol dehydrogenase associated with 17b-hydroxysteroid dehy-
drogenase activity from male rabbit liver. Arch Biochem Biophys 1986;249:
225–36.
[13] Antoun GR, Brglez I, Williamson DG. A 17b-hydroxysteroid dehydrogenase of
[39] Jez JM, Bennett MJ, Schlegel BP, Lewis M, Penning TM. Comparative anatomy of
the aldo-keto reductase superfamily. Biochem J 1997;326:625–36.
[40] Couture JF, Legrand P, Cantin L, Labrie F, Luu-The V, Breton R. Loop relaxation, a
female rabbit liver cytosol. Purification and characterization of multiple forms
mechanism that explains the reduced specificity of rabbit 20a-hydroxysteroid
of the enzyme. Biochem J 1985;225:383–90.
dehydrogenase, a member of the aldo-keto reductase superfamily. J Mol Biol
[14] Smirnov AN. Estrophilic 3a,3b,17b,20a-hydroxysteroid dehydrogenase from
2004;339:89–102.
rabbit liver-I. Isolation and purification. J Steroid Biochem 1990;36:609–16.
[15] Klein J, Thomas H, Post K, Wo¨rner W, Oesch F. Dihydrodiol dehydrogenase
activities of rabbit liver are associated with hydroxysteroid dehydrogenases
and aldo-keto reductases. Eur J Biochem 1992;205:1155–62.
[41] Sanli G, Dudley JI, Blaber M. Structural biology of the aldo-keto reductase
family of enzymes: catalysis and cofactor binding. Cell Biochem Biophys
2003;38:79–101.
[42] Penning TM. Molecular determinants of steroid recognition and catalysis in
[16] Endo S, Matsunaga T, Fujimoto A, Kumada S, Arai Y, Miura Y, et al. Characteri-
zation of rabbit morphine 6-dehydrogenase and two NAD+-dependent
aldo-keto reductases. Lessons from 3a-hydroxysteroid dehydrogenase. J Ste-
roid Biochem Mol Biol 1999;69:211–25.
3a(17
b)-hydroxysteroid dehydrogenases. Arch Biochem Biophys 2013;529:
[43] Lacy WR, Washenick KJ, Cook RG, Dunbar BS. Molecular cloning and expres-
131–9.
sion of an abundant rabbit ovarian protein with 20a-hydroxysteroid dehy-
[17] Platt K, Oesch F. An improved synthesis of trans-5,6-dihydroxy-1,3-cyclohex-
adiene (trans-1,2-dihydroxy-1,2-dihydrobenzene). Synthesis 1977;7:449–50.
[18] Doorn JA, Srivastava SK, Petersen DR. Aldose reductase catalyzes reduction of
the lipid peroxidation product 4-oxonon-2-enal. Chem Res Toxicol 2003;16:
1418–23.
drogenase activity. Mol Endocrinol 1993;7:58–66.
[44] Wintergalen N, Thole HH, Galla HJ, Schlegel W. Prostaglandin E2 9-reductase
from corpus luteum of pseudopregnant rabbit is a member of the aldo-keto
reductase superfamily featuring 20a-hydroxysteroid dehydrogenase activity.
Eur J Biochem 1995;234:264–70.
[19] Doyle MP, Yan M. Attempted synthesis of casbene by intramolecular cyclo-
propanation. ARKIVOC 2002;viii:180–5.
[45] Endo S, Arai Y, Hara A, Kitade Y, Bunai Y, El-Kabbani O, et al. Substrate
specificity and inhibitor sensitivity of rabbit 20a-hydroxysteroid dehydroge-
[20] Cone EJ, Gorodetzky CW, Yeh SY. Biosynthesis, isolation and identification of
nase. Biol Pharm Bull 2013;36:1514–8.
6
b
-hydroxynaltrexone, a major human metabolite of naltrexone. J Pharm Sci
[46] Tanaka Y, Nishikawa Y, Matsuda K, Yamazaki M, Hayashi R. Purification and
some properties of ketone reductase forming an active metabolite of sodium
2-[4-(2-oxocyclopentylmethyl)-phenyl]propionate dihydrate (loxoprofen so-
dium), a new anti-inflammatory agent, in rabbit liver cytosol. Chem Pharm
Bull (Tokyo) 1984;32:1040–8.
[47] Imamura Y, Nozaki Y, Otagiri M. Purification and characterization of befunolol
reductase from rabbit liver. Chem Pharm Bull (Tokyo) 1989;37:3338–42.
[48] Yamano S, Ichinose F, Todaka T, Toki S. Purification and characterization of two
major forms of naloxone reductase from rabbit liver cytosol, new members of
aldo-keto reductase superfamily. Biol Pharm Bull 1999;22:1038–46.
[49] Imamura Y, Koga T, Otagiri M, Satoh K, Hara A. Carbonyl reductase purified
from rabbit liver is not the product of a carbonyl reductase gene (RCBR5 or
RCBR6) cloned from the rabbit liver cDNA library. Biol Pharm Bull
1998;21:76–8.
1975;64:618–21.
[21] Chatterjie N, Inturrisi CE. Stereospecific synthesis of the 6b-hydroxy meta-
bolites of naltrexone and naloxone. J Med Chem 1975;18:490–2.
[22] Tajima K, Hashizaki M, Yamamoto K, Mizutani T. Metabolism of 3-tert-butyl-
4-hydroxyanisole by horseradish peroxidase and hydrogen peroxide. Drug
Metab Dispos 1992;20:816–20.
[23] Tajima K, Hashizaki M, Yamamoto K, Mizutani T. Carbonyl reduction of 6-tert-
butyl-2,3-epoxy-1,4-benzoquinone, a metabolite of 3-tert-butyl-4-hydroxyani-
sole, by rat liver microsomes and cytosol. Drug Metab Dispos 1994;22:815–8.
[24] Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual.
2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory; 1989.
[25] Endo S, Matsunaga T, Kumada S, Fujimoto A, Ohno S, El-Kabbani O, et al.
Characterization of rabbit aldose reductase-like protein with 3b-hydroxys-
teroid dehydrogenase activity. Arch Biochem Biophys 2012;527:23–30.
[26] Ohno S, Matsui M, Yokogawa T, Nakamura M, Hosoya T, Hiramatsu T, et al.
Site-selective post-translational modification of proteins using an unnatural
amino acid, 3-azidotyrosine. J Biochem 2007;141:335–43.
[27] Pare´s X, Julia` P. Isoenzymes of alcohol dehydrogenase in retinoid metabolism.
Methods Enzymol 1990;189:436–41.
[50] Felsted RL, Richter DR, Jones DM, Bachur NR. Isolation and characterization of
rabbit liver xenobiotic carbonyl reductases. Biochem Pharmacol
1980;29:1503–16.
[51] Wong PY. Purification and partial characterization of prostaglandin D2 11-keto
reductase in rabbit liver. Biochim Biophys Acta 1981;659:169–78.
[52] Suzuki T, Fujii Y, Miyano M, Chen LY, Takahashi T, Watanabe K. cDNA cloning,
expression, and mutagenesis study of liver-type prostaglandin F synthase. J
Biol Chem 1999;274:241–8.
[28] Watanabe K, Iguchi Y, Iguchi S, Arai Y, Hayaishi O, Roberts II LJ. Stereospecific
conversion of prostaglandin D2 to (5Z,13E)-(15S)-9
a,-11b,15-trihydroxy-
prosta-5,13-dien-1-oic acid (9 ,11 -prostaglandin F2) and of prostaglandin
a
b
[53] Suzuki-Yamamoto T, Nishizawa M, Fukui M, Okuda-Ashitaka E, Nakajima T, Ito
S, Watanabe KAT cDNA cloning expression and characterization of human
prostaglandinsynthase. FEBS Lett 1999;462:335–40.
H2 to prostaglandin F2 by bovine lung prostaglandin F synthase. Proc Natl
a
Acad Sci USA 1986;83:1583–7.
[29] Endo S, Matsunaga T, Ohta C, Soda M, Kanamori A, Kitade Y, et al. Roles of rat
and human aldo-keto reductases in metabolism of farnesol and geranylger-
aniol. Chem Biol Interact 2011;191:261–8.
[54] Esterbauer H, Schaur RJ, Zollner H. Chemistry and biochemistry of 4-hydro-
xynonenal, malonaldehyde and related aldehydes. Free Radic Biol Med
1991;11:81–128.
[30] Tanaka N, Aoki K, Ishikura S, Nagano M, Imamura Y, Hara A, et al. Molecular
basis for peroxisomal localization of tetrameric carbonyl reductase. Structure
2008;16:388–97.
[55] Kroh LW, Fiedler T, Wagner J. a-Dicarbonyl compounds–key intermediates for
the formation of carbohydrate-based melanoidins. Ann
N Y Acad Sci
2008;1126:210–5.
[31] Shibata T, Iio K, Kawai Y, Shibata N, Kawaguchi M, Toi S, et al. Identification of a
lipid peroxidation product as a potential trigger of the p53 pathway. J Biol
Chem 2006;281:1196–204.
[56] Igosheva N, Lorz C, O‘Conner E, Glover V, Mehmet H. Isatin, an endogenous
monoamine oxidase inhibitor, triggers a dose- and time-dependent switch
from apoptosis to necrosis in human neuroblastoma cells. Neurochem Int
2005;47:216–24.
[32] Higaki Y, Usami N, Shintani S, Ishikura S, El-Kabbani O, Hara A. Selective and
potent inhibitors of human 20
that metabolizes neurosteroids derived from progesterone. Chem Biol Interact
2003;143–144:503–13.
a
-hydroxysteroid dehydrogenase (AKR1C1)
[57] Medvedev A, Igosheva N, Crumeyrolle-Arias M, Glover V. Isatin: role in stress
and anxiety. Stress 2005;8:175–83.
[58] Ortwerth BJ, Speaker JA, Prabhakaram M, Lopez MG, Li EY, Feather MS.
[33] Byrns MC, Jin Y, Penning TM. Inhibitors of type 5 17
b
-hydroxysteroid dehy-
Ascorbic acid glycation: the reactions of
Res 1994;58:665–74.
L-threose in lens tissue. Exp Eye
drogenase (AKR1C3): overview and structural insights. J Steroid Biochem Mol
Biol 2011;125:95–104.
[59] Ortwerth BJ, James H, Simpson G, Linetsky M. The generation of superoxide
anions in glycation reactions with sugars, osones, and 3-deoxyosones. Bio-
chem Biophys Res Commun 1998;245:161–5.
[34] Kanazu T, Shinoda M, Nakayama T, Deyashiki Y, Hara A, Sawada H. Aldehyde
reductase is a major protein associated with 3-deoxyglucosone reductase
activity in rat, pig and human livers. Biochem J 1991;279:903–6.
[35] Veech RL, Eggleston LV, Krebs HA. The redox state of free nicotinamideadenine
dinucleotide phosphate in the cytoplasm of rat liver. Biochem J 1969;115:
609–19.
[36] Sherbet DP, Papari-Zareei M, Khan N, Sharma KK, Brandmaier A, Rambally S,
et al. Cofactors, redox state, and directional preferences of hydroxysteroid
dehydrogenases. Mol Cell Endocrinol 2007;265–266:83–8.
[60] Devamanoharan PS, Varma SD. Studies on L-threose as substrate for aldose
reductase: a possible role in preventing protein glycation. Mol Cell Biochem
1996;159:123–7.
[61] Endo S, Matsunaga T, Kuragano T, Ohno S, Kitade Y, Tajima K, et al. Properties
and tissue distribution of a novel aldo-keto reductase encoding in a rat gene
(Akr1b10). Arch Biochem Biophys 2010;503:230–7.
[62] Endo S, Matsunaga T, Horie K, Tajima K, Bunai Y, Carbone V, et al. Enzymatic
characteristics of an aldo-keto reductase family protein (AKR1C15) and its
localization in rat tissues. Arch Biochem Biophys 2007;465:136–47.
[37] Ishikura S, Matsumoto K, Sanai M, Horie K, Matsunaga T, Tajima K, et al.
Molecular cloning of a novel type of rat cytoplasmic 17b-hydroxysteroid