Abnormal Behavior in a Chromosome-
Engineered Mouse Model for Human
15q11-13 Duplication Seen in Autism
Jin Nakatani,1,10Kota Tamada,1,2,10Fumiyuki Hatanaka,1,3Satoko Ise,4Hisashi Ohta,4Kiyoshi Inoue,1Shozo Tomonaga,1
Yasuhito Watanabe,1,2Yeun Jun Chung,5Ruby Banerjee,5Kazuya Iwamoto,6Tadafumi Kato,6,7Makoto Okazawa,1
Kenta Yamauchi,8Koichi Tanda,8Keizo Takao,8,9Tsuyoshi Miyakawa,8,9Allan Bradley,5and Toru Takumi1,3,7,*
1Osaka Bioscience Institute, Suita, Osaka 565-0874, Japan
2Kyoto University Graduate School of Biostudies
3Department of Molecular Neuroscience
Kyoto University Graduate School of Medicine, Sakyo, Kyoto 606-8501, Japan
4Tsukuba Research Institute, Banyu Pharmaceutical Co. Ltd., Tsukuba, Ibaraki 300-2611, Japan
5The Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK
6Brain Science Institute, RIKEN, Wako, Saitama 351-0198, Japan
7Graduate School of Biomedical Sciences, Hiroshima University, Minami, Hiroshima 734-8553, Japan
8Frontier Technology Center, Kyoto University Graduate School of Medicine, Sakyo, Kyoto 606-8501, Japan
9Division of Systems Medicine, Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Aichi 470-1192, Japan
10These authors contributed equally to this work
Substantial evidence suggests that chromosomal
abnormalities contribute to the risk of autism. The
duplication of human chromosome 15q11-13 is
known to be the most frequent cytogenetic abnor-
mality in autism. We have modeled this genetic
change in mice by using chromosome engineering
to generate a 6.3 Mb duplication of the conserved
linkage group on mouse chromosome 7. Mice with
apaternal duplicationdisplaypoor socialinteraction,
behavioral inflexibility, abnormal ultrasonic vocaliza-
tions, and correlates of anxiety. An increased MBII52
snoRNA within the duplicated region, affecting the
serotonin 2c receptor (5-HT2cR), correlates with
altered intracellular Ca2+responses elicited by a
5-HT2cR agonist in neurons of mice with a paternal
duplication. This chromosome-engineered mouse
model for autism seems to replicate various aspects
of human autistic phenotypes and validates the rele-
vance of the human chromosome abnormality. This
model will facilitate forward genetics of develop-
mental brain disorders and serve as an invaluable
tool for therapeutic development.
Autism is a common and heterogeneous neuropsychiatric
disorder with manifestations of deficit in social interaction,
impaired communication, and repetitive behavior or restricted
interest (Volkmar and Pauls, 2003). Its definition has been
autism-related disorders such as Asperger disorder and Rett’s
syndrome (DiCicco-Bloom et al., 2006; Geschwind and Levitt,
2007; Lord et al., 2000; Veenstra-VanderWeele et al., 2004).
Autism is now considered a developmental brain disease (Bel-
monte et al., 2004; DiCicco-Bloom et al., 2006; Dykens et al.,
2004; Geschwind and Levitt, 2007; Maestrini et al., 2000; Veen-
stra-VanderWeele and Cook, 2004; Vorstman et al., 2006). The
first signs of autism appear at around 6 months, full diagnosis
is usually made at 3 years, and symptoms usually persist
throughout life. Autism is one of the most heritable neuropsychi-
atric disorders (Geschwind and Levitt, 2007), suggesting that
genetic factors play an important role in its etiology (Vorstman
et al., 2006). Candidate chromosomal regions and specific
genes have been investigated (Belmonte et al., 2004; Folstein
and Rosen-Sheidley, 2001; Persico and Bourgeron, 2006; Pol-
leux and Lauder, 2004). There are likely to be de novo mutations,
chromosomal abnormalities and common genetic variants that
contribute to the genetic etiologies of autism (Abrahams and
Geschwind, 2008; Geschwind, 2008; Geschwind and Levitt,
2007; Ramocki and Zoghbi, 2008). Several knockout mice
have also been reported as ‘‘putative’’ autistic models, as
judged from their phenotypes, but the molecular mechanism
responsible for the pathophysiology of autism is far from
Abnormalities of chromosomes are thought to account for
10% to 20% of autism cases (Beaudet, 2007). A recent study
has established de novo germline mutations including copy
number variants (CNVs) as a more significant risk factor for
ASD than previously recognized (Sebat et al., 2007). Paternally
or maternally inherited deletions of human chromosome
15q11-13 occur quite frequently, when they affect the imprinted
region this is recognized as Prader-Willi syndrome or Angelman
to autism spectrumdisorder(ASD),including
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1235
syndrome, respectively (Nicholls and Knepper, 2001). Duplica-
tion of the same region is the only recurrent cytogenetic aberra-
tion associated with autism, occurring in up to 5% of autism
cases (Belmonte et al., 2004; Bolton et al., 2004; Cook and
2001; Lord et al., 2000; Maestrini et al., 2000; Veenstra-Vander-
Weele et al., 2004; Veenstra-VanderWeele and Cook, 2004;
Vorstman et al., 2006).
On the basis of conserved human/mouse linkage, we have
generated mice with a 6.3 Mb duplication of mouse chromo-
some 7mirroring thehuman chromosome 15q11-13duplication.
This mouse model displays several phenotypes that recapitulate
aspects of the human condition and provides mechanistic
insight into the disease.
Construction of a 6.3 Mb Duplication
on Mouse Chromosome 7
mouse chromosome 7 (Figure 1). Chromosomal engineering
(van der Weyden and Bradley, 2006) was used to construct an
interstitial duplication of mouse chromosome 7 corresponding to
the region between common breakpoints in human chromosome
15q11-13. Sequential rounds of insertional gene targeting were
used to insert the selection cassettes and loxP sites required for
chromosome engineering proximal to the Herc2 and distal to
Figure 1. Human Chromosome 15q11-13 and Mouse Chromosome 7
Schematic representation of the genomic regions in the human and mouse genomes. Details of conserved linkage in human 15q11-13 and mouse chromosome
7 are shown. The paternally, maternally expressed, and nonimprinting genes were labeled with blue, red, and green, respectively. The two arrowheads (BP)
represent the common breakpoints, and the two arrows represent the targeting sites of 2 loxP sequences. Genomic segments that show linkage conservation
(i.e., identical gene order) in humans and mice are connected by dark shading if the gene orders are in the same direction relative to their respective centromeres.
If the gene orders in the syntenic segments are in opposite orientations, they are connected by light shading.
1236 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.
transiently transfected with a Cre expression plasmid to induce
recombination between the loxP sites, which generated clones
with a balanced duplication (Dp) and deletion (Df) (Figure 2C). Re-
dine (HAT) selection at a frequency of 26 3 10?7per electropo-
rated cell and confirmed via Southern blot analysis (Figures 2B
and 2D) and fluorescence in situ hybridization (FISH) (Figure 2E).
The deletion and duplication alleles were transmitted and estab-
lished in the germ line by standard procedures, and the expected
increase in genomic copy number of this region was confirmed
rial artificial chromosome (BAC) microarray (Figure 2F).
Increased Gene Expression of Duplicated Genes
from both females and males. Mice with a maternally (matDp/+)
and paternally (patDp/+) inherited duplication bred normally and
were fertile. The patDp/+ male mice began to show an increase
in body weight compared to wild-type (WT) mice after 15 weeks,
and the body weight of patDp/+ was significantly greater than
that of the WT after 20 weeks (data not shown).
that of brains at postnatal day 0 (P0), P7, and P14 to screen for
in H&E-stained sections of the olfactory bulb, cerebral cortex,
hippocampus, amygdala, corpus callosum, and cerebellum
either macroscopically or at the microscopic level (Figure S1
available online). The number of Purkinje cells in the cerebellum
was not significantly different between mice with the duplication
and WT mice (Figure S2). Bodian staining was also performed,
and this did not reveal any significant abnormality in the cortex,
hippocampus, amygdala, and cerebellum (data not shown).
Figure 2. Engineering an Interstitial Duplication on Chromosome 7
(A) Insertional double targeting and genomic coordinates, NCBI build m37. S, SacI; T, Tth111I.
(B) Southern blot analysis of ES cell DNA samples.
(C) Cre/loxP recombination generates duplication and deletion chromosomes.
(D) Southern blot analysis of ES cell DNA confirming the duplication.
(E) Confirmation by FISH. The probes used are shown on the left. The red and green bars represent the probes located within and outside the duplicated region,
respectively. The white arrow and arrowhead represent the Dp and Df alleles, respectively.
(F) A BAC array-CGH profile of chromosome 7 from mice with the duplication. Log2-transformed hybridization ratios of duplicated mouse DNA versus WT DNA
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1237
The 6.3 Mb duplication includes the region of genomic
imprinting. The relative expression levels of genes in the dupli-
cated region are expected to vary depending on whether they
are imprinted and on their mode of inheritance (Figure 3A).
Therefore, gene expression was assessed by quantitative RT-
PCR in the brains from patDp/+ and matDp/+ mice, and the
results were normalized to those in WT mice (Figure 3B). In the
adult brain, Ndn, Snrpn, Ube3a, Gabrb3, Gabra5, and Herc2
genes in the duplicated region were highly expressed, whereas
the expression of others was relatively less abundant. The
messenger RNA (mRNA) levels of the paternally expressed
genes, Ndn, and Snrpn, were increased more than 2-fold in the
patDp/+ mice. Unexpectedly, Ndn also exhibited increased
Snrpn showed the expected level (no change). The maternally
expressed gene, Ube3a, showed an approximately 2-fold
increase in matDp/+ mice. The expression of Atp10a did not
show any significant difference between patDp/+ and matDp/+
mice. The nonimprinting genes, GABAAreceptor subunits and
Herc2, showed the expected 1.5-fold increase in mice with the
duplication. Chrna7 and Tubgcp5, both of which are located
outside the duplicated region, did not show any significant
phenotypes reflected those in the brain (Figure S3). The expres-
sion of each gene in various areas of the adult mouse brains was
examined by in situ hybridization (Figure 3C). Alterations in the
expression patterns of these genes in patDp/+ and matDp/+
brains were unremarkable, although the expression level of
Snrpn in the hippocampus seemed to be higher in the patDp/+
mice, whereas that of Ube3a seemed to be higher in the
DNA methylation is an epigenetic modification in imprinted
regions and is found in 15q11-13 (Nicholls and Knepper, 2001).
locus. This region is methylated (Me) or unmethylated (UnMe) in
maternal and paternal alleles, respectively. We thus examined
allele-specific methylation by Southern blotting with a Snurf
probe by using methylation-sensitive (BssHII) and -insensitive
(HpaI) restriction enzymes. The ratios of Me and UnMe bands
were 1:2 and 2:1 in patDp/+ and matDp/+ mice, respectively,
Figure 3. Gene Expression in Mice with the Duplication
(A) Expected gene expression levels in wild-type (WT), paternal duplication (patDp/+), and maternal duplication (matDp/+) mice.
(B) mRNA expression in the mouseadult brain of the listed genes analyzed by quantitative RT-PCR. The relative expression levels ofpatDp/+ (n = 4) and matDp/+
(n = 4) were compared with WT (n = 7) normalized to 1.0. Blue, red, and green indicate paternally expressed, maternally expressed, and nonimprinted genes,
respectively. Dotted lines show the boundaries of the chromosomal rearrangement. Error bars represent the standard error of the mean (SEM). **p < 0.0001,
*p < 0.05.
(C) Snrpn, Ube3a, and Gabra5 mRNA expression in the adult mouse brain (top row) and hippocampus (other rows) detected by in situ hybridization. Scale bars
represent 2 mm (top row) and 200 mm (other rows).
(D) Methylation analysis by Southern blotting. Me, methylated; UnMe, unmethylated.
1238 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.
in contrast to 1:1 in the WT (Figure 3D). These results suggest
that allele-specific methylation is conserved in the mice with
the duplicated allele.
patDp/+ Mice Display Social Abnormalities
To analyze the effect of the chromosomal duplication on
behavior, we performed a comprehensive battery of behavioral
tests (Crawley, 2007; Takao et al., 2007; Yamasaki et al.,
2008). We observed significant differences between the WT
and mice with a duplication in the several tests described below
(Tables S1 and S2). The diagnosis of autism is based on behav-
ioral criteria (Volkmar and Pauls, 2003). Therefore, a valid mouse
model should reflect behavioral symptoms, including impair-
ment in social interaction (Crawley, 2004; Moy et al., 2006).
A three-chamber social interaction test (Crawley, 2004; Nadler
et al., 2004) was performed (Figure 4A). The mouse to be tested
was placed in the central chamber and could move freely among
the three chambers. A stranger mouse was placed in one of the
side chambers in a wire cage, and only a cage was placed in the
opposite chamber. WT mice tended to contact the stranger
mouse, and the time spent with the stranger mouse in the quad-
rant location depicted by the line in Figure 4A was significantly
higher than the time spent in the corresponding location in the
opposite chamber with the empty cage (Figure 4B). In contrast,
the patDp/+ mice exhibited no significant difference in time
spent between the quadrant spaces of either side (Figure 4B).
These phenotypes were also observed in mice with a different
background under similar experimental conditions (Figure S4).
To confirm that these results are due to specific changes in
social behavior, we further performed the three-chamber test
under different conditions. First, we assessed the reaction of
mice to a novel inanimate object. Both patDp/+ and WT mice
spent more time around the cage with a novel inanimate object
compared with the empty cage, and no significant difference
between patDp/+ and WT mice was observed (Figures 4C and
S5A). Second, the simultaneous interactions with a novel mouse
and another novel object were compared. WT mice spent more
time around the cage with a novel mouse than with a novel
object, whereas patDp/+ mice showed no significant difference
in time spent around the cages with a novel mouse and object
(Figures 4D and S5B). Third, the interactions of the mice with
a novel and familiar mouse were also compared. In WT mice,
although the time spent around the cages was not significantly
altered, the number of entries around the novel mouse tended
Figure 4. patDp/+ Mice Show Social Abnormalities
(A–E) Three-chamber test.
(A) Schematic representation of the three-chambered apparatus. The quadrant-like spaces between the full and dotted lines were used for quantitative analysis.
(B) A stranger mouse was restricted in one of the side chambers in a wire cage, and only an empty wire cage was placed in the opposite chamber. Comparison of
time spent in the quadrant spaces between ‘‘Stranger’’ and ‘‘Cage’’ for WT (n = 14) and patDp/+ mice (n = 13) is shown. Error bars represent the SEM. *p < 0.05.
(C)A novel objectA (adodecahedralpole) wasplaced inacage inthechamberonone side, and noobject inthechamberontheotherside. Both WT and patDp/+
mice spent more time around the cage with a novel object. n = 11. **p < 0.001.
with the subject (test mouse) in a cage in the chamber on the other side. WT mice spent more time around the stranger mouse. n = 11. *p < 0.05.
(E) A novel stranger mouse (C57BL/6J) is placed in a cage in the chamber on one side and a familiar mouse that was used in a previous test in D is placed in the
chamber on the other side. n = 11. These data were evaluated by the t test.
(F) Maternal separation-induced ultrasonic vocalizations at P5, P7, P14, and P21 (or P22). n = 32, 40, 40, and 16, respectively for patDp/+; n = 24, 39, 39, and 12,
respectively for the WT. **p < 0.005. For (B)–(F) error bars represent the SEM.
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1239
to be greater than that of the familiar mouse (p = 0.0541),
whereas in patDp/+ mice, no significant difference between
the novel and familiar mouse was found (Figure 4E and S5C).
These results suggest that WT mice are more interested in
a novel mouse than a novel inanimate object, but that patDp/+
mice have decreased sociability compared with the WT, which
may be analogous to the impairment in appropriate social inter-
action often seen in autistic patients (Crawley, 2004). On the
other hand, matDp/+ mice were indistinguishable from WT
olfactory driven, we examined the olfactory system of the mice
anatomically (by immunohistochemistry) and functionally (by
olfactory habituation/dishabituation test), and we excluded any
defects of the olfactory system in patDp/+ mice (data not
To measure behavioral flexibility, we habitually trained mice
and then analyzed their responses to a change in routine in
a reversal task by using the Morris water maze test and the
Barnes maze test, which have been generally validated for
spatial learning and memory (Crawley, 2007; Miyakawa et al.,
the mouse swims to find a hidden platform. Mice were trained to
locate the correct platform to escape from the water. Both
patDp/+ and WT mice learned the target quadrant (TA in Figures
5A–5C), suggesting that there was no impairment in spatial
learning in patDp/+ mice. When the target platform was then
moved to the opposite area (TA in Figure 5D), WT mice spent
significantly more time in the new TA quadrant compared with
the opposite quadrant (OP) (Figure 5E). On the other hand,
patDp/+ mice exhibited no difference between time in the TA
and OP (Figure 5F).
The Barnes maze is a circular white platform with 12 holes
(Figures 5G and 5J). One of the holes exits into a dark box called
the target initially placed at 0 degrees (Figure 5G). Mice were
trained to locate the correct hole to exit into the escape box.
Both patDp/+ and WT mice learned to identify the target at the
0 degree point, and there was no observable difference between
the two mice (Figures 5H and 5I), again suggesting that there
was no impairment in spatial learning of patDp/+ mice. When
the target was moved to the opposite side (Figure 5J), both
patDp/+ and WT mice could find the target; however, compared
with the WT, patDp/+ mice stayed less in the new target position
and more in the 180 or ±150 degree position, which is the posi-
tion of the original target or in its direction (Figure 5L). The time
spent between the target and the 180 degree position was
significantly different in WT mice (Figure 5K), whereas there
was no significant difference in patDp/+ mice (Figure 5L). The
Barnes maze test was also performed in mice with a different
background, and the results were consistent (Figures S7A and
S7B). Conversely, matDp/+ mice did not show any significant
change even in reversal learning compared with WT mice
(Figures S8A and S8B). These results suggest that patDp/+
mice do not respond as flexibly as WT and matDp/+ mice to
a change in situation, which may be comparable to the inflexi-
bility in routine that is characteristic of autism (Crawley, 2004),
although we should acknowledge that it is far from clear how
cognitive deficits in reversal learning are related to the behav-
ioral deficits in autism even in humans and even less clear
from mouse to humans (Geurts et al., 2009). Furthermore, we
found the lack of reversal deficits in the T maze test (data not
shown). Perhaps reversal deficits are only apparent during aver-
sively motivated escape behaviors and not appetitively moti-
vated approach behaviors.
To see impairment in communicative behavior, we measured
ultrasonic vocalizations (USVs) of neonatal mice that were sepa-
rated from their dams (Figure 4F). These USVs are thought to be
distress signals (Crawley, 2007) and may be related to commu-
nication between a dam and her pups (Crawley, 2004). In WT
mice, the USVs have a normal developmental course, emerging
soon after birth, peaking at around P5, and then decreasing to
almost zero at around P14, the time of eye opening when the
development of alternative communication may occur (Noirot,
1966). The USVs emitted by patDp/+ pups at P7 and P14
were markedly greater than those of WT pups (Figure 4F). In
the patDp/+ pups, the peak in the numbers of USVs seemed
to be delayed, and the USVs were still present at P14, when
those of WT mice had disappeared, suggesting that patDp/+
mice may be developmentally abnormal in comparison with
the WT. Detailed frequency analysis revealed that most of the
USVs at P7 and P14 emitted by patDp/+ pups were mainly in
the 50–70 kHz frequency range, with some over 70 kHz, the
latter of which was not seen in WT pups (Figure S9). No differ-
ence in USVs was observed between matDp/+ and WT pups
(Figure S8C), suggesting that communicative development
between dam and neonatal mice in patDp/+ is different from
that in WT or matDp/+ pups. This larger number of USVs in
patDp/+ pups may reflect higher anxiety and fear in response
to stress (Crawley, 2007).
We therefore examined vocalizations in adult animals, where
the effects of anxiety and novel environments may be more
controlled. Since adult mice emitted both audible and ultrasonic
vocalizations, we measured vocalizations consisting of both
frequencies. In a resident-intruder paradigm, the total number
of vocalizations ranging from both audible and ultrasonic bands
emitted by pairs of patDp/+ mice was significantly decreased
compared with that of a WT pair (Figure S10). The vocalizations
ranging in the ultrasonic bands in pairs of patDp/+ mice also
tended to be lower than those in pairs of WT mice. The behavior
between resident and intruder mice in each genotype was indis-
tinguishable between the genotypes. These results suggest that
vocal communication between pairs of patDp/+ mice is
decreased compared with a WT pair.
We observed the fear-related behaviors of patDp/+ mice by
conducting another test, the cued and contextual conditioning
task, by using a 60 dB white noise tone and a mild foot shock
(Figures S11A–S11C). No significant difference in the freezing
rate between patDp/+ and WT mice was seen during condi-
tioning training (Figure S11A) or in the same contextual environ-
ment after 24 hr (Figure S11B). However, patDp/+ mice showed
did the WT controls, especially during the first 3 min in the
absence of the cue (Figure S11C). matDp/+ mice displayed no
difference from WT mice (Figures S12A–S12C). These results
suggest that the patDp/+ mice show a generalized fear. Addi-
tionally, we conducted the elevated plus maze test to examine
anxiety (Figures S11D–S11G). As compared with the WT, the
1240 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.
Figure 5. patDp/+ Mice Show Behavioral Inflexibility in the Morris Water Maze and Barnes Maze Tests
(A–F) Morris water maze test; n = 20 for both genotypes. White bar, WT; black bar, patDp/+.
(A)Theconfigurationofthefourquadrantsintheprobetestaftertheoriginaltraining(TA,target quadrant;OP,oppositequadrant;AR,adjacent rightquadrant;AL,
adjacent left quadrant).
(B and C) Probe test after the original training. Upper panels indicate averaged swimming traces of the swim pattern for WT (B) and patDp/+ (C) mice. Warmer
color represents more time spent. Lower panels show the quadrant occupancy for WT (B) and patDp/+ (C) mice. Both WT and patDp/+ mice showed significantly
more time spent in the target quadrant compared with the other quadrants [WT, F(3,76) = 12.86, p < 0.0001; patDp/+, F(3,76) = 13.31, p < 0.0001; Newman-Keuls
post hoc comparison (trained quadrant more than all the other quadrants); p < 0.01 for both genotypes].
(D) The configuration of the four quadrants in the reversal probe test.
(E and F) Reversal probe test. Upper panels indicate averaged swimming traces of the swim pattern for WT (E) and patDp/+ (F) mice. Lower panels show the
quadrant occupancy for WT (E) and patDp/+ (F) mice. While WT mice spent significantly more time in the reversed target quadrant, patDp/+ mice showed no
significant difference in the time spent between the quadrants [WT, F(3,76) = 8.20, p < 0.0001; patDp/+, F(3,76) = 2.40, p = 0.0745; Neuman-Keuls post hoc
comparison (trained quadrant more than all the other quadrants); WT, p < 0.01; patDp/+, p > 0.05].
(G–L) Barnes maze test, n = 22 for both genotypes. White bar, WT; black bar, patDp/+.
(G) The target position in the Barnes maze original probe test. The hole at 0 degrees is the correct hole chosen as the target.
(H and I) Both genotypes could learn the target position spatially in the original probe test [WT, F(11,252) = 25.47, p < 0.0001; patDp/+, F(11,252) = 32.27,
p < 0.0001; Bonferroni post hoc comparison (time spent around the target position more than those of all the other holes), both genotypes, p < 0.01].
(J) The target position in the Barnes maze reversal probe test. The target at 0 degrees is moved to the opposite position. The original target position is labeled in
red, at 180 degrees, and the new target position is labeled in blue, at 0 degrees.
(K and L) While WT mice could learn the new target position flexibly, patDp/+ mice could not respond as flexibly as WT mice [WT, F(11,252) = 29,08, p < 0.0001;
patDp/+, F(11,252) = 16.04, p < 0.0001; Bonferroni post hoc comparison (target versus 180 degrees), WT, p < 0.01; patDp/+, p > 0.05]. *p < 0.01; n.s., not signif-
icant (p > 0.05). Furthermore, time spent around the 180 degree position and its neighboring 150 degree position was increased in patDp/+ mice compared to the
WT (180 degrees, p < 0.1; 150 degrees, p < 0.05).
For (B), (C), (E), (F), (H), (I), (K), and (L), error bars represent the SEM.
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1241
patDp/+ mice showed no significant difference in distance trav-
eled (Figure S11D), whereas the number of entries into the arms
and time in the open arms were significantly decreased in the
patDp/+ mice (Figures S11E and S11F), but not in the matDp/+
mice (Figures S12D–S12G). These results suggest that the
patDp/+ mice show increased anxiety, a feature common in
autistic individuals (Crawley, 2004).
5-HT2c Receptor Signaling Is Altered
in patDp/+ Neurons
Our behavioral tests demonstrated that patDp/+ mice show
abnormal behaviors. The primary benefit of the model mouse
system is that it allows us to study abnormality at the molecular
level. To demonstrate this possibility, we examined one molec-
ular candidate, i.e., a small nucleolar RNA (snoRNA). It has
been reported that a brain-specific snoRNA, HBII52, the human
ortholog of MBII52, plays a role in posttranscriptional modifica-
tion of the serotonin 2c receptor (5-HT2cR), a G protein-coupled
receptor (GPCR) (Kishore and Stamm, 2006), which may cause
amino acid substitutions in the second intracellular domain of
this receptor. The 5-HT2cR is the only GPCR that has been
shown to undergo physiologically important editing of its pre-
mRNA by adenosine deamination (A-to-I editing), resulting in
amino acid substitutions (Seeburg, 2002). We examined
MBII52 RNA expression in the brain by RNA blot hybridization
(Figure 6A). Since the locus including MBII52 is maternally im-
printed, the expression of MBII52 in the patDp/+ mouse brains
was approximately twice as much as that in the WT or matDp/+
potential sites that are located in the second intracellular
Figure 6. [Ca2+]iResponse by a 5-HT2cR Agonist in Neurons
(A) Northern blot analysis of MBII52. Quantitative data are shown in the right panel, where MBII52 expression in WT is defined as 1.
(B–D) The effect of WAY 161503 on [Ca2+]iin primary cultured neurons. Representative images (responding cells are indicated by an arrowhead) and average
responses under various concentrations of agonist are shown in (B) and (C), respectively. Averaged data for the concentration-dependent effect of WAY
161503 are indicated in (D). Error bars represent the SEM. n = 17 for patDp/+, n = 15 for matDp/+, n = 18 for WT. **p < 0.001.
1242 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.
domain. RNA editing ratios in patDp/+ at the A and B sites were
significantly higher than those in the WT, and patDp/+ editing
ratios at the D site were higher than matDp/+, whereas no signif-
icant difference was found for the editing frequency at the E and
C sites among the three types of mice (Figure S13).
Because 5-HT2cR induces an increase in the intracellular
calcium level ([Ca2+]i) via G proteins coupled to phospholipase
C, we asked whether altered amounts of MBII52 would affect
the [Ca2+]i response via altered coupling efficiency between
5-HT2cR and G proteins. To analyze the serotonergic signals
in neurons derived from mouse brains, we examined the effects
of 5-HT2cR on [Ca2+]iin primary cultured neurons by using mi-
crospectrofluorimetric techniques and the fluorescent indicator
Fura-2 (Figure 6B). A specific agonist for 5-HT2cR, WAY
161503, induced an increase in [Ca2+]i. The response to 100
nM WAY 161503 in patDp/+ neurons was significantly higher
than that in the WT (Figures 6C and 6D). These results demon-
strate that substantial alterations in the amount of MBII52 RNA
of the patDp/+ mice resulted in a significantly increased [Ca2+]i
response to 5-HT2cR signaling, suggesting that this alteration
in serotonergic signaling may contribute to the abnormal
behavior seen in the patDp/+ mice.
Idealanimalmodels ofhumanneuropsychiatric disorders should
not only phenocopy relevant human symptoms, but the pheno-
types should also be based on similar underlying mechanisms
acting both physiologically and genetically (Crawley, 2004).
Several kinds of animal models for autism have been reported
(Moy et al., 2006; Murcia et al., 2005; Persico and Bourgeron,
2006). Knockouts or knockin of single candidate genes, such
as genes in the oxytocin-vasopressin system, dishevelled-1
as possible autistic model mice (DiCicco-Bloom et al., 2006; Ja-
main et al., 2008; Kwon et al., 2006; Lijam et al., 1997; Moretti
et al., 2005; Tabuchi et al., 2007; Winslow and Insel, 2002;
Young, 2007). Ours mirrors a chromosomal abnormality found
in human autistic patients. In this regard, the chromosome-engi-
neered mouse described here is a model mouse for autism that
parallels both phenotypic and genotypic aspects of the human
In the rotarod test, patDp/+ mice exhibited a significantly
greater improvement of rotarod performance than did WT mice
(Figure S14). This result may simply mean that patDp/+ mice
possess higher motor coordination/learning ability compared
with WT mice, but taken together with the results of the reversal
learning, it can be interpreted that patDp/+ show better stereo-
typic behavior. This motor stereotypy or better performance in
repetitive tests of motor coordination has also been reported in
other models (Caston et al., 1998; Kwon et al., 2006). In addition
to the major symptoms, there are several associated manifesta-
tions of emotional behavior in autism, such as anxiety, fear, and
depression. Indeed patDp/+ mice displayed these signs in the
cued and contextual conditioning fear test, the elevated plus
maze test, and the Porsolt forced swim test (Figures S11 and
S15). Mao et al. reported that a patient with paternal duplicated
15q11-13 displayed depression and anxiety in addition to signif-
icant behavioral problems and obesity (Mao et al., 2000).
Furthermore, in the eight-arm radial maze test, we noticed
restriction, several mice did not seem to be eager to eat food. In
addition, the latency of patDp/+ mice to approach food was
findings might reflect the increase in latency to feed observed in
the novelty-suppressed feeding (NSF) test (Santarelli et al.,
2003). These behavioral phenotypes may imply that patDp/+
mice have greater fear and tend to freeze in novel environments
or have a lack of desire.
In this study, mice with a paternal duplication showed
abnormal phenotypes compared with WT mice. Reports on
human autism associated with a paternal duplication have
been accumulating (Bolton et al.,2004; Mao et al., 2000; Mohan-
das et al., 1999; Roberts et al., 2002; Veltman et al., 2005),
although it has also been reported that maternal duplication of
15q11-13 causes autism in humans (Cook et al., 1997). Provided
that autistic patients with the chromosome 15q11-13 duplica-
tions are the small affected cases compared with overall autistic
patients, one should re-evaluate more clinical cases with the use
of currently available high-resolution genome analysis tech-
niques such as array CGH, as well as multiple oligonucleotide
array platforms (Lee and Lupski, 2006). Some epigenetic
controls may be different between human and mouse. Although
the methylation status revealed by analyzing one probe around
the IC region in this study seems to be conserved also in the
mouse, methylation in other regions remains unknown. Epige-
netic, developmental, and environmental influences may affect
marked variability in phenotypic expression (Veltman et al.,
The link between social behaviors in rodents and social
behavior in humans is difficult to establish. Our model would
provide a powerful tool to explore its mechanism. It has been
reported that serotonin may be involved in the pathophysiology
of autism, because serotonin plays a role as a growth factor
in the immature brain (Bonnin et al., 2007; Riccio et al., 2008).
Increased serotonergic activity during development could
damage the neurocircuitry involved in emotional responses to
social stress and may have relevance to the symptoms of autism
(Whitaker-Azmitia, 2005). The 5-HT2cR studied here, mapped to
the X chromosome, may be a candidate molecule for human
genetic studies of autism, and its ligand may be a potential
lead for therapeutic targets. Another intriguing hypothesis is
imbalance between excitatory and inhibitory neural signals at
the developmental stages (Dykens et al., 2004; Levitt et al.,
2004; Polleux and Lauder, 2004; Rubenstein and Merzenich,
2003). In this respect, a cluster of the GABAAreceptor subunits
in the duplicated region and its relevance to development is of
particular interest for further study. It remains possible that other
genes in this duplicated region and their downstream effects
may cause abnormal behavior. Systematic approaches such
as using a series of BACs tiled across the region to make trans-
genic mice will help to resolve these questions. Our model
mouse will be valuable not only for therapeutic studies but also
provides a starting point for more detailed genetic analysis
directed toward understanding the etiology of developmental
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1243
A Chromosome-Engineered Mouse Model
The detailed procedure of the Cre/loxP chromosomal engineering system was
described previously (van der Weyden and Bradley, 2006; Zheng et al., 1999).
Genomic DNA was derived from male 129S5 mice. The 50hprt (hypoxanthine
phosphoribosyl transferase) library vector carries the neomycin resistance
gene for gene targeting, a loxP site, the 50hprt minigene for chromosome engi-
neering, and a Tyrosinase minigene for coat color tagging. The 30hprt library
backbone contains the puromycin resistance gene, a loxP site, 30hprt, and an
Agouti transgene under the control of the K14 promotor. Each rearrangement
requires the successive targeting of two end points with complementary halves
of the Hprt minigene and different positive selection marker. By recombination,
the Hprt minigene is reconstituted so that cell with rearranged chromosomes
can be selected with HAT media. The 50hprt and 30hprt libraries were screened
with a 440 bp fragment between mouse Mkrn3 and Frat3 genes and a 930 bp
fragment between Herc2 and Shyc genes, respectively. Two targeting vectors
by electroporation, confirming the structure of the recombinant chromosome at
double-targeted ES cells were used to induce the rearrangement. The Cre
expression vector pOG231 was electroporated into these cells, and the recom-
bination products were selected with HAT medium. The clones carrying the
duplication were injected into 3.5 day blastocysts from C57BL/6-TyrcBrd/cBrd
mice. Chimaeras that are generated from blastocyst injection are mated with
approved by the Osaka Bioscience Institute Animal Research Committee.
CGH by BAC Microarray
The mouse whole-genome BAC array used in this study contained 2803
unique BAC clones from mouse genomic libraries spaced at 1 Mb intervals
(Chung et al., 2004). Detailed conditions are included in the Supplemental
Quantitative Real-Time Reverse Transcription PCR
The quantitative assays for mRNA expression were described previously
(Yamamoto et al., 2005). TaqMan Low Density Array (Applied Biosystems),
which contained mouse Tubgcp5, Herc2, P, Gabrg3, Gabra5, Gabrb3,
Atp10a, Ube3a, Magel2, Mkrn3, Chrna7, and 18S rRNA (internal control), was
examined with an ABI PRISM 7900HT Sequence Detection System (Applied
Biosystems). Each quantification of relative RNA levels by the SYBR Green
real-time PCR technology was done as described previously (Akashi and
Takumi, 2005). The PCR primers and detailed procedures are included in the
Supplemental Experimental Procedures.
Three-Chambered Social Interaction
Social testing apparatus consisted of a rectangular, three-chambered box and
a lid with an infrared video camera (Nadler et al., 2004) (Ohara & Co., Tokyo).
Each chamber was 20 3 40 3 22 cm, and the dividing walls were made from
clear Plexiglas, with small square openings (5 3 3 cm) allowing access into
each chamber. An unfamiliar C57BL/6J male (stranger) that had no prior
of stranger in the left versus right side chamber was systematically alternated
between trials. The stranger mouse was enclosed in a small, round wire
cage, which allowed olfactory, visual, auditory, and tactile contacts but did
not allow sexual and deep contacts. The subject mouse was first placed in
the middle chamber and allowed to explore the entire social test box for
around wire cage by a camera, which is attached at the top of box. More
Morris Water Task
The visible platform, hidden platform, probe test, and reversal probe test
components of the Morris water task were conducted in a circular pool,
1.0 m in diamater (Ohara & Co.). Detailed conditions are included in the
Supplemental Experimental Procedures.
Barnes Maze Task
The Barnes maze task was conducted on ‘‘dry land,’’ a white circular surface,
1.0 m in diameter, with 12 holes equally spaced around the perimeter (Miya-
kawa et al., 2001) (Ohara & Co.). Detailed conditions are included in the
Supplemental Experimental Procedures.
less steel cylinder (size 7.5 cm diameter 3 7 cm height) on the COOL PLATE
(NCP-2215, Nisshin Rika Co., Ltd.) which maintained temperature of the
cylinder at 24?C in a sound proof room (AT-81, RION Co., Ltd.). The number
of vocalizations was measured for 5 min. More detailed conditions are
included in the Supplemental Experimental Procedures.
Calcium Measurement in Neuronal Cell Culture
The procedurefor primaryculture of neurons was described previously(Yoshi-
mura et al., 2006). The neurons were prepared from embryonic mice brain at
E16 and plated onto poly-L-lysine-coated glass bottom dishes. All measure-
ments were performed within 7–9 days from preparation. For measurement
of the intracellular calcium, primary cultured neurons were loaded with 5 mM
Fura-2 acetoxymethyl ester (Dojindo) at room temperature for 30 min. Cells
tion (135 mM NaCl, 5 mM KCl, 2 mM CaCl2, 2 mM MgCl2, 10 mM HEPES, and
10 mM glucose, adjusted at pH 7.4 with NaOH) by gravity. This chamber was
mounted on the stage of an inverted fluorescence microscope (Axiovert 135,
Zeiss). Various concentrations (0.1 nM, 1 nM, 10 nM, 100 nM, and 1000 nM)
of WAY 161503 (Rosenzweig-Lipson et al., 2006) (Tocris) were perfused for
3 min with a 10 min interval. With a digital image analysis system (MetaFluor,
Molecular Devices), the fluorescence ratio (340 nm/380 nm) for each neuron
was analyzed. For data analysis, the neurons displaying above 0.02 on the
intensity (D 340 nm/380 nm) at 1 mM WAY 161503 were selected.
Statistical analysis was conducted with StatView (SAS institute). Data were
analyzed by two-way ANOVA, or two-way repeated-measures ANOVA, or
one-way ANOVA followed by Bonferroni-Dunn test unless noted otherwise.
Values in tables and graphs were expressed as mean ±SEM.
All the detailed procedures were included in the Supplemental Experimental
Supplemental Data include Supplemental Experimental Procedures, 15
figures, and two tables and can be found with this article online at http://
T.T. acknowledges M. Young for discussions that led to start this project and
authors thank K. Sakimura and M. Abe for their help in the initial stage of this
study; N. Nakai, A. Yamamoto, T. Sudo, F. Law, A. Beasley, E. Grau, T. Ham-
ilton, L. Davis, H. Kitson, H. Ogino, and R. Takayama for technical assistance;
comments; and all members of the Takumi laboratory. Wealso thankK. Nakao
and members of Animal Resource Unit, RIKEN Center for Developmental
for the Promotion of Science fellowship. This work was supported in part by
a grant from the Grants-in-Aid for Scientific Research on Priority Areas-
Research on Pathomechanisms of Brain Disorders from the Ministry of Educa-
tion, Culture, Sports, Science and Technology, the Neuroinformatics Japan
Center, the Institute for Bioinformatics Research and Development, and
Core Research for Evolutional Science and Technology of the Japan Science
and Technology Agency, and by research grants from the Mitsubishi
1244 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.
Foundation, the Mother and Child Health Foundation, the Mitsubishi Pharma
Research Foundation, the Takeda Science Foundation, the Astellas Founda-
tion for Research on Metabolic Disorders, Sony Corporation, Nippon Boeh-
ringer Ingelheim Co., Ltd, and the Wellcome Trust.
Received: April 11, 2008
Revised: September 9, 2008
Accepted: April 3, 2009
Published: June 25, 2009
the threshold of a new neurobiology. Nat. Rev. Genet. 9, 341–355.
Akashi, M., and Takumi, T. (2005). The orphan nuclear receptor RORalpha
regulates circadian transcription of the mammalian core-clock Bmal1. Nat.
Struct. Mol. Biol. 12, 441–448.
Beaudet, A.L. (2007). Autism: highly heritable but not inherited. Nat. Med. 13,
Belmonte, M.K., Cook, E.H., Jr., Anderson, G.M., Rubenstein, J.L., Green-
ough, W.T., Beckel-Mitchener, A., Courchesne, E., Boulanger, L.M., Powell,
S.B., Levitt, P.R., et al. (2004). Autism as a disorder of neural information pro-
cessing: directions for research and targets for therapy. Mol. Psychiatry 9,
Bolton, P.F., Veltman, M.W., Weisblatt, E., Holmes, J.R., Thomas, N.S.,
Youings, S.A., Thompson, R.J., Roberts, S.E., Dennis, N.R., Browne, C.E.,
et al. (2004). Chromosome 15q11-13 abnormalities and other medical condi-
tions in individuals with autism spectrum disorders. Psychiatr. Genet. 14,
sci. 10, 588–597.
Caston, J., Yon, E., Mellier, D., Godfrey, H.P., Delhaye-bouchaud, N., and
Mariani, J. (1998). An animal model of autism: behavioural studies in the GS
guinea-pig. Eur. J. Neurosci. 10, 2677–2684.
Chung, Y.J., Jonkers, J., Kitson, H., Fiegler, H., Humphray, S., Scott, C., Hunt,
S., Yu, Y., Nishijima, I., Velds, A., et al. (2004). A whole-genome mouse BAC
microarray with 1-Mb resolution for analysis of DNA copy number changes
by array comparative genomic hybridization. Genome Res. 14, 188–196.
Cook, E.H., Jr., Lindgren, V., Leventhal, B.L., Courchesne, R., Lincoln, A.,
Shulman, C., Lord, C., and Courchesne, E. (1997). Autism or atypical autism
in maternally but not paternally derived proximal 15q duplication. Am. J.
Hum. Genet. 60, 928–934.
Cook, E.H., Jr., and Scherer, S.W. (2008). Copy-number variations associated
with neuropsychiatric conditions. Nature 455, 919–923.
Crawley, J.N. (2004). Designing mouse behavioral tasks relevant to autistic-
like behaviors. Ment. Retard. Dev. Disabil. Res. Rev. 10, 248–258.
Crawley, J.N. (2007). What’s Wrong with My Mouse? Second Edition (New
York: John Wiley & Sons, Inc.).
DiCicco-Bloom, E., Lord, C., Zwaigenbaum, L., Courchesne, E., Dager, S.R.,
Schmitz, C., Schultz, R.T., Crawley, J., and Young, L.J. (2006). The develop-
mentalneurobiology ofautism spectrumdisorder.J.Neurosci.26,6897–6906.
Dykens, E.M., Sutcliffe, J.S., and Levitt, P. (2004). Autism and 15q11-q13
disorders: behavioral, genetic, and pathophysiological issues. Ment. Retard.
Dev. Disabil. Res. Rev. 10, 284–291.
Folstein, S.E., and Rosen-Sheidley, B. (2001). Genetics of autism: complex
aetiology for a heterogeneous disorder. Nat. Rev. Genet. 2, 943–955.
Geschwind, D.H. (2008). Autism: many genes, common pathways? Cell 135,
Geschwind, D.H., and Levitt, P. (2007). Autism spectrum disorders: develop-
mental disconnection syndromes. Curr. Opin. Neurobiol. 17, 103–111.
Geurts, H.M., Corbett, B., and Solomon, M. (2009). The paradox of cognitive
flexibility in autism. Trends Cogn. Sci. 13, 74–82.
Jamain, S., Radyushkin, K., Hammerschmidt, K., Granon, S., Boretius, S.,
Varoqueaux, F., Ramanantsoa, N., Gallego, J., Ronnenberg, A., Winter, D.,
et al. (2008). Reduced social interaction and ultrasonic communication in
a mouse model of monogenic heritable autism. Proc. Natl. Acad. Sci. USA
Kishore, S., and Stamm, S. (2006). The snoRNA HBII-52 regulates alternative
splicing of the serotonin receptor 2C. Science 311, 230–232.
Kwon, C.H., Luikart, B.W., Powell, C.M., Zhou, J., Matheny, S.A., Zhang, W.,
Li, Y., Baker, S.J., and Parada, L.F. (2006). Pten regulates neuronal arboriza-
tion and social interaction in mice. Neuron 50, 377–388.
Lee, J.A., and Lupski, J.R. (2006). Genomic rearrangements and gene copy-
number alterations as a cause of nervous system disorders. Neuron 52,
Levitt, P., Eagleson, K.L., and Powell, E.M. (2004). Regulation of neocortical
interneuron development and the implications for neurodevelopmental disor-
ders. Trends Neurosci. 27, 400–406.
Lijam, N., Paylor, R., McDonald, M.P., Crawley, J.N., Deng, C.X., Herrup, K.,
Stevens, K.E., Maccaferri, G., McBain, C.J., Sussman, D.J., et al. (1997).
Social interaction and sensorimotor gating abnormalities in mice lacking
Dvl1. Cell 90, 895–905.
Lord, C., Cook, E.H., Leventhal, B.L., and Amaral, D.G. (2000). Autism spec-
trum disorders. Neuron 28, 355–363.
Maestrini, E., Paul, A., Monaco, A.P., and Bailey, A. (2000). Identifying autism
susceptibility genes. Neuron 28, 19–24.
Mao, R., Jalal, S.M., Snow, K., Michels, V.V., Szabo, S.M., and Babovic-Vuk-
sanovic, D. (2000). Characteristics of two cases with dup(15)(q11.2-q12): one
of maternal and one of paternal origin. Genet. Med. 2, 131–135.
(2001). Neurogranin null mutant mice display performance deficits on spatial
learning tasks with anxiety related components. Hippocampus 11, 763–775.
Mohandas, T.K., Park, J.P., Spellman, R.A., Filiano, J.J., Mamourian, A.C.,
Hawk, A.B., Belloni, D.R., Noll, W.W., and Moeschler, J.B. (1999). Paternally
derived de novo interstitial duplication of proximal 15q in a patient with devel-
opmental delay. Am. J. Med. Genet. 82, 294–300.
Moretti, P., Bouwknecht, J.A., Teague, R., Paylor, R., and Zoghbi, H.Y. (2005).
Abnormalities of social interactions and home-cage behavior in a mouse
model of Rett syndrome. Hum. Mol. Genet. 14, 205–220.
Moy, S.S., Nadler, J.J., Magnuson, T.R., and Crawley, J.N. (2006). Mouse
models of autism spectrum disorders: the challenge for behavioral genetics.
Am. J. Med. Genet. C. Semin. Med. Genet. 142, 40–51.
Murcia, C.L., Gulden, F., and Herrup, K. (2005). A question of balance:
a proposal for new mouse models of autism. Int. J. Dev. Neurosci. 23, 265–
Nadler, J.J., Moy, S.S., Dold, G., Trang, D., Simmons, N., Perez, A., Young,
N.B., Barbaro, R.P., Piven, J., Magnuson, T.R., et al. (2004). Automated appa-
ratus for quantitation of social approach behaviors in mice. Genes Brain
Behav. 3, 303–314.
Nicholls, R.D., and Knepper, J.L. (2001). Genome organization, function, and
imprinting in Prader-Willi and Angelman syndromes. Annu. Rev. Genomics
Hum. Genet. 2, 153–175.
Noirot, E. (1966). Ultra-sounds in young rodents. I. Changes with age in albino
mice. Anim. Behav. 14, 459–462.
Persico, A.M., and Bourgeron, T. (2006). Searching for ways out of the autism
maze: genetic, epigenetic and environmental clues. Trends Neurosci. 29, 349–
Polleux, F., and Lauder, J.M. (2004). Toward a developmental neurobiology of
autism. Ment. Retard. Dev. Disabil. Res. Rev. 10, 303–317.
Ramocki, M.B., and Zoghbi, H.Y. (2008). Failure of neuronal homeostasis
results in common neuropsychiatric phenotypes. Nature 455, 912–918.
Riccio, O., Potter, G., Walzer, C., Vallet, P., Szabo, G., Vutskits, L., Kiss, J.Z.,
and Dayer, A.G. (2008). Excess of serotonin affects embryonic interneuron
Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc. 1245
migration through activation of the serotonin receptor 6. Mol. Psychiatry 14, Download full-text
Roberts, S.E., Dennis, N.R., Browne, C.E., Willatt, L., Woods, G., Cross, I.,
Jacobs, P.A., and Thomas, S. (2002). Characterisation of interstitial duplica-
tions and triplications of chromosome 15q11-q13. Hum. Genet. 110, 227–234.
Rosenzweig-Lipson, S., Zhang, J., Mazandarani, H., Harrison, B.L., Sabb, A.,
Sabalski, J., Stack, G., Welmaker,G., Barrett, J.E., and Dunlop, J. (2006). Anti-
obesity-like effects of the 5-HT2C receptor agonist WAY-161503. Brain Res.
Rubenstein, J.L., and Merzenich, M.M. (2003). Model of autism: increased
ratio of excitation/inhibition in key neural systems. Genes Brain Behav. 2,
Santarelli, L., Saxe, M., Gross, C., Surget, A., Battaglia, F., Dulawa, S.,
Weisstaub, N., Lee, J., Duman, R., Arancio, O., et al. (2003). Requirement of
hippocampal neurogenesis for the behavioral effects of antidepressants.
Science 301, 805–809.
Sebat, J., Lakshmi, B., Malhotra, D., Troge, J., Lese-Martin, C., Walsh, T.,
Yamrom,B.,Yoon,S.,Krasnitz,A.,Kendall, J.,etal.(2007).Strong association
of de novo copy number mutations with autism. Science 316, 445–449.
Seeburg, P.H. (2002). A-to-I editing: new and old sites, functions and specula-
tions. Neuron 35, 17–20.
Tabuchi, K., Blundell, J., Etherton, M.R., Hammer, R.E., Liu, X., Powell, C.M.,
and Sudhof, T.C. (2007). A neuroligin-3 mutation implicated in autism
increases inhibitory synaptic transmission in mice. Science 318, 71–76.
Takao, K., Yamasaki, N., and Miyakawa, T. (2007). Impact of brain-behavior
phenotypying of genetically-engineered mice on research of neuropsychiatric
disorders. Neurosci. Res. 58, 124–132.
van der Weyden, L., and Bradley, A. (2006). Mouse chromosome engineering
for modeling human disease. Annu. Rev. Genomics Hum. Genet. 7, 247–276.
Veenstra-VanderWeele, J., and Cook, E.H., Jr. (2004). Molecular genetics of
autism spectrum disorder. Mol. Psychiatry 9, 819–832.
Veenstra-VanderWeele, J., Christian, S.L., and Cook, E.H., Jr. (2004). Autism
as a paradigmatic complex genetic disorder. Annu. Rev. Genomics Hum.
Genet. 5, 379–405.
Veltman, M.W., Thompson, R.J., Craig, E.E., Dennis, N.R., Roberts, S.E.,
Moore, V., Brown, J.A., and Bolton, P.F. (2005). A paternally inherited duplica-
tion in the Prader-Willi/Angelman syndrome critical region: a case and family
study. J. Autism Dev. Disord. 35, 117–127.
Volkmar, F.R., and Pauls, D. (2003). Autism. Lancet 362, 1133–1141.
Vorstman, J.A., Staal, W.G., van Daalen, E., van Engeland, H., Hochstenbach,
P.F., and Franke, L. (2006). Identification of novel autism candidate regions
through analysis of reported cytogenetic abnormalities associated with
autism. Mol. Psychiatry 11, 18–28.
Whitaker-Azmitia, P.M. (2005). Behavioral and cellular consequences of
increasing serotonergic activity during brain development: a role in autism?
Int. J. Dev. Neurosci. 23, 75–83.
Winslow, J.T., and Insel, T.R. (2002). The social deficits of the oxytocin
knockout mouse. Neuropeptides 36, 221–229.
Yamamoto, T., Nakahata, Y., Tanaka, M., Yoshida, M., Soma, H., Shinohara,
K., Yasuda, A., Mamine, T., and Takumi, T. (2005). Acute physical stress
elevates mouse period1 mRNA expression in mouse peripheral tissues via
a glucocorticoid-responsive element. J. Biol. Chem. 280, 42036–42043.
Yamasaki, N., Maekawa, M., Kobayashi, K., Kajii, Y., Maeda, J., Soma, M.,
Takao, K., Tanda, K., Ohira, K., Toyama, K., et al. (2008). Alpha-CaMKII defi-
ciency causes immature dentate gyrus, a novel candidate endophenotype of
psychiatric disorders. Mol. Brain 1, 6.
Yoshimura, A., Fujii, R., Watanabe, Y., Okabe, S., Fukui, K., and Takumi, T.
(2006). Myosin-Va facilitates the accumulation of mRNA/protein complex in
dendritic spines. Curr. Biol. 16, 2345–2351.
Zheng, B., Mills, A.A., and Bradley, A. (1999). A system for rapid generation of
coat color-tagged knockouts and defined chromosomal rearrangements in
mice. Nucleic Acids Res. 27, 2354–2360.
1246 Cell 137, 1235–1246, June 26, 2009 ª2009 Elsevier Inc.