Substrate specificity of a mouse aldo-keto reductase (AKR1C12)

Article abstract of DOI:10.1248/bpb.29.2488

AKR1C12, a mouse member of the aldo-keto reductase (AKR) superfamily, is highly expressed in the stomach and is identical to a protein encoded in an interleukin-3-regulated gene in mouse myeloid cells, but its function remains unknown. In this study, the recombinant AKR1C12 was purified to homogeneity and the specificity for coenzymes and substrates was examined at a physiological pH of 7.4. The enzyme reduced various α-dicarbonyl compounds, several ketosteroids, aldehydes and some ketones using NADH as the preferred coenzyme. In the reverse reaction, the enzyme showed coenzyme preference for NAD ⁺, and oxidized 3α-, 17β- and 20α-hydroxysteroids, and non-steroidal aliphatic and alicyclic alcohols, of which many hydroxysteroids and geranylgeraniol were good substrates, exhibiting low K _m and high k_cat/K_m values. The results, together with the intracellular high ratio of NAD⁺/NADH, suggest that AKR1C12 functions as a dehydrogenase for the endogenous hydroxysteroids and geranylgeraniol in mouse stomach and myeloid cells.

Full text of DOI:10.1248/bpb.29.2488

2488
Notes
Biol. Pharm. Bull. 29(12) 2488—2492 (2006)
Vol. 29, No. 12
Substrate Specificity of a Mouse Aldo-Keto Reductase (AKR1C12)
Satoshi ENDO,a Kengo MATSUMOTO,a Toshiyuki MATSUNAGA,a Shuhei ISHIKURA,a,b Kazuo TAJIMA,c
d
,a
*
Ossama El-KABBANI, and Akira HARA
a Laboratory of Biochemistry, Gifu Pharmaceutical University; Mitahora-higashi, Gifu 502–8585, Japan: b Hospital for
Sick Children; Toronto, Ontario M5G 1X8, Canada: c Faculty of Pharmaceutical Sciences, Hokuriku University;
Kanazawa 920–1181, Japan: and d Department of Medicinal Chemistry, Victorian College of Pharmacy, Monash
University; Parkville, Victoria 3052, Australia.
Received August 22, 2006; accepted September 29, 2006; published online October 5, 2006
AKR1C12, a mouse member of the aldo-keto reductase (AKR) superfamily, is highly expressed in the stom-
ach and is identical to a protein encoded in an interleukin-3-regulated gene in mouse myeloid cells, but its func-
tion remains unknown. In this study, the recombinant AKR1C12 was purified to homogeneity and the specificity
for coenzymes and substrates was examined at a physiological pH of 7.4. The enzyme reduced various a-dicar-
bonyl compounds, several ketosteroids, aldehydes and some ketones using NADH as the preferred coenzyme. In
the reverse reaction, the enzyme showed coenzyme preference for NADꢀ, and oxidized 3a-, 17b- and 20a-hydroxy-
steroids, and non-steroidal aliphatic and alicyclic alcohols, of which many hydroxysteroids and geranylgeraniol
were good substrates, exhibiting low Km and high kcat/Km values. The results, together with the intracellular high
ratio of NADꢀ/NADH, suggest that AKR1C12 functions as a dehydrogenase for the endogenous hydroxysteroids
and geranylgeraniol in mouse stomach and myeloid cells.
Key words aldo-keto reductase superfamily; AKR1C12; coenzyme specificity; hydroxysteroid dehydrogenase; geranylgeraniol
The aldo-keto reductase (AKR) superfamily is a rapidly enzyme also plays a role in the myeloid cell differentiation.13)
growing group of NAD(P)(H)-dependent oxidoreductases However, there has been only one report on the properties of
that metabolize carbohydrates, steroids, prostaglandins, and AKR1C12, in which two xenobiotic carbonyl compounds,
other endogenous aldehydes and ketones, as well as xenobi- 9,10-phenanthrenequinone (PQ) and 4-nitrobenzaldehyde,
otic compounds.1,2) Members of this superfamily were classi- were found to be good substrates.11) Thus, the endogenous
fied into 14 families, and each family is subdivided into sev- substrates of the enzyme remain unknown, and its ability to
eral subfamilies based on their amino acid sequence similari- oxidize alcohol substrates has not been studied. In this study,
ties. Among the subfamilies, AKR1C is the largest one, we have examined the coenzyme and substrate specificity of
which includes mammalian hydroxysteroid dehydrogenases recombinant AKR1C12 in both reduction and oxidation di-
(HSDs), prostaglandin F synthases and related proteins. To rections at physiological pH in order to elucidate the function
date, four members of this subfamily were identified in hu- related to the high expression levels in the stomach and
mans: 20a-HSD (AKR1C1), 3a-HSD type 3 (AKR1C2), myeloid cell differentiation.
17b-HSD type 5 (also called 3a-HSD type 2 and
prostaglandin F synthase; AKR1C3) and 3a-HSD type 1 MATERIALS AND METHODS
(AKR1C4).3,4) These members exhibit dihydrodiol dehydro-
genase and carbonyl reductase activities for dihydrodiols of
cDNA Isolation and Enzyme Purification The cDNA
aromatic hydrocarbons and xenobiotic carbonyl compounds, (nucleotides 1-977) for AKR1C12 was amplified from a total
and their genes share similar structural organization and are RNA of an ICR mouse liver by reverse transcription (RT)-
clustered on chromosome 10p15. In mice, a cluster of genes PCR. The preparation of the total RNA and RT were pre-
for eight members of the AKR1C subfamily was found on formed as described previously.14) In the PCR, Pfu DNA
mouse chromosome 13.5) Three of these mouse members are polymerase was used together with a sense primer (5ꢀ-AT-
20a-HSD (AKR1C18),6) 3a-HSD (AKR1C14)7) and 17b- GAGCTCCAAACAGCACTA) and antisense primer (5ꢀ-
HSD type 5 (AKR1C6),8) two are dual functional 3a(17b)- CCATGTTAATATTCCTCCGA), which were designed based
HSD (AKR1C20)9) and 3(17)a-HSD (AKR1C21),10) and the on the sequence of the cDNA for AKRa.13) The PCR prod-
others are reported to be aldo-keto reductases (AKR1C12,11) ucts were subcloned into the pCR T7/CT-TOPO vectors (In-
AKR1C1311) and AKR1C1912)). The five mouse HSDs are vitrogen), and the expression constructs were transfected into
NADP(H)-dependent, whereas the three mouse aldo-keto re- Escherichia coli BL21 (DE3) pLysS according to the proto-
ductases show dual coenzyme specificity for NADPH and col described by the manufacturer. The sequences of the in-
NADH or preference for NADH. The expression of the eight serts were verified by DNA sequencing with a CEQ2000XL
mouse enzymes in the tissue differs between themselves.5) DNA sequencer (Beckman Coulter).
AKR1C12 and AKR1C13 are unique in their high expression
The E. coli cells were cultured in a LB medium containing
in epithelial cells of mouse stomach and have been proposed ampicillin (50 mg/ml) at 37 °C until the absorbance at 600 nm
to be involved in detoxification of xenobiotic carbonyl com- reached 0.5. Then isopropyl-1-thio-b-D-galactopyranoside
pounds in the stomach.11) Interestingly, AKR1C12 is identi- (1 mM) was added, and the culture was continued for 24 h at
cal to a protein (AKRa) that is encoded in an interleukin-3- 20 °C. The cells were collected and the extract was prepared
regulated gene in mouse myeloid cells, suggesting that the as described previously.14) The recombinant AKR1C12 was
∗ To whom correspondence should be addressed. e-mail: hara@gifu-pu.ac.jp
© 2006 Pharmaceutical Society of Japan

December 2006
2489
purified at 4 °C by three column chromatography steps. The cause only one gene for AKR1C12 has been identified by ge-
E. coli extract (50 ml) was directly applied to a Sephadex G- nomic analysis5) and an cDNA for AKR1C12 with a replace-
100 column (5ꢁ90 cm) equilibrated with buffer A (10 mM ment of the amino acid at position 81 (accession no.
Tris–HCl, pH 8.0, 5 mM 2-mercaptoethanol, 1 mM EDTA and BC063780) is also isolated from a FVB/N mouse colon. Re-
20% glycerol). The enzyme fractions were concentrated by combinant AKR1C12 was highly expressed from the cDNA
ultrafiltration using an YM-10 membrane (Millipore), and in the E. coli cells, and detected as a major 36-kDa protein
applied to a Q-Sepharose column (2ꢁ20 cm) equilibrated by SDS-PAGE analysis of the cell extract (Fig. 1). The ex-
with Buffer A. The enzyme was eluted with a linear gradient tract showed NADPH-linked reductase activity (0.15 U/mg)
of 0—0.1 M NaCl, dialyzed against Buffer A without EDTA, towards 10 mM PQ at pH 7.4. When NADH was employed as
and then applied to a hydroxylapatite column (1.5ꢁ10 cm) the coenzyme, higher reductase activity towards PQ
equilibrated with the same buffer. Finally, the enzyme was (0.25 U/mg) was observed, as reported for mouse stomach
eluted with a linear gradient of 0—0.1 M potassium phos- AKR1C12.11) The cell extract also exhibited NADH-linked
phate in the buffer, concentrated by the ultrafiltration, and reductase activity towards 0.1 mM S-camphorquinone
stored at ꢂ20 °C. Protein concentration was determined by (0.09 U/mg) and NADꢃ-linked dehydrogenase activities
the method of Bradford15) using bovine serum albumin as the towards 1 mM S-tetralol (0.18 U/mg) and S-indan-1-ol
standard.
(0.19 U/mg). The control E. coli cells transfected with the
Assay of Enzyme Activity Reductase and dehydroge- vector alone showed both NADPH- and NADH-dependent
nase activities of the recombinant enzyme were assayed by reductase activities towards PQ (approximately 0.03 U/mg),
measuring the rate of change in NAD(P)H absorbance (at whereas no significant dehydrogenase activity towards S-
340 nm) and its fluorescence (at 455 nm with an excitation tetralol was observed. Therefore, the recombinant protein
wavelength of 340 nm), respectively. The standard reaction was purified by assaying NADꢃ-linked S-tetralol dehydroge-
mixture for the reductase activity consisted of 0.1 M potas- nase activity. The final preparation with a specific activity of
sium phosphate, pH 7.4, 0.1 mM NAD(P)H, substrate and en- 0.35 U/mg was greater than 97% pure (Fig. 1), and the purifi-
zyme, in a total volume of 2.0 ml. The dehydrogenase activ- cation yield was 40% (26 mg/l of the cultured cells).
ity was determined with 1 mM NAD(P)ꢃ and 1 mM (S)-(ꢃ)-
Coenzyme Specificity In the previous study on
1,2,3,4-tetrahydro-1-naphthol (S-tetralol) as the coenzyme AKR1C12, only the Km values of AKR1C12 for NADPH and
and substrate, respectively, in the standard reaction mixture. NADH were determined using PQ as the substrate at pH 6.5,
TBE and trans-benzene dihydrodiol were synthesized as de- but the coenzyme specificity in the reverse reaction was not
scribed previously.16,17) Other substrates were of the highest examined.11) It was reported that the time course of PQ re-
grade that could be obtained commercially. One unit (U) of duction by NAD(P)H-dependent reductases is proceeded by
enzyme activity was defined as the amount that catalyzes the redox cycling of the quinone and its hydroquinone until
reduction or formation of 1 mmol NAD(P)H per min at NADPH is completely oxidized.18—20) Indeed, similar phe-
25 °C. The pH dependency of the oxidoreduction catalyzed nomenon was observed in the NAD(P)H-linked PQ reduction
by AKR1C12 was examined with 0.1 M potassium phosphate by the purified AKR1C12 (data not shown). Therefore, we
buffers (pH 5.5—11.0), which were prepared by mixing the re-examined the coenzyme specificity by determining the Km
solutions of phosphoric acid and tripotassium phosphate. The and kcat values for the coenzymes in both S-camphorquinone
apparent Km and kcat values were determined over a range of reduction and S-tetralol oxidation at a physiological pH of
five substrate concentrations at a saturating concentration of 7.4. The enzyme showed coenzyme preference for NADH
coenzyme by fitting the initial velocities to the Michaelis– (Kmꢄ4.0 mM and kcat/Kmꢄ12.6 minꢂ1 mMꢂ1) over NADPH
Menten equation. The kinetic constants and IC50 values are (Kmꢄ70 mM and kcat/Kmꢄ1.7 minꢂ1 mMꢂ1). The Km value for
expressed as the means of two determinations.
Product Identification To identify products of the re-
duction and oxidation of steroidal substrates, the reaction
was conducted in a 2.0 ml system containing 0.1 mM
NAD(H), substrate (20—50 mM), enzyme (3 mg), and 0.1 M
potassium phosphate, pH 6.5. The substrate and products
were extracted into 6 ml ethyl acetate at 30 min after the reac-
tion was started at 37 °C, and analyzed by liquid chromatog-
raphy/mass spectrometry (LC/MS) as described previously.6)
RESULTS AND DISCUSSION
Expression and Purification of AKR1C12 The amino
acid sequence deduced from the isolated cDNA for
AKR1C12 differs from that for mouse stomach AKR1C1211)
only by two amino acids at positions 81 and 179, but is the
Fig. 1. SDS-PAGE of E. coli Cell Extracts and Purified Recombinant
AKR1C12
same as those for AKRa of BALB/c mouse13) and a transcript
of the AKR1C12 gene of CL57BL/6J mice (accession no.
NM_013777). Although the mouse strain is not specified for
the mouse stomach AKR1C12, the difference in the amino
The gel was stained with 0.2% Coomassie Brilliant Blue. Lanes: a, the extract
(25 mg) of E. coli cells transfected with the expression vector alone; b, the extract
(25 mg) of E. coli cells transfected with the expression vector harboring AKR1C12
cDNA; and c, the purified recombinant AKR1C12 (3 mg). The positions of molecular
acid sequences is probably due to the strain difference, be- mass markers are indicated on the right of the panel.

2490
Vol. 29, No. 12
Table 1. Substrate Specificity in the Reduction Catalyzed by AKR1C12
Km
(mM)
kcat
kcat/Km
Substrate
Non-steroids
ꢂ1
(minꢂ1
)
(minꢂ1 mM
)
PQ
0.7
5.1
7.7
17
16
24
25
113
36
161
115
144
93
321
838
55
48
61
73
36
49
44
51
14
20
11
13
5.6
16
12
78
9.4
8.6
4.3
2.3
2.0
1.8
0.45
0.39
0.12
0.09
0.09
0.06
0.05
0.01
S-Camphorquinone
Isatin
R-Camphorquinone
1-Phenyl isatin
2,3-Heptanedione
Phenyl-1,2-propanedione
Dimethyl-2-oxoglutarate
2,3-Hexanedione
Pyridine-4-aldehyde
1-Nonanal
Fig. 2. Effects of Oxidized and Reduced Coenzymes on Oxidoreductase
Activities of AKR1C12
TBE
(a) Inhibition of NADH-linked (closed symbol) and NADPH-linked (open symbol)
S-camphorquinone reductase activities by NAD(P)ꢃ. The activity was assayed in pres-
ence of the given concentrations of NADꢃ (triangle) or NADPꢃ (circle). The concentra-
tions of NADH and NADPH were the same (0.1 mM), while the concentrations of S-
camphorquinone for NADH-linked and NADPH-linked activities were 0.05 mM and
0.2 mM, respectively, to maintain their saturated concentrations. (b) Inhibition of
NADꢃ-linked (closed symbol) and NADPꢃ-linked (open symbol) dehydrogenase activ-
ities by NAD(P)H. The activity was assayed in the presence of NADH (triangle) or
NADPH (circle) with 1.0 mM S-tetralol as the substrate. The concentrations of NADꢃ
and NADPꢃ were 2.0 mM and 0.1 mM, respectively.
4-Nitrobenzaldehyde
2,3-Pentanedione
Cyclohexanone
Ketosteroids
5b-Pregnan-20a-ol-3-one
17a-Hydroxyprogesterone
4-Androsten-3a-ol-17-one
5b-Androstane-3,17-dione
4-Androstene-3,17-dione
5b-Pregnane-3,20-dione
5b-Androstan-3a-ol-17-one
3.3
14
35
0.39
1.5
2.5
0.52
0.26
0.25
0.30
0.12
0.11
0.07
0.04
0.03
0.02
0.02
12
7.8
7.5
13
S-camphorquinone (5.1 mM) in the presence of 0.1 mM
NADH was also lower than that (16 mM) determined using
0.1 mM NADPH. Even in the oxidation of 1 mM S-tetralol,
NADꢃ (Kmꢄ13 mM and kcat/Kmꢄ13.4 minꢂ1 mMꢂ1) was a bet- 5b-androstane-3,17-dione, three products, 5b-androstan-
ter coenzyme than NADPꢃ (Kmꢄ331 mM and kcat/Kmꢄ 17b-ol-3-one, 5b-androstane-3a,17b-diol and 5b-androstan-
0.04 minꢂ1 mMꢂ1), and the Km values for S-tetralol in the 3a-ol-17-one, were formed. The results suggest that the en-
presence of 1 mM NADꢃ and NADPꢃ were 90 and 530 mM, zyme exhibits 3a-, 17b- and 20a-HSD activities.
respectively. Thus, AKR1C12 exhibits a pronounced prefer-
ence for NAD(H) at a physiological pH 7.4.
In the reverse reaction using NADꢃ as the coenzyme,
AKR1C12 oxidized various non-steroidal alicyclic alcohols
Because the predominant forms of nucleotide coenzymes and long-chain aliphatic alcohols, of which geranylgeraniol
in mammalian cells are NADꢃ and NADPH,21,22) it is gener- (GGOH) was the best substrate, exhibiting the lowest Km and
ally believed that NAD(H)-dependent dehydrogenases func- highest kcat/Km values (Table 2). The enzyme oxidized 20a-,
tion mainly in the oxidative direction, whereas the NADPH- 17b- and 3a-hydroxysteroids, and the oxidized products of
dependent enzymes function as reductases.23) NADꢃ, but not 4-pregnen-20a-ol-3-one, 5a-androstan-17b-ol-3-one and 4-
NADPꢃ, significantly inhibited both NADPH- and NADH- androsten-3a-ol-17-one were identified as progesterone, 5a-
linked reductase activities of AKR1C12 (Fig. 2), and less androstane-3,17-dione and 4-androstene-3,17-dione, respec-
than 20% of the activities were retained in the presence of a tively, by the LC/MS analysis. Although the kcat values for
cellular NADꢃ concentration of 2 mM.22) In case of the oxida- the hydroxysteroids were low, kcat/Km values for many hy-
tion of S-tetralol using the coenzyme cellular concentrations droxysteroids were relatively high due to their low Km values
of NADꢃ and NADPꢃ, the NADPꢃ-linked activity was sig- when compared with those for non-steroidal alicyclic alco-
nificantly inhibited by NADH, whereas the inhibition of the hols. This suggests that the hydroxysteroids are endogenous
NADꢃ-linked activity by NADH and NADPH was relatively substrates of AKR1C12.
low. These results indicate that AKR1C12, which was previ-
ously described as a reductase,11) acts as a NADꢃ-dependent AKR1C Subfamily The broad substrate specificity of
dehydrogenase in vivo. AKR1C12 for 3a-, 17b- and 20a-hydroxysteroids is clearly
Comparison with Other Rodent Enzymes in the
Substrate Specificity The recombinant AKR1C12 re- distinct from NADP(H)-dependent mouse HSDs (AKR1C6,
duced various a-dicarbonyl compounds, aldehydes and some AKR1C14, AKR1C18, AKR1C20 and AKR1C21), which
ketones, in addition to the known substrates, PQ and 4-ni- catalyze the oxidoreduction of hydroxyl groups at one or two
trobenzaldehyde, in the presence of NADH as the coenzyme positions on the steroid substrate.6—10) AKR1C12 is also dif-
(Table 1). The enzyme also reduced 3-, 17- and 20-ketos- ferent from NAD(H)-preferring mouse aldo-keto reductases
teroids at low rates, although no significant activity was ob- (AKR1C13 and AKR1C19), which do not show significant
served towards short-chain aliphatic aldehydes (1-propanal, HSD activity.11,12) On the other hand, the substrate specificity
1-butanal and hexanal) and other ketones (4-nitroacetophe- of AKR1C12 for both steroids and non-steroidal alicyclic al-
none, 4-benzoylpyridine, propiophenone, acetone, mena- cohols are similar to that of AKR1C24, which has recently
dione and PGD2). The reduced products of 5b-androstan-3a- been identified as a novel type of NAD(H)-preferring 17b-
ol-17-one and 17a-hydroxyprogesterone were identified as HSD.24) The pH optima of the NADH-linked S-camphor-
5b-androstane-3a,17b-diol and 4-pregnene-17a,20a-diol-3- quinone reductase and NADꢃ-linked S-tetralol dehydro-
one, respectively, by the LC/MS analysis. In the reduction of genase activities of AKR1C12 were observed around 6.0 and

December 2006
2491
Table 2. Substrate Specificity in the Oxidation Catalyzed by AKR1C12
there is no study on the enzymes responsible for these metab-
olisms of GGOH. The present study is the first report that
identifies the GGOH-metabolizing enzyme with high cat-
alytic activity. GGOH is contained in many plants, which are
ingested as herbal medicines and dietary foods.31) The exoge-
nous GGOH is probably diffused into the cells as demon-
strated in previous studies on its biological actions and me-
tabolism using cultured cells.25—31,33) The high expression of
AKR1C12 and AKR1C24 in the rodent gastrointestinal tract,
together with their high catalytic activity towards GGOH,
suggests their roles in the metabolism of ingested GGOH to
prevent its biological effects on the cells of the tissues. On
the other hand, AKR1C12 is identical to AKRa that is en-
coded in the interleukin-3-regulated gene in mouse myeloid
cells, in which the enzyme is thought to be involved in the
production or catabolism of autocrine factors that promote
the proliferation and/or lineage commitment of early myeloid
progenitors.13) The newly found substrates, GGOH and the
hydroxysteroids, of the enzyme may be the autocrine factors
required in the differentiation of myeloid cells, although fur-
ther studies are needed to understand the significance of the
up-regulation of the enzyme expression in hematopoiesis.
Km
(mM)
kcat
kcat/Km
Substrate
Alicyclic alcohols
ꢂ1
(minꢂ1
)
(minꢂ1 mM
)
S-Indan-1-ol
S-Tetralol
cis-Benzene dihydrodiol
trans-Benzene dihydrodiol
2-Cyclohexen-1-ol
Prostaglandin F2a
63
90
800
760
1080
100
13
14
24
9.9
13
0.20
0.16
0.03
0.01
0.01
0.001
0.12
Aliphatic alcohols
Geranylgeraniol
Farnesol
Nerol
Geraniol
0.4
4.2
41
0.34
1.7
2.1
0.85
0.40
0.05
0.03
42
1.1
1-Nonanol
240
0.14
0.0006
20a-Hydroxysteroids
5-Pregnene-3b,20a-diol
5a-Pregnan-20a-ol-3-one
4-Pregnen-20a-ol-3-one
5b-Pregnan-20a-ol-3-one
17b-Hydroxysteroids
5a-Androstan-17b-ol-3-one
5b-Androstane-3a,17b-diol
5b-Androstan-17b-ol-3-one
Testosterone
2.2
2.5
5.5
5.0
1.2
1.2
1.7
0.14
0.55
0.48
0.31
0.03
10
24
13
36
32
2.3
3.8
1.5
3.9
1.1
0.23
0.16
0.12
0.11
0.03
17b-Estradiol
REFERENCES
3a-Hydroxysteroids
4-Androsten-3a-ol-17-one
4-Pregnen-3a-ol-20-one
5b-Androstan-3a-ol-17-one
35
9.0
10
2.5
0.66
0.15
0.07
0.07
0.02
1) Jez J. M., Flynn T. G., Penning T. M., Biochem. Pharmacol., 54, 639—
647 (1997).
2) Hyndman D., Bauman D. R., Heredia V. V., Penning T. M., Chem. Biol.
Interact., 143—144, 499—525 (2003).
3) Penning T. M., Endocr. Rev., 18, 281—305 (1997).
4) Matsunaga T., Shintani S., Hara A., Drug Metab. Pharmacokin., 21,
1—18 (2006).
5) Vergnes L., Phan J., Stolz A., Reue K., J. Lipid Res., 44, 503—511
(2003).
6) Ishikura S., Nakajima S., Kaneko T., Shintani S., Usami N., Yamamoto
I., Carbone V., El-Kabbani O., Hara A., “Enzymology and Molecular
Biology of Carbonyl Metabolism 12,” eds. by Weiner H., Plapp B.,
Lindahl R., Maser E., Purdue University Press, West Lafayette, 2005,
pp. 341—351.
7) Ishikura S., Nakajima S., Carbone V., El-Kabbani O., Hara A., “Enzy-
mology and Molecular Biology of Carbonyl Metabolism 12,” eds. by
Weiner H., Plapp B., Lindahl R., Maser E., Purdue University Press,
West Lafayette, 2005, pp. 324—331.
10.5, respectively, which are also the same as those of
AKR1C24. Additionally, the NADꢃ-linked S-tetralol dehy-
drogenase of AKR1C12 was inhibited by hexestrol, dienstrol,
zearalenone and genistein (IC50 values were 1.5, 5.0, 4.7 and
40 mM, respectively), which show similar inhibitory potency
for AKR1C24. Furthermore, the amino acid sequences of
AKR1C12 and AKR1C24 are 91% identical. Therefore,
AKR1C12 and AKR1C24 are identical enzyme species.
Role of AKR1C12 It has been proposed that AKR1C24
plays a role in the inactivation of androgen and estrogen, and
the production of active progesterone in the gastrointestinal
tract and other tissues of rats.24) The tissue expression pat-
terns of the mRNAs for AKR1C2424) and AKR1C125) are al-
most the same. The identification of AKR1C12 with the
mouse counterpart of AKR1C24 further supports the pro-
posed physiological role of this enzyme species in the steroid
metabolism. In addition, we emphasize another role of this
enzyme in the metabolism of GGOH. As AKR1C12 did,
AKR1C24 efficiently oxidized GGOH (Kmꢄ0.6 mM and
kcat/Kmꢄ2.8 minꢂ1 mMꢂ1) in our preliminary study. GGOH,
an intermediate product in the mevalonate pathway, is uti-
lized for protein prenylation, a post-translational maturation
of diverse proteins involved in cell proliferation.25—27) The
exposure of GGOH to the cultured cells also results in the in-
hibition of cell proliferation,28) apoptosis,29,30) and activation
of peroxisome proliferators-activated receptors (a and g).31)
In contrast to the biological actions of GGOH, literature on
its metabolism is little. GGOH is suggested to be metabo-
lized to its diphosphate product by a kinase,25) and to ger-
anylgeranoic acid via geranylgeranial by aldehyde and alco-
hol dehydrogenases.32) In rat thymocytes, GGOH is metabo-
lized to (R)-2,3-dihydrogeranylgeranoic acid.33) However,
8) Deyashiki Y., Ohshima K., Nakanishi M., Sato K., Matsuura K., Hara
A., J. Biol. Chem., 270, 10461—10467 (1995).
9) Matsumoto K., Endo S., Ishikura S., Matsunaga T., Tajima K., El-Kab-
bani O., Hara A., Biol. Pharm. Bull., 29, 539—542 (2006).
10) Ishikura S., Usami N., Nakajima S., Kameyama A., Shiraishi H., Car-
bone V., El-Kabbani O., Hara A., Biol. Pharm. Bull., 27, 1939—1945
(2004).
11) Ikeda S., Okuda-Ashitaka E., Masu Y., Suzuki T., Watanabe K.,
Nakano M., Shingu K., Ito S., FEBS Lett., 459, 433—437 (1999).
12) Ishikura S., Horie, K., Sanai M., Matsumoto K., Hara A., Biol. Pharm.
Bull., 28, 1075—1078 (2005).
13) Du Y., Tsai S., Keller J. R., Williams S. C., J. Biol. Chem., 275,
6724—6732 (2000).
14) Nakanishi M., Deyashiki Y., Ohshima K., Hara A., Eur. J. Biochem.,
228, 381—387 (1995).
15) Bradford M. M., Anal. Biochem., 72, 248—254 (1976).
16) Tajima K., Hashizaki M., Yamamoto K., Mizutani T., Drug. Metab.
Dispos., 20, 816—820 (1992).
17) Platt K. L., Oesch F., Synthesis, 7, 449—450 (1977).
18) Jarabak J., Arch. Biochem. Biophys., 291, 334—338 (1991).
19) Jarabak J., Arch. Biochem. Biophys., 292, 239—243 (1992).
20) Shimada H., Oginuma M., Hara A., Imamura Y., Chem. Res. Toxicol.,
17, 1145—1150 (2004).
21) Williamson J. R., Corkey B. E., Method Enzymol., 13, 434—513
(1969).

2492
Vol. 29, No. 12
22) Jones D. P., J. Chromatogr., 225, 446—449 (1981).
23) Labrie F., Luu-The V., Lin S.-X., Labrie C., Simard J., Breton R.,
Bélanger A., Steroids, 62, 148—158 (1997).
24) Ishikura S., Matsumoto K., Sanai M., Horie K., Matsunaga T., Tajima
K., El-Kabbani O., Hara A., J. Biochm. (Tokyo), 139, 1053—1063
(2006).
Sakai I., Ohsawa S., Nakaya K., Anticancer Res., 17, 1051—1057
(1997).
30) Takeda Y., Nakao K., Nakata K., Kawakami A., Ida H., Ichikawa T.,
Shigeno M., Kajiya Y., Hamasaki K., Kato Y., Eguchi K., Jpn. J. Can-
cer Res., 92, 918—925 (2001).
31) Takahashi N., Kawada T., Goto T., Yamamoto T., Taimatsu A., Matsui
N., Kimura K., Saito M., Hosokawa M., Miyashita K., Fushiki T.,
FEBS Lett., 514, 315—322 (2002).
25) Crick D. C., Andres D. A., Waechter, C. J., Biochem. Biophys. Res.
Commun., 237, 483—487 (1997).
26) Zhang F. L., Casey P. J., Annu. Rev. Biochem., 65, 241—269 (1996).
27) Bifulco M., Life Sci., 77, 1740—1749 (2005).
32) Fliesler S. J, Schroepfer G. J., Jr., J. Biol. Chem., 258, 15062—15070
(1983).
28) Burke Y. D., Stark M. J., Roach S. L., Sen S.E., Crowell P. L., Lipids,
32, 151—156 (1997).
33) Kodaira Y., Usui K., Kon I., Sagami H., J. Biochm. (Tokyo), 132,
327—334 (2002).
29) Ohizumi H., Masuda Y., Yoda M., Hashimoto S., Aiuchi T., Nakajo S.,