Structure-function modeling of the interactions of N-alkyl-N-hydroxyanilines with rat hepatic aryl sulfotransferase IV

Article abstract of DOI:10.1021/tx990184z

Although previous investigations have clearly shown that N-hydroxy arylamines and N-hydroxy heterocyclic amines are substrates for sulfotransferases, relatively little is known about which structural features of the N-hydroxy arylamines are important for sulfation to occur. The purpose of this investigation was to determine the extent to which secondary N-alkyl-N-hydroxy arylamines interact with aryl sulfotransferase (AST) IV (also known as tyrosine-ester sulfotransferase or ST1A1) and to evaluate these interactions using molecular modeling techniques. AST IV is a major cytosolic sulfotransferase in the rat, and it catalyzes the sulfation of various phenols, benzylic alcohols, arylhydroxamic acids, oximes, and primary N-hydroxy arylamines. In this study, three secondary N-hydroxy arylamines, N-hydroxy-N-methylaniline, N-ethyl-N-hydroxyaniline, and N-hydroxy-N-n-propylaniline, were found to be substrates for the purified rat hepatic AST IV. However, when the N-alkyl substituent was an n-butyl group (i.e., N-n-butyl-N-hydroxyaniline), the interaction with the enzyme changed from that of a substrate to competitive inhibition. This change in specificity was further explored through the construction and use of a model for AST IV based on mouse estrogen sulfotransferase, an enzyme whose crystal structure has been previously determined to high resolution. Molecular modeling techniques were used to dock each of the above N-hydroxy arylamines into the active site of the homology model of AST IV and determine optimum ligand geometries. The results of these experiments indicated that steric constraints on the orientation of binding of secondary N-alkyl-N-hydroxy arylamines at the active site of AST IV play a significant role in determining the nature of the interaction of the enzyme with these compounds.

Full text of DOI:10.1021/tx990184z

Chem. Res. Toxicol. 2000, 13, 1251-1258
1251
Str u ctu r e-F u n ction Mod elin g of th e In ter a ction s of
N-Alk yl-N-h yd r oxya n ilin es w ith Ra t Hep a tic Ar yl
Su lfotr a n sfer a se IV
Roberta S. King,† Vyas Sharma,‡ Lars C. Pedersen,§ Yoshimitsu Kakuta,§
Masahiko Negishi,§ and Michael W. Duffel*,‡
Division of Medicinal and Natural Products Chemistry, College of Pharmacy, The University of
Iowa, Iowa City, Iowa 52242, Department of Biomedical Sciences, College of Pharmacy, University
of Rhode Island, Kingston, Rhode Island 02881, and Pharmacogenetics Section, Laboratory of
Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences,
National Institutes of Health, Research Triangle Park, North Carolina 27709
Received November 3, 1999
Although previous investigations have clearly shown that N-hydroxy arylamines and
N-hydroxy heterocyclic amines are substrates for sulfotransferases, relatively little is known
about which structural features of the N-hydroxy arylamines are important for sulfation to
occur. The purpose of this investigation was to determine the extent to which secondary N-alkyl-
N-hydroxy arylamines interact with aryl sulfotransferase (AST) IV (also known as tyrosine-
ester sulfotransferase or ST1A1) and to evaluate these interactions using molecular modeling
techniques. AST IV is a major cytosolic sulfotransferase in the rat, and it catalyzes the sulfation
of various phenols, benzylic alcohols, arylhydroxamic acids, oximes, and primary N-hydroxy
arylamines. In this study, three secondary N-hydroxy arylamines, N-hydroxy-N-methylaniline,
N-ethyl-N-hydroxyaniline, and N-hydroxy-N-n-propylaniline, were found to be substrates for
the purified rat hepatic AST IV. However, when the N-alkyl substituent was an n-butyl group
(i.e., N-n-butyl-N-hydroxyaniline), the interaction with the enzyme changed from that of a
substrate to competitive inhibition. This change in specificity was further explored through
the construction and use of a model for AST IV based on mouse estrogen sulfotransferase, an
enzyme whose crystal structure has been previously determined to high resolution. Molecular
modeling techniques were used to dock each of the above N-hydroxy arylamines into the active
site of the homology model of AST IV and determine optimum ligand geometries. The results
of these experiments indicated that steric constraints on the orientation of binding of secondary
N-alkyl-N-hydroxy arylamines at the active site of AST IV play a significant role in determining
the nature of the interaction of the enzyme with these compounds.
ity of the sulfated metabolites. As in the sulfation of
primary N-hydroxy arylamines and N-hydroxy hetero-
cyclic amines, sulfation of N-alkyl-N-hydroxy arylamines
could constitute either detoxification or metabolic activa-
tion depending upon the properties of the product of the
reaction. In the case of N-OH-MAB, the sulfated metabo-
lite was more reactive than N-OH-MAB itself (2, 3).
In tr od u ction
N-Alkyl-N-hydroxy arylamines are most often encoun-
tered as intermediary metabolites formed through N-
oxidation of N-alkyl arylamines catalyzed by the flavin-
containing monooxygenases (1). Although relatively few
studies have been published on the sulfation of N-alkyl-
N-hydroxy arylamines, it has been reported that both
N-hydroxy-N-methyl-4-aminoazobenzene (N-OH-MAB)1
and N-ethyl-N-hydroxy-4-aminoazobenzene (N-OH-EAB)
were converted to reactive species by a sulfotransferase
enzyme in rat liver cytosol (2, 3). The limited number of
investigations on the sulfation of secondary N-hydroxy
arylamines may be due in part to the difficulty of
synthesis of these compounds and in part to the instabil-
Exposure to N-alkyl arylamines can occur either
directly from the environment or indirectly due to
intermediary metabolism of arylamines and nitroaro-
matic compounds. Primary, secondary, and tertiary aryl-
amines can be interconverted by enzymatic N-methyla-
tion and N-dealkylation reactions (4). Such N-methyla-
tion reactions are catalyzed by an S-adenosylmethionine-
dependent N-methyltransferase (4), while the N-dealkyla-
tion of secondary and tertiary arylamines may involve
either cytochrome P450 or flavin-containing monooxyge-
nases (5). The presence of these interconversion reactions
increases the possibility that an N-hydroxy arylamine can
be formed through metabolism of an arylamine.
* To whom correspondence should be addressed: Division of Me-
dicinal and Natural Products Chemistry, College of Pharmacy, The
University of Iowa, Iowa City, IA 52242. Telephone: (319) 335-8840.
Fax: (319) 335-8766. E-mail: michael-duffel@uiowa.edu.
† University of Rhode Island.
‡ The University of Iowa.
§ National Institutes of Health.
While previously published investigations clearly show
that N-hydroxy arylamines and N-hydroxy heterocyclic
amines are substrates for sulfotransferases (2, 6, 7), as
reviewed in ref 8, none of these studies have attempted
1 Abbreviations: AST, aryl sulfotransferase; N-OH-EAB, N-hydroxy-
N-ethyl-4-aminoazobenzene; EST, estrogen sulfotransferase; N-OH-
MAB, N-hydroxy-N-methyl-4-aminoazobenzene; PAPS, 3′-phospho-
adenosine 5′-phosphosulfate; PAP, adenosine 3′,5′-diphosphate.
10.1021/tx990184z CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/21/2000

1252 Chem. Res. Toxicol., Vol. 13, No. 12, 2000
King et al.
either with a Chromatotron apparatus (Harrison Research)
utilizing Silica gel 60 PF-254 (EM Science, Gibbstown, NJ ) or
with flash column chromatography using Davisil silica gel
(grade 633, 200-425 mesh, Aldrich Chemical Co., Milwaukee,
WI).
Rea gen ts a n d Ch em ica ls. 2-Naphthol, 1-naphthalenemeth-
anol, N-ethyl-N-methylaniline, N,N-diethylaniline, N,N-di-n-
butylaniline, ethyl iodide, 1-iodopropane, m-chloroperoxybenzoic
acid, sodium cyanoborohydride, iron(III)nitrate nonahydrate,
methanol-d4 (99.8 at. % D), and 4,7-diphenyl-1,10-phenanthro-
line were purchased from Aldrich Chemical Co. Aniline was
obtained from Mallinckrodt Inc. (Paris, KY). L-Ascorbic acid, zinc
metal dust, 37% aqueous formaldehyde, and sodium borohydride
were from Fisher Scientific (Pittsburgh, PA). 2-Mercaptoethanol
was obtained from Sigma Chemical Co. (St. Louis, MO). Amber
Reacti-Vials (total volume of 0.1 or 0.3 mL, fitted with Teflon/
silicone disks) were obtained from Pierce Chemical Co. (Rock-
ford, IL). N,N-Di-n-propylaniline was prepared by alkylation of
aniline with 1-iodopropane (10). All other assay reagents and
buffer components were from commercial sources. Ca u t ion :
N-Alkyl-N-hydroxy arylamines should be handled with ap-
propriate safety precautions to avoid exposure.
F igu r e 1. Summary of structures and nomenclature of the
N-alkyl-N-hydroxy arylamines that have been investigated.
N-Hydroxy-N-methylaniline (1), N-ethyl-N-hydroxyaniline (2),
N-hydroxy-N-n-propylaniline (3), and N-n-butyl-N-hydroxy-
aniline (4) are shown.
to determine which structural features of the N-hydroxy
arylamines may be important for sulfation. Furthermore,
only one report has addressed the sulfation of secondary
N-alkyl-N-hydroxy arylamines (2).
The purpose of this investigation was to explore the
potential for interactions of highly purified secondary
N-hydroxy arylamines with aryl sulfotransferase (AST)
IV2 through combination of kinetic studies on catalysis
and inhibition with homology modeling and ligand dock-
ing studies. AST IV is a major cytosolic sulfotransferase
in the livers of male rats and has been shown to be a
critical enzyme involved in 2-acetylaminofluorene-medi-
ated hepatocarcinogenesis (9). Recent studies also indi-
cate that the enzyme isolated from rat liver also catalyzes
sulfation of primary N-hydroxy arylamines (7). The
studies described in this paper begin to address the
hypothesis that structural features of secondary N-alkyl-
N-hydroxy arylamines can influence the type of interac-
tion that is observed. The N-alkyl-N-hydroxyaniline
derivatives that were chosen for this study are seen in
Figure 1. These compounds provide a systematic exten-
sion of the alkyl carbon chain on the N-hydroxy aryl-
amine for investigating the possibility for steric and/or
hydrophobic interactions at the active site of AST IV. The
docking of the secondary N-alkyl-N-hydroxy arylamines
into the active site of an AST IV homology model was
used to evaluate the role of the optimum ligand geometry
in the sulfation of these compounds catalyzed by AST IV.
Such studies with model N-hydroxy arylamines begin
to provide some of the fundamental structure-activity
relationships necessary for improving prediction of the
relative rates of formation of sulfuric acid esters from this
class of xenobiotics. These investigations on the specifi-
city of a major rat hepatic sulfotransferase are particu-
larly relevant in light of the extensive use of the rat in
studies on hepatocarcinogenesis of arylamines, N-hy-
droxy arylamines, and related compounds.
Syn th esis of N-Alk yl-N-h yd r oxya n ilin es. All of the N-
alkyl-N-hydroxyanilines shown in Figure 1 were synthesized by
Cope elimination (11) of the corresponding tertiary arylamine
N-oxides under argon at 12-15 mmHg. N-Methyl-N-hydroxy-
aniline (1) and N-ethyl-N-hydroxyaniline (2) were isolated
according to an extraction procedure described previously (12).
N-Alkyl-N-hydroxy arylamines 3 and 4 were obtained directly
from the condensate collected from the pyrolysis reaction. All
four N-alkyl-N-hydroxyanilines were isolated as their oxalic acid
salts (13) and were stable to storage under argon at -20 °C for
up to 1 month. Compounds 1-4 were characterized by 1H NMR
spectroscopy as well as by their ability to reduce iron(III) to iron-
(II). The high concentration of oxalate in the oxalate salts made
1H NMR the most desirable analytical method for characteriza-
tion. In addition, it was clear from UV and NMR spectra that
compounds 1-4 contained none of the potential oxidation
products or other reaction byproducts that can be derived from
N-hydroxy arylamines. The characterization of compounds 1-4
by 1H NMR is summarized as follows. N-Hydroxy-N-methyl-
aniline (1) oxalate (80 MHz, CD3OD): δ 3.05 (3H, d, J ) 3.75
Hz), 7.28 [5H, m, (d,d,s)]. N-Ethyl-N-hydroxyaniline (2) oxalate
(360 MHz, CD3OD): δ 1.23 (3H, t, J ) 7.08 Hz), 3.45 (2H, q, J
) 7.02 Hz), 7.10 (1H, t, J ) 7.2 Hz), 7.28 (2H, d, J ) 7.6 Hz),
7.34 (2H, d, J ) 7.1 Hz). N-Hydroxy-N-n-propylaniline (3)
oxalate (360 MHz, CD3OD): δ 1.27 (3H, t, J ) 7.45 Hz), 1.99
(2H, sextet, J ) 7.46 Hz), 3.64 (2H, t, J ) 7.46 Hz), 7.40 (1H, t,
J ) 7.13 Hz), 7.58 (2H, d, J ) 7.48 Hz), 7.63 (2H, d, J ) 7.14
Hz). N-n-Butyl-N-hydroxyaniline (4) oxalate (360 MHz, CD3-
OD): δ 1.17 (3H, t, J ) 7.04 Hz), 1.29 (2H, m), 3.60 (4H, m),
7.14 (1H, t, J ) 7.4 Hz), 7.12 (2H, d, J ) 7.2 Hz), 7.15 (2H, d,
J ) 7.5 Hz).
Syn th esis of Ter tia r y Am in e N-Oxid es. The tertiary
arylamine N-oxides utilized for the above preparations of
secondary N-hydroxy arylamines were synthesized from the
corresponding tertiary arylamines by oxidation with m-chlo-
roperoxybenzoic acid in chloroform (14, 15). These reactions
were complete in 1-3 h at 4 °C, and the isolated yields were
60-88%. N-Ethyl-N-methylaniline N-oxide and N,N-diethyl-
aniline N-oxide were extracted from the reaction mixture
according to the procedure described in ref 15, thus forming the
hydrochloride salts. Since N,N-di-n-propylaniline N-oxide and
N,N-di-n-butylaniline N-oxide did not partition into aqueous
hydrochloric acid, the N-oxide products were isolated after
repetitive washing with aqueous 1.0 M sodium bicarbonate to
remove the benzoic acid coproduct. The chloroform layer was
dried with anhydrous magnesium sulfate and filtered and the
solvent removed by rotary evaporation. In the case of N,N-di-
n-propylaniline N-oxide, the product was then isolated as the
hydrochloride by adding ethereal HCl to an ether solution of
the N-oxide. N,N-Di-n-butylaniline N-oxide was used as the
Exp er im en ta l P r oced u r es
Gen er a l. NMR spectra were obtained on a Bruker WM-360
spectrometer with tetramethylsilane as the internal standard.
HPLC analyses were carried out with an Econosphere C18
column (5 µm, 4.6 mm × 250 mm) obtained from Alltech
Associates (Deerfield, IL). Analyses by thin-layer chromatog-
raphy (TLC) were conducted on silica gel (Chromagram sheets,
2 mm, with a fluorescence indicator, Eastman Kodak Co.,
Rochester, NY). Preparative chromatography was conducted
2
Although the Human Genome Organization has adopted nomen-
clature for the human sulfotransferases, a systematic extension of that
nomenclature to sulfotransferases in the rat as well as other species
has not yet been accomplished.

Sulfation of N-Alkyl-N-hydroxy Arylamines
Chem. Res. Toxicol., Vol. 13, No. 12, 2000 1253
crystalline solid isolated directly after removal of the chloroform
by rotary evaporation. The characterization of the N-oxide
precursors of compounds 1-4 by 1H NMR is summarized as
follows. N-Ethyl-N-methylaniline N-oxide HCl (80 MHz, CD3-
OD): δ 1.25 (3H, t, J ) 7.1 Hz), 3.07 (3H, s), 4.21 (2H, q, J )
7.1 Hz), 4.95 (1H, broad s), 7.69 (3H, m), 7.95 (2H, d, J ) 7.41
Hz). N,N-Diethylaniline N-oxide HCl (360 MHz, CD3OD): δ 1.20
(6H, t, J ) 7.08 Hz), 4.15 (2H, sextet, J ) 6.93 Hz), 4.27 (2H,
sextet, J ) 6.90 Hz), 4.84 (1H, s), 7.68 (3H, m), 7.83 (2H, d, J )
7.53). N,N-Di-n-propylaniline N-oxide HCl (360 MHz, CD3OD):
δ 0.94 (6H, t, J ) 4.46 Hz), 1.35 (2H, m), 1.82 (2H, m), 4.05
(2H, m), 4.19 (2H, m), 7.65 (3H, m), 7.86 (2H, d, J ) 7.47 Hz).
N,N-Di-n-butylaniline N-oxide (360 MHz, CD3OD): δ 0.89 (6H,
t, J ) 7.23 Hz), 1.31 (6H, m), 1.80 (2H, m), 4.09 (2H, m), 4.24
(2H, m), 7.65 (3H, m), 7.88 (2H, d, J ) 7.51 Hz).
Qu a n tita tion of N-Alk yl-N-h yd r oxya n ilin es by Ir on (III)-
Red u cin g Equ iva len ts. Before use in the kinetic assays
described below, the concentrations of N-alkyl-N-hydroxyaniline
derivatives in solution were determined by a colorimetric assay
that was previously developed for quantitation of p-chloro-N-
methyl-N-hydroxyaniline (5). The following reagents were mixed
in a glass test tube: 600 µL of 4,7-diphenyl-1,10-phenanthroline
(0.5 mg/mL in ethanol), 100 µL of 1.0 M sodium acetate (pH
5.6), 20 µL of 0.1 M iron(III) nitrate, a N-hydroxyarylamine
solution, and water to make a final volume of 800 µL. The
reaction was started by addition of either N-hydroxy arylamine
or iron reagent, and after 1 min, the reaction was stopped by
adding 25 µL of 0.02 M phosphoric acid. The phosphate ions
stopped the reaction by complexing with the unreacted ferric
ions (16), thereby preventing further color formation. After the
reaction was terminated, the concentration of the N-hydroxy
arylamine in the sample was determined from the absorbance
at 535 nm after subtraction of the absorbance for a blank sample
containing all reagents except the N-hydroxy arylamine. An
extinction coefficient of 19 250 M-1 cm-1 per reducing equivalent
(5) was used, and calculations of concentrations of N-hydroxy
arylamines were based on two reducing equivalents per mol-
ecule of N-hydroxy arylamine.
M potassium phosphate at pH 7.0. Following incubation for 2
min at 37 °C for temperature equilibration, the reaction was
initiated by addition of enzyme and conducted for 10 min at 37
°C. Because the sulfuric acid esters of the N-hydroxy arylamines
were expected to be unstable (2, 3), an assay that measures the
level of formation of the coproduct, PAP, was utilized for
determination of the progress of the AST IV-catalyzed reaction
(26). Control assays were carried out with all assay components
except the N-hydroxy arylamine to ensure that only N-hydroxy
arylamine-dependent formation of PAP was assessed. PAPS was
synthesized and purified as described previously (7).
Ca lcu la tion of Kin etic Con sta n ts. Apparent Km and Vmax
values were calculated from a nonlinear least-squares fit of the
kinetic data to the Michaelis-Menten equation (27). At least
four substrate concentrations with two or three repetitions at
each concentration were utilized over a range of substrate
concentrations 2-3-fold above and below the Km. The Ki value
for N-n-butyl-N-hydroxyaniline was calculated by a nonlinear
least-squares fit of the data to equations modeling reversible
inhibition (27); the best fit of the data was to the equation for
competitive inhibition.
Con str u ction of th e AST IV Hom ology Mod el. The
structure of AST IV was modeled using the structure of mouse
estrogen sulfotransferase (EST) (28) as a template by fitting the
AST IV sequence into the electron density map of EST. This
AST IV model was refined using XPLOR 3.85 (29) by an initial
positional refinement followed by simulated annealing using the
structure factors originally obtained from the X-ray diffraction
data collected on EST. The homology model also contained the
coordinates of PAP and estradiol. Subsequent structural ma-
nipulations on the homology model were performed using
molecular modeling package SYBYL 6.4 (Tripos Associates, St.
Louis, MO). The structure of a proposed mimic of the transition
state intermediate of the sulfotransferase reaction, PAP-
vanadate (30), was extracted from PDB file 1BO6 containing
coordinates of the EST-PAP-vanadate complex, and the PAP-
vanadate structure was modified by replacing the vanadate with
a sulfuryl group. This resulted in a PAPS molecule that retained
the geometry of the adenosine 3′,5′-diphosphate present in the
homology model derived from the EST-PAP complex. The PAPS
molecule was then fit into the position occupied by PAP in the
active site of the AST IV homology model using the Multifit
option in SYBYL 6.4. The enzyme was given Kollman all-atom
charges, and Gasteiger-Hu¨ckel charges were used for the
ligands. The AST IV homology model with PAPS in its active
site was used as the receptor in the docking studies.
Dock in g of Liga n d s. The automated docking package
Flexidock, a module of SYBYL 6.4, was used in the docking
studies. Individual N-hydroxy arylamine ligands were con-
structed, given Gasteiger-Hu¨ckel charges, and optimized by
energy minimization using the Powel algorithm. These ligands
were fit into the position occupied by estradiol in the AST IV
homology model using the Multifit option in SYBYL 6.4. The
oxygen atom and the aromatic carbon atoms of the ligand were
overlapped on the 3R oxygen and aromatic ring of estradiol,
respectively. The resultant ligand structure was used as the
initial conformation for the docking experiments. These experi-
ments provided a set of solutions that represented the optimum
ligand geometry in the active site of the homology model.
Representations of these optimum ligand geometries in Figures
3 and 4 were obtained through the use of MOLSCRIPT (31) and
RASTER3D (32).
P u r ifica tion of Hep a tic Ar yl Su lfotr a n sfer a se IV. AST
IV was purified to apparent homogeneity from the livers of male
Sprague-Dawley rats (300-350 g) by a published procedure (17).
The activity of the enzyme at pH 5.5 with 2-naphthol as a
substrate (18) was used to follow the purification, and the
specific activity of the homogeneous enzyme was 800-1000 nmol
of 2-naphthyl sulfate formed min-1 (mg of protein)-1. The
homogeneity of the enzyme was initially assessed with SDS-
PAGE (19), whereby the preparation exhibited one band of
protein with an Mr of 33 500 (Coomassie blue staining). How-
ever, SDS-PAGE by itself will not clearly distinguish AST IV
from two other sulfotransferases that have been previously
characterized with the arylhydroxamic acid N-hydroxy-2-acetyl-
aminofluorene and named N-hydroxy arylamine sulfotrans-
ferases (20-23). Despite significant differences in amino acid
sequences and the masses of subunits, these enzymes are often
not well separated on SDS-PAGE (20, 21, 23). Therefore, we
confirmed the identity of the enzyme in these studies by using
a previously described procedure (7) to compare HPLC tryptic
peptide maps of the highly purified native AST IV with a
recombinant form of the enzyme (24). The peptide maps (data
not shown) conclusively indicated that the enzyme isolated from
liver was AST IV. Protein concentrations were determined (25)
with bovine serum albumin as the standard, and typical
purifications from the liver tissue of four rats yielded 1-1.5 mg
of homogeneous AST IV.
Resu lts
Kin etic Assa y P r oced u r e. Enzyme assays were conducted
essentially as described previously for primary N-hydroxy
arylamines (7). This included bubbling each component of the
reaction mixture with argon and conducting the enzymatic
reactions in sealed, argon-filled, amber-colored reaction vials.
Reaction mixtures with a total volume of 0.03 mL contained 0.2
mM PAPS, 8.3 mM 2-mercaptoethanol, 1-2 µg of AST IV,
varying concentrations of the N-hydroxy arylamine, and 0.25
Sta bility of N-Alk yl-N-h yd r oxya n ilin es u n d er As-
sa y Con d ition s. We have previously reported (7) that
primary N-hydroxy arylamines are susceptible to oxida-
tive degradation during sulfotransferase assays unless
steps are taken to remove dissolved oxygen from the
reaction solution. Therefore, we examined the stability
of a representative N-alkyl-N-hydroxyaniline under the

1254 Chem. Res. Toxicol., Vol. 13, No. 12, 2000
King et al.
Ta ble 1. Sp ecificity of Ar yl Su lfotr a n sfer a se IV for
N-Alk yl-N-h yd r oxya n ilin esa
N-alkyl-N-
Km(app) Vmax [nmol min-1
kcat/Km
Ki
(µM)
hydroxyaniline (µM) (mg of protein)-1] (mM-1 min-1
)
1
2
3
4
119 ( 15
650 ( 75
161 ( 35
-
85 ( 5
273 ( 12
98 ( 6
-
24.2 ( 3.4
14.2 ( 1.8
20.6 ( 4.7
-
-
-
-
66 ( 11
a
The structures of the compounds examined are shown in
Figure 1. Kinetic constants were determined in argon-saturated
reaction mixtures at pH 7.0 in 0.25 M potassium phosphate with
varying concentrations of the N-alkyl-N-hydroxyaniline, 0.2 mM
PAPS, 8.3 mM 2-mercaptoethanol, and 1.5 µg of AST IV in a total
volume of 30 µL. Reaction mixtures were incubated at 37 °C for
10 min and analyzed by HPLC as indicated in Experimental
Procedures.
conditions used for enzymatic assays. Accordingly, ul-
traviolet spectroscopy (data not shown) revealed that
N-n-butyl-N-hydroxyaniline was chemically stable under
the conditions of argon saturation used in the assays for
sulfotransferase activity with these compounds. There-
fore, the anaerobic conditions previously developed for
primary N-hydroxy arylamines were suitable for main-
tenance of N-alkyl-N-hydroxyanilines during enzymatic
assays with AST IV.
In ter a ction of N-Alk yl-N-h yd r oxya n ilin es w ith
AST IV. Each N-alkyl-N-hydroxyaniline was examined
for its ability to serve as either a substrate for or an
inhibitor of aryl sulfotransferase IV. N-Hydroxy-N-me-
thylaniline (1), N-ethyl-N-hydroxyaniline (2), and N-
hydroxy-N-n-propylaniline (3) were found to be sub-
strates for the purified enzyme (Table 1 and Figure 2).
The catalytic efficiency of the enzyme with these sub-
strates, as judged by the values of kcat/Km, varied by less
than 2-fold for the compounds with N-methyl, N-ethyl,
and N-propyl substituents on N-hydroxyaniline. How-
ever, a dramatic change occurred upon lengthening the
N-alkyl substituent from N-propyl to N-butyl. N-n-Butyl-
N-hydroxyaniline (4) was not a substrate for aryl sul-
fotransferase IV, but it was instead a reversible com-
petitive inhibitor of the AST IV-catalyzed sulfation of
1-naphthalenemethanol.
F igu r e 2. Initial velocity data for AST IV-catalyzed sulfation
of N-hydroxy-N-methylaniline, N-ethyl-N-hydroxyaniline, and
N-hydroxy-N-n-propylaniline. Assays were carried out as de-
scribed in Table 1. Data points are observed values, and the
lines represent the results of nonlinear curve fitting as described
in Experimental Procedures.
Assessm en t of th e P oten tia l for Ir r ever sible In -
h ib it ion of AST IV b y a Secon d a r y N-Hyd r oxy
Ar yla m in e. Previous reports have indicated a time-
dependent irreversible inactivation of AST IV with N-hy-
droxy-2-acetylaminofluorene (33) and N-hydroxyaniline
(7). Therefore, we examined the potential for one of the
secondary N-hydroxy arylamines to irreversibly inhibit
the enzyme. AST IV (10 µg) was incubated at 37 °C in
0.25 M potassium phosphate at pH 7.0 either without or
with 1 mM N-ethyl-N-hydroxyaniline and 0.2 mM PAPS
for up to 21 min. At 7 min intervals, aliquots were diluted
into a standard assay for sulfation of 2-naphthol at pH
5.5 (34). The specific activities of the aliquots of the
enzyme remained within 10% of the control value through-
out the preincubation, regardless of the presence or
absence of PAPS (data not shown). This result indicated
that neither N-ethyl-N-hydroxyaniline nor the sulfuric
acid ester of N-ethyl-N-hydroxyaniline covalently inac-
tivated the purified protein. The sulfuric acid ester of
N-ethyl-N-hydroxyaniline failed to act as an irreversible
inhibitor of the enzyme, yet it was nevertheless an
unstable reaction product. Attempts to isolate and/or
quantitate this product only led to mixtures of degrada-
tion products. Therefore, although the reaction product
was unstable in our attempts at its isolation, it did not
react with the AST IV in such a way as to impair the
catalytic function of the enzyme.
Mod elin g AST IV by Hom ology to Mou se Estr o-
gen Su lfotr a n sfer a se (EST). The dramatic change in
the interaction of N-n-alkyl-N-hydroxyarylamines with
AST IV as the n-alkyl group was extended could not be
explained on the basis of changes in hydrophobic char-
acteristics, and it suggested that steric and/or electro-
static interactions at the active site of the enzyme were
critical. In the absence of a crystal structure for AST IV,
we developed a model for the enzyme based upon its
strong homology to EST. As described in detail in
Experimental Procedures, the homology model was pro-
duced by manual alignment of the amino acid sequence
of AST IV with the electron density map of EST-PAP,
followed by structural refinement and energy minimiza-
tion of the model. The structure of a proposed transition
state mimic, an EST-PAP-vanadate complex, was used
as a template for construction of PAPS and its placement
at the active site of the AST IV homology model.

Sulfation of N-Alkyl-N-hydroxy Arylamines
Chem. Res. Toxicol., Vol. 13, No. 12, 2000 1255
F igu r e 3. Active site region of a model of AST IV constructed by homology to EST. (A) The homology model of AST with estradiol
coordinates removed and a sulfuryl group replacing the PAP-vanadate complex as described in Experimental Procedures. (B) Starting
conformation of N-methyl-N-hydroxyaniline for energy minimization with Flexidock. The oxygen atom and aromatic carbons of
N-methyl-N-hydroxyaniline were aligned with coordinates corresponding to the positions of the 3R oxygen and aromatic carbons of
estradiol in the crystal structure of the EST-PAP complex.
F igu r e 4. Optimum geometries of N-alkyl-N-hydroxyanilines calculated for ligand docking within the active site of the homology
model for AST IV. (A) N-Hydroxy-N-methylaniline, N-ethyl-N-hydroxyaniline, and N-hydroxy-N-n-propylaniline are shown. (B) N-n-
Butyl-N-hydroxyaniline is shown.
Likewise, the crystal structure of the EST-PAP-vana-
date complex and an EST-PAP-estradiol complex were
utilized to assign a starting orientation of the oxygen
atom and the aromatic ring of each N-hydroxy arylamine
for ligand docking computations. These calculations,
carried out with Flexidock as described below, were then
used to determine an optimum ligand geometry for the
N-hydroxy arylamine in relation to the amino acid
residues at the active site of the homology model of AST
IV. A representation of the active site region of the model
of AST IV based on the crystal structure of EST is seen
in Figure 3.
Or ien ta tion of Dock ed Liga n d s in th e Active Site
Mod el. The optimum geometry for the binding of each
of the four N-alkyl-N-hydroxy arylamines to the homol-
ogy model of AST IV was determined. The algorithm that
was used for this optimization, Flexidock, took into
account van der Waals, electrostatic, and torsional energy
terms to calculate a single energy term describing the
ligand-receptor interaction. The bond lengths and bond
angles were kept constant during the docking procedure.
As seen in Figure 4A, the orientations of the N-hy-
droxyarylamine molecules that served as substrates were
similar. Namely, the alkyl substituent groups occupied
a pocket formed by Pro 43, His 104, Phe 138, and Tyr
165 in the homology model. However, in the case of N-n-
butyl-N-hydroxyaniline (Figure 4B), this pocket was
unable to accommodate the n-butyl substituent, and the
orientation of the inhibitor changed so that the aromatic
ring of the inhibitor was accommodated in this region
rather than the N-alkyl substituent. As noted below, one
result of this dramatic change in orientation was to place
the oxygen of the N-OH group a much greater distance
from both the sulfuryl group and the critical His 104.

1256 Chem. Res. Toxicol., Vol. 13, No. 12, 2000
King et al.
Ta ble 2. Ca lcu la ted En er gies a n d Bon d Dista n ces for
N-Alk yl-N-h yd r oxya n ilin es Dock ed in to th e Active Site
Mod el for AST IV
might expect a consistent increase in the catalytic ef-
ficiency of the enzyme as the alkyl chain length on the
N-hydroxy nitrogen was increased. However, this was not
observed. The N-methyl, N-ethyl, and N-n-propyl deriva-
tives of N-hydroxyaniline (compounds 1-3 in Table 1)
were substrates for AST IV, while the N-butyl derivative
was found to be a reversible competitive inhibitor of the
purified enzyme. It is also noteworthy that previously
published kinetic constants (7) for N-hydroxyaniline [i.e.,
Km ) 230 ( 18 µM, Vmax ) 43 ( 1 nmol min-1 (mg of
protein)-1, and kcat/Km ) 6.3 mM-1 min-1] are also
consistent with the results in Table 1. That is, AST IV
can catalyze sulfation of an unsubstituted N-hydroxy
arylamine as well as an N-hydroxy arylamine bearing a
methyl, ethyl, or n-propyl substituent on the nitrogen
with relatively small effects on the overall catalytic
efficiency of the reaction. The similarities in catalytic
efficiencies observed for the enzyme with these substrates
may in part be due to the ability of the active site to
accommodate multiple binding orientations of these
N-hydroxy arylamines, of which some orientations are
not catalytically productive. Nevertheless, for those N-
hydroxy arylamines that are substrates, these noncata-
lytic orientations are in equilibrium with an orientation
that is suitable for sulfuryl transfer.
While compounds 1-3 are capable of binding to the
active site of AST IV in an orientation that is appropriate
for catalysis, the n-butyl substituent seen in compound
4 imparts a specificity of binding that precludes any
orientation of the molecule at the active site that is
favorable for sulfuryl transfer to the oxygen of the
N-hydroxy arylamine. The fact that the Ki of 4 (66 µM)
is 2-10-fold lower than the Km’s of 1-3 (119-650 µM)
suggests that the steric effects primarily involve orienta-
tion of the molecule in relation to catalysis rather than
simply the hydrophobic characteristics of substrate bind-
ing. Thus, the lack of a strong trend in kinetic parameters
among compounds 1-3 as substrates, as well as the
observation that the n-butyl derivative became an inhibi-
tor of the enzyme, points to a complex relationship
between steric and hydrophobic factors in this class of
molecules.
binding
energy
terma
distance from
the oxygen of
distance from
the oxygen of
N-alkyl-N-
the ligand to the
the ligand to the
hydroxyaniline (kcal/mol) sulfur of PAPS (Å) nitrogen of His 104 (Å)
1
2
-1636.7
-1636.3
-1441.0
-1434.0
-1237.0
-944.0
1.886
1.886
2.142
2.021
2.095
2.710
3.204
3.205
1.911
2.414
3.282
5.089
3
4
a
All binding energies and bond distances for binding of N-alkyl-
N-hydroxyanilines to the homology model for AST IV were
calculated as described in Experimental Procedures. Two very
similar docking positions were found for compounds 1 and 2, while
single solutions to the docking procedure were obtained for
compounds 3 and 4.
The calculated values of the binding energy for each
of the Flexidock solutions and of the corresponding
distances between the oxygen atom of the docked ligand
and both the sulfur atom of PAPS and the nitrogen atom
of His 104 are shown in Table 2. The binding energy
terms calculated for compounds 1-3 were lower in energy
(-1447 ( 240 kcal/mol) than that calculated for N-n-
butyl-N-hydroxyaniline (-944 kcal/mol). Although this
represents a difference between substrates 1-3 and the
inhibitor, 4, the distances from the oxygen atom of the
N-hydroxy arylamine to the sulfur atom of PAPS and the
closest nitrogen of the key histidine residue are also
critical parameters for use of the model in predicting
catalytic function. The oxygen-sulfur distance is 2.006
( 0.118 Å for the calculated geometries of N-methyl-,
N-ethyl-, and N-n-propyl-N-hydroxyaniline (compounds
1-3). In contrast, the distance from the oxygen of N-n-
butyl-N-hydroxyaniline to the sulfur of PAPS (2.710 Å)
is significantly greater than that observed for the three
substrates. Likewise, the distance between the nitrogen
of His 104 and the oxygen of the N-alkyl-N-hydroxy-
aniline is much greater for the inhibitor (5.089 Å) than
for substrates 1-3 (2.803 ( 0.612 Å).
Discu ssion
The increased hydrophobicity of 4 relative to those of
1-3 most likely enhances the ability of the xenobiotic to
bind to the active site of AST IV, while steric effects of
the longer alkyl substituent alter the orientation of the
binding so that catalysis does not occur. Molecular
docking studies also support this hypothesis. It is seen
that in the case of AST IV substrates 1-3, the alkyl chain
on the N-hydroxy nitrogen of the docked ligands is
accommodated in a pocket of four amino acids comprising
Pro 43, His 104, Phe 138, and Tyr 165. Due to the steric
constraints on this pocket, it is unable to accommodate
the n-butyl chain of the N-n-butyl-N-hydroxyaniline.
On the basis of studies performed on mouse estrogen
sulfotransferase, it has been suggested that after the
proton on the oxygen atom of a substrate has been
abstracted by catalytic His 108, there is an in-line attack
by this oxygen on the sulfur atom of PAPS (30). The
residue corresponding to His 108 of mouse estrogen
sulfotransferase is His 104 in AST IV, and the sequence
of catalytic events outlined for mouse estrogen sul-
fotransferase based on its crystal structure (30) is
consistent with the kinetic mechanism proposed earlier
for AST IV based on initial velocity studies in solution
(35). From the crystal structure of a PAP-vanadate-
Drugs, carcinogens, and other xenobiotics containing
various organic functional groups are known to interact
with aryl (phenol) sulfotransferases such as AST IV, yet
within each class of similar substrates or inhibitors,
enzyme activity is affected to varying degrees by elec-
tronic, steric, and lipophilic characteristics (35-39). For
example, the sulfation of phenols and aryl oximes cata-
lyzed by AST IV is affected by the electronic contribution
of aromatic ring substituents; however, the magnitude
and direction of this effect are not consistent between the
two different structural classes (35, 38). The AST IV-
catalyzed sulfation of benzylic alcohols is affected by the
lipophilicity of the substrate (37) and by steric effects
related to the configurations of chiral benzylic alcohols
(39). In addition, aldehydes and carboxylic acids corre-
sponding to the higher oxidation states of benzylic
alcohols act as inhibitors of the enzyme (39).
In the studies presented here, AST IV was found to
catalyze the sulfation of secondary N-hydroxy aryl-
amines. N-Alkyl-N-hydroxy arylamines 1-4 were chosen
for investigation of the possibility of steric and hydro-
phobic interactions at the active site of AST IV. On the
basis of previous work with other substrate classes, one

Sulfation of N-Alkyl-N-hydroxy Arylamines
Chem. Res. Toxicol., Vol. 13, No. 12, 2000 1257
estrogen sulfotransferase complex, the distance from the
position of the oxygen of estradiol to the vanadium atom
in the PAP-vanadate complex (analogous to the sulfur
atom in PAPS) is 2.3 Å, and the distance from the oxygen
to the nitrogen of His 108 is 2.9 Å (30). In the AST IV
homology model (Table 2), the computed distances be-
tween the oxygen atom of the docked ligand and both the
sulfur atom of PAPS (1.9-2.1 Å) and the nitrogen atom
of His 104 (1.9-3.3 Å) are consistent with a catalytic role
for His 104, and they also suggest distinctions between
favorable and unfavorable binding orientations for ca-
talysis, with a calculated O-S distance of 2.7 Å and an
O-N distance of 5.1 Å for the inhibitor N-n-butyl-N-
hydroxyaniline (Table 2). According to this model, the
change in the docked ligand geometry of N-n-butyl-N-
hydroxyaniline results in an unfavorable orientation for
catalysis. Therefore, although N-n-butyl-N-hydroxy-
aniline binds well to the active site of AST IV as seen by
the Ki value, its inability to accept a sulfuryl group from
PAPS is predicted by the homology model and docking
experiments.
of aryl sulfotransferases for this class of xenobiotic
metabolites.
Ack n ow led gm en t. This investigation was supported
in part by U.S. Public Health Service Grant CA38683
awarded by the National Cancer Institute, Department
of Health and Human Services, and by a predoctoral
fellowship to R.S.K. through a National Institutes of
Health Training Grant in the Pharmacological Sciences
(GM07069).
Refer en ces
(1) Ziegler, D. M. (1984) in Drug Metabolism and Drug Toxicity
(Mitchell, J . R., and Horning, M. G., Eds.) pp 33-53, Raven Press,
New York.
(2) Kadlubar, F. F., Miller, J . A., and Miller, E. C. (1976) Hepatic
metabolism of N-hydroxy-N-methyl-4-aminoazobenzene and other
N-hydroxy arylamines to reactive sulfuric acid esters. Cancer Res.
36, 2350-2359.
(3) Kadlubar, F. F., Ketterer, B., Flammang, T. J ., and Christodoul-
ides, L. (1980) Formation of 3-glutathione-5-yl-N-methyl-4-ami-
noazobenzene and inhibition of aminoazo dye-nucleic acid binding
in vitro by reaction of glutathione with metabolically-generated
N-methyl-4-aminoazobenzene-N-sulfate. Chem.-Biol. Interact. 31,
265-278.
Our experimental data on sulfation of secondary N-
hydroxyanilines are explained well by a model of the
active site of AST IV based upon homology to the crystal
structure of mouse estrogen sulfotransferase. The se-
quences of the two enzymes are 50% identical, and as
noted above, there is evidence suggesting that the
catalytic mechanisms of sulfuryl transfer are essentially
the same in the mouse estrogen sulfotransferase and the
rat hepatic AST IV. â-Estradiol serves as a substrate for
both enzymes, although the apparent Km with mouse
estrogen sulfotransferase (40) is approximately 100-fold
less than the apparent Km obtained with AST IV under
similar reaction conditions (J . Sheng and M. Duffel,
unpublished results); maximal velocities for sulfation of
â-estradiol catalyzed by the two enzymes are comparable
under the same conditions of buffer, pH, and tempera-
ture. These similarities provide a degree of confidence
that the homology model is relevant to catalytic function.
Nonetheless, there are differences between the two
enzymes that warrant further investigation with respect
to their potential role in catalytic function. For example,
the mouse estrogen sulfotransferase exists as a monomer
in solution, while AST IV is a dimer. The potential for
this structural difference to be manifested in alterations
in catalytic specificity remains to be determined. Thus,
even though this homology model predicts the complex
behavior of the secondary N-hydroxy arylamines pre-
sented here, it may undergo further refinement as other
substrate classes and structural aspects of the enzyme
are explored. Such investigations are in progress.
In conclusion, our finding that AST IV is able to
catalyze the sulfation of N-alkyl-N-hydroxy arylamines
contributes to a more complete understanding of the
metabolic pathways available to these intermediary
metabolites. Moreover, these results provide an indica-
tion that steric factors relating to the substituent on the
N-hydroxy nitrogen can dramatically influence the na-
ture of the interaction with the aryl sulfotransferase.
Although these observations provide the first quantita-
tive exploration of the role of these structural character-
istics in the sulfation of secondary N-hydroxy arylamines,
further studies with larger sets of compounds and other
hepatic sulfotransferases will undoubtedly be necessary
to provide a more complete understanding of the catalytic
process and more detailed predictions of the specificity
(4) Ziegler, D. M., Ansher, S. S., Nagata, T., Kadlubar, F. F., and
J akoby, W. B. (1988) N-Methylation: Potential mechanism for
metabolic activation of carcinogenic primary arylamines. Proc.
Natl. Acad. Sci. U.S.A. 85, 2514-2517.
(5) Prough, R. A., and Ziegler, D. M. (1977) The relative participation
of liver microsomal amine oxidase and cytochrome P-450 in
N-demethylation reactions. Arch. Biochem. Biophys. 180, 363-
373.
(6) Duffel, M. W., Modi, R. B., and King, R. (1992) Interactions of a
Primary N-Hydroxy Arylamine with Rat Hepatic Aryl Sulfotrans-
ferase IV. Drug Metab. Dispos. 20, 339-342.
(7) King, R. S., and Duffel, M. W. (1997) Oxidation-dependent
inactivation of aryl sulfotransferase IV by primary N-hydroxy
arylamines during in vitro assays. Carcinogenesis 18, 843-849.
(8) Glatt, H. (1997) Sulfation and sulfotransferases 4: bioactivation
of mutagens via sulfation. FASEB J . 11, 314-321.
(9) Ringer, D. P., Norton, T. R., and Howell, B. A. (1990) 2-Acetyl-
aminofluorene-mediated alteration in the level of liver aryl
sulfotransferase IV during rat hepatocarcinogenesis. Cancer Res.
50, 5301-5307.
(10) Zander, A. (1883) Ueber die specifischen volumina einiger allyl-,
propyl- und verwandter verbindungen. Liebigs Ann. Chem. 214,
138-171.
(11) Cope, A. C., LeBel, N. A., Lee, H.-H., and Moore, W. R. (1957)
Amine oxides. III. Selective formation of olefins from unsym-
metrical amine oxides and quaternary ammonium hydroxides. J .
Am. Chem. Soc. 79, 4720-4729.
(12) Kadlubar, F. F., McKee, E. M., and Ziegler, D. M. (1973) Reduced
pyridine nucleotide-dependent N-hydroxy amine oxidase and
reductase activities of hepatic microsomes. Arch. Biochem. Bio-
phys. 156, 46-57.
(13) Rogers, M. A. (1955) Aliphatic hydroxylamines. Part I. Prepara-
tion. J . Chem. Soc., 769-772.
(14) Craig, J . C., and Purushothaman, K. K. (1970) An improved
preparation of teriary amine N-oxides. J . Org. Chem. 35, 1721-
1722.
(15) Muntz, R. L., Pirkle, W. H., and Paul, I. C. (1972) Determination
of absolute configuration of (-)-N-ethyl-N-methylaniline oxide by
X-ray analysis: the molecular geometry of an N-oxide salt. J .
Chem. Soc., Perkin Trans. 2, 483-488.
(16) Tsen, C. C. (1961) An improved spectrophotometric method for
the detection of tocopherols using 4,7-diphenyl-1,10-phenanthro-
line. Anal. Chem. 33, 849-851.
(17) Duffel, M. W., Binder, T. P., Hosie, L., Baden, H., Sanders, J . A.,
Knapp, S. A., and Baron, J . (1991) Purification, immunochemical
characterization, and immunohistochemical localization of rat
hepatic aryl sulfotransferase IV. Mol. Pharmacol. 40, 36-44.
(18) Sekura, R. D., and J akoby, W. B. (1981) Aryl sulfotransferase IV
from rat liver. Arch. Biochem. Biophys. 211, 352-359.
(19) Laemmli, U. K. (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680-685.
(20) Gong, D. W., Ozawa, S., Yamazoe, Y., and Kato, R. (1991)
Purification of hepatic N-hydroxyarylamine sulfotransferases and
their regulation by growth hormone and thyroid hormone in rats.
J . Biol. Chem. 110, 226-231.

1258 Chem. Res. Toxicol., Vol. 13, No. 12, 2000
King et al.
(21) Nagata, K., Ozawa, S., Yamazoe, Y., and Kato, R. (1993) Isolation
and expression of a cDNA encoding a male-specific rat sulfotrans-
ferase that catalyzes activation of N-hydroxy-2-acetylaminofluo-
rene. J . Biol. Chem. 268, 24720-24725.
(31) Kraulis, P. J . (1991) MOLSCRIPT: A program to produce both
detailed and schematic plots of protein structures. J . Appl.
Crystallogr. 24, 946-950.
(32) Merritt, E. A., and Bacon, D. J . (1997) Raster3D: Photorealistic
(22) Yamazoe, Y., Nagata, K., Ozawa, S., and Kato, R. (1994) Struc-
tural similarity and diversity of sulfotransferases. Chem.-Biol.
Interact. 92, 107-117.
(23) Kiehlbauch, C. C., Lam, Y. F., and Ringer, D. P. (1995) Ho-
modimeric and heterodimeric aryl sulfotransferases catalyze the
sulfuric acid esterification of N-hydroxy-2-acetylaminofluorene.
J . Biol. Chem. 270, 18941-18947.
(24) Chen, X., Yang, Y.-S., Zheng, Y., Martin, B. M., Duffel, M. W.,
and J akoby, W. B. (1992) Tyrosine-ester sulfotransferase from
rat liver: bacterial expression and identification. Protein Expres-
sion Purif. 3, 421-426.
(25) Bensadoun, A., and Weinstein, D. (1976) Assay of proteins in the
presence of interfering materials. Anal. Biochem. 70, 241-250.
(26) Duffel, M. W., Binder, T. P., and Rao, S. I. (1989) Assay of purified
aryl sulfotransferase suitable for reactions yielding unstable
sulfuric acid esters. Anal. Biochem. 183, 320-324.
(27) Perella, F. W. (1988) EZ-FIT: A practical curve-fitting microcom-
puter program for the analysis of enzyme kinetic data on IBM-
PC compatible computers. Anal. Biochem. 174, 437-447.
(28) Kakuta, Y., Pedersen, L. G., Carter, C. W., Negishi, M., and
Pedersen, L. C. (1997) Crystal structure of estrogen sulphotrans-
ferase. Nat. Struct. Biol. 4, 904-908.
(29) Brunger, A. T., Kuriyan, J ., and Karplus, M. (1987) Crystal-
lographic R-factor refinement by molecular dynamics. Science 235,
458-460.
(30) Kakuta, Y., Petrotchenko, E. V., Pedersen, L. C., and Negishi,
M. (1998) The sulfuryl transfer mechanism. Crystal structure of
a vanadate complex of estrogen sulfotransferase and mutational
analysis. J . Biol. Chem. 273, 27325-27330.
molecular graphics. Methods Enzymol. 277, 505-524.
(33) Ringer, D. P., Norton, T. R., and Self, R. R. (1992) Reaction
product inactivation of aryl sulfotransferase IV following electro-
philic substitution by the sulfuric acid ester of N-hydroxy-2-
acetylaminofluorene. Carcinogenesis 13, 107-112.
(34) Sekura, R. D., Duffel, M. W., and J akoby, W. B. (1981) Aryl
sulfotransferases. Methods Enzymol. 77, 197-206.
(35) Duffel, M. W., and J akoby, W. B. (1981) On the mechanism of
aryl sulfotransferase. J . Biol. Chem. 256, 11123-11127.
(36) Campbell, N. R., Van Loon, J . A., Sundaram, R. S., Ames, M. M.,
and Hansch, C. (1987) Human and rat liver phenol sulfotrans-
ferase: structure-activity relationships for phenolic substrates.
Mol. Pharmacol. 32, 813-819.
(37) Binder, T. P., and Duffel, M. W. (1988) Sulfation of benzylic
alcohols catalyzed by aryl sulfotransferase IV. Mol. Pharmacol.
33, 477-479.
(38) Mangold, J . B., McCann, D. J ., and Spina, A. (1993) Aryl
sulfotransferase IV-catalyzed sulfation of aryl oximes: steric and
substituent effects. Biochim. Biophys. Acta 1163, 217-222.
(39) Duffel, M. W. (1994) Molecular specificity of aryl sulfotransferase
IV (tyrosine-ester sulfotransferase) for xenobiotic substrates and
inhibitors. Chem.-Biol. Interact. 92, 3-14.
(40) Kakuta, Y., Pedersen, L. C., Chae, K., Song, W.-C., Leblanc, D.,
London, R., Carter, C. W., and Negishi, M. (1998) Mouse steroid
sulfotransferases: substrate specificity and preliminary X-ray
crystallographic analysis. Biochem. Pharmacol. 55, 313-317.
TX990184Z