Purification and characterization of oxidoreductases-catalyzing carbonyl reduction of the tobacco-specific nitrosamine 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK) in human liver cytosol

Article abstract of DOI:10.1080/00498250050119826

1. Four enzymes were purified to homogeneity from human liver cytosol and were demonstrated to be responsible for carbonyl reduction of the tobacco-specific nitrosamine 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK). 2. Carbonyl reductase (EC 1.1.1.184), a member of the short-chain dehydrogenase/reductase (SDR) superfamily, was compared with three isoenzymes of the aldo-keto reductase (AKR) superfamily in terms of enzyme kinetics, co-substrate dependence and inhibition pattern. 3. AKR1C1, 1C2 and 1C4, previously designated as dihydrodiol dehydrogenases (DD1, DD2 and DD4), showed lower K(m) (0.2, 0.3 and 0.8 mM respectively) than did carbonyl reductase (7 mM), whereas carbonyl reductase exhibited the highest enzyme efficiency (V(max)/K(max)) for NNK. Multiplication of enzyme efficiencies with the relative quantities of individual enzymes in cytosol resulted in a rough estimate of their contributions to total alcohol metabolite formation. These were ~ 60% for carbonyl reductase, 20% each for AKR1C1 and 1C2, and 1% for AKR1C4. 4. Except for AKR1C4, the enzymes had a strong preference for NADPH over NADH, and the highest activities were measured with an NADPH-regenerating system. Carbonyl reductase activity was extensively inhibited by menadione, rutin and quercitrin, whereas medroxyprogesterone acetate, phenolphthalein and flufenamic acid were potent inhibitors of AKR1C1, 1C2 and 1C4. 5. In conclusion, cytosolic members of the SDR and AKR superfamilies contribute to reductive NNK detoxification in human liver, the enzymes responsible being carbonyl reductase and aldoketo reductases of the AKR1C subfamily.

Full text of DOI:10.1080/00498250050119826

xenobiotica, 2000, vol. 30, no. 8, 755±769
Puri®cation and characterization of
oxidoreductases-catalyzing carbonyl reduction
of the tobacco-speci®c nitrosamine
4-methylnitrosamino-1-(3-pyridyl)-1-butanone
(NNK) in human liver cytosol
Œ
A. ATALLA , U. BREYER-PFAFF and E. MASER *
‹
‹
‹
University of Marburg, D-35033 Marburg, Germany
Department of Pharmacology and Toxicology, School of Medicine, Philipps-
Œ
Department of Toxicology, University of Tubingen, D-72074 Tubingen, Germany
Received 25 February 2000
1. Four enzymes were puri®ed to homogeneity from human liver cytosol and were
demonstrated to be responsible for carbonyl reduction of the tobacco-speci®c nitrosamine
4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK).
}
2. Carbonyl reductase (EC 1.1.1.184), a member of the short-chain dehydrogenase
reductase (SDR) superfamily, was compared with three isoenzymes of the aldo-keto
reductase (AKR) superfamily in terms of enzyme kinetics, co-substrate dependence and
inhibition pattern.
3. AKR1C1, 1C2 and 1C4, previously designated as dihydrodiol dehydrogenases
(DD1, DD2 and DD4), showed lower K (0.2, 0.3 and 0.8 mm respectively) than did
m
carbonyl reductase (7 mm), whereas carbonyl reductase exhibited the highest enzyme
}
quantities of individual enzymes in cytosol resulted in a rough estimate of their
eæciency (V
K ) for NNK. Multiplication of enzyme eæciencies with the relative
max
m
C
contributions to total alcohol metabolite formation. These were 60% for carbonyl
reductase, 20% each for AKR1C1 and 1C2, and 1% for AKR1C4.
4. Except for AKR1C4, the enzymes had a strong preference for NADPH over
NADH, and the highest activities were measured with an NADPH-regenerating system.
Carbonyl reductase activity was extensively inhibited by menadione, rutin and quercitrin,
whereas medroxyprogesterone acetate, phenolphthalein and ¯ufenamic acid were potent
inhibitors of AKR1C1, 1C2 and 1C4.
5. In conclusion, cytosolic members of the SDR and AKR superfamilies contribute to
reductive NNK detoxi®cation in human liver, the enzymes responsible being carbonyl
reductase and aldoketo reductases of the AKR1C subfamily.
Introduction
The tobacco-speci®c nitrosamine 4-methylnitrosamino-1-(3-pyridyl)-1-
butanone (NNK) is among the most abundant carcinogens in tobacco and tobacco
smoke (Hecht and Hoåmann 1988, Fischer et al. 1990). Its carcinogenic potency in
mouse, rat and hamster and its speci®city for the lung has strongly implicated NNK
as an important etiological factor in the causation of lung cancer in human smokers
(Hecht and Hoåmann 1988, 1989).
NNK is a procarcinogen that requires metabolic activation by cytochrome P450
(CYP) enzymes to express its mutagenic and carcinogenic properties (Hecht and
Hoåmann 1988, Crespi et al. 1991). This activation involves the a-hydroxylation
of the two carbons bound to the N-nitroso group leading to the formation of reac-
tive electrophilic species that methylate and pyridyloxobutylate purine bases of
DNA (Murphy et al. 1990b). The pathway competing with NNK activation is
!
* Author for correspondence; e-mail: maser mailer.uni-marburg.de
}
’
http: www.tandf.co.uk journals
Xenobiotica ISSN 0049-8254 print ISSN 1366-5928 online
}}
2000 Taylor & Francis Ltd
}

756
A. Atalla et al.
NNK carbonyl reduction. Like NNK, the resulting N-nitroso alcohol 4-
(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) has also been proposed to
be carcinogenic in animals (Hecht 1994). Nevertheless, carbonyl reduction of NNK
is important since it initiates the detoxi®cation of NNK by providing the hydroxy
butyl function necessary for glucuronosylation and ®nal excretion (Hecht 1994).
Pyridine N-oxidation of NNAL and NNK is also considered as a detoxi®cation
reaction, since NNK N-oxides are only one-tenth as carcinogenic as NNK, and
metabolic studies in laboratory animals failed to demonstrate substantial amounts of
metabolic activation products of these compounds (Castonguay et al. 1983a, Hecht
1994, 1996). However, NNAL N-oxide and NNK N-oxide are relatively minor
NNK metabolites in the urine of smokers and smokeless tobacco users (Carmella et
al. 1997), such that detoxi®cation of NNK occurs mainly by carbonyl reduction and
glucuronosylation in humans (Carmella et al. 1997, 1995, 1993, Hecht et al. 1993a).
The extent of the competing pathways a-hydroxylation versus carbonyl
reduction of NNK that occur in diåerent species and tissues may depend on the
composition of enzymes present in the tissues and may be related to the
organospeci®city of and individual susceptibility to NNK-induced cancerogenesis.
Therefore, to understand human susceptibility to NNK, it is important to identify
the enzymes that catalyze the activation and detoxi®cation reactions.
Several studies have addressed the question of which CYP enzymes mediate the
activation of NNK. In human, strong evidence indicates involvement of CYP1A2,
2A6 and 3A4 in NNK activation (reviewed in Hecht 1994, 1998, Smith et al. 1997).
However, it was estimated that 39±100% of the NNK dose was excreted as NNAL
and NNAL-glucuronide in smoker’s urine (Carmella et al. 1993). Although carbonyl
reduction constitutes the major pathway of NNK metabolism in vivo and several
tissues in vitro (reviewed in Maser 1998, 1997), research on the enzymes catalyzing
this reaction has largely been neglected (Hecht et al. 1980).
Two major groups of enzymes can catalyze the reduction of xenobiotic ketones
in mammals (Wermuth 1985, Flynn and Green 1993, Maser 1995). One constitutes
}
the family of short-chain dehydrogenases reductases (SDR), which comprises
membrane-bound as well as soluble enzymes (Jornvall et al. 1995). A member of this
family, 11b-hydroxysteroid dehydrogenase 1 (11b-HSD 1; EC 1.1.1.146), was
identi®ed as a microsomal NNK reductase in mouse liver and lung (Maser 1998,
Maser et al. 1996). A cytosolic SDR enzyme designated carbonyl reductase (CR; EC
1.1.1.184) occurs in all human tissues investigated (Wermuth 1985) and it can
reduce a broad spectrum of ketones and quinones (Wermuth 1985, Wermuth 1981).
On puri®cation, it is separated into three forms (CR I±III) that diåer in their pI, but
which exhibit very similar enzymatic properties (Wermuth 1981, Inazu et al. 1992).
Recently, NNK carbonyl reduction has been demonstrated in cytosol from mouse
liver and lung (Atalla and Maser 1999) as well as human liver and lung (Maser et al.
2000). Experiments with various inhibitors pointed to the participation of CR in this
reaction.
For the second superfamily of carbonyl-reducing enzymes, the aldoketo
reductases (AKR), a systematic nomenclature has been proposed on the basis of
amino acid sequence identities (Jez et al. 1997). The members of the various families
and subfamilies bear no structural similarities to the SDR proteins in spite of
overlapping substrate speci®cities (Flynn and Green 1993, Maser 1995). Human
AKR capable of reducing xenobiotic ketones are primarily found in the AKR1C
subfamily. They use NADPH as a co-substrate as does CR, but in contrast to the

Puri®cation of NNK oxidoreductases from human liver cytosol
757
latter they can catalyse eæciently the oxidation of aliphatic alcohols and diols in the
+
presence of NADP . This reaction has led to their designation as dihydrodiol
dehydrogenases (DD) (Vogel et al. 1980, Hara et al. 1990). Fractionation of human
liver cytosol led to the puri®cation of AKR1C1 (DD1 of Hara et al. 1990) with
activity towards 20a-hydroxysteroids, AKR1C2 (DD2 of Hara et al. 1990),
previously characterized as a bile-acid-binding protein (Stolz et al. 1993) and
AKR1C4 (DD4 of Hara et al. 1990) that had been described as human liver
chlordecone reductase (Molowa et al. 1986) and was later obtained in three
molecular forms as 3a-hydroxysteroid dehydrogenase (Takikawa et al. 1990). The
C
AKR have molecular masses of 37 kDa (Hara et al. 1996a), whereas those of the
SDR are 30 kDa (Wermuth et al. 1988).
C
A procedure for the parallel puri®cation of the three CR isoforms and of
AKR1C1, 1C2 and 1C4 from human liver has been developed (Breyer-Pfaå and Nill
2000). In the present investigation, this procedure was used to obtain a complete
survey of all cytosolic enzymes participating in NNAL formation in human liver by
measuring the NNK reductase activity in each puri®cation step. Human liver was
selected for our studies, since, in contrast to other tissues, all possible cytosolic
NNK carbonyl-reducing enzymes could be expected in this organ. Moreover, NNK
is known to distribute throughout the entire body with the liver being the major site
of NNK detoxi®cation. According to our results, NNK was a substrate for CR as
well as for aldoketo reductases, but the kinetic parameters varied widely among
enzymes.
Some of the results have been presented in abstract form (Atalla et al. 1999).
Materials and methods
Materials
Human liver samples (female 25 years of age, female 54, male 70) were obtained from Professor Dr
W. Lauchart, Department of Surgery, University of Tubingen, Germany. These were either samples
from livers excluded from transplantation for medical reasons or excess normal tissue obtained on partial
hepatectomy for tumour metastases. Pieces (5±10 g) were stored at 80 °C for up to 7 years, which did
®
not aåect speci®c enzyme activities towards standard substrates (4-benzoylpyridine and (S)-(-)-1-
indanol).
Human recombinant CR was generously provided by Professor Dr B. Wermuth, Inselspital, Bern,
Switzerland (Bohren et al. 1994).
Chemicals
NNK (& 99% purity) was from Campro Scienti®c (Emmerich, Germany); NNAL was a gift from
D. Hoåmann (American Health Foundation, Valhalla, NY, USA). Pyridine nucleotide co-substrates
and glucose 6-phosphate dehydrogenase were from Boehringer (Mannheim, Germany), Sephadex G-
100, Q Sepharose, CM Sepharose, DEAE Sepharose, Polybuåer Exchanger PBE 94, and the Polybuåers
74 and 96 from Pharmacia (Freiburg, Germany), hydroxyapatite Bio-Gel HTP from Bio-Rad (Munich,
Germany), Reactive Red 120 Agarose, (S)-(-)- and (R)-( )-1-indanol, medroxyprogesterone acetate,
®
¯ufenamic acid, quercitrin, rutin, dexamethasone, phenolphthalein and the molecular weight standard
Dalton Mark VII-L for protein electrophoresis from Sigma-Aldrich (Deisenhofen, Germany), 4-
benzoylpyridine and menadione from Merck (Darmstadt, Germany), a 50-ml ultra®ltration chamber and
®lters YM10 from Amicon (Witten, Germany), the carrier ampholytes Servalyt 6±9 and 9±11 from Serva
(Heidelberg, Germany), and Resolyte 3.5±10 and 6±9.5 from Merck.
Enzyme isolation
The preparation of liver cytosol and its fractionation by gel ®ltration, ion exchange, hydroxy apatite
and aænity chromatography followed by chromatofocusing were largely carried out by a modi®cation
(Breyer-Pfaå and Nill 2000) of a procedure of Hara et al. (1990) (®gures 1 and 2). Fractions were tested
for AKR activity by measuring the NADPH production as re¯ected by an increase of absorbance at
339 nm on incubation with 0.5 mm (S)-(-)-1-indanol. CR activity was followed by the decrease of

758
A. Atalla et al.
Figure 1. Puri®cation scheme of the aldoketo reductases AKR1C1, 1C2 and 1C4, and the isoforms
I±III of CR. For details, see the Materials and methods. Q, Q Sepharose; CM, CM Sepharose;
HA, Bio-Gel HTP; R, Reactive Red 120 Agarose; PBE, polybuåer exchanger; DE, DEAE
Sepharose; *, Q3-CM; AKR, aldoketo reductase; CR, carbonyl reductase.
NADPH absorbance at 339 nm with 1 mm 4-benzoylpyridine (Breyer-Pfaå and Nill 2000). Minor
modi®cations were introduced in the present investigation (®gure 1, table 1). Thus, the Q3 fraction
resulting from Q Sepharose fractionation of the gel ®ltrate was applied to 10 ml CM Sepharose in 10 mm
sodium phosphate pH 6.5 and the fraction not adsorbed (Q3-CM) was applied to 3 ml Reactive Red 120
Agarose. After 17 h at 4 °C, a ®rst fraction with enzymatic activity (Q3-CM-R1) could be eluted with
10 mm Tris-HCl, pH 7.5, in 20% glycerol, whereas a second fraction (Q3-CM-R2) was obtained after
adding 2.5 m NaCl to the eluent. On chromatofocusing on PBE 94 of these two fractions by the procedure
of Breyer-Pfaå and Nill (2000), the ®rst one was separated into CR II and III (eluted at pH 7.1 and 6.9
respectively), and the second one into CR II and several partially overlapping fractions displaying (S)-
1-indanol oxidase activity. Of these, only the ®rst one (Q3-CM-R2-PBE2, eluted at pH 6.3) served for
puri®cation by DEAE Sepharose chromatography to obtain pure AKR1C4 (Breyer-Pfaå and Nill 2000).
Enzyme identi®cation
Except for CR III, the enzymes were homogeneous on SDS-gel electrophoresis (Laemmli 1970) with
apparent molecular masses of 36 kDa for the aldoketo reductases and 32±33 kDa for the CR (®gure 3). In
isoelectric focussing, pI was 8.0 for AKR1C1, 7.25 for AKR1C2 and 6.3 for AKR1C4 in accordance with
Hara et al. (1996b). Two more acidic isoforms of AKR1C4 were eluted from PBE 94 with decreasing pH
of the polybuåer (Breyer-Pfaå and Nill 2000), but these were not obtained in pure form. CR I and II
puri®ed to homogeneity had pI 7.8 and 7.3 in accordance with Wermuth (1981), while CR III with a
¯
pI 6.7 showed an admixture on electrophoresis.
¯
The identities of the enzymes were further con®rmed by partial amino acid sequence analysis (4base
lab, Reutlingen, Germany). Sequences of tryptic fragments were identical with published sequences for
AKR1C2 (amino acids 305±318; Hara et al. 1996b), for CR I from a Q2 fraction and CR II from a Q3
fraction (amino acids 113±119; Wermuth et al. 1988). In addition, speci®c activities towards standard
substrates and results of inhibition experiments were in accordance with published data (Breyer-Pfaå
and Nill 2000).
Assay of NNK carbonyl reduction
The assay was performed by incubating 20 ll enzyme preparation (corresponding to 1±5 lg protein)
+
and 100 ll of an NADPH-regenerating system (®nal concentrations: NADP , 0.2 mm; glucose 6-
}
phosphate, 2 mm; glucose 6-phosphate dehydrogenase, 1 U ml; MgCl , 8 mm) in 100 mm sodium
#
phosphate buåer, pH 7.4, at 37 °C. The reaction was started by adding 10 ll NNK to ®nal concentrations
ranging from 0.5 to 8 mm (CR) or 0.05 to 4 mm (AKR) in a total volume of 400 ll. The reaction was
stopped after 30 min by denaturating the protein for 3 min at 100 °C before cooling on ice. The samples
were centrifuged in an Eppendorf centrifuge at 8000g for 5 min to sediment organic material, and 200 ll
supernatant served for HPLC analysis.

Puri®cation of NNK oxidoreductases from human liver cytosol
759
Figure 2. Elution patterns from Q Sepharose. (upper) Total proteins (UV, E ) and reductase
#)!
activity towards NNK (*); (lower) 4-benzoylpyridine reductase (+) and (S)-indanol oxidase (*)
activities. Fractions combined were 12±26 (Q1), 43±62 (Q2) and 63±73 (Q3).
For determining the co-substrate dependency, 0.5 mm NADH or NADPH instead of the NADPH-
regenerating system were employed. The pH dependence of NNK carbonyl reduction was tested
between pH 4.5 and 11.5 by using a universal buåer containing acetic acid, MES, KH PO , Tris and
#
%
glycine in a concentration of 30 mm each.
Control experiments with heat-inactivated enzyme fractions were performed to determine non-
enzymatic substrate conversion.
Metabolite determination by HPLC
NNK and its reduced alcohol metabolite NNAL were detected on a reverse-phase HPLC system
(Merck Hitachi L-6220 Intelligent Pump, Merck Hitachi AS-2000 A Autosampler, Merck Hitachi L-
4000 A UV Detector), using a 4.6 mm 25 cm LiChrosphere 100 RP18 column (Merck, Germany) and
}
¬
a Shimadzu C-R6A Chromatopac integrator. With an eluent of 25% acetonitrile (v v) in 10 mm sodium
C
}
phosphate buåer, pH 7.4, at a ¯ow rate of 0.5 ml min, NNK elutes at 11.0 min and NNAL at 6.8 min.
Substances were monitored at 230 nm. The identity of NNK and NNAL was determined by injecting
authentic standards. Concentrations of NNK and NNAL were calculated from peak areas of samples and
standards used for calibration curves.

760
A. Atalla et al.
Table 1. NNK carbonyl reductase activities and protein quantities of AKR1C1, AKR1C2, AKR1C4
and CR isoforms during the puri®cation procedure.
Protein
(mg)
Activity
(mU)
Speci®c activity
}
(mU NNAL mg)
Fraction or enzyme
Cytosol
Gel ®ltrate
Q1
Q1-CM
AKR1C1
Q2
Q2-CM
Q2-CM-HA1
CR I
810
113
6.0
2.6
0.6
31
9.2
1.3
0.16
0.15
3.7
1.7
45
26
785
650
16
8.8
6
294
124
81
1.0
5.8
2.6
3.4
10
9.5
14
62
180
160
7.8
11
3.6
4.7
3.8
200
66
29
23
28
CR II
Q2-CM-HA2
AKR1C2
Q3
Q3-CM
Q3-CM-R1
CR II
19
162
124
53
10
12
72
14
0.05
0.18
3.5
CR III
Q3-CM-R2
20
CR II
Q3-CM-R2-PBE2
AKR1C4
0.14
0.27
0.15
21
0.65
0.45
150
2.4
3.0
Values of puri®ed enzyme preparations are shown in bold.
Activities were determined using 0.5 mm NNK and an NADPH-
regenerating system.
Figure3. SDS-PAGE demonstratingthe homogeneity ofpuri®ed enzymes. (upper) Lanes 1, AKR1C1;
2, AKR1C2; 3, AKR1C4; M, molecular mass standard; (lower) lanes 1, CR I; 2, CR II from Q2;
3, CR II from Q3-CM-R1; 4, CR II from Q3-CM-R2; 5, CR III; M, molecular mass standard.
Protein (3±6 lg) was applied per lane.

Puri®cation of NNK oxidoreductases from human liver cytosol
761
Figure 4. Co-substrate dependency of puri®ed NNK carbonyl reducing enzymes. Enzyme activities
were assayed in a reaction mixture containing 0.5 mm NNK and either 0.5 mm NADH, 0.5 mm
NADPH or an NADPH-regenerating system (RS). Values represent the speci®c activity
³
(nmol mg min) and are the means SD of three-to-four determinations (for AKR1C4: n 2).
Upper panel, AKR1C1; second panel, AKR1C2; third panel, AKR1C4; lower panel, CR I.
}
}
¯
Inhibitor studies
Stock solutions of the inhibitors were added to the incubation mixtures. Inhibitors not suæciently
soluble in buåer were dissolved either in methanol or 0.03 m NaOH. Control velocities were determined
in the presence of appropriate quantities of the solvents. After a pre-incubation of 5 min, the reaction was
started by adding NNK.
Protein determination
Protein concentration was determined by the method of Bradford (1976) using bovine serum albumin
as standard.

762
A. Atalla et al.
³
Figure 5. Kinetics of NNAL formation from NNK by puri®ed enzymes. Values are means SD of
}
(mU mg), Cl (ll mg min).
}
}
three-to-four individual preparations. K (mm), V
m
max
int
Kinetic parameter estimations
Time and enzyme protein concentrations were chosen so that reaction velocities were time linear,
which was true at least for 1 h. Kinetic parameters were calculated according to the Michaelis±Menten
equation by using GraphPad Inplot kinetic computer software. In cases of substrate inhibition, a formula
}
that maximal substrate concentrations exceeded K at least two-fold. Otherwise, Cl was derived from
for non-competitive inhibition was used. Enzyme eæciency (Cl ) was calculated as V
K provided
int max
m
m int
the ratios of reaction rate substrate concentration at concentrations far below K . Speci®c activities are
}
expressed as mU mg (nmol NNAL mg min).
m
}
}
}
Results
Puri®cation of NNK carbonyl reducing enzymes
The fractionation scheme is outlined in ®gure 1. The initial separation on Q
Sepharose resulted in three enzymatically active fractions demonstrating that
several soluble enzymes participate in NNK reduction to NNAL in human liver
cytosol (®gure 2). NNK carbonyl reducing activity was detected in all fractions
}
exhibiting aldoketo reductase (AKR) and or CR activities with standard substrates
as shown, for example, in ®gure 2. During subsequent puri®cation, speci®c
activities increased (table 1) in parallel to those towards (S)-(-)-1-indanol on
puri®cation of AKR1C1, 1C2 and 1C4 and to those towards 4-benzoylpyridine on
puri®cation of the CR isoforms. The apparent homogeneity of the individual
enzymes in gel electrophoresis is shown in ®gure 3. An exception is CR III which,
in addition to the expected protein with a molecular mass of 32±33 kDa, obviously
showed an admixture of an autocatalytically modi®ed form of CR (Wermuth,
personal communication). No NNK reduction was observed in fractions devoid of

Puri®cation of NNK oxidoreductases from human liver cytosol
763
Figure 6. pH dependency of NNK carbonyl reductase activities of AKR1C1, AKR1C2 and carbonyl
reductase (CR I). Activities were determined with 0.5 mm NNK as the substrate and an NADPH-
regenerating system as co-substrate. Data points are the average of three determinations.
activity towards standard substrates. Determinations of kinetic parameters, pH
dependence and inhibitor sensitivities were carried out with CR I.
Cosubstrate requirement of puri®ed NNK reductases
Very little NNK reduction took place on incubating pure enzymes with NADH
as the electron donor, except for AKR1C4, which produced considerable quantities
of NNAL with NADH. Overall, NADPH was the preferred co-substrate, and
highest reaction rates were obtained with an NADPH-regenerating system (®gure
+
4), probably because NADP produced during the incubation with NADPH alone
can act as an inhibitor. All further experiments were carried out using the NADPH-
regenerating system.

764
A. Atalla et al.
Table 2. Eåect of inhibitors on carbonyl reduction mediated by AKR1C1, AKR1C2, AKR1C4
and CR I.
Enzyme
Inhibitor
Concentration (lm)
Inhibition (%)
AKR1C1
phenolphthalein
medroxyprogesterone acetate
¯ufenamic acid
ethanol
phenolphthalein
medroxyprogesterone acetate
¯ufenamic acid
ethanol
phenolphthalein
medroxyprogesterone acetate
dexamethasone
ethanol
menadione
rutin
5
2
10
5%
1
2
10
5%
10
10
73
48
51
70
26
60
66
42
71
72
75
27
81
63
43
30
AKR1C2
AKR1C4
50
5%
100
100
100
5%
Carbonyl reductase
quercitrin
ethanol
Activities were determined using 1 mm NNK and an NADPH-regenerating system. Values represent
!
the mean of two or three determinations. Deviations were 9%.
Absolute (uninhibited) values were: AKR1C1 12, AKR1C2 9 6, AKR1C4 6 5 and carbonyl
¯
¯
±
¯ ±
}
reductase 285 nmol NNAL mg min.
}
¯
Figure 7. Metabolic pathways of NNK. UDPGT, uridine diphosphate glucuronosyl transferase.
Kinetics of puri®ed NNK reductases
Substrate concentration dependencies of the enzymes with respect to NNAL
production from NNK and kinetic parameters are given in ®gure 5. Low K and
m
were obtained with the AKR, whereas with CR saturation of NNAL formation
V
was not apparent such that the kinetic parameters have to be regarded as estimates.
max
"
Slight substrate inhibition occurred with AKR1C1 at NNK concentrations 1 mm
(®gure 5). Carbonyl reduction eæciencies re¯ected by enzyme eæciencies (Cl )
int
}
were highest with CR (370 ll mg min) compared with those of the AKR. When
Cl for CR was calculated from the reaction rate substrate concentration ratio at
}
}
int
}
NNK concentrations of 0.5 or 1 mm, 450ll mg min was obtained. Interestingly,
400 ll mg min resulted from incubations of NNK with human recombinant CR by
}
}
}
following a similar approach (data not shown).

Puri®cation of NNK oxidoreductases from human liver cytosol
765
pH-dependence of puri®ed enzymes
AKR1C1 showed a broad optimum of NNK reduction rates at pH 6±7 and
C
AKR1C2 at pH 6.5±7.5 (®gure 6). Pure AKR1C4 was not available in quantities
C
suæcient for studying pH dependence. CR exhibited an optimum at pH 7.
C
Inhibition of puri®ed enzymes
Selected substances were employed to inhibit the isolated NNK carbonyl
reducing enzymes (table 2). NNK carbonyl reduction by AKR was potently
inhibited by medroxyprogesterone acetate, phenolphthalein and ¯ufenamic acid. In
addition, dexamethasone potently inhibited AKR1C4. The formation of NNAL by
CR was decreased most extensively by the competing substrate menadione, followed
by the speci®c inhibitors rutin and quercitrin. Ethanol considerably aåected all
cytosolic NNK carbonyl reducing enzymes at a concentration of 5%, with AKR1C1
exhibiting a residual activity of 30%.
Discussion
Carbonyl-reducing enzymes are generally regarded as part of the metabolic
systems protecting against toxic actions of endogenous and exogenous compounds.
For instance, CR was the main quinone reductase in human liver cytosol. By
transfer of two electrons the enzyme prevents one-electron reduction of quinones,
which may give rise to oxidative stress and thus the enzyme protects against indirect
toxic eåects (Wermuth et al. 1986). Likewise, aldehydes and ketones are mostly
more toxic than their corresponding alcohol metabolites (Wermuth 1985, Flynn and
Green 1993). Even if an alcohol can exert similar toxic eåects like the parent ketone,
as is the case with NNAL and NNK, its toxic potential will mostly be less due to its
easier elimination either in free form or following conjugation. Therefore, CR as
well as aldoketo reductases can be considered as contributing to the detoxi®cation of
an important human carcinogen (®gure 7).
In experiments with rat and monkey in vivo, NNK is rapidly reduced to NNAL,
and in urine from human tobacco users NNAL and its glucuronide, but neither
NNK nor NNK N-oxide can be detected (Hecht 1998). This is in accordance with
the biochemical properties of the enzymes mediating NNK conversion to NNAL,
because all of them, including microsomal 11b-HSD 1, prefer NADPH over
NADH as co-substrate (Wermuth 1981, Hara et al. 1990, Maser and Bannenberg
+
1994, Maser et al. 1996, Maser 1998). In view of the high NADPH NADP ratio in
}
cell cytosol (Veech et al. 1969), substrate reduction should predominate over the
reverse reaction.
At NNK doses in the lower lg kg range, rat excretes very low amounts of
}
NNAL and or NNAL-glucuronide, accounting for only 6±7% of the administered
}
dose (Morse et al. 1990, Richter and Tricker 1994). In monkey urine, NNAL plus
NNAL-glucuronides made up 21% (Hecht et al. 1993b) or 32% (Meger et al. 1999)
of the NNK dose. Interestingly, in smokers’ urine, the sum of NNAL and the two
NNAL-glucuronides represent 40±100% of the estimated NNK uptake with
cigarette mainstream smoke (Carmella et al. 1993), indicating large interindividual
}
variations in the expression of NNK carbonyl reducing enzymes and or NNAL-
glucuronosyltransferases. Considering the fact that only 16% of habitual smokers
develop lung cancer (Belinsky et al. 1997), these interindividual variations of NNK-
detoxifying enzymes are of major interest since they may be linked to NNK-induced
carcinogenesis. In view of the patterns of urinary metabolites, it may behypothesized

766
A. Atalla et al.
that during evolution the capability to detoxify NNK via carbonyl reduction and
glucuronosylation increased in the sequence rat±monkey±human.
Investigations on the in vitro metabolism of NNK have generally been performed
with tissue homogenates, subcellular fractions, cultured cells, cultured tissues and
isolated perfused tissues (Castonguay et al. 1983b, Murphy et al. 1990a, Richter et
al. 1998, Schrader et al. 1998, Schulze et al. 1998, 1996) but to a lesser extent with
puri®ed enzymes. With respect to carbonyl reduction, 11b-hydroxysteroid de-
hydrogenase 1 (EC 1.1.1.146), an enzyme physiologically responsible for the
interconversion of active 11-hydroxy-glucocorticoids to the inactive 11-oxo forms,
has previously been identi®ed as microsomal carbonyl reductase involved in the
reduction of NNK to NNAL (Maser et al. 1996, Maser 1997). In the present study,
four cytosolic enzymes have additionally been puri®ed and characterized in terms
of their NNK carbonyl-reducing activity. This multiplicity of NNK carbonyl
reducing enzymes has to be taken into account when calculating kinetic values for
NNAL formation in crude cellular fractions or whole tissues and assigning these
values to individual enzymes. For example, in studies on NNK metabolism with
isolated perfused liver, Schrader et al. (1998) observed a shift from a-hydroxylation
towards NNAL formation at high compared with low NNK concentrations,
re¯ecting the fact that several carbonyl reducing enzymes with diåerent aænities to
NNK are active in NNAL formation.
Estimates on the contribution of individual enzymes to total NNK reduction in
human liver cytosol can be based on enzyme eæciencies and relative enzyme
quantities. On the assumption that losses during puri®cation of cytosolic enzymes
are comparable for the aldoketo reductases and CR, ratios of enzyme quantities in
liver cytosol in the present investigation and in a previous report (Breyer-Pfaå and
C
Nill 2000) should be 3:3:1:1.5 for AKR1C1, AKR1C2, AKR1C4 and CR I±III
(table 1). Multiplication by the enzyme eæciencies (®gure 5) results in relative
C
contributions of
20, 20, 1 and 60% to NNK reduction by the four enzymes. No
indication was obtained of the occurrence of further cytosolic enzymes with NNK-
reducing capacity in human liver.
The inhibitor sensitivities of aldoketo reductases have in previous investigations
mainly been determined for dehydrogenase reactions. Nonetheless, the present data
on the inhibition of NNK carbonyl reduction (table 2) are in accordance with
reports on potent inhibition of AKR1C1, 1C2 and 1C4 by medroxyprogesterone
acetate (Hara et al. 1990), of AKR1C4 by dexamethasone (Deyashiki et al. 1992)and
of AKR1C1 and 1C2 by ¯ufenamic acid (Matsuura et al. 1997). The extensive
inhibitory eåect of the ¯avonoids rutin and quercitrin on NNK reduction by CR
corresponds to ®ndings with 4-nitrobenzaldehyde as substrate (Wermuth 1981). A
conspicuous ®nding was the eæcient NNK reduction by AKR1C4 with NADH as
a co-substrate (®gure 4). Whereas chlordecone reduction by chlordecone reductase
(later identi®ed as AKR1C4) was not supported by NADH (Molowa et al. 1986),the
dehydrogenation of (S)-1-indanol catalyzed by AKR1C4 proceeded at similar rates
+
with 0.25 mm NADP and 2 mm NAD as co-substrates (Hara et al. 1990). In
+
+
contrast, NAD was a much poorer co-substrate in the same reaction mediated by
AKR1C1 and 1C2 (Hara et al. 1990). Thus, in oxidative as well as in reductive
reactions, the preference for triphospho- over diphosphopyridine nucleotides seems
to be less distinct with AKR1C4 than with the other AKR.
Nodata are available on the product stereoselectivity of NNK carbonyl reduction
by individual enzymes. NNAL is a chiral alcohol and separation of its diastereomeric

Puri®cation of NNK oxidoreductases from human liver cytosol
767
glucuronides excreted in urine indicated species-speci®c diåerences in the pro-
}
duction and or conjugation of the enantiomers (reviewed in Hecht 1998). Their
absolute con®gurations were recently elucidated with the result that (R)-NNAL is
mainly produced in rat liver microsomes, but (S)-NNAL-glucuronide pre-
dominates in rat urine. The patas monkey given NNK intravenously and human
heavily using tobacco (snuæng) excreted more (R)- than (S)-NNAL-glucuronide in
urine (Hecht et al. 1997). Interestingly, the higher tumorigenicity of (R)- compared
}
with (S)-NNAL in the A J mouse lung has been related to a preferential metabolic
activation of (R)-NNAL and preferential glucuronidation of (S)-NNAL
(Upadhyayaet al. 1999). Therefore, further investigations on the stereoselectivity of
NNAL formation and disposition in man seem warranted.
In conclusion, four enzymes have been puri®ed from human liver cytosol to
homogeneity, AKR1C1, AKR1C2, AKR1C4 and CR, and it has been shown that
these enzymes are responsible for carbonyl reduction of the tobacco-speci®c
nitrosamine NNK. Considering that carbonyl reduction precedes glucuronosylation
and excretion of NNK, it is conceivable that these cytosolic members of the AKR
and SDR protein superfamilies contribute to the overall detoxi®cation of this
carcinogen.
Acknowledgements
This study was supported by grants (to E. Maser) from the Deutsche
}
Forschungsgemeinschaft (MA 1704 3±1), the Stiftung Verhalten und Umwelt
(Munich, Germany), and the European Commission (BIO4±97±2123).
References
Atalla, A., Breyer-Pfaff, U. and Maser, E., 1999, Carbonyl reductase and three aldo-keto reductases
responsible for NNK carbonyl reduction in human liver cytosol. Naunyn-Schmiedeberg’s Archives
of Pharmacology, Suppl. 359, R154.
Atalla, A. and Maser, E., 1999, Carbonyl reduction of the tobacco-speci®c nitrosamine 4-
(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in cytosol of mouse liver and lung.
Toxicology, 139, 155±166.
Belinsky, S. A., Swafford, B. S., Finch, G. L., Mitchell, C. E., Kelly, G., Hahn, F. F., Anderson,
M. W. and Nikula, K. J., 1997, Alterations in the K-ras and p53 genes in rat lung tumors.
Environmental Health Perspectives, 105 (suppl. 4), 901±906.
Bohren, K. M., Wermuth, B., Harrison, D., Ring, D., Petsko, G. A. and Abby, K. H., 1994,
Expression, crystallization and preliminary crystallographic analysis of human carbonyl reductase.
Journal of Molecular Biology, 244, 659±664.
Bradford, M. M., 1976, A rapid and sensitive method for the quantitation of microgram quantities of
proteins using the principle of dye-binding. Analytical Biochemistry, 72, 248±254.
Breyer-Pfaff, U. and Nill, K., 2000, High-aænity stereoselective reduction of the enantiomers of
ketotifen and of ketonic nortriptyline metabolites by aldo-keto reductases from human liver.
Biochemical Pharmacology, 59, 249±260.
Carmella, S. G., Akerkar, S. and Hecht, S. S., 1993, Metabolites of the tobacco-speci®c nitrosamine
4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in smokers’ urine. Cancer Research, 53,
721±724.
Carmella, S. G., Akerkar, S. A., Richie, J. P. and Hecht, S. S., 1995, Intraindividual and
interindividual diåerences in metabolites of the tobacco-speci®c lung carcinogen 4-
(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in smoker’s urine. Cancer Epidemiology
Biomarkers and Prevention, 4, 635±652.
Carmella, S. G., Borukhova, A., Akerkar, S. A. and Hecht, S. S., 1997, Analysis of human urine for
pyridine-N-oxide metabolites of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, a tobacco-
speci®c lung carcinogen. Cancer Epidemiology Biomarkers and Prevention, 6, 113±120.
Castonguay, A., Lin, D., Stoner, G. D., Radok, P., Furuya, K., Hecht, S. S., Schut, H. A. J. and
}
Klaunig, J. E., 1983a, Comparative carcinogenicity in A J mice and metabolism by cultured
mouse peripheral lung of N«-nitrosonornicotine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone, and their analogues. Cancer Research, 43, 1223±1229.

768
A. Atalla et al.
Castonguay, A., Stoner, G. D., Schut, H. A. and Hecht, S. S., 1983b, Metabolism of tobacco-speci®c
N-nitrosamines by cultured human tissues. Proceedings of the National Academy of Sciences, USA,
80, 6694±6697.
Crespi, C. L., Penman, B. W., Gelboin, H. V. and Gonzalez, F. J., 1991, A tobacco smoke-derived
nitrosamine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, is activated by multiple human
cytochrome P450s including the polymorphic human cytochrome P4502D6. Carcinogenesis, 12,
1197±1201.
Deyashiki, Y., Taniguchi, H., Amano, T., Nakayama, T., Hara, A. and Sawada, H., 1992, Structural
and functional comparison of two human liver dihydrodiol dehydrogenases associated with 3
alpha-hydroxysteroid dehydrogenase activity. Biochemical Journal, 282, 741±746.
Fischer, S., Spiegelhalder, B., Eisenbarth, J. and Preussmann, R., 1990, Investigations on the origin
of tobacco-speci®c nitrosamines in mainstream smoke of cigarettes. Carcinogenesis, 11, 723±730.
Flynn, T. G. and Green, N. C., 1993, The aldo-keto reductases: an overview. Advances in Experimental
Medicine and Biology, 328, 251±257.
Hara, A., Matsuura, K., Tamada, Y., Sato, K., Miyabe, Y., Deyashiki, Y. and Ishida, N., 1996a,
Relationship of human liver dihydrodiol dehydrogenases to hepatic bile-acid-binding protein and
an oxidoreductase of human colon cells. Biochemical Journal, 313, 373±376.
Hara, A., Matsuura, K., Tamada, Y., Sato, K., Miyabe, Y., Deyashiki, Y. and Ishida, N., 1996b,
Relationship of human liver dihydrodiol dehydrogenases to hepatic bile-acid-binding protein and
an oxidoreductase of human colon cells. Biochemical Journal, 313, 373±376.
Hara, A., Taniguchi, H., Nakayama, T. and Sawada, H., 1990, Puri®cation and properties of multiple
forms of dihydrodiol dehydrogenase from human liver. Journal of Biochemistry, Tokyo, 108,
250±254.
Hecht, S. S., 1994, Metabolic activation and detoxi®cation of tobacco-speci®c nitrosamines ± a model
for cancer prevention strategies. Drug Metabolism Reviews, 26, 373±390.
Hecht, S. S., 1996, Recent studies on mechanisms of bioactivation and detoxi®cation of 4-
(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), a tobacco-speci®c lung carcinogen.
Critical Reviews in Toxicology, 26, 163±181.
Hecht, S. S., 1998, Biochemistry, biology, and carcinogenicity of tobacco-speci®c N-nitrosamines.
Chemical Research in Toxicology, 11, 559±603.
Hecht, S. S., Carmella, S. G., Murphy, S. E., Akerkar, S., Brunnemann, K. D. and Hoffmann, D.,
1993a, A tobacco-speci®c lung carcinogen in the urine of men exposed to cigarette smoke. New
England Journal of Medicine, 329, 1543±1546.
Hecht, S. S. and Hoffmann, D., 1988, Tobacco-speci®c nitrosamines, an important group of
carcinogens in tobacco and tobacco smoke. Carcinogenesis, 9, 875±884.
Hecht, S. S. and Hoffmann, D., 1989, The relevance of tobacco-speci®c nitrosamines to human cancer.
Cancer Surveys, 8, 273±294.
Hecht, S. S., Spratt, T. E. and Trushin, N., 1997, Absolute con®guration of 4-(methylnitrosamino)-
1-(3-pyridyl)-1-butanol formed metabolically from 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone. Carcinogenesis, 18, 1851±1854.
Hecht, S. S., Trushin, N., Reid Quinn, C. A., Burak, E. S., Jones, A. B., Southers, J. L., Gombar,
C. T., Carmella, S. G., Anderson, L. M. and Rice, J. M., 1993b, Metabolism of the tobacco-
speci®c nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in the patas monkey:
pharmacokinetics and characterization of glucuronide metabolites. Carcinogenesis, 14, 229±236.
Hecht, S. S., Young, R. and Chen, C. B., 1980, Metabolism in the F344 rat of 4-(N-methyl-N-
nitrosamino)-1-(3-pyridyl)-1-butanone, a tobacco-speci®c carcinogen. Cancer Research, 40,
4144±4150.
Inazu, N., Ruepp, B., Wirth, H. and Wermuth, B., 1992, Carbonyl reductase from human testis:
puri®cation and comparison with carbonyl reductase from human brain and rat testis. Biochimica
et Biophysica Acta, 1116, 50±56.
Jez, J. M., Flynn, T. G. and Penning, T. M., 1997, A nomenclature system for the aldo-keto reductase
superfamily. Advances in Experimental Medicine and Biology, 414, 579±589.
Jornvall, H., Persson, B., Krook, M., Atrian, S., Gonzalez-Duarte, R., Jeffery, J. and Ghosh, D.,
}
1995, Short-chain dehydrogenases reductases (SDR). Biochemistry, 34, 6003±6013.
Laemmli, U. K., 1970, Cleavage of structural proteins during the assembly of the head of bacteriophage
T4. Nature, 227, 680±685.
Maser, E., 1995, Xenobiotic carbonyl reduction and physiological steroid oxidoreductionÐthe
pluripotency of several hydroxysteroid dehydrogenases. Biochemical Pharmacology, 49, 421±440.
Maser, E., 1997, Stress, hormonal changes, alcohol, food constituents and drugs: factors that advance the
incidence of tobacco smoke-related cancer? Trends in Pharmacological Sciences, 18, 270±275.
Maser, E., 1998, 11b-Hydroxysteroid dehydrogenase responsible for carbonyl reduction of the tobacco-
speci®c nitrosamine4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in mouse lung microsomes.
Cancer Research, 58, 2996±3003.
Maser, E. and Bannenberg, G., 1994, 11b-Hydroxysteroid dehydrogenase mediates reductive
metabolism of xenobiotic carbonyl compounds. Biochemical Pharmacology, 47, 1805±1812.

Puri®cation of NNK oxidoreductases from human liver cytosol
769
Maser, E., Richter, E. and Friebertshauser, J., 1996, The identi®cation of 11 beta-hydroxysteroid
dehydrogenase as carbonyl reductase of the tobacco-speci®c nitrosamine 4-(methylnitrosamino)-
1-(3-pyridyl)-1-butanone. European Journal of Biochemistry, 238, 484±489.
Maser, E., Stinner, B. and Atalla, A., 2000, Carbonyl reduction of 4-(methylnitrosamino)-1-(3-
pyridyl)-1-butanone (NNK) by cytosolic enzymes in human liver and lung. Cancer Letters, 148,
135±144.
Matsuura, K., Deyashiki, Y., Sato, K., Ishida, N., Miwa, G. and Hara, A., 1997, Identi®cation of
amino acid residues responsible for diåerences in substrate speci®city and inhibitor sensitivity
between two human liver dihydrodiol dehydrogenase isoenzymes by site-directed mutagenesis.
Biochemical Journal, 323, 61±64.
Meger, M., Richter, E., Zwickenpflug, W., Oehlmann, C., Hargaden, M. B., A-Rahim, Y. I. and
Vesell, E. S., 1999, Metabolism and disposition of 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone (NNK) in rhesus monkeys. Drug Metabolism and Disposition, 27, 471±478.
Molowa, D. T., Shayne, A. G. and Guzelian, P. S., 1986, Puri®cation and characterization of
chlordecone reductase from human liver. Journal of Biological Chemistry, 261, 12624±12627.
Morse, M. A., Eklind, K. I., Toussaint, M., Amin, S. G. and Chung, F. L., 1990, Characterization of
a glucuronide metabolite of 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and its
dose-dependent excretion in the urine of mice and rats. Carcinogenesis, 11, 1819±1823.
Murphy, S. E., Heiblum, R. and Trushin, N., 1990a, Comparative metabolism of N«-nitro-
sonornicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone by cultured F344 rat oral
tissue and esophagus. Cancer Research, 50, 4685±4691.
Murphy, S. E., Palomino, A., Hecht, S. S. and Hoffmann, D., 1990b, Dose±response study of DNA
and hemoglobin adduct formation by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in F344
rats. Cancer Research, 50, 5446±5452.
Richter, E., Engl, J. and Tricker, A. R., 1998, Metabolism of 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone (NNK) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) in precision-cut
human and rodent liver and lung tissue slices. Toxicological Sciences, 42 (suppl. 1), 183.
Richter, E. and Tricker, A. R., 1994, Nicotine inhibits the metabolic activation of the tobacco-speci®c
nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in rats. Carcinogenesis, 15,
1061±1064.
Schrader, E., Hirsch-Ernst, K. I., Richter, E. and Foth, H., 1998, Metabolism of 4-
(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in isolated rat lung and liver. Naunyn-
Schmiedeberg’s Archives of Pharmacology, 357, 336±343.
Schulze, J., Schrader, E., Foth, H., Kahl, G. F. and Richter, E., 1998, Eåect of nicotine or cotinine
on metabolism of 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK) in isolated rat lung and
liver. Naunyn-Schmiedeberg’s Archives of Pharmacology, 357, 344±350.
Schulze, J., Schlager, W., Wunsch, R. and Richter, E., 1996, Metabolism of 4-(methylnitrosamino)-
1-(3-pyridyl)-1-butanone (NNK) by hamster, mouse and rat intestine: relevance of species
diåerences. Carcinogenesis, 17, 1093±1099.
Smith, T. J., Liao, A. M., Liu, Y., Butler Jones, A., Anderson, L. M. and Yang, C. S., 1997,
Enzymes involved in the bioactivation of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in
patas monkey lung and liver microsomes. Carcinogenesis, 18, 1577±1584.
Stolz, A., Hammond, L., Lou, H., Takikawa, H., Ronk, M. and Shively, J. E., 1993, cDNA cloning
and expression of the human hepatic bile acid-binding protein. A member of the monomeric
reductase gene family. Journal of Biological Chemistry, 268, 10448±10457.
Takikawa, H., Stolz, A., Sugiyama, Y., Yoshida, H., Yamanaka, M. and Kaplowitz, N., 1990,
Relationship between the newly identi®ed bile acid binder and bile acid oxidoreductases in human
liver. Journal of Biological Chemistry, 265, 2132±2136.
Upadhyaya, P., Kenney, P. M., Hochalter, J. B., Wang, M. and Hecht, S. S., 1999, Tumorigenicity
and metabolism of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol enantiomers and metabolites
}
in the A J mouse. Carcinogenesis, 20, 1577±1582.
Veech, R. L., Eggleston, L. V. and Krebs, H. A., 1969, The redox state of free nicotinamide-adenine
dinucleotide phosphate in the cytoplasm of rat liver. Biochemical Journal, 115, 609±619.
Vogel, K., Bentley, P., Platt, K. L. and Oesch, F., 1980, Rat liver cytoplasmic dihydrodiol
dehydrogenase. Puri®cation to apparent homogeneity and properties. Journal of Biological
Chemistry, 255, 9621±9635.
Wermuth, B., 1981, Puri®cation and properties of an NADPH-dependent carbonyl reductase from
human brain. Journal of Biological Chemistry, 256, 1206±1213.
Wermuth, B., 1985, Aldo-keto reductases. In T. G. Flynn and H. Weiner (eds), Enzymology of
}
Carbonyl Metabolism: Aldehyde Dehydrogenase, Aldo Keto Reductase, and Alcohol Dehydrogenase
(New York: Alan R. Liss), pp. 209±230.
Wermuth, B., Bohren, K. M., Heinemann, G., von Wartburg, J. P. and Gabbay, K. H., 1988,
Human carbonyl reductase. Nucleotide sequence analysis of a cDNA and amino acid sequence of
the encoded protein. Journal of Biological Chemistry, 263, 16185±16188.
Wermuth, B., Platt, K. L., Seidel, A. and Oesch, F., 1986, Carbonyl reductase provides the enzymatic
basis of quinone detoxication in man. Biochemical Pharmacology, 35, 1277±1282.