Mutation Research 376 1997.

6170

Human acetyltransferase polymorphisms

Denis M. Grant ) , Nicola C. Hughes, Susan A. Janezic, Geoffrey H. Goodfellow,
Hillary J. Chen, Andrea Gaedigk, Violeta L. Yu, Ranbir Grewal
Diision of Clinical Pharmacology and Toxicology, Research Institute, The Hospital for Sick Children, 555 Uniersity Aenue, Toronto,
Ontario M5G 1X8, Canada

Abstract

Conjugation of primary amino and hydroxylamino groups with acetate, catalyzed by acetyl CoA-dependent arylamine
acetyltransferase NAT. enzymes, may play an important role in the intricate series of metabolic pathways that produce or
prevent toxicity following exposure to homo- and heterocyclic arylamine and hydrazine xenobiotics. Two independently
regulated and kinetically distinct human acetyltransferases are now known to exist, namely NAT1 and NAT2. Interindividual
variation in NAT2 function is associated with the classical isoniazid acetylation polymorphism which was discovered over
forty years ago. At last count, fifteen variant alleles at the NAT2 gene locus have been linked to the isoniazid acetylator
phenotype, and each of these can be identified in population studies using specific PCR-based genotyping tests. On the
other hand, NAT1 shows kinetic selectivity for compounds whose disposition is unrelated to the classical isoniazid
acetylation polymorphism. NAT1 expression is also phenotypically variable in human populations, at least in part due to
allelic differences at the NAT1 gene locus. Nine NAT1 variant alleles have been described to date, of which NAT1) 14 and
NAT1) 15 clearly produce defective NAT1 proteins and lead to functional impairment in the metabolism of NAT1-selective
substrates both in vivo and in vitro. On the other hand, it has been reported that the NAT1) 10 variant associates with
eleated NAT1 activity and increased risk for cancers of the bladder and colon. Because of the important toxicologic
consequences of allelic variation in NAT1 and NAT2 function for the metabolic activation of arylamine and heterocyclic
amine procarcinogens, further studies are needed to improve our understanding of the extent of NAT allelic variation, to
determine the functional capacity of each variant gene product, and to develop accurate methods of detecting them in
population and epidemiological studies.

Keywords: Pharmacogenetics; Acetylation; Acetyltransferase; Polymorphism; Metabolic activation; Arylamine carcinogen

1. Introduction mechanisms underlying the occurrence of observed

Two recent reviews have presented detailed de-
scriptions of the historical background to acetylation
pharmacogenetics, and of the development of tech-
niques to investigate the biochemical and molecular
)
Corresponding author. Tel.: q1 416. 813-5175; Fax: q1
416. 813-7562; e-mail: grant@sickkids.on.ca
interindividual variations in acetylation capacity in
human populations Meyer, 1995; Weber and Vatsis,
1995.. The purpose of this presentation is to briefly
summarize the current state of knowledge concern-
ing allelic variations in the human acetyltransferases
NAT1 and NAT2, and to place this information in
the context of the role that these enzymes may play
in altering susceptibility to amine-induced toxicity.

0027-5107r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.

PII S 0 0 2 7 - 5 1 0 7 9 7 . 0 0 0 2 6 - 2
62 D.M. Grant et al.r Mutation Research 376 (1997) 6170

2. Pathways of aromatic amine detoxication and produces non-toxic, stable N-acetates, N-

metabolic activation
A large body of both experimental and epidemio-
logic evidence has established the importance of
exposure to homo- and heterocyclic arylamine and
hydrazine drugs, environmental chemicals and food
pyrolysis products in the pathogenesis of toxicities
such as idiosyncratic drug reactions and cancers of
the bladder and colon. It is also quite clear that these
compounds are not by themselves toxic. Rather, they
first require metabolic activation by enzymes of
xenobiotic biotransformation to chemically reactive
electrophiles that can bind covalently to intracellular
nucleophiles such as DNA and proteins. This binding
may subsequently lead to either direct cytotoxicity or
to nucleotide substitutions that drive the cell towards
a more malignant phenotype Hein, 1988..
The complex interplay of potential reaction path-
ways undergone by aromatic and heterocyclic amines
is shown in Fig. 1, and is described here briefly.
N-conjugation of arylamines by the action of N-
acetyltransferases NATs . , UDP-glucurono -
syltransferases UGTs. or sulfotransferases SULTs.
glucuronides and N-sulfates. On the other hand,
N-oxidation of arylamines or N-acetamides, medi-
ated by cytochrome P-4501A2 CYP1A2 .
Guengerich, 1992. or by the peroxidase function of
prostaglandin H synthase PHS. Eling and Curtis,
1992. is a critical initial step in the metabolic activa-
tion process, producing more reactive hydroxyl-
amines or hydroxamic acids, respectively. Further
N-acetylation, O-acetylation and intramolecular
N,O-acetyltransfer also influence levels of chemi-
cally unstable acetoxy esters and N-acetyl-acetoxy
esters. These are generally considered to be ultimate
carcinogens that can spontaneously decompose to
electrophilic DNA- or protein-binding nitrenium ions
Boteju and Hanna, 1994.. Although NAT was ini-
tially discovered and characterized based on its abil-
ity to catalyze acetyl CoA-dependent conjugation on
nitrogen atoms to produce arylamides and hy-
drazides, it can also mediate O-acetylation OAT.
and, to a limited extent, N,O-acetyltransfer NOAT.
reactions Hein et al., 1993.. SULTs and UGTs can
also biotransform hydroxylamines and hydroxamic
acids formed by NATrCYP1A2 into sulfate or glu-

Fig. 1. Potential pathways of arylamine detoxication and metabolic activation. See text for details. Enzyme abbreviations are: NAT,
arylamine N-acetyltransferase; OAT, O-acetyltransferase catalysed by NATs.; NOAT, intramolecular N,O-acyltransferase also catalysed
by NATs.; CYP1A2, cytochrome P4501A2; PHS, prostaglandin H synthase; UGT, UDP-glucuronosyltransferase; SULT, sulfotransferase.
 D.M. Grant et al.r Mutation Research 376 (1997) 6170 63

curonide esters that decompose to nitrenium ions
Chou et al., 1995; Orzechowski et al., 1994.. Fi-
nally, microsomal deacetylases DAC. can influence
levels of substrates for activating pathways by re-
versing the acetylation reaction. Indeed, the tumori-
genicity of 2-acetylaminofluorene AAF. may re-
quire its deacetylation to the arylamine Aune et al.,
1985., which is then presumably activated by path-
ways such as those described above.
3. The human acetylation polymorphism
Drug-metabolizing enzymes possess both isoform
multiplicity and overlapping substrate selectivity to
ensure versatility in the organisms ability to bio-
transform most of the chemicals to which it is ex-
posed. However, another important determinant of
enzyme diversity is allelic variation, since the exis-
tence of allelic forms of a particular gene product
can lead to marked interindividual differences in the
metabolism of xenobiotics. Many such pharmacoge-
netic variants of drug metabolism Kalow and Grant,
1995a,b. are monogenic and polymorphic. That is,
allelic variation occurs at a single gene locus, and
produces distinguishable phenotypes with the rarest
one occurring at a frequency of at least 1% in the
population. Such phenotypes, usually derived from
drug pharmacokinetic or metabolite excretion param-
eters, subdivide populations into so-called extensive
metabolizer and poor metabolizer groups.
From the discovery in the 1950s that the inci-
dence of side effects from isoniazid was related to a
genetic polymorphism that impaired the drugs elim-
ination in about 5560% of Caucasian individuals,
in vivo and in vitro studies have shown that the
disposition of several arylamine and hydrazine drugs
and xenobiotics is affected by the same defect re-
viewed in Evans, 1992... This impairment is due to
a reduction in the rate of N-acetylation by a cytoso-
lic CoASAc:arylamine N-acetyltransferase enzyme
NAT; EC 2.3.1.5., in a two-step reaction involving

Table 1
a
Variant alleles at the human NAT2 gene locus
b
Allele Phenotype Nucleotide change Amino acid change Consequence
191 282 341 481 590 803 857 Expressed In vivo
c
yM yF y yK yT qD yB
NAT2) 4 Rapid G C T C G A G none wild-type.
NAT2) 5A Slow C T Ile 114 Thr x Vma x x NAT2
NAT2) 5B Slow C T G Lys 268 Arg
NAT2) 5C Slow C G
NAT2) 6A Slow T A Arg 197 Gln x stability x NAT2
NAT2) 6B Slow A
NAT2)7A Slow A Gly 286 Glu x stability x NAT2
NAT2)7B Slow T A affinity
NAT2) 12A Rapid G Lys 268 Arg None None
NAT2) 12B Rapid T G
NAT2) 13 Rapid T
NAT2) 14A Slow A Arg 64 Gln x stability x NAT2
NAT2) 14B Slow A T
NAT2) 17 Unknown A434 C Gln145 Pro Unknown Unknown
NAT2) 18 Unknown A845 C Lys 282 Thr Unknown Unknown
a
Adapted from Vatsis et al. 1995..
b
Only changes from the wild-type NAT2 ) 4 . sequence top line. are indicated. Nucleotide substitutions shown in boldfaced italics change
the amino acid indicated and have a functional consequence, those in regular italics change an amino acid with no or unknown effect on
NAT2 function, and those in regular font are silent. Thus substitutions at positions 191, 341, 590 and 857 are diagnostic for defective NAT2
function and hence for the slow acetylator phenotype.
c
Letters and symbols underneath positions of nucleotide changes indicate alterations in naturally-occurring restriction sites that accompany
the mutations indicated in the columns below. Negative sign indicates loss, and positive sign indicates gain, of the following sites:
M s MspI; F s FokI; K s KpnI; T s TaqI; D s DdeI; B s BamH I.
64 D.M. Grant et al.r Mutation Research 376 (1997) 6170

transfer of acetate from acetyl CoA to primary aro-

matic amino and hydrazino functional groups Grant
and Meyer, 1993.. The clinical and toxicological
consequences of the acetylation polymorphism have
been studied in much detail Evans, 1992.. Of partic-
ular relevance to the present discussion are reports of
statistically significant but often controversial associ-
ations between the isoniazid slow acetylator pheno-
type and the incidence of toxicities such as sulfo-
namide hypersensitivity Delomenie et al., 1994; Lee
et al., 1993; Shear et al., 1986; Wolkenstein et al.,
1995. and bladder cancer Cartwright et al., 1982;
Risch et al., 1995., and of the rapid acetylator pheno-
type with colorectal cancer Ladero et al., 1991;
Lang et al., 1986; Probst-Hensch et al., 1995..

4. Variant alleles at the NAT2 gene locus

Since the original cloning of two separate NAT

genes and the establishment of the human NAT2
gene locus as the site of the human isoniazid acetyla-
tion polymorphism Blum et al., 1990, 1991; Grant
et al., 1990, 1991., the study of NAT2 allelic vari-
ants has been an area of intensive investigation Bell
et al., 1993; Blum et al., 1991; Deguchi et al., 1990;
Hickman and Sim, 1991; Lin et al., 1993; Ohsako
and Deguchi, 1990; Vatsis et al., 1991.. To date, a
total of fifteen variant NAT2 alleles that correlate
with acetylator phenotype have been detected in
human populations reviewed in Vatsis et al., 1995.,
accounting for greater than 98% of all of the alleles
so far examined Table 1.. The wild-type NAT2) 4
allele Blum et al., 1990. is associated with the rapid
acetylator phenotype. Each of the allelic variants that
have been investigated at the functional level pos-
sesses a diagnostic combination of between one and
three nucleotide substitutions at positions 191, 282,
341, 481, 590, 803 and 857 within the 870 bp NAT2
gene coding region. All but one position 341. of
these seven possible sites for nucleotide substitutions
lead to alterations in naturally occurring restriction
endonuclease recognition sequences. As such, a
number of NAT2 genotyping tests have been devel-
oped, using PCR amplificationrrestriction endonu-
clease digestion combinations, that are able to accu-
rately predict acetylator phenotype as determined
using in vivo probes such as caffeine Fig. 2..
Fig. 2. A. Urinary caffeine test for NAT2 acetylator phenotyping.
Shown is a bimodal population frequency distribution of the
urinary AFMUr1X molar ratio expressed on a logarithmic scale.
determined following intake of 200 mg of caffeine by each of 131
healthy subjects from the Toronto population. B. Phenotype
genotype correlation study. The NAT2 acetylator genotype was
determined by a PCR amplificationrrestriction digestionrallele-
specific reamplification method as described Delomenie et al.,
1996., and compared with the caffeine-derived acetylator pheno-
type. Complete concordance between phenotype and genotype
was observed in this population sample.
More recently, we have refined our own genotyping
test to more accurately determine the phase of
potentially ambiguous mutations that could lead to
genotype ambiguities but would not lead to pheno-
et al., 1996.. Such
type misclassification Delomenie
methods are useful for determining allele frequencies
in phylogenetic studies, but may be unnecessarily
cumbersome if the goal is simply to predict pheno-
type in clinical settings. In the latter case, it is only
necessary to determine zygosity at any of the four
currently known nucleotide positions within the 870
bp NAT2 gene coding region nucleotides 191, 341,
590 and 857. where mutations produce defective
NAT2 function.
 D.M. Grant et al.r Mutation Research 376 (1997) 6170 65

Table 2 produce proteins with significantly reduced stabili-

Kinetic constants for sulfamethazine acetylation by NAT2 allelic ties Fig. 3., with no impairment of substrate affin-
variant proteins expressed in E. coli
ity. Indeed, although the Arg 197 Gln and Gly 286
Variant Km Vmax
Glu mutations caused by G 590A and G 857A in
m M. nmolrminrmg.
NAT2) 6 and NAT2)7, respectively, appear to in-
NAT2 4 wild-type. 100 7.2
crease rather than decrease substrate affinity Table
NAT2 5A 54 0.052
NAT2 5B 40 0.012 2., the proteins are sufficiently unstable Fig. 3. that
NAT2 5C - 0.002 these mutations still produce the slow acetylator
NAT2 6A 16 0.008 phenotype in vivo. However, in no case is the con-
NAT2 7B 7 5.0 tent of NAT2 immunoreactive protein decreased upon
NAT2 13 83 2.4
heterologous expression of the variant alleles in E.
NAT2 14A 115 1.8
NAT2 14B 103 1.3 coli or following incubations that lead to decreased
enzyme activity, contrary to observations in liver
cytosols from individuals with the slow acetylator
We Grant et al., 1994. and others Hein et al., phenotype Grant et al., 1990.. Presumably there are
1994; Hickman et al., 1995. have used heterologous systems to more rapidly degrade these mutant NAT2
expression systems in E. coli to investigate the proteins within human cells that are not present in
mechanisms by which many of these mutant alleles the artificial environment of the bacterial expression
may lead to previously observed reductions in NAT2 host.
protein content in human liver Grant et al., 1990.. There are significant interethnic differences in
The Ile114 Thr substitution produced by the T 341 C certain NAT2) allele frequencies Delomenie et al.,
mutation common to the three NAT2) 5 alleles re- 1996. Table 3.. For instance, the NAT2) 5 alleles
duces Vmax Table 2. without impairing enzyme T 341 C. are the most common of the NAT2 variants
affinity or stability Fig. 3., possibly by preventing in Caucasian populations but are rare in Orientals.
proper folding of the nascent polypeptide into its This difference in NAT2) 5 allele frequency in fact
functionally active tertiary structure. On the other accounts for a large part of the known interethnic
hand, the mutations causing Arg 197 Gln, Gly 286 difference in NAT2 acetylator phenotype between
Glu and Arg 64 Gln amino acid substitutions these two racial groups. On the other hand,

Fig. 3. In vitro stabilities of selected NAT2 allelic variants expressed in E. coli. Lysates were prepared from bacterial strains induced to
express recombinant allelic variant forms of NAT2 as described Dupret and Grant, 1992. and rates of acetylation of the NAT2-selective
substrate sulfamethazine were measured at the times indicated following incubation at 378C. The three NAT2 5 variants left panel. showed
similar stabilities to the wild-type NAT2 4, while NAT2 6A, NAT2 7B and NAT2 14A right panel. were markedly less stable.
66 D.M. Grant et al.r Mutation Research 376 (1997) 6170

Table 3
a
Interethnic comparison of NAT2 allele frequencies
Ethnic group no. of alleles. Rapid acetylator alleles Slow acetylator alleles G 857A

G 191A T 341 C G 590A Gly 286 Glu.11

Arg 64 Gln. Ile114 Thr. Arg 197 Gln.
Caucasian-Canadians 226. 0.240 0 0.470 0.230 0.060
Caucasian-Americans 842. 0.240 0 0.430 0.310 0.020
Caucasian-Europeans 868. 0.263 0 0.455 0.264 0.018
Indians 122. 0.257 0 0.330 0.380 0.033
African-Americans 428. 0.347 0.083 0.295 0.230 0.045
Native Africans: Gabonese 104. 0.270 0.086 0.404 0.221 0.019
Dogons 100. 0.250 0.050 0.300 0.370 0.030
Hispanics 296. 0.403 0.007 0.275 0.180 0.135
Amerindians: Panama 142. 0.740 0 0.021 0 0.239
Columbia 202. 0.653 0 0.104 0.035 0.208
Koreans 170. 0.681 0.011 0.018 0.180 0.110
Japanese 448. 0.671 0 0.014 0.220 0.095
Chinese 508. 0.532 0 0.051 0.296 0.121
Filipino 200. 0.395 0 0.065 0.360 0.180
Polynesians 50. 0.600 0 0.040 0.340 0.020
Inuit 180. 0.789 0 0.022 0.167 0.022
a
et al. 1996. with additional results from the authors laboratory.
Adapted from Delomenie

NAT2)7B G 857A. is somewhat more prevalent in
Orientals than in Caucasians. NAT2) 14 G 191A. is
common in both American and native African Blacks
but very low or undetectable in other ethnic groups.
Finally, our recent studies in a Canadian Inuit popu-
lation unpublished observations. indicate that this
ethnic group, which is likely of Oriental origin,
possesses the highest frequency 0.789. of the wild-
type NAT2) 4 allele so far observed.
5. Variation at the NAT1 gene locus
Cloning and expression studies Blum et al., 1990;
Grant et al., 1991. had revealed the existence of a
second human NAT enzyme, NAT1, which shows
kinetic selectivity for compounds, such as p-amino-
salicylic acid PAS., whose disposition is unrelated
to the classical isoniazid acetylation polymorphism
governed by the NAT2 gene locus. In vitro studies
Cribb et al., 1991; Grant et al., 1991. showed that
NAT1 and NAT2 were independently expressed but
that NAT1 expression was also variable in human
populations. In addition, random sequencing of am-
plified NAT1 genes from human genomic DNA sug-
gested that allelic variation at this gene locus indeed
exists Vatsis and Weber, 1993; Vatsis et al., 1995..
Eight NAT1 allelic variants have so far been reported
Table 4., and six additional forms are currently
being characterized at the functional level H.J. Lin,
personal communication.. However, the functional
significance of many of these sequence variations is
as yet unknown. We developed in vivo and in vitro
tests for NAT1 function in man using PAS as a
probe drug, and used them in a population pharma-
cogenetic study of a healthy sample from the Toronto
population Hughes et al., 1997.. We observed that
in vivo and in vitro indices of PAS acetylation were
highly variable, and allowed for the identification of
persons predicted to have either deficient or elevated
NAT1 function. One subject with severely defective
PAS acetylation possessed two novel NAT1 variants,
NAT1) 14 and NAT1) 15. NAT1) 14 encodes a
protein with a single amino acid change Arg 187
Gln. that decreases its affinity for PAS by 1520-fold
Fig. 4., while NAT1) 15 has a premature stop
codon Arg 187 Stop. that produces a truncated,
inactive protein. We also detected a third, previously
identified variant, NAT1) 11 Vatsis and Weber,
1993.. However, it is not clear whether NAT1) 11
 D.M. Grant et al.r Mutation Research 376 (1997) 6170 67

Table 4
a
Variant alleles at the human NAT1 gene locus
b
Allele Nucleotide change Amino acid change Consequence
y 344 y40 459 559 560 640 1075 1088 1091 1095 Expressed In vivo
NAT1) 4 C A G C G T ) T C wild-type
NAT1) 3 A none unknown
NAT1) 5 many unknown
NAT1) 10 A A none NAT1?
NAT1) 11 T T A G D9 A Ser 214Ala none x NAT1?
NAT1) 14 ? ? A A A Arg 187 Gln x affinity x NAT1
NAT1) 15 ? ? T A A Arg 187 Stop no NAT1 x NAT1
NAT1) 16 q3A A none unknown
a
From Vatsis et al. 1994, 1995., Hughes et al. 1997.. Six additional variant alleles not listed here H. Lin, personal communication.
contain nucleotide substitutions that alter single amino acid residues and three of these give rise to NAT1 proteins devoid of catalytic
activity.
b
Only changes from the wild-type NAT1) 4 . sequence top line. are indicated. Nucleotide substitutions shown in boldfaced italics change
the amino acid indicated and have a functional consequence, those in regular italics change an amino acid with no or unknown effect on
NAT1 function, and those in regular font are silent. The asterisk at position 1075 represents a trinucleotide repeat, 9 nucleotides of which
are deleted in the NAT1) 11 allele.

alters NAT1 function, since 1. the amino acid change
in this variant Ser 214Ala. does not alter the activity
of the expressed enzyme, and 2. subjects that are
heterozygous for this allele have only slightly lower
NAT1 phenotyping ratios than the average of the
remaining population Hughes et al., 1997.. This
allele possesses both 5X and 3X non-coding mutations
Table 4. whose functional significance is as yet
unknown. We have devised genotyping tests to de-
tect NAT1) 11, NAT1) 14 and NAT1) 15 in human
populations, and have found each at a low but signif-
icant frequency 0.01 to 0.03. in Caucasians.
On the other hand, recent studies suggest that
another previously published NAT1 allelic variant,
NAT1) 10, produces eleated NAT1 activity in
bladder and colon tissues Bell et al., 1995a., in-
creased carcinogen-DNA adduct binding in bladder
Badawi et al., 1995., and increased risk for colorec-
tal Bell et al., 1995b. and bladder cancers Taylor et
al., 1995.. These interesting preliminary observations

Fig. 4. A. Kinetics of p-aminosalicylic acid PAS. acetylation in vitro by whole blood lysates from a homozygous wild-type
NAT1) 4 r NAT1) 4 . subject and from a subject Subject 244. possessing the variant alleles NAT1) 14 and NAT1) 15. B. Kinetics of
PAS acetylation in vitro by recombinant NAT1 variants expressed in E. coli. The affinity of NAT1 14 for PAS is about 15-fold lower than
that of NAT1 4 both in blood and in the heterologous expression system. Subject 244 displays characteristics of NAT1 14 function in vivo
because the remaining allele in this individual, NAT1) 15, produces a truncated protein that is devoid of any catalytic function Hughes et
al., 1997..
68 D.M. Grant et al.r Mutation Research 376 (1997) 6170
require confirmation and further study, since the
functional significance of the mutations that occur in
the NAT1) 10 allele have also not yet been directly
established. NAT1) 10 has no coding region muta-
tions, but contains two nucleotide substitutions
T 1088A and C 1095A. in its 3X untranslated region
3X-UTR.Table 4.. The T 1088A mutation alters a
p o ly a d e n y la tio n sig n a l T A A T A A A
TAA A AAA., leading to the suggestion that
NAT1) 10 is associated with increased NAT1 activ-
ity due to altered transcript polyadenylation. Because
of the potentially important toxicologic implications
of allelic variation in NAT1 function that have been
mentioned above Badawi et al., 1995; Bell et al.,
1995a,b; Taylor et al., 1995., it will be important to
fully characterize the extent of allelic variation at the
NAT1 gene locus, to understand the true functional
consequences of each variant, and to have accurate
diagnostic tests for discriminating them in population
and epidemiological studies.
6. Toxic consequences of NAT allelic variation
Any mechanistic model for chemical-induced tox-
icities such as cancer must be able to explain ob-
served tissue-selectivity of the toxic event. In the
case of bladder cancer Fig. 5., one mechanism to
explain its tissue selectivity suggests that activating
and detoxifying enzymes compete in the liver for the
formation of metabolites which are then in some
fashion delivered to the bladder. In this scenario, in
the absence of liver N-acetylation i.e., in phenotypi-
cally slow acetylators for the liver- and gut-specific
NAT2., CYP1A2-mediated N-hydroxylation pre-
dominates, followed by glucuronidation and trans-
port to the bladder lumen where acid hydrolysis
generates a chemically reactive species. In the past,
however, such models suggested that the bladder
plays a rather passive role and did not consider the
possibility that biotransformation within the bladder
itself could also produce highly reactive metabolites.
Until recently, such models had also not taken into
account the existence of the independently regulated
NAT1, its ability to efficiently biotransform ary-
lamine carcinogens Hein et al., 1993., its distinct
substrate specificity from that of NAT2, its variabil-
ity in human populations and the likelihood that its
Fig. 5. A possible model for arylamine bioactivation in bladder
cancer. For an explanation, refer to the text. Abbreviations are as
in the legend to Fig. 1.
catalytic activity for certain substrates can be much
higher in bladder than that of NAT2. Recent studies
have provided evidence that human bladder is capa-
ble of both N-acetylating carcinogenic amines and
O-acetylating their hydroxylamine metabolites, and
that NAT1 is likely to be the enzyme that predomi-
nates in this tissue. Moreover, although we have not
been able to detect CYP1A2 activity in microsomes
from human bladder biopsies, prostaglandin H syn-
thase PHS. activity is detectable in all bladders so
far tested unpublished observations.. PHS has been
shown to be capable of activating aromatic amines to
DNA-binding species Eling and Curtis, 1992..
Thus the tendency for a given aromatic amine to
be converted into either innocuous metabolites or
into reactive DNA-binding species that produce can-
cer in target tissues will depend on 1. the catalytic
specificities of acetylating both NAT1 and NAT2.,
deacetylating, oxidizing CYP1A2 or PHS. and other
conjugating enzymes SULT, UGT. for the chemical
in question; 2. the chemical reactivities of the
metabolites produced; 3. the relative tissue levels of
the enzymes involved; and 4. both genetically and
environmentally based population variation in the
expression of each of these enzymes. The studies
conducted so far in this regard would suggest that
the role of NAT1 in predisposition to amine-induced
cancer may be greater than anticipated during earlier
epidemiological studies that had taken into account
 D.M. Grant et al.r Mutation Research 376 (1997) 6170
only NAT2-based variation. This may explain the
fact that the association between the NAT2 slow
acetylator phenotype and bladder cancer incidence is
not particularly strong, and several studies have found
no association. Indeed, in light of the model outlined
in Fig. 5, it is tempting to hypothesize that individu-
als with a more precisely defined metabolizer phe-
notype of combined high CYP1A2 or PHS., low
NAT2 and high NAT1 activity may be particularly at
risk for developing bladder cancer following expo-
sure to aromatic amine procarcinogens. Further
mechanistic and epidemiological studies are required
to more fully examine the roles of these interacting
pathways in producing altered risk for arylamine-in-
duced toxicity.
Acknowledgements
Original work from the authors laboratory was
supported by operating grants from the Medical Re-
search Council of Canada and the National Cancer
Institute of Canada, and by personal salary support
from the Medical Research Council of Canada, The
Natural Sciences and Engineering Research Council
of Canada, the Ontario Ministry of Health, and the
Research Training Committee of the Hospital for
Sick Children.
References
Aune, T., R.R. Vanderslice, J.E. Croft, E. Dybing, J.R. Bend and
R.M. Philpot 1985. Cancer Res., 45, 58595866.
Badawi, A.F., A. Hirvonen, D.A. Bell, N.P. Lang and F.F.
Kadlubar 1995. Role of aromatic amine acetyltransferases,
NAT1 and NAT2, in carcinogen-DNA adduct formation in the
human urinary bladder, Cancer Res., 55, 52305237.
Bell, D.A., J.A. Taylor, M.A. Butler, E.A. Stephens, J. Wiest,
L.H. Brubaker, F.F. Kadlubar and G.W. Lucier 1993. Geno-
typerphenotype discordance for human arylamine N-acetyl-
transferase NAT2. reveals a new slow-acetylator allele com-
mon in African-Americans, Carcinogenesis, 14, 16891692.
Bell, D.A., A.F. Badawi, N.P. Lang, K.F. Ilett, F.F. Kadlubar and
A. Hirvonen 1995a. Polymorphism in the N-acetyltransferase
1 NAT1. polyadenylation signal: association of NAT1) 10
allele with higher N-acetylation activity in bladder and colon
tissue, Cancer Res., 55, 52265229.
Bell, D.A., E.A. Stephens, T. Castranio, D.M. Umbach, M. Wat-
son, M. Deakin, J. Elder, C. Hendrickse, H. Duncan and R.C.
Strange 1995b. Polyadenylation polymorphism in the acetyl-
69
transferase 1 gene NAT1. increases risk of colorectal cancer,
Cancer Res., 55, 35373542.
Blum, M., D.M. Grant, O.W. McBride, M. Heim and U.A. Meyer
1990. Human arylamine N-acetyltransferase genes: isolation,
chromosomal localization and functional expression, DNA
Cell Biol., 9, 193203.
Blum, M., A. Demierre, D.M. Grant, M. Heim and U.A. Meyer
1991. Molecular mechanism of slow acetylation of drugs and
carcinogens in humans, Proc. Nat. Acad. Sci. USA., 88,
52375241.
Boteju, L.W. and P.E. Hanna 1994. Arylamine-nucleoside adduct
formation: evidence for arylnitrene involvement in the reac-
tions of an N-acetoxyarylamine, Chem. Res. Toxicol., 7,
684689.
Cartwright, R.A., H.J. Rogers, D. Barham-Hall, R.W. Glashan,
R.A. Ahmad, E. Higgins and M.A. Kahn 1982. Role of
N-acetyltransferase phenotypes in bladder carcinogenesis: a
pharmacogenetic epidemiological approach to bladder cancer,
Lancet, 2, 842846.
Chou, H.-C., N.P. Lang and F.F. Kadlubar 1995. Metabolic
activation of N-hydroxy arylamines and N-hydroxy hetero-
cyclic amines by human sulfotransferases., Cancer Res., 55,
525529.
Cribb, A.E., D.M. Grant, M.A. Miller and S.P. Spielberg 1991.
Expression of monomorphic arylamine N-acetyltransferase
NAT1. in human leukocytes, J. Pharmacol. Exp. Ther., 259,
12411246.
Deguchi, T., M. Mashimo and T. Suzuki 1990. Correlation
between acetylator phenotypes and genotypes of polymorphic
arylamine N-acetyltransferase in human liver, J. Biol. Chem.,
265, 1275712760.
Delomenie, C., D.M. Grant, P. Mathelier-Fusade, C. Jacomet, F.
Leynadier, E. Jacqz-Aigrain, W. Rozenbaum, R. Krishnamoor-
thy and J.-M. Dupret 1994. N-acetylation genotype and risk
of severe reactions to sulfonamides in AIDS patients, Br. J.
Clin. Pharmac., 38, 581582.
Delomenie, C., L. Sica, D.M. Grant, R. Krishnamoorthy and J.-M.
Dupret 1996. Genotyping of the polymorphic N-acetyltrans-
ferase NAT2 . gene locus in two native African populations,
Pharmacogenetics, 6, 177185.
Dupret, J.-M. and D.M. Grant 1992. Site-directed mutagenesis of
recombinant human arylamine N-acetyltransferase expressed
in E. coli: evidence for direct involvement of Cys 68 in the
catalytic mechanism of polymorphic human NAT2, J. Biol.
Chem., 267, 73817385.
Eling, T.E. and J.F. Curtis 1992. Xenobiotic metabolism by
prostaglandin H synthase, Pharmacol. Ther., 53, 261273.
Evans, D.A. 1992. N-acetyltransferase, in: W. Kalow Ed..,
Pharmacogenetics of Drug Metabolism, Pergamon Press, New
York, pp. 95178.
Grant, D.M., K. Morike, M. Eichelbaum and U.A. Meyer 1990.
Acetylation pharmacogenetics: the slow acetylator phenotype
is caused by decreased or absent arylamine N-acetyltrans-
ferase in human liver, J. Clin. Invest., 85, 968972.
Grant, D.M., M. Blum, M. Beer and U.A. Meyer 1991.
Monomorphic and polymorphic human arylamine N-
acetyltransferases: a comparison of liver isozymes and ex-
70 D.M. Grant et al.r Mutation Research 376 (1997) 6170
pressed products of two cloned genes, Mol. Pharmacol., 39,
184191.
Grant, D.M. and U.A. Meyer 1993. Arylamine N-acetyltrans-
ferase, in: A. Haeberli Ed.., Human Protein Data, VCH
Verlagsges., Weinheim.
Grant, D.M., D.A. Bell and R. Grewal 1994. Functional analysis
of variant alleles of human arylamine acetyltransferase NAT2
by heterologous expression in E. coli, 10th International
Symposium on Microsomes and Drug Oxidations, Toronto,
Canada, p. 259.
Guengerich, F.P. 1992. Metabolic activation of carcinogens,
Pharmacol. Ther., 54, 1761.
Hein, D.W. 1988. Acetylator genotype and arylamine-induced
carcinogenesis, Biochim. Biophys. Acta, 948, 3766.
Hein, D.W., M.A. Doll, T.D. Rustan, K. Gray, Y. Feng, R.J.
Ferguson and D.M. Grant 1993. Metabolic activation and
deactivation of arylamine carcinogens by recombinant human
NAT1 and polymorphic NAT2 acetyltransferases, Carcinogen-
esis, 14, 16331638.
Hein, D.W., R.J. Ferguson, M.A. Doll, T.D. Rustan and K. Gray
1994. Molecular genetics of human polymorphic N-acetyl-
transferase: enzymatic analysis of 15 recombinant wild-type,
mutant, and chimeric NAT2 allozymes, Hum. Mol. Genet., 3,
729734.
Hickman, D. and E. Sim 1991. N-acetyltransferase polymor-
phism: comparison of phenotype and genotype in humans,
Biochem. Pharmacol., 42, 10071014.
Hickman, D., J.R. Palamanda, J.D. Unadkat and E. Sim 1995.
Enzyme kinetic properties of human recombinant arylamine
N-acetyltransferase 2 allotypic variants expressed in Es-
cherichia coli, Biochem. Pharmacol., 50, 697703.
Hughes, N.C., K.L. McQueen, S.A. Janezic, M.A.S. Jewett and
D.M. Grant 1997. Identification and characterization of vari-
ant alleles of human acetyltransferase NAT1 with defective
function using p-aminosalicylate as an in vivorin vitro probe,
Pharmacogenetics, 7, in press.
Kalow, W. and D.M. Grant 1995a. Genetic variation of drug
metabolizing enzymes, in: G.M. Pacifici and G.N. Fracchia
Eds.., Advances in Drug Metabolism in Man, European
Commission DG XII Science, Research and Development,
Brussels, pp. 5784.
Kalow, W. and D.M. Grant 1995b. Pharmacogenetics, in: C.
Scriver, A. Beaudet, W. Sly and D. Valle Eds.., The Metabolic
and Molecular Bases of Inherited Disease, 7th Edition, Mc-
Graw-Hill Inc, New York, pp. 293326.
Ladero, J.M., J.F. Gonzalez, J. Benitez, E. Vargas, M.J. Fernan-
dez, W. Baki and M. Diaz-Rubio 1991. Acetylator polymor-
phism in human colorectal carcinoma, Cancer Res., 51, 2098
2100.
Lang, N.P., D.Z.J. Chu, C.F. Hunter, D.C. Kendall, T.J. Flam-
mang and F.F. Kadlubar 1986. Role of aromatic amine
acetyltransferase in human colorectal cancer, Arch. Surg., 121,
12591261.
Lee, B.L., D. Wong, N.L. Benowitz and P.M. Sullam 1993.
Altered patterns of drug metabolism in patients with acquired
immunodeficiency syndrome, Clin. Pharmacol. Ther., 53,
529535.
Lin, H.J., C.Y. Han, B.K. Lin and S. Hardy 1993. Slow acetyla-
tor mutations in the human polymorphic N-acetyltransferase
gene in 786 asians, blacks, hispanics, and whites application
to metabolic epidemiology, Am. J. Hum. Genet., 52, 827834.
Meyer, U.A. 1995. Molecular genetics of the acetyltransferases,
in: G.M. Pacifici and G.N. Fracchia Eds.., Advances in Drug
Metabolism in Man, European Commission DG XII Sci-
ence, Research and Development, Brussels, pp. 587606.
Ohsako, S. and T. Deguchi 1990. Cloning and expression of
cDNAs for polymorphic and monomorphic arylamine N-
acetyltransferases from human liver, J. Biol. Chem., 265,
46304634.
Orzechowski, A., D. Schrenk, B.S. Bock-Hennig and K.W. Bock
1994. Glucuronidation of carcinogenic arylamines and their
N-hydroxy derivatives by rat and human UDP-glucurono-
syltransferases of the UGT1 gene complex, Carcinogenesis,
15, 15491553.
Probst-Hensch, N.M., R.W. Haile, S.A. Ingles, M.P. Longnecker,
C.Y. Han, B.K. Lin, D.B. Lee, G.T. Sakamoto, H.D. Frankl
and E.R. Lee 1995. Acetylation polymorphism and preva-
lence of colorectal adenomas, Cancer Res., 55, 20172020.
Risch, A., D.M. Wallace, S. Bathers and E. Sim 1995. Slow
N-acetylation genotype is a susceptibility factor in occupa-
tional and smoking-related bladder cancer, Hum. Mol. Genet.,
4, 231236.
Shear, N.H., S.P. Spielberg, D.M. Grant, B.K. Tang and W.
Kalow 1986. Differences in metabolism of sulfonamides
predisposing to idiosyncratic toxicity, Ann. Intern. Med., 105,
179184.
Taylor, J.A., D. Umbach, E. Stephens, D. Paulson, C. Robertson,
J.L. Mohler and D.A. Bell 1995. Role of N-acetylation
polymorphism at NAT1 and NAT2 in smoking-associated
bladder cancer, Proc. Am. Assoc. Cancer Res., 36, 282.
Vatsis, K.P., K.J. Martell and W.W. Weber 1991. Diverse point
mutations in the human gene for polymorphic N-acetyltrans-
ferase, Proc. Nat. Acad. Sci. USA, 88, 63336337.
Vatsis, K.P. and W.W. Weber 1993. Structural heterogeneity of
Caucasian N-acetyltransferase at the NAT1 gene locus, Arch.
Biochem. Biophys., 301, 7176.
Vatsis, K.P., Y. Yao and W.W. Weber 1994. Allelic variation at
the NAT1 locus of N-acetyltransferase among Orientals, Am.
J. Hum. Genet., 55 Suppl., A340.
Vatsis, K.P., W.W. Weber, D.A. Bell, J.-M. Dupret, D.A. Price
Evans, D.M. Grant, D.W. Hein, H.J. Lin, U.A. Meyer, M.V.
Relling, E. Sim, T. Suzuki and Y. Yamazoe 1995. Nomencla-
ture for N-acetyltransferases, Pharmacogenetics, 5, 117.
Weber, W.W. and K.P. Vatsis 1995. Human N-acetyltransferases:
genetic polymorphism and metabolic profiles of the major
isoforms, in: G.M. Pacifici and G.N. Fracchia Eds.., Ad-
vances in Drug Metabolism in Man, European Commission
DG XII Science, Research and Development, Brussels, pp.
351405.
Wolkenstein, P., V. Carriere, D. Charue, S. Bastuji-Garin, J.
Revuz, J.-C. Roujeau, P. Beaune and M. Bagot 1995. A slow
acetylator genotype is a risk factor for sulfonamide-induced
toxic epidermal necrolysis and Stevens-Johnson syndrome,
Pharmacogenetics, 5, 255258.