BIOLOGY OF REPRODUCTION 72, 157–163 (2005)
Published online before print 8 September 2004.
Comparison of Germ Cell Mutagenicity in Male CYP2E1-Null and Wild-Type Mice
Treated with Acrylamide: Evidence Supporting a Glycidamide-Mediated Effect
B.I. Ghanayem,1,2K.L. Witt,3L. El-Hadri,2U. Hoffler,2G.E. Kissling,4M.D. Shelby,5and J.B. Bishop3
Laboratory of Pharmacology and Chemistry,2Toxicology Operations Branch,3Biostatistics Branch,4Office of Program
Development,5National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
Acrylamide is an animal carcinogen and probable human car-
cinogen present in appreciable amounts in heated carbohydrate-
rich foodstuffs. It is also a germ cell mutagen, inducing domi-
nant lethal mutations and heritable chromosomal translocations
in postmeiotic sperm of treated mice. Acrylamide’s affinity for
male germ cells has sometimes been overlooked in assessing its
toxicity and defining human health risks. Previous investigations
of acrylamide’s germ cell activity in mice showed stronger ef-
fects after repeated administration of low doses compared with
a single high dose, suggesting the possible involvement of a sta-
ble metabolite. A key oxidative metabolite of acrylamide is the
epoxide glycidamide, generated by cytochrome P4502E1
(CYP2E1).To explore the role of CYP2E1 metabolism in the germ
cell mutagenicity of acrylamide, CYP2E1-null and wild-type
male mice were treated by intraperitoneal injection with 0, 12.5,
25, or 50 mg acrylamide (5 ml saline)?1kg?1day?1for 5 con-
secutive days. At defined times after exposure, males were mat-
ed to untreated B6C3F1females. Females were killed in late ges-
tation and uterine contents were examined. Dose-related in-
creases in resorption moles (chromosomally aberrant embryos)
and decreases in the numbers of pregnant females and the pro-
portion of living fetuses were seen in females mated to acryl-
amide-treated wild-type mice. No changes in any fertility pa-
rameters were seen in females mated to acrylamide-treated
CYP2E1-null mice. Our results constitute the first unequivocal
demonstration that acrylamide-induced germ cell mutations in
male mice require CYP2E1-mediated epoxidation of acrylamide.
Thus, CYP2E1 polymorphisms in human populations, resulting
in variable enzyme metabolic activities, may produce differen-
tial susceptibilities to acrylamide toxicities.
acrylamide, CYP2E1-null mice, dominant lethals, environment,
male reproductive tract, reproductive toxicant, sperm
Acrylamide is a rodent carcinogen and probable human
carcinogen [1, 2], and a mutagen in mammalian somatic
cells in vitro and in vivo, particularly in assays that detect
induction of chromosomal damage [1, 3–7]. In addition,
acrylamide is a mammalian germ cell mutagen, inducing
high frequencies of dominant lethal mutations (generally
associated with chromosomal alterations that result in death
of embryos around the time of implantation) [8–12], heri-
1Correspondence: Burhan I. Ghanayem, National Institute of Environmen-
tal Health Sciences, P.O. Box 12233, MD B3-10, 111 Alexander Dr., Re-
search Triangle Park, NC 27709. FAX: 919 541 4632;
Received: 17 June 2004.
First decision: 8 July 2004.
Accepted: 27 August 2004.
? 2005 by the Society for the Study of Reproduction, Inc.
ISSN: 0006-3363. http://www.biolreprod.org
table chromosomal translocations [12–14], and specific lo-
cus mutations  in postmeiotic sperm and spermatogo-
nial stem cells  of male mice.
Human exposure to acrylamide occurs during manufac-
turing and through its use in polyacrylamide gels, as a
grouting agent and soil conditioner, and in stabilization of
tunnel and dam structures . The recent discovery of ac-
rylamide in grain-based and carbohydrate-rich foodstuffs
subjected to high heat during processing [16–19] has added
to concerns over the potential risks for detrimental health
effects resulting from human exposure to this chemical. An
FAO/WHO  panel estimated that the general public is
exposed to 0.3–0.8 ?g acrylamide kg?1day?1through food
intake; children have 2–3 times the exposures of adults,
when calculations are expressed on a bodyweight basis. The
U.S. Food and Drug Administration issued preliminary ex-
posure estimates of acrylamide from food sources of 0.43
?g kg?1day?1for adults and 1.06 ?g kg?1day?1for chil-
dren . The review by the European Commission of di-
etary acrylamide exposure data (http://europa.eu.int/comm/
food/fs/sc/scf/out131 en.pdf) reported estimated acrylam-
ide intakes of 10–100 ?g kg?1day?1by one group and
lower intakes, ranging from 0.36 to 2.1 ?g kg?1day?1, were
estimated by a second group. The recent discovery of ac-
rylamide in food has increased the need for improved as-
sessment of risk factors of human exposure to acrylamide,
not only because of the possible cancer risk for populations
or individuals who consume high amounts of acrylamide-
containing foods but also for the potential risk of increased
frequencies of germ cell mutations and resulting adverse
Dearfield et al. , in a review of the genotoxicity of
acrylamide, discussed three possible metabolic pathways
for acrylamide: radical-mediated polymerization, most ef-
ficiently conducted under anaerobic conditions and used to
generate polyacrylamide; Michael-type reactions resulting
from the alpha-beta unsaturated characteristics of the ac-
rylamide molecule; and oxidative metabolism of the acryl-
amide double bond to yield the epoxide, glycidamide. Gly-
cidamide is a relatively stable intermediate, having an in
vivo half-life of 1.5 h in the rat and demonstrating a pattern
of even distribution among tissues in mice and rats after
acrylamide dosing .
The affinity of acrylamide for germinal tissues was il-
lustrated by whole-body radiographic studies of the system-
ic distribution of radiolabeled acrylamide in male mice,
which showed intense accumulation of acrylamide, or a
metabolite, in testis, then epididymis, and finally, the glans
penis, during a period of 1 h to 9 days posttreatment .
The mechanism(s) by which acrylamide induces germ cell
mutagenic effects is not clear. Sega et al.  proposed that
alkylation of the sulfhydryl groups of sperm protamine fol-
lowing acrylamide exposure produced the observed germ
GHANAYEM ET AL.
ployed in studies 1–3, spermiogenic stages
exposed, and spermatogenic stages sam-
pled in each mating period. For example,
the dosing and mating periods employed
in study 1 sampled sperm a minimum of 7
days posttreatment (the shortest interval
between the end of dosing and the initia-
tion of cohabitation) up to a maximum of
17 days posttreatment (the longest interval,
spanning the time from the beginning of
dosing to the last day of cohabitation).
Dark hatched areas on the spermatogene-
sis timeline indicate the spermiogenic pe-
riod of greatest sensitivity to acrylamide
(Days 5–8); light hatched areas indicate
periods of lower sensitivity and nonhatch-
ed portions of the Stages sampled bars in-
dicate nonsensitive spermiogenic stages
Dosing and mating schemes em-
cell chromosomal damage in treated male mice. Subsequent
studies of genetic effects, DNA adduct formation, and un-
scheduled DNA synthesis in spermiogenic cells after ac-
rylamide exposure led to the proposal of a glycidamide-
mediated mutagenic pathway via DNA or protamine alkyl-
ation [24–26]. Consistent with this hypothesis, glycidamide
has been shown to induce dominant lethal mutations in
male mice, with the most sensitive stages being spermato-
zoa of the testis and epididymis ; these are the same
stages that show the highest sensitivity to dominant lethal
induction by acrylamide.
Investigation of the oxidative metabolism of acrylamide
in mice by Sumner et al.  using13C-NMR to compare
urinary metabolites in CYP2E1?/? (CYP2E1-null) versus
CYP2E1?/? (wild-type) mice treated with13C-acrylamide
demonstrated that oxidative metabolism of acrylamide in
wild-type mice via CYP2E1 generated the reactive epoxide
glycidamide. In contrast, they identified no urinary metab-
olites originating from the epoxidation of acrylamide in the
urine of CYP2E1-null mice. Sumner et al.  concluded
that metabolism of the parent compound was exclusively
routed through direct glutathione conjugation, no detectable
glycidamide was formed in the absence of CYP2E1, and
there appeared to be no alternative pathways for the oxi-
dative metabolism of acrylamide.
Our hypothesis centers on the premise that glycidamide
is responsible for acrylamide-induced germ cell mutations
in mice. The current investigations were undertaken to as-
sess the role of acrylamide epoxidation to glycidamide in
the induction of dominant lethal mutations in male mouse
germ cells using CYP2E1-null and wild-type mice.
MATERIALS AND METHODS
Acrylamide (CAS no. 79-06-1; ?99.5% pure) was manufactured by
Fluka Chemie GmbH and purchased from Sigma-Aldrich Laborchemika-
lien GmbH (Milwaukee, WI). All dosing solutions were made daily by
mixing acrylamide with normal saline (0.9%); dosing volume was 5 ml/
kg body weight.
CYP2E1?/? (CYP2E1-null) and CYP2E1?/? (wild-type) mouse
strains were obtained from a colony developed in the laboratory of Dr.
Frank Gonzalez (National Cancer Institute, Bethesda, MD) , and main-
tained by inbreeding at Charles River Laboratories (Wilmington, MA) (see
Hoffler et al.  for details of stock development and maintenance).
Nullizygosity of the CYP2E1-null mice was confirmed using Western blot
analysis as previously described . Male wild-type and CYP2E1-null
mice were approximately 8 wk old and ranged in weight from 20 to 24 g
at the beginning of each study. Female B6C3F1(Taconic Laboratories,
Germantown, NY) were also 8–9 wk old and weighed 20–24 g at the
beginning of the study. All animals were housed in controlled environment
facilities with a 12L:12D cycle and were fed National Institutes of Health
(NIH) #31 diet and tap water. Both food and water were available ad
libitum throughout the experiments. All animals were acclimated for a
minimum of 1 wk before the start of the studies and all animal care and
experimental procedures were conducted in strict accordance with NIH
animal care and use guidelines (National Research Council’s Guide for
Care and Use of Laboratory Animals, 1996, National Academy of Sci-
For the three studies described below, acrylamide was administered to
groups of CYP2E1-null and wild-type male mice by intraperitoneal (i.p.)
injection at dose levels of 0, 12.5, 25, or 50 mg (5 ml saline)?1kg?1day?1
for 5 consecutive days. Matching vehicle controls were treated with 5 ml/
kg saline/day (i.p.), for 5 consecutive days. The i.p. injection was selected
for the route of administration because this study was designed to inves-
tigate the role of CYP2E1 metabolism in the germ cell damage observed
after acrylamide treatment, and i.p. injection was the route used in pre-
vious dominant lethal experiments with acrylamide  and glycidamide
. The number and distribution of mice varied among treatment groups
and experiments and was dependent primarily on the availability of age-,
weight-, and sex-matched CYP2E1-null and wild-type mice. In all exper-
iments described below, male mice were mated to untreated virgin B6C3F1
female mice (3 females/male in study 1; 2 females/male in studies 2 and
3). Mating schemes are presented in Figure 1. The experimental proce-
dures followed in the three studies are described below.
Untreated male CYP2E1-null and wild-type mice (30 per genotype)
were each cohabited with three untreated B6C3F1female mice for a period
of 5 days. At 48 h after the end of this mating period, the 30 males within
ACRYLAMIDE GERM CELL EFFECTS ARE CYP2E1-DEPENDENT
females mated by treatment group, genotype, dose of acrylamide, and
mating period for each of the three studies. The numbers at the top of
each bar are the actual numbers of pregnant females out of total number
of females cohabited per treatment group. *, Significantly less than control
at P ? 0.05; †, significantly different from same dose CYP2E1-null treat-
ment group at P ? 0.05.
Percentage (%) of pregnant females out of the total number of FIG. 3.
treatment group, genotype, dose of acrylamide, and mating period for
each of the three studies. *, Significantly less than control at P ? 0.05;
†, significantly less than same dose CYP2E1-null treatment group at
P ? 0.05.
Mean number of implantations per pregnant female (?SEM) by
each genotype were subdivided into two groups. One group of 10 males/
genotype received saline (i.p., once daily for 5 consecutive days) and the
other group of 20 males/genotype received 50 mg/kg acrylamide (i.p.,
once daily for 5 consecutive days). One week after treatments ended,
males were cohabited with three females each, for a second round of
mating, for a period of 5 days. Approximately 13 days after the last day
of cohabitation, each female was humanely killed with CO2/O2and uterine
contents were carefully examined for the following endpoints: total num-
ber of implantation sites, number of live fetuses, number of resorption
moles, number of early and late dead embryos, and number of dead fetuses
. Live fetuses were humanely killed after examination.
Because the 50 mg/kg dose of acrylamide used in study 1 produced
unacceptable levels of sterility in treated wild-type males, an investigation
was conducted, using only wild-type mice, to identify lower doses of ac-
rylamide that would induce dominant lethal mutations while maintaining
fertility at consistently high levels. Thus, wild-type mice were treated with
acrylamide at 0 (5 mice), 12.5 (12 mice), or 25 (13 mice) mg kg?1day?1
for 5 consecutive days, as described above. Forty-eight hours after the last
dose of acrylamide, each of the males was cohabited with two females for
a period of 5 days. At the end of the 5-day mating period, females were
removed and replaced with two females for a second 5-day mating period.
Approximately 13 days after the last day of cohabitation (for each of the
two mating periods), females were humanely killed with CO2/O2and uter-
ine contents were carefully examined as described above for study 1. Live
fetuses were humanely killed after examination.
Results of study 2 indicated that 12.5 and 25 mg/kg acrylamide were
appropriate dose levels for use in a more definitive third study. Thus,
CYP2E1-null and wild-type male mice were treated with acrylamide doses
of 0, 12.5, or 25 mg kg?1day?1for 5 consecutive days, as described above
for study 2. The saline-vehicle control groups consisted of 8 CYP2E1-null
and 11 wild-type males; the two acrylamide dose groups consisted of 12
CYP2E1-null and 13 wild-type males each. Forty-eight hours after the last
dose of acrylamide, each male was mated to two females for a period of
5 days. At the end of the 5-day mating period, females were removed and
replaced with two females for a second 5-day mating period. Approxi-
mately 13 days after the end of each of the two cohabitation periods,
females were humanely killed with CO2/O2and uterine contents were
carefully examined as described above for study 1. Live fetuses were hu-
manely killed after examination.
Extrabinomial variability of proportions of adverse events was tested
using the binomial variance test and Tarone test for binomial overdisper-
sion . Extrabinomial variability of proportions can arise when propor-
tions vary substantially among females mated to each male, in addition to
varying among males; when detected, it must be addressed in the statistical
analyses. Extrabinomial variability was present only for resorptions in
wild-type controls in the first mating period of study 3. In addition, number
of implantations, percentage of resorptions, and percentage of live fetuses
were not normally distributed. Transformations, such as the arcsine trans-
formation, did not improve normality . Therefore, nonparametric
methods were used to make comparisons among dose groups and among
genotype groups, and the litter was considered the unit of analysis. Krus-
kal-Wallis analysis of variance was applied to litter counts and percent-
ages, and where significant differences across groups were found, pairs of
groups were compared using the Mann-Whitney U-test . Percentages
of pregnant females were compared across groups using a chi-square test
in study 1 . In studies 2 and 3, dose-related trends were tested using
the Jonckheere-Terpstra test . Significant overall group differences in
percentage of pregnant females were followed by Fisher exact test to iden-
tify which pairs of groups differed . In studies 2 and 3, Week 1 out-
comes were compared with Week 2 outcomes by considering the male as
the unit of analysis. For each male, the total numbers of implantations,
percentages of resorptions, and percentage of live fetuses were calculated
for each week. These outcomes were then compared between Week 1 and
Week 2 using the Wilcoxon signed-ranks test, which takes advantage of
the paired nature of the data .
The fertility parameters most pertinent to the evaluation
of the dominant lethality of acrylamide are percentage (%)
pregnant females (Fig. 2), the mean number of implants per
pregnant female (Fig. 3), percentage live fetuses per preg-
nant female (Fig. 4), and percentage resorptions per preg-
nant female (Fig. 5). None of these four fertility parameters
GHANAYEM ET AL.
per pregnant female (?SEM) by treatment group, genotype, dose of ac-
rylamide, and mating period for each of the three studies. *, Significantly
less than control at P ? 0.05; †, significantly less than same dose CYP2E1-
null treatment group at P ? 0.05.
Percentage (%) of live fetuses out of the total implantation sites
sites per pregnant female (?SEM) by treatment group, genotype, dose of
acrylamide, and mating period for each of the three studies. *, Signifi-
cantly greater than control at P ? 0.05; †, significantly greater than same
dose CYP2E1-null treatment group at P ? 0.05.
Percentage (%) of resorption moles out of the total implantation
were significantly altered in females mated to CYP2E1-null
male mice treated with 12.5, 25, or 50 mg acrylamide kg?1
day?1for 5 days. Thus, the combined results from the three
independent studies that comprised this investigation clear-
ly demonstrated that exposure to acrylamide for 5 consec-
utive days, at doses up to 50 mg kg?1day?1, had no effect
on the fertility of CYP2E1-null male mice. In contrast, ac-
rylamide exposure induced marked reproductive effects in
wild-type male mice.
In study 1, only 3 of 60 females mated to wild-type
males treated with 50 mg/kg acrylamide for 5 days became
pregnant (Fig. 2); the data from studies 2 and 3 showed
that females mated to wild-type males treated with lower
doses of acrylamide, 12.5 or 25 mg kg?1day?1, did not
show significant reductions in pregnancies, compared with
the corresponding control groups (Fig. 2). The proportion
of females that became pregnant after cohabitation with ac-
rylamide-treated males was the least sensitive of the four
indicators of the reproductive toxicity of acrylamide.
Implantations per Female
Treatment of wild-type male mice with acrylamide re-
sulted in a reduction in the mean number of implantations
per pregnant female in some dose groups or mating periods
(Fig. 3). In study 1, the three pregnant females mated to
wild-type males treated with 50 mg/kg acrylamide had a
significantly reduced mean number of implantations (P ?
0.01) compared with the two wild-type control groups as
well as the acrylamide-treated CYP2E1-null group. In study
2, no effect on the number of implantations was noted in
females mated to acrylamide-treated wild-type males dur-
ing the first mating period; however, females mated to 12.5
or 25 mg/kg acrylamide-treated males during the second
mating period had significant reductions (P ? 0.01) in the
number of implantation sites compared with females mated
to the vehicle controls. These results indicate greater ac-
rylamide-induced damage to sperm exposed during the ear-
lier developmental stages (spermatids) sampled in the sec-
ond mating period (Fig. 1). The same pattern of reduced
implantation sites in females mated in the second mating
period to wild-type males exposed to acrylamide was seen
in study 3. Thus, acrylamide, at doses of 12.5–50 mg kg?1
day?1for 5 consecutive days, produced dose-related de-
creases in the mean number of implantation sites per preg-
nant female in the second mating period (Fig. 3).
Percentage Live Fetuses per Female
Acrylamide treatment reduced the percentage live fetus-
es per pregnant female (percentage of live fetuses out of
the total number of implantations) in particular dose groups
and mating periods (Fig. 4). The percentage live fetuses per
pregnant female was reduced from approximately 96% in
the control groups to 44% in the females mated to wild-
type males treated with 50 mg/kg acrylamide (P ? 0.01).
In study 2 (wild-type mice only), the reductions in per-
centage live fetuses were greater in matings that occurred
during the first period, compared with those that occurred
during the second mating period. This contrasts with the
data on implantation reduction, which showed greater ef-
fects in the second mating period. However, in our expe-
rience with dominant lethal tests with a variety of chemi-
cals, effects on implantation data tend to be more variable
than effects on resorption and live fetus data, and thus the
latter endpoints are more reliable indicators of the most
sensitive treatment window. In general, the results for mat-
ing periods 1 and 2 are consistent with each other and with
the literature and indicate critical sensitivity of condensed
spermatids and early epididymal spermatozoa.
ACRYLAMIDE GERM CELL EFFECTS ARE CYP2E1-DEPENDENT
percentage resorptions (Fig. 5). Error bars indicate SEM. Live fetuses are
replaced by resorption moles (embryonic death) as dose of acrylamide
increases. A similar pattern is seen in both mating periods, although the
effects are slightly greater in mating period 1, which included the entire
period of greatest sperm stage sensitivity.
Relationship between the percentage live fetuses (Fig. 4) and the
The clearest indicator of dominant lethality is the per-
centage resorptions (mean proportions of implants/female
that are resorptions) in females mated to treated males, and
this parameter was dramatically altered in our studies: dose-
related increases in percentage resorptions were seen in fe-
males mated to acrylamide-treated wild-type males (Fig. 5).
No significant increases in resorptions were observed in
females mated to CYP2E1-null males treated with 12.5–50
mg acrylamide kg?1day?1for 5 consecutive days in any
of the three studies (Fig. 5). The inverse relationship be-
tween percentage resorptions and percentage live fetuses in
the studies reported here clearly demonstrates that resorp-
tion replaced living fetuses in females mated to wild-type
mice treated with acrylamide in a dose-dependent manner
The contrasting observations of increased levels of dom-
inant lethal mutations in acrylamide-treated wild-type mice
compared with the lack of reproductive effects in acryl-
amide-treated CYP2E1-null mice demonstrate that in vivo
metabolism of acrylamide to the epoxide intermediate, gly-
cidamide, is a prerequisite for the induction of dominant
lethal mutations in postmeiotic germ cells of male mice.
The CYP2E1 pathway is the only pathway by which ac-
rylamide is oxidatively metabolized to glycidamide in wild-
type mice . In the absence of a functional CYP2E1
enzyme, metabolism of acrylamide in mice is directed pri-
marily through glutathione conjugation . Further con-
firmation of the requirement for a functional CYP2E1 en-
zyme in the metabolism of acrylamide to glycidamide
comes from studies in which we treated CYP2E1-null and
wild-type mice with acrylamide and used13C NMR spec-
troscopy to identify urinary metabolites (urine collected
over a 24-h period following a single i.p. dose of 50 mg
acrylamide/kg). We detected no glycidamide or metabolites
of glycidamide in the urine of CYP2E1-null mice, and all
the detected urinary metabolites of acrylamide in these
mice were derived from glutathione conjugation of the par-
ent molecule [28, 37]. In contrast, a significant portion of
acrylamide administered to wild-type mice was eliminated
in the urine as glycidamide and other metabolites derived
from this epoxide [28, 37]. Collectively, these results dem-
onstrate that negligible formation of glycidamide occurs in
CYP2E1-null mice treated with acrylamide.
Mature sperm through late spermatids are the spermi-
ogenic stages sensitive to dominant lethal induction by ac-
rylamide, with the period of greatest sensitivity being ap-
proximately 5–8 days after dosing [8, 12, 13, 38]. The mat-
ing schemes employed in this study overlapped in this pe-
riod of greatest sensitivity (Fig. 1). This may be why
comparative evaluations of some reproductive endpoints in
the first and second mating periods in acrylamide-treated
wild-type mice did not always reveal significant differenc-
es. The inverse relationship between percentage live fetuses
and percentage resorptions expected in a positive dominant
lethal assay was clearly apparent in the data from both the
Week 1 and Week 2 matings in study 3 (Fig. 6). Both of
these critical endpoints were more affected in the Week 1
matings that spanned the entire period of greatest sensitiv-
ity, consistent with previously published data [8, 27].
The present studies support the hypothesis that the ep-
oxide intermediate of acrylamide, glycidamide, is the ulti-
mate germ cell mutagen and may act by binding to nucle-
ophilic sites in chromatin of late spermatids and early sper-
matozoa. Our experiments do not identify the location of
the binding site, however. Whether glycidamide binds to
the sulfhydryl groups of cysteine-rich protamines, to DNA,
or both remains a question to be answered. The parent com-
pound acrylamide should theoretically be able to directly
alkylate DNA through Michael-type reactions, by virtue of
the double bond linking the alpha and beta carbons ,
without a requirement for metabolism to the glycidamide
intermediate. However, measurements of acrylamide ad-
ducts in sperm DNA and sperm protamine showed acryl-
amide to be only weakly effective in forming adducts in
sperm DNA in vivo but quite efficient at alkylating cyste-
ine-rich protamine in vitro . Extending this observation,
Sega  showed that the total amount of sperm head al-
kylation measured in mice after acrylamide exposure close-
ly matched the amount of sperm protamine alkylation; al-
kylation of DNA after acrylamide dosing accounted for less
than 5% of the total sperm head alkylation. Together, these
adduct data strongly indicate a significant role for prot-
amine alkylation in the induction of dominant lethal mu-
tations in mice by acrylamide.
Acrylamide-induced genetic alterations in sperm might
indeed arise through a mechanism that doesn’t involve di-
rect alkylation of DNA. Sega  suggested that DNA
breakage in late spermatids and early spermatozoa could be
induced by protamine alkylation and consequent chromatin
strand distortion. Because these later spermiogenic stages
are no longer DNA repair competent, any DNA damage
induced at these stages would persist. It is during this pe-
riod of spermatogenesis that histones are replaced by sim-
pler, cysteine-rich protamines and the increasingly com-
pacted DNA becomes rapidly inaccessible to interaction
with chemicals. Thus, because acrylamide is most effective
at inducing genetic damage in late spermatids and early
spermatozoa, it appears more likely that it acts via forma-
tion of protamine adducts than through direct DNA alkyl-
On the other hand, although acrylamide is a weak al-
GHANAYEM ET AL.
kylator of DNA in vivo , metabolism of acrylamide to
glycidamide may enhance alkylation of the relatively weak-
er nucleophilic sites in DNA. Wang et al.  have shown
that acrylonitrile, another CYP2E1 substrate, can alkylate
the sulfhydryl groups of cysteine but only its epoxide me-
tabolite, glycidonitrile, can react with nucleic acids. In ex-
periments in our laboratory, we detected a significant in-
crease in glycidamide-derived DNA adducts in whole testes
from wild-type mice treated with acrylamide, confirming
that glycidamide can form DNA adducts in mixed-cell tes-
ticular tissue of mice; negligible levels of these DNA ad-
ducts were detected in CYP2E1-null mice exposed to ac-
rylamide (unpublished data). It is important to note that the
testicular tissue we examined was from a mix of cell types,
not just sperm. In addition, recent preliminary results from
our laboratory showed that, while high levels of acrylamide
were detectable in plasma of treated CYP2E1-null mice,
negligible levels were found in the plasma of wild-type
mice, suggesting that distribution and persistence of the
parent molecule occurs (unpublished data). However, be-
cause we saw no induction of dominant lethal mutations in
CYP2E1-null mice in the current study, it appears that even
if acrylamide does alkylate DNA or protamines directly,
this action is ineffective in inducing detectable germ cell
To our knowledge, there has been no pharmacokinetic
studies in CYP2E1-null mice designed to measure the dis-
tribution of unmetabolized acrylamide. The earlier studies
of Marlowe et al.  in wild-type mice showed radiola-
beled acrylamide moving from the site of administration to
the testes and eventually to the glans penis in a fashion
comparable with movement of sperm, but the radiolabel
may have represented glycidamide rather than acrylamide;
this experiment was not designed to distinguish the parent
compound from the metabolites. Thus, the degree of direct
testicular exposure to acrylamide as well as acrylamide’s
ability to alkylate protamine in vivo remains unresolved at
this time. However, results from our comparative dominant
lethal study using wild-type and CYP2E1-null mice are not
contradictory to the hypothesis  that acrylamide is me-
tabolized in the liver to glycidamide and that, although
some metabolite might bind to liver DNA, some is trans-
ported to the testes, where it binds to accessible sperm
DNA or, more likely, to sperm protamines, resulting in
chromosomal damage that leads to death of early postim-
plantation conceptuses. It should be emphasized that
CYP2E1 is present in many tissues, including the testes [40,
41], and that in situ metabolism of acrylamide to glycidam-
ide in the testis may occur.
The confirmed presence of acrylamide in common car-
bohydrate foodstuffs, cooked at high temperatures, implies
potential widespread exposure of the general population to
acrylamide via the consumption of acrylamide-containing
foods [19, 20, 42]. Estimates of human exposure range
from 0.3 to 100 ?g acrylamide (kg body weight)?1day?1,
as compiled from three large population studies in Europe
and the United States [19, 20, http://europa.eu.int/comm/
food/fs/sc/scf/out131 en.pdf]. The U.S. National Toxicolo-
gy Program’s Center for the Evaluation of Risks to Human
Reproduction recently issued an expert panel report assess-
ing acrylamide exposure in humans and the potential for
acrylamide-induced genetic damage in germ cells [http://
cerhr.niehs.nih.gov/news/acrylamide/final report.pdf]. This
report provides details on exposure levels in the general
population as well as in occupationally exposed groups and
smokers. Reviewing all the available data, panel members
agreed that the general nonsmoking population is exposed
to approximately 0.5–1.0 ?g acrylamide (kg body
weight)?1day?1; estimated exposures in children 2–5 yr of
age were 2–3 times the adult levels when expressed as a
body weight ratio [page 6 of the report]. Although these
estimated typical human exposure levels are markedly low-
er than the levels in treated laboratory animals, the panel
cautioned that dose-response information for heritable ef-
fects in humans or animals is limited [page 151 of the re-
port]. Additional evidence of human exposure comes from
the identification of hemoglobin adducts of acrylamide in
workers in surfactants production and the textile industry
. Most recently, glycidamide was identified in the urine
of humans exposed to low levels of acrylamide , in-
dicating that humans are capable of metabolizing acryl-
amide, presumably via CYP2E1, to glycidamide. Because
our animal data and data from human studies  both
implicate a critical role for CYP2E1 enzymes in conversion
of acrylamide to the mutagenically active epoxide inter-
mediate, glycidamide, a consideration of polymorphic en-
zyme variants in humans is integral to the determination of
human risk of germ cell damage from low-dose chronic
acrylamide exposure. Variable metabolic capacities linked
to genetic polymorphisms in the CYP2E1 gene produce
differences in the ability to metabolize a number of drugs,
environmental pollutants, and other agents, and thus indi-
vidual risks vary [45–48].
In conclusion, the current work confirmed that acryl-
amide is a potent inducer of dominant lethal mutations in
male mice and that this effect is directly related to dose.
Furthermore, our results provide the first direct demonstra-
tion that this dominant lethal effect of acrylamide is de-
pendent on acrylamide epoxidation to glycidamide by
The authors are grateful to Sue Edelstein of Image Associates for her
patience and expert assistance in preparing the figures for this manuscript.
In addition, we extend our sincere thanks to Drs. Paul Foster and Melissa
Rhodes, NIEHS, and Dr. Sally Darney, EPA, for thorough and thoughtful
reviews of this manuscript.
1. Dearfield K, Abernathy C, Ottley M, Brantner J, Hayes P. Acrylamide:
its metabolism, developmental and reproductive effects, genotoxicity,
and carcinogenicity. Mutat Res 1988; 195:45–77.
2. International Agency for Research on Cancer. Some Industrial Chem-
icals. In: IARC Monographs on the Evaluation of Carcinogen Risks
to Humans, vol. 60. Lyon, France: International Agency for Research
on Cancer; 1994:389–433.
3. Adler I-D, Ingwersen I, Kliesch U, Tarras AE. Clastogenic effects of
acrylamide in mouse bone marrow cells. Mutat Res 1988; 301:249–
4. Knaap A, Kramers P, Voogd C, Bergkamp W, Groot M, Langebroek
P, Mout H, van der Stel J, Verharen H. Mutagenic activity of acryl-
amide in eukaryotic systems but not in bacteria. Mutagenesis 1988;
5. Kligerman A, Atwater A, Bryant M, Erexson G, Kwanyuen P, Dear-
field K. Cytogenetic studies of ethyl acrylate using C57BL/6 mice.
Mutagenesis 1991; 6:137–151.
6. Lahdetie J, Suutari A, Sjoblom T. The spermatid micronucleus test
with the dissection technique detects the germ cell mutagenicity of
acrylamide in rat meiotic cells. Mutat Res 1994; 309:255–262.
7. Dearfield K, Douglas G, Ehling U, Moore M, Sega G, Brusick D.
Acrylamide: a review of its genotoxicity and an assessment of heri-
table genetic risk. Mutat Res 1995; 330:71–99.
8. Shelby M, Cain K, Hughes L, Braden P, Generoso W. Dominant lethal
effects of acrylamide in male mice. Mutat Res 1986; 173:35–40.
9. Ehling U, Neuhauser-Klaus A. Reevaluation of the induction of spe-
163 Download full-text
ACRYLAMIDE GERM CELL EFFECTS ARE CYP2E1-DEPENDENT
cific-locus mutations in spermatogonia of the mouse by acrylamide.
Mutat Res 1992; 283:185–191.
10. Gutierrez-Espeleta G, Hughes L, Piegorsch W, Shelby M, Generoso
W. Acrylamide: dermal exposure produces genetic damage in male
mouse germ cells. Fundam Appl Toxicol 1992; 18:189–192.
11. Chapin R, Fail P, George J, Grizzle T, Heindel J, Harry G, Collins B,
Teague J. The reproductive and neural toxicities of acrylamide and
three analogues in Swiss mice, evaluated using the continuous breed-
ing protocol. Fundam Appl Toxicol 1995; 27:9–24.
12. Marchetti F, Lowe X, Bishop J, Wyrobek AJ. Induction of chromo-
somal aberrations in mouse zygotes by acrylamide treatment of male
germ cells and their correlation with dominant lethality and heritable
translocations. Environ Mol Mutagen 1997; 30:410–417.
13. Shelby M, Cain K, Cornett C, Generoso W. Acrylamide: induction of
heritable translocations in male mice. Environ Mol Mutagen 1987; 9:
14. Adler I-D, Reitmer P, Schmoller R, Schriever-Schwemmer G. Dose-
response for heritable translocations induced by acrylamide in sper-
matids of mice. Mutat Res 1994; 309:285–291.
15. Russell L, Hunsicker P, Cacheiro N, Generoso W. Induction of spe-
cific-locus mutations in male germ cells of the mouse by acrylamide
monomer. Mutat Res 1991; 262:101–107.
16. Rosen J, Hellenas K. Analysis of acrylamide in cooked foods by liq-
uid chromatography tandem mass spectrometry. Analyst 2002; 127:
17. Tareke E, Rydberg P, Karlsson P, Eriksson S, Tornqvist M. Analysis
of acrylamide, a carcinogen formed in heated foodstuffs. J Agric Food
Chem 2002; 50:4998–5006.
18. Weiss G. Cancer risks. Acrylamide in food: uncharted territory. Sci-
ence 2002; 297:27.
19. FAO/WHO. Health implications of acrylamide in food. In: Geneva,
Switzerland: FAO/WHO; 2002:http://www.who.int/fsf/Acrylamide/
20. DiNovi M, Howard D. The updated exposure assessment for acryl-
amide. In: Acrylamide in Food: Update—Scientific Issues, Uncertain-
ties, and Research Strategies. FDA/CFSAN; 2004.
21. Segerback D, Calleman C, Schroeder J, Costa L, Faustman E. For-
mation of N-7-(2-carbamoyl-2-hydroxyethyl)guanine in DNA of the
mouse and the rat following intraperitoneal administration of
[14C]acrylamide. Carcinogenesis 1995; 16:1161–1165.
22. Marlowe C, Clark M, Mast R, Friedman M, Waddell W. The distri-
bution of [14C]acrylamide in male and pregnant Swiss-Webster mice
studied by whole-body autoradiography. Toxicol Appl Pharmacol
23. Sega G, Alcota R, Tancongco C, Brimer P. Acrylamide binding to the
DNA and protamine of spermiogenic stages in the mouse and its re-
lationship to genetic damage. Mutat Res 1989; 216:221–230.
24. Sega G, Generoso E. Measurement of DNA breakage in specific germ-
cell stages of male mice exposed to acrylamide, using an alkaline-
elution procedure. Mutat Res 1990; 242:79–87.
25. Sega G, Generoso E, Brimer P. Acrylamide exposure induces a de-
layed unscheduled DNA synthesis in germ cells of male mice that is
correlated with the temporal pattern of adduct formation in testis
DNA. Environ Mol Mutagen 1990; 16:137–142.
26. Sega G. Adducts in sperm protamine and DNA vs. mutation frequen-
cy. Prog Clin Biol Res 1991; 372:521–530.
27. Generoso W, Sega G, Lockhart A, Hughes L, Cain K, Cacheiro N,
Shelby M. Dominant lethal mutations, heritable translocations, and
unscheduled DNA synthesis induced in male mouse germ cells by
glycidamide, a metabolite of acrylamide. Mutat Res 1996; 371:175–
28. Sumner SCJ, Fennell T, Moore T, Chanas B, Gonzalez F, Ghanayem
B. Role of cytochrome P450 2E1 in the metabolism of acrylamide
and acrylonitrile in mice. Chem Res Toxicol 1999; 12:1110–1116.
29. Lee S, Buters J, Pineau T, Fernadex-Salguero P, Gonzalez F. Role of
CYP2E1 in the hepatotoxicity of acetaminophen. J Biol Chem 1996;
30. Hoffler U, El-Masri H, Ghanayem B. Cytochrome P450 2E1
(CYP2E1) is the principal enzyme responsible for urethane metabo-
lism: comparative studies using CY2E1-null and wild-type mice. J
Pharmacol Exper Therapeut 2003; 305:557–564.
31. Wang H, Chanas B, Ghanayem B. Cytochrome P450 2E1 (CYP2E1)
is essential for acrylonitrile metabolism to cyanide: comparative stud-
ies using CYP2E1-null and wild-type mice. Drug Metab Dispos 2002;
32. Lockhart A-M, Bishop JB, Piegorsch WW. Issues regarding data ac-
quisition and analysis in the dominant lethal assay. Proc Am Statist
Assoc 1991; Biopharm Sec: 234–237.
33. Lockhart A, Piegorsch W, Bishop J. Assessing overdispersion and
dose-response in the male dominant lethal assay. Mutat Res 1992;
34. Snedecor G, Cochran W. Statistical Methods. Ames, IA: The Iowa
State University Press; 1967.
35. Jonckheere A. A distribution-free k-sample test against ordered alter-
natives. Biometrika 1954; 41:133–145.
36. Fleiss J. Statistical Methods for Rates and Proportions. New York:
John Wiley and Sons; 1973.
37. Ghanayem BI, Wang H, Sumner SCJ. Using cytochrome P450 gene
knock-out mice to study chemical metabolism, toxicity, and carcino-
genicity. Toxicol Pathol 2000; 28:839–850.
38. Adler I-D, Baumgartner A, Gonda H, Friedman M, Skerhut M. 1-
Aminobenzotriazole inhibits acrylamide-induced dominant lethal ef-
fects in spermatids of male mice. Mutagenesis 2000; 15:133–136.
39. MacWilliams J. Acrylamide. In: M Grayson DE (ed.), Kirk-Othmer/
Encyclopedia of Chemical Technology, vol. 1, 3rd ed. New York: John
Wiley and Sons; 1978: 298–311.
40. Forkert P-G, Lash L, Nadeau V, Tardif R, Simmonds A. Metabolism
and toxicity of trichloroethylene in epididymis and testis. Toxcol Appl
Pharmacol 2002; 182:244–254.
41. Healy L, Pluta L, Recio L. Expression and distribution of cytochrome
P450 2E1 in B6C3F1 mouse liver and testes. Chem Biol Interact
42. Friedman M. Chemistry, biochemistry, and safety of acrylamide. A
review. J Agric Food Chem 2003; 51:4504–4526.
43. Schettgen T, Broding HC, Angerer J, Drexler H. Hemoglobin adducts
of ethylene oxide, propylene oxide, acrylonitrile and acrylamide-bio-
markers in occupational and environmental medicine. Toxicol Lett
44. Fennell TR, Snyder R, Burgess JP, Friedman MA. Metabolism and
hemoglobin adducts of [1,2,3–13] acrylamide in humans. In: Program
of the 43rd Annual Meeting of the Society of Toxicology; 2004; Bal-
timore, MD. Abstract 845.
45. Liu T, Chung Y, Wang P, Chi C, Hsieh L. Safrole-DNA adducts in
human peripheral blood—an association with areca quid chewing and
CYP2E1 polymorphisms. Mutat Res 2004; 559:59–66.
46. Suzuki S, Muroishi Y, Nakanishi I, Oda Y. Relationship between ge-
netic polymorphisms of drug-metabolizing enzymes (CYP1A1,
CYP2E1, GSTM1, and NAT2), drinking habits, histological subtypes,
and p53 gene point mutations in Japanese patients with gastric cancer.
J Gastroenterol 2004; 39:220–230.
47. Canalle R, Burim R, Tone L, Takahashi C. Genetic polymorphisms
and susceptibility to childhood acute lymphoblastic leukemia. Environ
Mol Mutagen 2004; 43:100–109.
48. Ishikawa H, Yamamoto H, Tian Y, Kawano M, Yamauchi T, Yokoy-
ama K. Evidence of association of the CYP2E1 genetic polymorphism
with micronuclei frequency in human peripheral blood. Mutat Res