Screening for ENU-Induced Mutations in Mice That Result in Aberrant
Kristin M. Hamre, Daniel Goldowitz, and
University of Tennessee Health Science Center
Douglas B. Matthews
University of Memphis
One way to investigate the genetic underpinnings of ethanol-related phenotypes is to create random
mutations and screen the mutagenized mice for their behavioral phenotypes. The purposes of this article
are to assess the efficacy of a novel high throughput screen to detect known strain differences and to
provide evidence of the ability of this screen to detect phenodeviants, as illustrated by two new lines of
mutant mice. All mice were tested for the following phenotypes after a dose of 2.25 g/kg of ethanol:
ataxia, anxiolytic response, locomotor activity, core body temperature, and blood ethanol concentration,
as well as ethanol consumption based on a two-bottle choice test. The authors obtained several baseline
measures that allowed for the detection of phenodeviants on these measures as well. To validate this
screen, A/J, DBA/2J, and C57BL/6J mouse strains were tested, and previously reported strain differences
were found in all phenotypes except ethanol-induced hypothermia. Additionally, two mutant pedigrees
were identified: 7TNJ, which exhibited abnormal ethanol-induced locomotor activity, and 112TNR,
which exhibited an enhanced ability on the rotarod. These data demonstrate the efficacy of this screen
to detect known as well as novel phenotypic differences.
Keywords: locomotor activation, anxiolytic response, ataxia, ethanol consumption, N-ethyl nitrosourea
The genetic influences on ethanol consumption and ethanol-
induced behaviors or responses have been widely reported in both
human and animal literature (see Dick & Foroud, 2003; Enoch &
Goldman, 2001; Foroud & Li, 1999, for reviews). This work has
translated into a significant effort to identify the genes that under-
lie phenotypic differences in alcohol consumption or responses.
Although various candidate genes have been hypothesized to me-
diate these differences (for examples see Fehr, Shirley, Belknap,
Crabbe, & Buck, 2002; Pandey, Carr, Heilig, Ilveskoski, & Thiele,
2003; Thiele & Badia-Elder, 2003), only one has been unambig-
uously identified (Fehr et al., 2002; Shirley, Walter, Reilly, Fehr,
& Buck, 2004).
A variety of approaches have been used in an attempt to identify
genes that affect ethanol consumption and related phenotypes.
Analyses of knockout or transgenic animals have examined spe-
cific candidate genes to determine whether they play critical roles
in ethanol consumption or ethanol-mediated responses. This gene-
driven approach has typically focused on neurotransmitters and
their receptors (e.g., Connor, Young, Lawford, Ritchie, & Noble,
2002; Fehr, Rademacher, & Buck, 2003; Hallikainen et al., 2003;
Pandey et al., 2003) or on ethanol-metabolizing enzymes (see
Eriksson et al., 2001, for review). A limitation of this approach is
its focus on specific genes, making it therefore initially biased in
nature in that it ignores the discovery of new genes or the study of
various other potentially crucial genes, such as transcription fac-
tors. Another approach is the use of high throughput analyses of
gene products, that is, transcriptomic or proteomic approaches, but
such studies are strictly correlative in nature. An alternative
phenotype-driven approach, for example, the analysis of quantita-
tive trait loci (QTL), has the advantage of being unbiased. In this
approach, the investigators work from the phenotype back to the
gene rather than determining whether or not a particular gene
influences a phenotype. This approach has been widely and suc-
cessfully used to identify chromosomal regions that contribute to
ethanol-related behaviors and ethanol consumption (see Crabbe,
Phillips, Buck, Cunningham, & Belknap, 1999, for review). The
drawback to this approach has been that the conclusive identifica-
tion of the responsible gene (or genes) within a chromosomal
region has been an arduous task, and thus only a small number
have currently been identified (Fehr et al., 2002; Shirley et al.,
Recently another approach, chemically induced mutagenesis,
which can ultimately lead to the identification of genes responsible
for ethanol consumption or ethanol-mediated responses, has begun
Kristin M. Hamre, Daniel Goldowitz, and Sarah Wilkinson, Department
of Anatomy and Neurobiology, University of Tennessee Health Science
Center; Douglas B. Matthews, Department of Psychology, University of
This article was supported by National Institute of Mental Health Grants
MH61971 (to Daniel Goldowitz, Kristin M. Hamre, and Douglas B.
Matthews), R25 MH-066890 (to Daniel Goldowitz), U01-AA-13503 (to
Daniel Goldowitz), AA014588 (to Douglas B. Matthews), and AA13509
(to Douglas B. Matthews).
We thank Kevin Wilkinson for technical support; John Crabbe for input
and advice on the design of this screen; Gene Rinchik for assistance in the
design of this project; and Jay Snoddy, Elissa Chesler, Eric Baker, and
Leslie Galloway for statistical analyses through Mutrack. We also thank
Thomas Albright for assistance with conducting the analyses and the
Kinsey Institute for Research in Sex, Gender, and Reproduction for allow-
ing access to these data.
Correspondence concerning this article should be addressed to Kristin
M. Hamre, Department of Anatomy and Neurobiology, University of
Tennessee Health Science Center, 855 Monroe Avenue, Memphis, TN
38163. E-mail: email@example.com
2007, Vol. 121, No. 4, 665–678
Copyright 2007 by the American Psychological Association
to be used. With the help of the chemical mutagen N-ethyl nitro-
sourea (ENU), single base pair mutations are made randomly
throughout the genome (Balling, 2001; Brown & Balling, 2001;
Goldowitz et al., 2004; Justice et al., 2000; Nolan et al., 2000).
Ultimately, through breeding of offspring from mutagenized par-
ents, lines of mice are created that are homozygous for one or more
mutations. Whether a mutant line is of interest is determined by the
screens that the mutagenized offspring undergo. This technique is
also an unbiased, phenotype-driven approach as in QTL analysis
but additionally yields single-gene mutations that can, in theory, be
more readily identified and ultimately provide proof of the impor-
tance of a given gene in a phenotypic outcome. Going from mutant
DNA to an identified gene has become a much easier process with
the completed sequence of the mouse genome, availability of a
whole genome readout with microarrays, and bioinformatic tools
that can help in identifying gene candidates (e.g., Zhang, Kirov, &
Snoddy, 2005). A notable success in the production of interesting
ENU-induced mutants has been the identification of genes impor-
tant to mammalian circadian rhythm (Kapfhamer et al., 2002;
Vitaterna et al., 1994). However, many mutations caused no phe-
notypic outliers on the phenotypic screens that were administered,
necessitating the evaluation of many lines of mice with rapid
screening. An additional complication is the robustness of the
screen, as measured by its sensitivity to detect outliers and its
reproducibility across mice of the same genotype. If the statistical
decision to identify outliers is lax, then deficiencies in robustness
can lead to many false positives, thereby resulting in a significant
waste of resources (technician time and mice). On the other hand,
if the statistical cut-off for identifying phenodeviants is stringent,
one could lose mutants (false negatives).
The rapid testing of mutagenized, and potentially knockout,
mice has two general requirements. First, each mouse is tested for
multiple measures to maximize the ability to detect the potential
effect of a given gene on as many phenotypes as experimentally
feasible. Second, to most efficiently screen large numbers of mice,
the screening parameters need to be, at times, condensed. For
example, ataxia measures are evaluated in a single day of baseline
testing followed by a single day of ethanol exposure. A recent
series of seminal papers have emphasized the potential impact that
environmental manipulations can exert on behavioral measures
(e.g., Crabbe, Wahlsten, & Dudek, 1999; Wahlsten et al., 2003)
and particularly the effect of prior testing on subsequent testing in
alcohol-related responses (Camarini & Hodge, 2004). Thus, it is
imperative to determine whether the testing of multiple measures,
as well as shortened screening parameters, obscures the ability to
detect differences and to obtain values consistent with those pre-
viously reported in the literature.
Thus, we developed a novel high-throughput screen to evaluate
ethanol-mediated phenotypes as well as ethanol consumption.
There were three requirements for each phenotype examined, with
the exception of ethanol consumption. First, each phenotype as-
sessed a different ethanol-induced response (e.g., changes in core
body temperature vs. locomotor activation), although there was
clearly some overlap (e.g., both the elevated plus maze and the
activity chamber measured locomotor activity). Use of multiple
tests allowed for the broadest range of phenotypes to be examined.
Second, phenotypes were assessed at similar doses of ethanol to
maximize the efficiency of the screening process. Third, this
screen is part of a larger effort (in this case, the Tennessee Mouse
Genome Consortium [TMGC]) to examine mutagenized mice. The
tests in the ethanol screen were chosen to be similar to screens
used in other parts of the TMGC screen. The use of similar tests by
different screening facilities provides the potential for testing at
different sites within TMGC as well as for comparison across
One of the goals of a screen is to be able to accurately detect a
mutant mouse without missing potential mutants (false negatives),
as well as spuriously misidentifying mice as mutants (false posi-
tives). Testing parameters have repeatedly been shown to influence
this ability, and thus we included the analysis of strains with
known differences to determine which phenotypes could be accu-
rately analyzed in a screening paradigm. However, in all instances,
any mutant identified needs to be more fully examined to ensure
that the phenotype is replicable and consistent. To demonstrate this
subsequent testing, we discuss the identification of a mutant ped-
igree, 7TNJ, and then its further testing for confirmation of the
There are two purposes to this report. The first purpose is to
describe this novel high-throughput screen and test its efficacy
for use in the analysis of ethanol consumption and ethanol
responses in mutagenized mice. Evaluation of the screen is
accomplished by testing standard inbred strains of mice and
comparing the results with previously published data to deter-
mine whether previously reported strain differences are found.
If these data are consistent with data determined elsewhere,
then it demonstrates that the testing parameters are adequate for
detecting differences among mutagenized mice. The ethanol
responses include motor impairment on an accelerating rotarod,
anxiolytic response in an elevated plus maze, locomotor acti-
vation in an activity chamber, changes in body temperature,
measurement of blood ethanol concentration (BEC), and etha-
nol self-administration in a two-bottle choice test. In this initial
study, three standard inbred strains were tested in this screen.
The three strains—C57BL/6J (hereafter called B6), DBA/2J
(hereafter called D2), and A/J—were chosen because they have
been shown to exhibit differences in several of the traits being
assessed in this screen (Belknap, Crabbe, & Young, 1993;
Crabbe, Belknap, Mitchell, & Crawshaw, 1994, Crabbe, Galla-
her, Phillips, & Belknap, 1994; Crabbe, Johnson, Gray, Ko-
sobud, & Young, 1982; Crabbe, Metten, Cameron, & Wahlsten,
2005; Crawshaw, Wallace, Christensen, & Crabbe, 2001; Craw-
shaw et al., 1997; Demarest, McCaughran, Mahjubi, Cipp, &
Hitzemann, 1999; Dudek & Tritto, 1994; Koyner, Demarest,
McCaughran, Cipp, & Hitzemann, 2000; Melo, Shendure, Po-
ciask, & Silver, 1996; Moore & Kakihana, 1978; O’Connor,
Crawshaw, Kosobud, Bedichek, & Crabbe, 1989; Peirce, Derr,
Shendure, Kolata, & Silver, 1998; Rustay, Wahlsten, & Crabbe,
The second purpose is to describe two pedigrees of mice that
were found to exhibit abnormal phenotypes in this screen. One of
the mutant pedigrees, called 7TNJ, exhibited enhanced ethanol-
induced locomotor activity. This phenotype is specific to the
screen because no other abnormalities were detected in this pedi-
gree. The second pedigree exhibited an enhanced ability to navi-
gate the rotarod, which is observed both at baseline and after
ethanol exposure. This second pedigree highlights one of the
advantages of this screen, in that abnormalities can be observed
both in baseline behaviors and following ethanol exposure.
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
Mice were individually housed in cages and allowed a minimum
of 1 week to acclimate to the testing environment. Mice were
maintained on a 12-hr light–dark cycle and given food and water
ad libitum. All mice were tested following the Animal Care and
Use guidelines of the University of Memphis, which follow the
guidelines of the National Institutes of Health (1986). Mice were
housed and tested within the same large room (30 ft ? 30 ft), and
the housing area and testing areas were separated by over 20 ft.
Mice were tested between 8 a.m. and 3 p.m. All mice were tested
at 60–70 days of age.
As mentioned above, to assess the ability of this screen to
replicate previous data, we examined the B6, D2, and A/J strains
of mice in this study. The following numbers of mice were used:
10 male and 10 female D2 mice were tested for all screens, the
same numbers of A/J mice were tested except that samples for
BEC from 3 females were lost, and 9 male and 9 female B6 mice
were tested for all screens. Although a rare occurrence, data from
a particular mouse were excluded for a particular test if there were
any equipment malfunctions or if the mouse failed to perform the
test (e.g., if the mouse jumped off the elevated plus maze). All
three strains of mice were obtained from Jackson Laboratory (Bar
The ENU-mutagenesis strategy employed for the generation of
mutagenized mice has been previously described (Goldowitz et al.,
2004; Rinchik & Carpenter, 1999). In this strategy, pedigrees were
derived from F1 males from mutagenized males and females
carrying a chromosome inversion that “marks” a specific region of
the genome. In this mutagenesis screen we “marked” the targeted
chromosome by using a chromosome inversion line to mate to F1
mice derived from mutagenized F0 mice mated to another inbred
line (typically, B6). The inversion contained a marker that was
visible (e.g., hairy ears) in heterozygous mice. In the F2 progeny,
there was an allele from the inversion stock and from the mu-
tagenized line. The inversion inhibited crossovers in future gener-
ations so that subsequent breeding proceeded to produce homozy-
gous mutants and heterozygotes maintained across the inversion.
The physical marker moved with the inversion and allowed for the
identification of heterozygous mice. When the inversion was ho-
mozygous, lethality followed and this genetic combination was
therefore not present. In contrast, mutant mice possessed two
copies of the chromosomal region from the mutagenized progen-
itor and thus appeared wild-type for the inversion. A selective
breeding scheme was followed that allowed for the generation and
identification of genetically homozygous test class mutant mice by
the third generation (i.e., G3) of breeding. Subsequent generations
carried the abnormal recessive mutation generated by the ENU-
mutagen. Additional information on this breeding scheme can be
found on the TMGC Web site at www.tnmouse.org/neuro-
mutagenesis/mutagenesis.html. Homozygous, recessive mutants
were identified by a physical or molecular marker. Mutants were
compared with nonmutagenized mice processed in the same breed-
ing scheme or other potential mutants on the same background.
The background for the two mutants described in this article is
C3H-C57BL/6J mixed background for the 7TNJ and C3.BliA-
C57BL/6J mixed background for the 112TNR.
Definition of Terms
As described above, each offspring from a mutagenized male
carried a different mutation. The resulting inbred descendents from
that offspring were defined as a pedigree. All mice from the
original mutagenized male as well as from other mutagenized
males of the same background were grouped together as a family.
Any mouse that met the criterion defined below for an outlier was
defined as a mutant or its synonymous term, phenodeviant.
Parameters and Sequence of the Screen
Each mouse underwent the complete screen in the sequence
described below and listed in Table 1. The screen was divided into
two sections: one that measured responses of each mouse to a
single bolus of ethanol and one that measured ethanol consump-
tion. The first part occurred during the 1st week of testing, and the
second part occurred during the 2nd week of testing.
Responses to an acute low dose of ethanol.
screening measured the responses to a single bolus of ethanol.
Each testing session had groups of 3–6 mice. In this article, we
first describe the sequence of the screen, then each testing appa-
ratus, and finally the measures recorded from each test.
The 1st day of screening, baseline scores on the accelerating
rotarod were obtained. Each mouse was given 10 consecutive trials
on the rotarod and then returned to the home cage. During the 2nd
day of testing, the response of each mouse to ethanol was mea-
sured on a variety of behaviors. The mice were transported to the
testing area, weighed, and allowed 20 min to become acclimated.
Subsequently, each mouse was given an intraperitoneal injection
of 2.25 g/kg of ethanol (20% v/v in saline). Then 10 min after the
ethanol injection, mice were tested for the anxiolytic response in
an elevated plus maze. Mice were given 5 min to explore the
elevated plus maze. Then each mouse was moved immediately into
an activity chamber, and the locomotor activity was measured for
the next 10 min. After the activity test, mice were given three
consecutive tests on the accelerating rotarod. Each mouse was then
observed and scored with the Majchrowicz scale (Majchrowicz,
1975; Majchrowicz & Hunt, 1976) and subsequently allowed to sit
in its home cage. At 60 min after the ethanol injection, the mice’s
core body temperature was measured, and blood was taken to
measure their BEC. After the blood collection, testing was com-
pleted for the day, and mice were returned to the cage rack. On the
The initial week of
Sequence of the Screen
Length of test
Elevated plus maze
Core body temperature
Two-bottle choice (3%)
Two-bottle choice (10%)
aTime (in min) after injection of a single dose of ethanol at 2.25 g/kg.
BEC ? blood ethanol concentration.
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
3rd day, the baseline temperature was recorded. This temperature
was recorded at the same time of day as the ethanol-induced
temperature was measured on the previous day. Mice were then
left undisturbed until the Friday of the 1st testing week.
The accelerating rotarod was from Columbus
Instruments (Columbus, OH). Mice were placed on the rotarod
when it was slowly rotating at the initial speed of 5 rpm. The
rotarod accelerated 25 revolutions during each minute of the test.
The dowel was 3 in. in diameter and was covered with 220-grit
sandpaper to minimize the ability of the mouse to “cling” to the
rod as it rotated. Each mouse was allowed to remain on the rod in
each trial until it fell off. On this test, the length of time the mouse
was able to stay on the rod and the speed of rotation when the
mouse fell were both recorded. For the baseline score, 10 trials
were administered, whereas only 3 trials were given after the
ethanol exposure. After each trial a short (approximately 1 min)
rest period was provided. For the analysis of baseline score, only
the last 3 of the 10 trials were used to figure the mean. For the
analysis of ethanol-induced ataxia, all 3 of the trials were aver-
aged. The mean ethanol score was subtracted from the mean
baseline score to determine a difference score.
Elevated plus maze.
The elevated plus maze was approxi-
mately 1 ft above the floor and contained four 51-cm-long ?
11.5-cm-wide arms arranged at right angles (Columbus Instru-
ments). The closed arms had opaque walls 30 cm high, extending
the length of the arm. The maze was constructed from Plexiglas.
Each arm of the plus maze contained two photobeams, and the
center of the maze was equipped with a single photobeam; pho-
tobeam signals were fed into a personal computer for automated
monitoring. Each maze was located inside a cubicle that limited
the view of subjects to the surroundings or to other mice under-
going behavioral testing. At the time of testing, each animal was
placed in the center of the maze facing an open arm. Several
measures were collected: the amount of time spent in the (a) open
arm, (b) closed arm, and (c) middle and the number of entries into
(d) open arms and (e) closed arms. Also calculated were the
percentage of entries into the (a) open arms and (b) closed arms
and the percentage of spent time in the (c) open arms and (d)
Locomotor activity was measured in a 17 in.
? 17 in. activity chamber from Med Associates (St. Albans, VT)
interfaced with a Celeron computer. Three 16-beam arrays tracked
the movement of the mouse. In this 10-min test, the distance
traveled was measured in two 5-min time intervals. Analyses were
conducted on both the total activity for the entire 10-min test and
on the activity in each block of 5 min. However, more emphasis
was placed on the evaluation of locomotor activity during the
second 5-min block because mice were typically more active
during the first few minutes as they habituated to the novel
Core body temperature and BEC.
measured with a digital thermometer from Physitemp (Clifton, NJ)
with a probe with a 1/8–in.-diameter tip that was appropriate for
use with mice. The baseline temperature measurements of the mice
were obtained 24 hr after the ethanol temperature measurement.
Consistency in the timing of measuring temperature was necessary
due to circadian fluctuations (Crawshaw et al., 1997, 2001). Im-
mediately after the temperature reading, tail blood was taken into
specialized capillary tubes supplied by Analox Instruments
Core body temperature was
(Lunenberg, MA), and an Analox alcohol analyzer was used to
determine the BEC of the blood.
Ethanol consumption as measured with the two-bottle choice
Beginning on Friday of the 1st week of testing, all mice
were weighed and given the modified two-bottle choice test to
measure ethanol consumption. From Friday to Monday, mice were
given two bottles, one containing water and the other containing
3% alcohol solution in water to allow mice to adjust to the taste
and smell of the alcohol. From Monday morning to Friday morn-
ing of the 2nd week (96 hr), mice were given water and 10%
ethanol solution in water. On its cage, each mouse had two
graduated cylinder tubes, with rubber stoppers and sipper tubes,
containing these solutions. The location of the tube with the
ethanol solution (right vs. left) was switched each day to avoid side
bias. Each day, the amount of liquid that was consumed from each
tube was recorded and the tubes refilled. Analyses were conducted
on the data obtained during the 4 days when the 10% ethanol
solution was given. Four measures were obtained or determined
and analyzed: (a) total amount of 10% ethanol solution consumed,
(b) total amount of water consumed during those 4 days, (c) grams
of 10% ethanol consumed per kilogram of body weight of the
mouse per day (g/kg/day), and (d) preference ratio, which was
determined by dividing the amount of 10% ethanol consumed by
the total amount of liquid consumed during the last 4 days. Grams
per kilogram per day is a standard measure that was included to
control for differences in consumption due to differences in body
size (e.g., larger mice would be expected to consume more than
smaller mice), whereas the preference ratio was included to control
for differences in liquid consumption.
For the assessment of strain differences, we compared the three
mouse strains for each behavioral measure using a one-way anal-
ysis of variance with Newman-Keuls post-hoc tests (Winer,
Brown, & Michels, 1991). We also compared genders within
genotype using a Student’s t test (Winer et al., 1991). All behaviors
were compared for gender differences, but only those with signif-
icant effects were reported. For the rotarod test, we used paired t
tests to compare the two values for each mouse across the last three
In the analysis of mutagenized mice, the statistics package that
was used for the analysis was SAS implemented in our custom-
ized, Web-based data repository and analytic toolbox called
Mutrack. The following criteria were used to assign a particular
pedigree as mutant or phenodeviant on a specific trait:
1.The mean for the pedigree had to be over 1.7 standard
deviations from the mean, which was the cut-off point on
a bell curve where 90% of the mice fell within the bell
curve; anything outside that point represented the 5% by
which were outliers either above or below.
2.Three out of 4 mice had to exhibit this abnormal pheno-
3.After the initial 4–6 mice were tested, an additional 12
mice from the same pedigree were tested, and the phe-
notype had to be observed in the successive cohorts.
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
4. At least one of the successive cohorts had to be from the
next generation to demonstrate that the trait was herita-
The two pedigrees described in this report met all criteria for being
identified as a mutant. More information, including further statis-
tical analyses, can be found at the following website: www
Confirmation of Mutant Phenotype by Means of
Additional Screening for Ethanol-Induced Locomotor
To confirm the ethanol reactivity and to assess the response of
7TNJ mice at different ethanol doses, we gave an additional
activity test to a separate group of mice. The controls used in this
test were the 1TNH, which were chosen because they were non-
mutagenized and had a pedigree mean that was only 0.1 standard
deviation from the family mean. Both males and females were
tested from both strains.
In this test, mice were given 3 days of testing: The 1st day they
received no injections so that they could acclimate to the activity
chamber, and the 2nd and 3rd days they were given an injection of
the appropriate solutions. For half of the mice, on the 2nd day they
were given isovolumetric saline, and on the 3rd day they were
given a dose of ethanol, whereas for the other half of the mice the
sequence was reversed. Two doses of ethanol were used in this
study: 1.5 g/kg and 2.25 g/kg, although each mouse was given only
one dose. Each mouse was tested on 2 consecutive weeks, with the
dose and sequence of testing identical on both weeks.
Immediately after injection, mice were placed in the activity
chambers described above. Mice remained in the chambers for the
30-min test. Both horizontal and vertical movements were re-
corded. Movement was recorded in 5-min bins. Scores were re-
corded for both weeks and for saline and ethanol conditions.
Additionally, a difference score, saline–ethanol, was computed.
An analysis of variance was used to statistically compare the two
groups, and the Neuman-Keuls post-hoc test was used to identify
which particular groups differed.
We assessed the ability of this screen to detect differential
responses to ethanol or differences in ethanol consumption by
examining three standard inbred strains of mice. The tests and the
order in which they were administered are shown in Table 1.
Performance on an accelerating rotarod provided a standard
means to assess the degree of motor coordination in mice at
baseline and the degree of ethanol-induced ataxia following etha-
nol exposure (Rustay et al., 2003). In the current test, both the
length of time that the mice were able to remain on the rotarod and
the related speed (data not shown) were determined for mice at
baseline and 25–30 min following ethanol exposure. To determine
whether 10 trials were sufficient for the behavior to asymptote in
the basal condition, we analyzed the values for each of the last 3
trials across the strains using paired t tests. No significant differ-
ences were found between any of the values, and thus stable
performance was observed (for example, in the comparison be-
tween Trials 8 and 10, for B6, p ? .431; for D2, p ? .197; and for
AJ, p ? .83).
Strain differences were detected in the rotarod test both at
baseline and after alcohol exposure. At baseline, there was an
effect of strain, F(2, 56) ? 50.7, p ? .000001, with B6 mice able
to remain on the rotarod significantly longer than either D2 or A/J
mice (see Figure 1). Similar effects were observed after ethanol
exposure, although the differences were less pronounced. B6 mice,
F(2, 56) ? 7.1, p ? .002, had significantly higher scores than did
D2 (p ? .01) and A/J (p ? .05) mice, as shown in Figure 1. The
D2 and A/J mice were not significantly different on any measure
in the rotarod test. As a consequence of the other two measures, the
difference scores (baseline–ethanol) had similar strain effects
(data not shown).
Elevated Plus Maze
The elevated plus maze was used to measure the anxiolytic
response of the mice (Lister, 1987). The typical interpretation is
that mice that spent more time in the open arms of the plus maze
and made more entries into the open arms of the plus maze had
lower levels of anxiety. For example, a previous study has shown
that acute ethanol exposure increased the amount of time mice
spent in the open arms during a test session (Lister, 1987). How-
ever, this potential anxiolytic construct can be confounded by other
behaviors, such as reduction in locomotor activity produced by
ethanol exposure. In the present study, mice were examined only
following ethanol exposure. As can be seen in Table 2, strain
differences in the anxiolytic response to ethanol were observed.
All of the measures showed similar effects, as exemplified by the
amount of time in the open arms and the percentage of entries into
the open arms. In all measures, A/J mice exhibited higher mea-
mice are found in their latency to fall off an accelerating rotarod at baseline
and following ethanol exposure. Mean (? SEM) length of time on the
rotarod is shown in seconds during the baseline condition (left) and after a
dose of 2.25 g/kg of ethanol (right). In the baseline test, B6 mice had a
significantly longer latency than did either of the other two strains (p ?
.000001). Similarly, B6 mice had a longer latency to fall off the rotarod
following ethanol exposure compared with D2 (p ? .01) and A/J (p ? .05)
mice. D2 and A/J mice were not significantly different.
Strain differences in A/J, C57BL/6J (B6), and DBA/2J (D2)
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
sures of anxiety than did either B6 or D2 mice. In other words, A/J
mice spent significantly less time in the open arms of the maze
than did either of the other two strains, F(2, 73) ? 10.2, p ? .0002,
and made fewer entries into the open arms, as shown by the
percentage of entries into the open arms, F(2, 73) ? 6.4, p ? .003.
However, one caveat to this study is that A/J mice have been
shown to be significantly less active than either of the other two
strains by others (e.g., Crabbe, Gallaher, et al., 1994) as well as in
the present study. B6 and D2 mice were not significantly different
in any of the measures.
A concern with the elevated plus maze test is the interpretation
of mice that either remain in the middle or never enter either an
open or closed arm (Wahlsten et al., 2003). In the present study, all
mice moved out of the middle of the maze, and over 80% of the
mice of each genotype made entries into both types of arms.
Low doses of ethanol have been shown to stimulate locomotor
activation (Crabbe et al., 1982, Crabbe, Gallaher, et al., 1994;
Demarest et al., 1999; Dudek & Tritto, 1994; Koyner et al., 2000).
In the present screen, locomotor activity was assessed in an activ-
ity chamber for a 10-min test that began 15 min after ethanol
injection. Strain differences were detected (see Figure 2): Both B6
and D2 mice had significantly higher levels of locomotor activity
than did the A/J mice, F(2, 70) ? 17.6, p ? .000001. There was
also a trend toward a significant difference between B6 and D2
mice (p ? .10). When the activity was analyzed by bins, analysis
of the second 5-min bin demonstrated that the differences between
both B6 and D2 mice and A/J mice remained, F(2, 73) ? 14.1, p ?
.00001, whereas a significant difference between B6 and D2 mice
was uncovered (p ? .05; see Figure 2).
It is interesting that gender differences were observed in the A/J
and D2 mice but not in the B6 mice. In both the A/J and D2 strains,
the males were more active than the females. Among A/J mice,
males moved a mean of 632 (? 140) cm during the test, whereas
females moved a mean of 301 (? 293) cm during the same 10-min
trial, t(18) ? 2.30, p ? .05. Among D2 mice, males moved a mean
of 2,741 (? 450) cm in the activity chamber, whereas females
moved a mean of 1,490 (? 256) cm during the 10 min, t(18) ?
2.79, p ? .05.
Moderate doses of ethanol are known to cause a decrease in core
body temperature (Crabbe, Belknap, et al., 1994; Crawshaw et al.,
1997, 2001; Moore & Kakihana, 1978; O’Connor et al., 1989). In
the present study, temperature was measured at 60 min following
ethanol injection and again at 24 hr after injection. The 24-hr time
point was used as the measure of baseline temperature. The tem-
(B6) and DBA/2J (D2) mice in an activity chamber following a dose of
2.25 g/kg of ethanol. Mice were tested in the chamber for 10 min, and the
test began immediately after the elevated plus maze test, which was 15 min
after injection of alcohol. The three bars on the left show the mean (?
SEM) total distance moved in centimeters for the entire duration of the test,
and the bars on the right show the mean (? SEM) distance moved in
centimeters for the second 5 min of the test. In both instances, A/J mice
were significantly less active compared with either the B6 or D2 mice (p ?
.00001). Comparisons between B6 and D2 mice detected a significant
difference (p ? .05) only in the second bin (right).
Strain differences in the locomotor activity of A/J, C57BL/6J
Mean (? SEM) Temperature Measures (All in Degrees Celsius), Elevated Plus Maze Measures,
and Blood Ethanol Concentration (BEC; in mg/dl) for the Three Mouse Strains
Temperature after alcohola
Percent time openc
Percent entries opene
Number of entries openf
37.62 ? 0.11
37.01 ? 0.18
0.68 ? 0.15
21.5 ? 5.8d
37.7 ? 5.4d
4.5 ? 0.7d
159.7 ? 10.3
38.10 ? 0.52
36.94 ? 0.93
1.11 ? 0.19
52.7 ? 4.9
60.4 ? 4.6
15.8 ? 2.4
192.6 ? 8.9
37.94 ? 0.11
36.97 ? 0.69
0.97 ? 0.18
56.9 ? 6.4
56.0 ? 6.0
13.9 ? 2.6
307.0 ? 16.6
aTemperature 60 min after injection of 2.25 g/kg of ethanol.
bDifference in basal temperature - temperature after alcohol.
cPercentage of time in the open arms of the elevated plus maze after 2.5 g/kg of ethanol.
dIn the elevated plus maze, A/J mice spent significantly less time in the open arms of the plus maze and made
fewer numbers of and a smaller percentage of entries into the open arms (p ? .01).
ePercentage of entries in the open arms of the elevated plus maze after 2.5 g/kg of ethanol.
fNumber of entries in the open arms of the elevated plus maze after 2.5 g/kg of ethanol.
B6 ? C57BL/6J mouse strain; D2 ? DBA/2J mouse strain.
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
perature data, shown in Table 2, indicated that baseline tempera-
tures were significantly different between the B6 and A/J mice,
F(2, 73) ? 6.8, p ? .002, and close to significant between the D2
and A/J mice (p ? .06). All strains exhibited the previously
reported ethanol-induced temperature decrease after ethanol expo-
sure. However, when the ethanol-induced temperatures were com-
pared among the three strains, no significant differences were
observed (see Table 2). Furthermore, because the baseline temper-
atures were different between the strains, the most appropriate way
to analyze ethanol-induced changes was to examine the difference
scores (baseline minus ethanol), as shown in Table 2. In the
analyses of the difference scores, there was a trend toward differ-
ences among the strains (p ? .1), although it did not reach
statistical significance. Thus, previously reported strain differences
in ethanol-induced hypothermia were not found in this testing
paradigm. It is possible that the multiple-test regimen influenced
the temperature measures and obscured our ability to detect strain-
specific differences in ethanol-induced hypothermia. Further re-
search should address this possibility.
BEC was determined 1 hr following ethanol injection to provide
a control for the assessment of appropriate ethanol exposure as
well as to provide an initial indication of whether the previously
identified strain differences in ethanol metabolism (Grisel, Metten,
Wenger, Merrill, & Crabbe, 2002) were found. However, it is
recognized that a true examination of differences in ethanol me-
tabolism would require more than a single time point, as used in
the present study. As seen in Table 2, the highest level of BEC was
found in the D2 mice, an intermediate level was found in the B6
mice, and the lowest level was found in the A/J mice. All three
strains were significantly different in BEC levels: D2 mice’s were
significantly higher than that of both other strains, F(2, 70) ? 37.2,
p ? .000001, whereas B6 mice had significantly higher BECs than
did A/J mice (p ? .05).
A two-bottle choice test was used to assess whether mice
preferentially consumed an ethanol-containing solution over un-
adulterated water. In the present experiment, mice were given 3
days of acclimation with a low, 3% ethanol solution followed by
4 days of testing at a higher, 10% ethanol solution. Statistical
analyses were conducted on only the latter 4 days. Several mea-
sures were analyzed, including grams/kilogram/day measures and
preference ratio measures, and all yielded consistent results. As
shown in Figure 3 (top), B6 mice consumed significantly greater
amounts of ethanol than did either the D2 or A/J mice, for
grams/kilogram/day, F(2, 56) ? 116.7, p ? .0000001; for prefer-
ence ratio, F(2, 56) ? 203.0, p ? .0000001. The D2 and A/J mice
consumed equivalent amounts of ethanol.
Previous data has shown that female mice consumed higher
amounts of ethanol than did males of the same strain (Gill, De-
saulniers, Desjardins, & Lake, 1998; Melo et al., 1996; Peirce et
al., 1998). Thus, it was of interest to compare whether gender
differences could also be detected in the present two-bottle choice
regimen. As shown in Figure 3, differences could be detected, but
not for all strains. Comparison of B6 males and females demon-
strated that the females consumed significantly more than did the
males, t(18) ? 6.78, p ? .0001. A similar outcome was observed
for the D2 mice, although the differences were less dramatic,
t(18) ? 2.48, p ? .05. No differences based on sex were observed
in A/J mice.
Analysis of ENU-Induced Mutants
The criterion for defining a pedigree as mutant was that three
quarters of the mice tested must be at or greater than 1.7 standard
and DBA/2J (D2) mice. Ethanol consumption was measured in a two-bottle
choice test. Means were calculated from the portion of the test where mice
were given 10% (v/v) ethanol. The top graph shows the preference ratio
calculated as the percentage of ethanol solution consumed out of the total
liquid consumed. B6 mice consumed significantly more ethanol than did
either of the other two strains (p ? .0000001). The bottom graph shows the
strain differences in a second measure, as well as the gender differences in
ethanol consumption. Mean (? SEM) g/kg/day of ethanol consumed by
mice of each sex is shown for the three strains. For all strains, males are
shown in black bars, and females are shown in white bars. Strain compar-
isons again showed that the B6 mice drank significantly more than did the
other two strains (p ? .0000001). Comparisons between sexes within each
strain showed that for both B6 (p ? .0001)) and D2 (p ? .05) but not A/J
mice, females drank significantly more ethanol than did males of the same
Comparison of ethanol consumption in A/J, C57BL/6J (B6)
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
deviations from the family mean. Additionally, the phenotype
must be present in mice in the next generation. The mean of the
family was defined as the “mean of the mean,” which was calcu-
lated by determining the mean for each pedigree and averaging the
means of all pedigrees to obtain a family mean and standard
deviation. The mean for any mutant pedigree was not included in
the calculation for determining the family mean and standard
deviation. All data in this section involved standard deviation
instead of standard error of the mean.
7TNJ mutant mice.
The 7TNJ pedigree exhibited increased
locomotor activity in the activity test, which was a 10-min test that
began 15 min after an injection of 2.25 g/kg of ethanol. This
phenotype was specific to the activity chamber, because all other
measures after the ethanol exposure as well as measures of ethanol
consumption were within 2 standard deviations for mice in this
family (see Table 3).
For total activity, the 7TNJ mice were 3.5 standard deviations
away from the mean. This result highlighted the heightened loco-
motor activity of the mice in this pedigree (see Table 3). The initial
activity in the activity chamber had been defined as reflecting the
habituation to the novel environment. Thus, we analyzed the data
not only as a total but also in 5-min bins. In the initial 5-min
period, the 7TNJ mice moved over 1,630 cm, whereas the family
mean was only 560, which showed that the 7TNJ mice were 3
standard deviations away from the mean. In the second 5-min bin,
the differences were more pronounced, with the pedigree mean
being 321 whereas that for 7TNJ was 1,290, supporting the con-
clusion that the hyperactivity was not just a function of initial
exploration of a novel environment (see Figure 4).
To further characterize the locomotor response of the 7TNJ
mice to ethanol, an additional cohort of mutant and control 1TNH
mice were given expanded testing. Mice were given ethanol at
either the same dose as used in the initial screen, 2.25 g/kg, or the
lower dose of 1.5 g/kg; movement was compared with the baseline
distance traveled following injection of saline. The difference
scores (ethanol–saline) are shown in Figure 5. At the 1.5 g/kg dose
(see top panel), the 7TNJ mice exhibited a mild locomotor stim-
ulant effect, whereas the controls exhibited similar locomotor
levels both with and without alcohol. At the 2.25 g/kg dose (see
bottom panel), the response of the 7TNJ mice to ethanol was
significantly different from that of the 1TNH mice, F(3, 42) ? 8.9,
p ? .001. Post-hoc analyses showed that in Week 1 mutants were
significantly different from controls (p ? .05), although the dif-
ference was more extreme in the comparison of Week 2 values
(p ? .001). Thus, the phenotype of abnormal locomotor activity
that was detected in the initial screen was confirmed in the addi-
tional screening. This confirmation further demonstrated that the
screen is an effective tool for detecting outliers that can then be
more rigorously tested and characterized.
112TNR mutant mice.
The 112TNR pedigree exhibited an
abnormal enhanced ability to remain on the rotarod (see Figure 6),
as measured by latency to fall off the accelerating rotarod. Similar
results were obtained in the analysis of rotarod speed (data not
shown). The abnormal rotarod ability was a highly robust pheno-
type, because all but 1 of the 18 112TNR mice tested were over 2.0
standard deviations from the family mean. The abnormal pheno-
type was specific to the rotarod, because 112TNR mice exhibited
normal behavior in all other tests in this screen and in BEC (see
Table 3). Furthermore, this abnormal phenotype had been shown
to be heritable, as evidenced by its presence in multiple genera-
tions of mice.
During baseline, the 112TNR mice remained on the rotarod for
an average of 85.8 ? 16.8 s, whereas the mean for the pedigree as
a whole was 42.0 ? 6.6 s. The next nearest pedigree had a mean
of 65.7 s, or almost 20 s shorter than that for 112TNR. Thus, in the
baseline measures, the 112TNR mice were over 6.6 standard
deviations from the mean. After ethanol, a similar scenario was
found, with the mice from this pedigree also remaining on the
rotarod for extended lengths of time (see Figure 6). Again the mice
were over 6.4 standard deviations from the mean and were able to
stay on the rotarod over twice as long as the average mouse on this
background. However, this ethanol-induced rotarod phenotype ap-
peared to be due primarily to the enhanced ability of these mice to
remain on the rotarod rather than an insensitivity to the ataxic
effects of ethanol, as shown by the difference scores (baseline–
ethanol), which indicate a large drop in performance after ethanol
Comparison of the Mutant Pedigrees 112TNR and 7TNJ With the Means of Their Respective TNR and TNJ Families
Test TNR family112TNR TNJ family 7TNJ
Percent time opena
Percent entries openb
Number of entries openc
Rotarod after ethanol
Temperature after ethanol
53.0 ? 11.9
56.9 ? 15.8
7.2 ? 4.9
42.0 ? 6.6
28.8 ? 4.4
12.41 ? 8.16
1,350 ? 401
38.05 ? 0.33
36.37 ? 0.60
231 ? 33
.41 ? .13
7.81 ? 2.59
61.4 ? 36.4
68.5 ? 37.8
4.5 ? 5.4
85.7 ? 16.8
57.5 ? 16.6
28.28 ? 21.45
883 ? 667
38.59 ? 0.32
35.69 ? 0.69
219 ? 24
.42 ? .22
8.78 ? 4.43
52.1 ? 12.8
57.7 ? 14.2
9.1 ? 7.2
34.3 ? 14.0
27.7 ? 10.2
5.08 ? 8.86
898 ? 583
37.99 ? 0.39
36.55 ? 0.68
222 ? 42
.37 ? .14
6.87 ? 2.71
52.0 ? 27.6
58.9 ? 33.5
14.1 ? 10.6
42.7 ? 12.5
33.7 ? 17.0
9.02 ? 16.36
2,921 ? 1,700
38.20 ? 0.61
37.24 ? 0.67
203 ? 47
.47 ? .21
9.80 ? 4.84
aPercentage of time in the open arms of the elevated plus maze.
bPercentage of entries into the open arms of the elevated plus maze.
cNumber of entries into the open arms of the elevated plus maze.
dTotal distance moved in centimeters in the entire 10 min of the trial.
Values are given in mean ? standard deviation. BEC ? blood ethanol concentration.
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
in the 112TRN mice (see Table 3). Thus, we identified a pedigree
of mice that exhibits abnormal enhanced locomotor ability on the
In the evaluation of mice to detect outliers, there is a need to test
a large number of mice in as short a time period and as reliably as
possible. In the present study, we describe a screen that has been
designed to detect differences in commonly assessed ethanol-
related behaviors. Three standard inbred strains of mice—C57BL/
6J, DBA/2J, and A/J—were tested to determine the efficacy with
which this screen can detect strain differences. Strain differences
were found on many of the measures that were tested. This screen,
therefore, is an effective tool to detect differential responses to
ethanol and differences in ethanol consumption. Additionally, we
were able to identify two pedigrees that met the criteria to be
deemed as mutants: consistency, replicability, and heritability. One
pedigree, 7TNJ, exhibited abnormal ethanol-induced locomotor
activity. A second pedigree, 112TNR, exhibited enhanced motor
coordination, as shown by an enhanced ability to remain on an
From the analysis of standard inbred strains, we addressed the
issue of whether this screen could detect previously reported strain
differences. In general we found that known strain differences
were recapitulated in the present study, although there were a few
The rotarod is one of the measures that most accurately and
robustly replicated previous findings. A comparison with recent
results from Rustay et al. (2003), who reported the behavior of
different strains of mice on the accelerating rotarod, demonstrates
that although the parameters of the testing varied between studies,
the results were highly consistent. The effects of alcohol were
similar even though Rustay et al. used a dose of 2 g/kg whereas the
present screen used a dose of 2.25 g/kg. A similar magnitude of
decrement and enhancement was found in all three strains of mice,
with highly consistent strain values obtained in both studies. The
baseline measures also showed comparable results between the
two studies. Thus, the present screen was not only able to detect
similar genetic differences in rotarod performance but was also
able to obtain comparable values (Crabbe et al., 2005; Rustay et
Ethanol consumption as determined by the two-bottle choice
test has been a frequently studied measure in rodents. Strain
differences in ethanol consumption have been reliably and consis-
tently reported since the early studies of McClearn and Rodgers
(1959). The consistent finding is that B6 mice consume signifi-
cantly more ethanol than do either the D2 or A/J mice. Further-
more, female mice have been shown to consume higher levels of
ethanol than do male mice (Gill et al., 1998; Melo et al., 1996;
Peirce et al., 1998), which is consistent with the results shown in
the present study. Although the present study used an abbreviated
version of the two-bottle choice test, both the strain and gender
differences were observed.
Only two measures either partly replicated previous findings or
showed results in the same direction, although with a different
magnitude. First, the ability to detect strain differences in BEC was
preserved, with D2 mice having higher BECs than did B6 mice,
which was the same direction as previously found (Grisel et al.,
2002). However, although the differences were in the same direc-
tion, previous studies have shown that the D2 mice typically have
family. The pedigrees within the TNJ family are listed along the bottom. The solid line marks the family mean,
and the two dashed lines mark 2.0 standard deviations above and below the mean. Each mouse within the
pedigree is shown by the ? symbol, and all mice from a single pedigree are shown in one column. The mice
in the 7TNJ pedigree are within the gray shading. The graph shows the distance traveled in the second of the
two 5-min bins. As can be seen, the 7TNJ mice exhibited heightened locomotor activity following ethanol
exposure compared with that of other pedigrees in this family.
Comparison of the locomotor activity as shown by distance traveled (in cm) of mice in the TNJ
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
a BEC value that is closer to that of the B6 mice (e.g., see Grisel
et al., 2002) than was obtained in the present study. Differences in
the methods used to measure the BECs (Analox vs. gas chromato-
graph) or the dose administered may account for the differences in
BECs in the two studies.
The second example of a partially replicated finding is the
examination of locomotor activity. In the test of locomotor activity
following ethanol exposure, strain differences have been found
between B6 and D2 mice (Crabbe et al., 1982; Crabbe, Gallaher,
et al., 1994; Koyner et al., 2000; Phillips & Dudek, 1991), as well
as lower activity in the A/J mice compared with that in other
strains (Crabbe, Gallaher, et al., 1994). In the B6–D2 comparison,
D2 mice showed significantly higher activity in the activity cham-
ber following ethanol exposure than did B6 mice (Crabbe et al.,
1982; Crabbe, Gallaher, et al., 1994; Koyner et al., 2000; Phillips
& Dudek, 1991). In the present study, the activity levels were
higher in D2 versus B6 mice, with a trend toward a significant
difference in the total activity measures and a significant differ-
ence in the analysis of the second 5-min bin. One explanation for
the modest difference observed in the present study is that at
higher ethanol doses, similar to the one used in the present study,
the differences in activity levels between B6 and D2 mice are less
extreme (Crabbe, Gallaher, et al., 1994). The behavior of A/J mice
compared with that of the other two strains is consistent with what
has been previously reported, where A/J mice exhibit significantly
less locomotor activity following ethanol exposure (Crabbe, Gal-
laher, et al., 1994).
Differences in baseline, as well as ethanol-induced, tempera-
tures among mouse strains have been previously reported (Crabbe,
Belknap, et al., 1994; Crawshaw et al., 1997, 2001; Moore &
Kakihana, 1978; O’Connor et al., 1989). The differences between
basal and ethanol-induced temperatures (see Table 1) are similar to
those previously reported (e.g., Crabbe, Belknap, et al., 1994). The
screen used was able to detect ethanol-induced hypothermia in all
mouse strains. However, the differences between strains were not
statistically different in the present study. In previous studies,
either higher doses of ethanol or multiple testing regimens were
used to better analyze this variable (Crabbe, Belknap, et al., 1994;
Crawshaw et al., 1997, 2001; Moore & Kakihana, 1978; O’Connor
et al., 1989). Moreover, the exposure to ethanol on Day 2 could
have impacted the baseline values obtained on Day 3, and the
multiple tests that were given on Day 2 could have impacted the
ethanol-induced temperature drop. Thus, measurement of temper-
ature differences following ethanol exposure is one of the variables
that are likely not accurately measured in this type of screen.
The elevated plus maze has been used in numerous studies to
demonstrate genetic differences in anxiolytic response to alcohol,
ranging from the study of selectively bred lines (Boehm, Reed,
McKinnon, & Phillips, 2002) to the analyses of specific knockout
mice (Boehm, Peden, Chang, Harris, & Blednov, 2003; Fee et al.,
2004). Previously, strain differences in the elevated plus maze
have been explored in mice without ethanol, with considerable
variability in the results (Crabbe, Gallaher, et al., 1999; Wahlsten
et al., 2003). However, to our knowledge this is the first use of this
test to examine these three strains of mice following ethanol
exposure. What has been shown is that without ethanol, both B6
and D2 mice spend approximately one third of the time in the open
arms, whereas A/J mice spend approximately 20%–40% of the
time in the open arms (Crabbe, Gallaher, et al., 1999; Wahlsten et
al., 2003). In the present study, we demonstrated that following
ethanol injection both B6 and D2 spent slightly over 50% of the
time in the open arms of the elevated plus, whereas A/J mice
continued to spend approximately 20% of the time. The small
amount of time that the A/J mice spent in the open arms suggests
that A/J mice have a diminished anxiolytic response following a
dose of 2.25 g/kg of ethanol compared with that of B6 and D2
mice. However, without the testing of naive mice in the elevated
plus maze, this conclusion is tentative. Consequently, initial iden-
tification of mutants in this behavior would require additional
observed in mutant 7TNJ mice compared with that in the 1TNH controls in
expanded testing. The difference score is calculated by subtracting the
distance moved at baseline from the distance moved after ethanol exposure.
The top graph shows the difference scores (in centimeters) for both 1TNH
mice and 7TNJ mice during the 1st week (open bars) and 2nd week (striped
bars) of testing following a dose of 1.5 g/kg of ethanol. As is observed, the
controls show little locomotor activation, whereas the mutants exhibit a
modest, nonsignificant increase in locomotor activity. The bottom graph
shows the difference scores for the two lines during both weeks following
the higher dose of 2.25 g/kg of ethanol. Week 1 values for both pedigrees
are shown in white bars, and the values for Week 2 are shown in striped
bars. At this dose, the 7TNJ mice exhibit a significant increase in loco-
motor activity, whereas the controls exhibit a decrease in activity (p ?
An abnormal locomotor response following ethanol exposure is
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
One of the caveats in the present procedure is that there is no
baseline test for either locomotor activation in the activity chamber or
anxiolysis in the elevated plus maze. Thus, baseline activity measures
could confound the interpretation of the results. This confound was
addressed in two ways. First, other members of the TMGC examined
separate groups of genetically identical mice for each screen. Testing
in the elevated plus maze was not repeated in the same cohort of mice
because of the concern that the anxiolytic effect of ethanol would be
obscured by multiple exposures in the maze. However, separate
cohorts of genetically identical mice were tested for baseline locomo-
tor behavior in the elevated plus maze in the drug abuse and general
behavior screens. Any pedigrees identified as outliers in the ethanol
screen were cross-checked with this baseline activity in the other
screens to assess their normality in these phenotypes. Second, any
pedigree that exhibited an abnormal response in these measures was
given more extensive testing to ensure that differences were not
simply due to baseline locomotor differences. An example of this
expanded testing is shown for 7TNJ.
seconds) until a mouse falls off the accelerating rotarod. The pedigrees within the TNR family are listed along
the bottom. The solid line marks the family mean, and the two dashed lines mark 2.0 standard deviations above
and below the mean. The graph shows the data for each mouse tested (? symbol), with all mice from a single
pedigree shown in a single column. The mice in the 112TNR pedigree are highlighted by the gray shading. The
top graph shows the time during the training trials and is the mean for the last 3 of the 10 trials. The bottom graph
shows the mean time the mice are able to stay on the rotarod for the 3 trials following a dose of 2.25 g/kg of
ethanol. As can be seen, the mice in the 112TNR pedigree exhibit an enhanced ability to remain on the
Comparison of the motor coordination ability in mice in the TNR family is shown by the time (in
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
The importance of the conditions of testing has been repeatedly
documented throughout the years (e.g., Bouwknecht, van der
Gugten, Groenink, Olivier, & Paylor, 2004; Camarini & Hodge,
2004; McIlwain, Merriweather, Yuva-Paylor, & Paylor, 2001).
The debate on this issue was reignited within the last few years by
two papers: one by Crabbe, Wahlsten, et al. (1999) and the other
by Wahlsten et al. (2003). In the present study, multiple measures
(rotarod, locomotor activity, elevated plus maze, temperature, and
BEC) were taken after a single injection of alcohol. However, for
the most part, values obtained were reflective of what was previ-
ously reported in the literature. Additionally, the shortened data
collection strategy also proved no barrier to replicating known
data. Taken together, these data provide additional evidence of the
robustness of these phenotypes (Crabbe, Wahlsten, et al., 1999;
Bouwknecht et al., 2004).
Studies of ethanol-induced differences in locomotor activity
have involved various mouse models, including strains of mice and
the recombinant inbred lines generated from the respective paren-
tal strains, or selectively bred lines (Boehm et al., 2002; Demarest
et al., 1999; Hitzemann, Cipp, Demarest, Mahjubi, & McCaugh-
ran, 1998; Phillips et al., 2002), as well as knockouts and trans-
genics (Kralic et al., 2003; Risinger, Freeman, Greengard, &
Fienberg, 2001; Rubinstein et al., 1997). In the analyses of strain
differences, several QTLs that underlie this difference—such as on
chromosomes 1, 2, 4, and 6 (Demarest et al., 1999; Hitzemann et
al., 1998)—have been identified. Several knockouts have been
shown to affect ethanol-induced locomotor activity by both in-
creasing and decreasing the amount of movement. Examples of
knockouts that show increased ethanol-induced activity levels are
the DARPP molecule in the dopaminergic pathway (Risinger et al.,
2001), the GABAA receptor alpha subunit (Kralic et al., 2003),
and the dopamine D4 receptor (Rubinstein et al., 1997). It will be
of interest to determine whether the gene that underlies the mutant
phenotype in 7TNJ mice is found in any of these QTLs or involves
alterations in any of the same pathways.
The rotarod has been used to assess whether mice have diffi-
culty with balance and motor coordination and has typically iden-
tified mutations in genetic pathways that disrupt their ability to
remain on the rotarod, or an ataxic response. However, several
recent studies have described several lines of knockout mice that
also have enhanced abilities on the rotarod. Two examples are the
knockout of Rac3, which is a member of the Rho GTPase family
(Corbetta et al., 2005), and the knockout of the Pcmt1 enzyme
involved in protein repair (Vitali & Clarke, 2004). Alterations in
several neurotransmitters or their receptors also have been shown
to enhance rotarod performance, as shown by knockouts of the
alpha 1d-adrenergic receptor (Mishima et al., 2004), of the central
nicotinic acetylcholine receptors (Marubio & Paylor, 2004), and of
the histidine decarboxylase enzyme, which also results in abnor-
mal levels of acetylcholine (Dere et al., 2004). It would be of
interest to determine whether the 112TNR mice also have abnor-
malities within any of these pathways or whether this represents a
novel genetic pathway that enhances rotarod performance.
In conclusion, the success of this screen in detecting known
strain differences demonstrates that it is a useful tool for assessing
potential genetic differences in a fast and reliable manner. Neither
the modifications in the parameters of each test nor the adminis-
tration of multiple tests to each mouse obscured the ability to
detect known strain differences. Moreover, this screen has the
ability to detect outliers among the mutagenized pedigrees in
ethanol-related phenotypes, such as that found in 7TNJ, and in
baseline measures, as shown in 112TNR. Subsequent research on
these pedigrees is needed to determine the genetic basis for the
Balling, R. (2001). ENU mutagenesis: Analyzing gene function in mice.
Annual Review of Genomics and Human Genetics, 2, 463–492.
Belknap, J. K., Crabbe, J. C., & Young, E. R. (1993). Voluntary consump-
tion of ethanol in 15 inbred mouse strains. Psychopharmacology (Ber-
lin), 112, 503–510.
Boehm, S. L., II, Peden, L., Chang, R., Harris, R. A., & Blednov, Y. A.
(2003). Deletion of the fyn-kinase gene alters behavioral sensitivity to
ethanol. Alcoholism: Clinical and Experimental Research, 27, 1033–
Boehm, S. L., II, Reed, C. L., McKinnon, C. S., & Phillips, T. J. (2002).
Shared genes influence sensitivity to the effects of ethanol on locomotor
and anxiety-like behaviors, and the stress axis. Psychopharmacology
(Berlin), 161, 54–63.
Bouwknecht, J. A., van der Gugten, J., Groenink, L., Olivier, B., & Paylor,
R. (2004). Effects of repeated testing in two inbred strains on flesinoxan
dose-response curves in three mouse models for anxiety. European
Journal of Pharmacology, 494, 35–44.
Brown, S. D., & Balling, R. (2001). Systematic approaches to mouse
mutagenesis. Current Opinion in Genetics & Development, 11, 268–
Camarini, R., & Hodge, C. W. (2004). Ethanol preexposure increases
ethanol self-administration in C57BL/6J and DBA/2J mice. Pharmacol-
ogy, Biochemistry, and Behavior, 79, 623–632.
Connor, J. P., Young, R. M., Lawford, B. R., Ritchie, T. L., & Noble, E. P.
(2002). D(2) dopamine receptor (DRD2) polymorphism is associated
with severity of alcohol dependence. European Psychiatry, 17, 17–23.
Corbetta, S., Gualdoni, S., Albertinazzi, C., Paris, S., Croci, L., Consalez,
G. G., & de Curtis, I. (2005). Generation and characterization of Rac3
knockout mice. Molecular and Cellular Biology, 25, 5763–5776.
Crabbe, J. C., Belknap, J. K., Mitchell, S. R., & Crawshaw, L. I. (1994).
Quantitative trait loci mapping of genes that influence the sensitivity and
tolerance to ethanol-induced hypothermia in BXD recombinant inbred
mice. Journal of Pharmacology and Experimental Therapeutics, 269,
Crabbe, J. C., Gallaher, E. S., Phillips, T. J., & Belknap, J. K. (1994).
Genetic determinants of sensitivity to ethanol in inbred mice. Behavioral
Neuroscience, 108, 186–195.
Crabbe, J. C., Johnson, N. A., Gray, D. K., Kosobud, A., & Young, E. R.
(1982). Biphasic effects of ethanol on open field activity: Sensitivity and
tolerance to C57BL/6N and DBA/2N mice. Journal of Comparative and
Physiological Psychology, 96, 440–451.
Crabbe, J. C., Metten, P., Cameron, A. J., & Wahlsten, D. (2005). An
analysis of the genetics of alcohol intoxication in inbred mice. Neuro-
science and Biobehavioral Reviews, 28, 785–802.
Crabbe, J. C., Phillips, T. J., Buck, K. J., Cunningham, C. L., & Belknap,
J. K. (1999). Identifying genes for alcohol and drug sensitivity: Recent
progress and future directions. Trends in Neuroscience, 22, 173–179.
Crabbe, J. C., Wahlsten, D., & Dudek, B. C. (1999, June 4). Genetics of
mouse behavior: Interactions with laboratory environment. Science, 284,
Crawshaw, L. I., Wallace, H. L., Christensen, R., & Crabbe, J. C. (2001).
Influence of ethanol on thermoregulation: Mapping quantitative trait
loci. Physiological Genomics, 7, 159–169.
Crawshaw, L. I., Wallace, H. L., Crabbe, J. C., Ramos, C., Duerr, J., &
O’Connor, C. S. (1997). Ethanol disrupts and decreases the regulated
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS
body temperature differentially in C57BL/6J and DBA/2J mice. Amer-
ican Journal of Physiology, 273, R696–702.
Demarest, K., McCaughran, J., Jr., Mahjubi, E., Cipp, L., & Hitzemann, R.
(1999). Identification of an acute ethanol response quantitative trait
locus on mouse chromosome 2. Journal of Neuroscience, 19, 549–561.
Dere, E., De Souza-Silva, M. A., Spieler, R. E., Lin, J. S., Ohtsu, H., Haas,
H. L., & Huston, J. P. (2004). Changes in motoric, exploratory and
emotional behaviours and neuronal acetylcholine content and 5-HT
turnover in histidine decarboxylase-KO mice. European Journal of
Neuroscience, 20, 1051–1058.
Dick, D. M., & Foroud, T. (2003). Candidate genes for alcohol depen-
dence: A review of genetic evidence from human studies. Alcoholism:
Clinical and Experimental Research, 27, 868–879.
Dudek, B. C., & Tritto, T. (1994). Biometrical genetic analysis of ethanol’s
psychomotor stimulant effect. Alcoholism: Clinical and Experimental
Research, 18, 956–963.
Enoch, M. A., & Goldman, D. (2001). The genetics of alcoholism and
alcohol abuse. Current Psychiatry Reports, 3, 144–151.
Eriksson, C. J., Fukunaga, T., Sarkola, T., Chen, W. J., Chen, C. C., Ju,
J. M., et al. (2001). Functional relevance of human adh polymorphism.
Alcoholism: Clinical and Experimental Research, 25, 157S–163S.
Fee, J. R., Sparta, D. R., Knapp, D. J., Breese, G. R., Picker, M. J., &
Thiele, T. E. (2004). Predictors of high ethanol consumption in RIIbeta
knockout mice: Assessment of anxiety and ethanol-induced sedation.
Alcoholism: Clinical and Experimental Research, 28, 1459–1468.
Fehr, C., Rademacher, B. L., & Buck, K. J. (2003). Evaluation of the
glutamate decarboxylase genes Gad1 and Gad2 as candidate genes for
acute ethanol withdrawal severity in mice. Progress in Neuro-
psychopharmacology & Biological Psychiatry, 27, 467–472.
Fehr, C., Shirley, R. L., Belknap, J. K., Crabbe, J. C., & Buck, K. J. (2002).
Congenic mapping of alcohol and pentobarbital withdrawal liability loci
to a ?1 centimorgan interval of murine chromosome 4: Identification of
Mpdz as a candidate gene. Journal of Neuroscience, 22, 3730–3738.
Foroud, T., & Li, T. K. (1999). Genetics of alcoholism: A review of recent
studies in human and animal models. American Journal on Addictions,
Gill, K., Desaulniers, N., Desjardins, P., & Lake, K. (1998). Alcohol
preference in AXB/BXA recombinant inbred mice: Gender differences
and gender-specific quantitative trait loci. Mammalian Genome, 9, 929–
Goldowitz, D., Frankel, W. N., Takahashi, J. S., Holtz-Vitaterna, M., Bult,
C., Kibbe, W. A., et al. (2004). Large-scale mutagenesis of the mouse to
understand the genetic bases of nervous system structure and function.
Molecular Brain Research, 132, 105–115.
Grisel, J. E., Metten, P., Wenger, C. D., Merrill, C. M., & Crabbe, J. C.
(2002). Mapping of quantitative trait loci underlying ethanol metabolism
in BXD recombinant inbred mouse strains. Alcoholism: Clinical and
Experimental Research, 26, 610–616.
Hallikainen, T., Hietala, J., Kauhanen, J., Pohjalainen, T., Syvalahti, E.,
Salonen, J. T., & Tiihonen, J. (2003). Ethanol consumption and DRD2
gene TaqI a polymorphism among socially drinking males. American
Journal of Medical Genetics, 119A, 152–155.
Hitzemann, R., Cipp, L., Demarest, K., Mahjubi, E., & McCaughran, J., Jr.
(1998). Genetics of ethanol-induced locomotor activation: Detection of
QTLs in a C57BL/6J x DBA/2J F2 intercross. Mammalian Genome, 9,
Justice, M. J., Carpenter, D. A., Favor, J., Neuhauser-Klaus, A., Hrabe de
Angelis, M., Soewarto, D., et al. (2000). Effects of ENU dosage on
mouse strains. Mammalian Genome, 11, 484–488.
Kapfhamer, D., Valladares, O., Sun, Y., Nolan, P. M., Rux, J. J., Arnold,
S. E., et al. (2002). Mutations in Rab3a alter circadian period and
homeostatic response to sleep loss in the mouse. Nature Genetics, 32,
Koyner, J., Demarest, K., McCaughran, J., Jr., Cipp, L., & Hitzemann, R.
(2000). Identification and time dependence of quantitative trait loci for
basal locomotor activity in the BXD recombinant inbred series and a
B6D2 F2 intercross. Behavior Genetics, 30, 159–170.
Kralic, J. E., Wheeler, M., Renzi, K., Ferguson, C., O’Buckley, T. K.,
Grobin, A. C., et al. (2003). Deletion of GABAA receptor alpha 1
subunit-containing receptors alters responses to ethanol and other anes-
thetics. Journal of Pharmacology and Experimental Therapeutics, 305,
Lister, R. G. (1987). The use of a plus-maze to measure anxiety in the
mouse. Psychopharmacology (Berlin), 92, 180–185.
Majchrowicz, E. (1975). Induction of physical dependence upon ethanol
and the associated behavioral changes in rats. Psychopharmacologia, 43,
Majchrowicz, E., & Hunt, W. A. (1976). Temporal relationship of the
induction of tolerance and physical dependence after continuous intox-
ication with maximum tolerable doses of ethanol in rats. Psychophar-
macology (Berlin), 50, 107–112.
Marubio, L. M., & Paylor, R. (2004). Impaired passive avoidance learning
in mice lacking central neuronal nicotinic acetylcholine receptors. Neu-
roscience, 129, 575–582.
McClearn, G. E., & Rodgers, D. A. (1959). Differences in alcohol prefer-
ence among inbred strains of mice. Quarterly Journal of Studies on
Alcoholism, 20, 691–695.
McIlwain, K. L., Merriweather, M. Y., Yuva-Paylor, L. A., & Paylor, R.
(2001). The use of behavioral test batteries: Effects of training history.
Physiology & Behavior, 73, 705–717.
Melo, J. A., Shendure, J., Pociask, K., & Silver, L. M. (1996). Identifica-
tion of sex-specific quantitative trait loci controlling alcohol preference
in C57BL/ 6 mice. Nature Genetics, 13, 147–153.
Mishima, K., Tanoue, A., Tsuda, M., Hasebe, N., Fukue, Y., Egashira, N.,
et al. (2004). Characteristics of behavioral abnormalities in alpha1d-
adrenoceptors deficient mice. Behavioural Brain Research, 152, 365–
Moore, J. A., & Kakihana, R. (1978). Ethanol-induced hypothermia in
mice: Influence of genotype on development of tolerance. Life Sciences,
National Institutes of Health (1986). Guide for the care and use of labo-
ratory animals (DHEW Publication No. 86–23). Washington, DC: U.S.
Government Printing Office.
Nolan, P. M., Peters, J., Strivens, M., Rogers, D., Hagan, J., Spurr, N., et
al. (2000). A systematic, genome-wide, phenotype-driven mutagenesis
programme for gene function studies in the mouse. Nature Genetics, 25,
O’Connor, C. S., Crawshaw, L. I., Kosobud, A., Bedichek, R. C., &
Crabbe, J. C. (1989). The effect of ethanol on behavioral temperature
regulation in mice. Pharmacology, Biochemistry, and Behavior, 33,
Pandey, S. C., Carr, L. G., Heilig, M., Ilveskoski, E., & Thiele, T. E.
(2003). Neuropeptide y and alcoholism: Genetic, molecular, and phar-
macological evidence. Alcoholism: Clinical and Experimental Research,
Peirce, J. L., Derr, R., Shendure, J., Kolata, T., & Silver, L. M. (1998). A
major influence of sex-specific loci on alcohol preference in C57Bl/6
and DBA/2 inbred mice. Mammalian Genome, 9, 942–948.
Phillips, T. J., & Dudek, B. C. (1991). Locomotor activity responses to
ethanol in selectively bred long- and short-sleep mice, two inbred mouse
strains, and their F1 hybrids. Alcoholism: Clinical and Experimental
Research, 15, 255–261.
Phillips, T. J., Shen, E. H., McKinnon, C. S., Burkhart-Kasch, S., Lessov,
C. N., & Palmer, A. A. (2002). Forward, relaxed, and reverse selection
for reduced and enhanced sensitivity to ethanol’s locomotor stimulant
effects in mice. Alcoholism: Clinical and Experimental Research, 26,
Rinchik, E. M., & Carpenter, D. A. (1999). N-ethyl-N-nitrosourea mu-
SCREENING FOR ETHANOL PHENOTYPES IN ENU MUTANTS
tagenesis of a 6- to 11-cM subregion of the Fah-Hbb interval of mouse
chromosome 7: Completed testing of 4557 gametes and deletion map-
ping and complementation analysis of 31 mutations. Genetics, 152,
Risinger, F. O., Freeman, P. A., Greengard, P., & Fienberg, A. A. (2001).
Motivational effects of ethanol in DARPP-32 knockout mice. Journal of
Neuroscience, 21, 340–348.
Rubinstein, M., Phillips, T. J., Bunzow, J. R., Falzone, T. L., Dziewcza-
polski. G., Zhang, G., et al. (1997). Mice lacking dopamine D4 receptors
are supersensitive to ethanol, cocaine, and methamphetamine. Cell, 90,
Rustay, N. R., Wahlsten, D., & Crabbe, J. C. (2003). Assessment of genetic
susceptibility to ethanol intoxication in mice. Proceedings of the Na-
tional Academy of Sciences of the United States of America, 100,
Shirley, R. L., Walter, N. A., Reilly, M. T., Fehr, C., & Buck, K. J. (2004).
Mpdz is a quantitative trait gene for drug withdrawal seizures. Nature
Neuroscience, 7, 699–700.
Thiele, T. E., & Badia-Elder, N. E. (2003). A role for neuropeptide Y in
alcohol intake control: Evidence from human and animal research.
Physiology & Behavior, 79, 95–101.
Vitali, R., & Clarke, S. (2004). Improved rotorod performance and hyper-
activity in mice deficient in a protein repair methyltransferase. Behav-
ioural Brain Research, 153, 129–141.
Vitaterna, M. H., King, D. P., Chang, A. M., Kornhauser, J. M., Lowrey,
P. L., McDonald, J. D., et al. (1994, April 29). Mutagenesis and mapping
of a mouse gene, Clock, essential for circadian behavior. Science, 264,
Wahlsten, D., Metten, P., Phillips, T. J., Boehm, S. L., II, Burkhart-Kasch.
S., Dorow, J., et al. (2003). Different data from different labs: Lessons
from studies of gene-environment interaction. Journal of Neurobiology,
Winer, B. J., Brown, D. R., & Michels, K. M. (1991). Statistical principles
in experimental design (3rd ed.) Boston: McGraw Hill.
Zhang, B., Kirov, S., & Snoddy, J. (2005). WebGestalt: An integrated
system for exploring gene sets in various biological contexts. Nucleic
Acids Research, 33(Web Server issue), W741–748.
Received June 26, 2006
Revision received March 15, 2007
Accepted March 22, 2007 ?
HAMRE, GOLDOWITZ, WILKINSON, AND MATTHEWS