Identification of Arx transcriptional targets
in the developing basal forebrain
Carl T. Fulp1, Ginam Cho2, Eric D. Marsh3, Ilya M. Nasrallah2, Patricia A. Labosky4
and Jeffrey A. Golden1,2,?
1Neuroscience Graduate Group, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA,
2Department of Pathology and3Division of Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA 19104,
USA and4Vanderbilt Center for Stem Cell Biology, Vanderbilt University Medical Center, Nashville, TN 37232, USA
Received June 4, 2008; Revised August 1, 2008; Accepted August 27, 2008
Mutations in the aristaless-related homeobox (ARX) gene are associated with multiple neurologic disorders in
humans. Studies in mice indicate Arx plays a role in neuronal progenitor proliferation and development of the
cerebral cortex, thalamus, hippocampus, striatum, and olfactory bulbs. Specific defects associated with Arx
loss of function include abnormal interneuron migration and subtype differentiation. How disruptions in ARX
result in human disease and how loss of Arx in mice results in these phenotypes remains poorly understood.
To gain insight into the biological functions of Arx, we performed a genome-wide expression screen to identify
transcriptional changes within the subpallium in the absence of Arx. We have identified 84 genes whose
expression was dysregulated in the absence of Arx. This population was enriched in genes involved in cell
migration, axonal guidance, neurogenesis, and regulation of transcription and includes genes implicated in
autism, epilepsy, and mental retardation; all features recognized in patients with ARX mutations. Additionally,
we found Arx directly repressed three of the identified transcription factors: Lmo1, Ebf3 and Shox2. To further
to compare the Arx gene regulatory network (GRN) to the Dlx1/2 GRN and interneuron transcriptome. These ana-
lyses identifieda subsetofgenes in the ArxGRN thatare sharedwith thatoftheDlx1/2 GRNand thatareenriched
in the interneuron transcriptome. These data indicate Arx plays multiple roles in forebrain development, both
dependent and independent of Dlx1/2, and thus provides further insights into the understanding of the mechan-
isms underlying the pathology of mental retardation and epilepsy phenotypes resulting from ARX mutations.
Mutations in aristaless-related homeobox (ARX) result in at
least nine distinct neurological disorders. They can be categor-
ized as (i) those diseases that include structural malformations:
hydrancephaly with abnormal/ambiguous genitalia (HYD-AG;
OMIM #300215) (1), X-linked lissencephaly with ambiguous
genitalia (XLAG; OMIM #300215) (1–5), and agenesis of
the corpus callosum with abnormal genitalia (ACC-AG;
OMIM #300004; also known as Proud’s syndrome) (1), and
(ii) those diseases that include no obvious structural malfor-
mations: infantile epileptic-dyskinetic encephalopathy (IEDE;
OMIM #308350) (6), early infantile epileptic encephalopathy
syndrome; OMIM #308350) (7), West syndrome (WS; also
known as X-linked infantile spasm syndrome; OMIM
#308350) (8–10), X-linked myoclonic epilepsy with mental
retardation and spasticity (XMEMRS; OMIM #300432)
(8,11), Partington syndrome (PRTS; OMIM #309510) (8,12)
and X-linked mental retardation with or without seizures
(NS-XLMR; OMIM #300419) (8,13–18).
The ARX protein contains a paired-related homeodomain, a
highly conserved octapeptide domain, a C-terminal aristaless
domain and four polyalanine (polyA) tracts (19). In general,
?To whomcorrespondence shouldbeaddressed at:Department ofPathology,AbramsonResearch Center, Room516C, Children’sHospitalofPhiladelphia,
3615 Civic Center Boulevard, Philadelphia, PA 19104, USA. Tel: þ1 2155904307; Fax: þ1 2155903709; Email: email@example.com
# 2008 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/
licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is
Human Molecular Genetics, 2008, Vol. 17, No. 23
Advance Access published on September 16, 2008
the malformation phenotypes are associated with protein trun-
cation mutations and missense mutations in the homeobox
domain, whereas the non-malformation phenotypes are associ-
ated with missense mutations outside of the homeobox or an
expansion in the first or second polyA tracts (1,20).
Mice deficient for Arx exhibit a complex, pleiotrophic phe-
notype, generally reflective of the expression domains of Arx.
Within the forebrain Arx is expressed in the developing gangli-
onic eminences (GEs), the cerebral cortical ventricular zone
(VZ), the striatonigral neurons of the striatum, the hippo-
campus, theventral thalamus
(2,13,21–24). In the cerebral cortex, hypothalamus and olfac-
tory bulbs, Arx is primarily localized in GABAergic inter-
neurons where its expression is maintained throughout
adulthood (2,21,23,24). During early development (E9.0–
E14.5) Arx is expressed in Emx2þ cortical VZ cells, appar-
ently under the transcriptional control of Pax6 (25); likely in
progenitor cells that will give rise to the excitatory neurons
of the cortex (21,23). Curiously, colabeling of Arx has not
been observed in mature excitatory neurons, potentially due
to rapid downregulation of expression upon exit from the
cell cycle (21,23). Outside the forebrain, it is expressed in
the floor plate of the developing spinal cord, somites, testes,
pancreas and skeletal muscle (2,8,13,23,26,27).
Arx-deficient mice exhibit decreased proliferation of corti-
cal neuroepithelial cells, dysgenesis of the thalamus, hippo-
campus, striatum and olfactory bulbs, as well as, abnormal
interneuron migration and subtype differentiation (2,28).
How loss of Arx in mice results in these phenotypes is
unknown but must reflect alterations in Arx-dependent tran-
Given that Arx is a transcription factor, defining the tran-
scriptional targets and gene regulatory networks (GRNs) in
which Arx participates is essential to further elucidating its
role in normal brain development and how disruptions in
ARX lead to human disease. Using Arx deficient mice, com-
bined with expression profiling, we have defined the GRN
regulated by Arx within the subpallium, validated a subset
of these target genes and used these data to gain new insights
into the genetic regulation of interneuron development.
Generation of ARX mutant mice
To generate mice carrying a floxed Arx allele we used the
strategy summarized in Figure 1A (also see Materials and
Methods). Arxflox/þand Arxflox/Y(Arx is on the X-chromosome)
mice were bred to homozygosity (Arxflox/floxfemale mice) and
mated with Pou3f4-Cre male mice which express Cre through-
out the pallium and subpallium at E14.5 (kindly provided by
Dr E.B. Crenshaw, Children’s Hospital of Philadelphia)
(Fig. 1B) (29,30). Western blots using an antibody specific to
the C-terminus of Arx confirmed the absence of Arx protein
in the GEs dissected from conditional mutant mice (Fig. 1C).
At E18.5, the whole brain of conditional mutant mice was
smaller than that of the wild-type and the olfactory bulbs
were nearly absent, similar to that observed by others in Arx
mutant mouse (Fig. 1D) (2,24). Immunohistochemistry using
the C-terminal a-Arx antibody detected a near complete loss
of Arx expression in the forebrain of Arx2/ymice (Fig. 1E).
directed to calbindin (Calb1) showed a loss of the normal distri-
bution of Calb1þ interneurons throughout the cerebral cortex,
being restricted to the subcortical and SVZ regions in Arx2/
y;Pou3f4-Cre mice (Fig. 1F), again similar to that previously
reported in Arx mutant mice (2,28).
Identification of putative Arx target genes
The strong subpallial expression of Arx, along with the well-
defined phenotype in GE-derived cells in Arx mutant mice,
prompted us to focus on the role of Arx in subpallial develop-
ment. Thus, all subsequent studies were based on a genome-
wide microarray analysis of the subpallia microdissected
from E14.5 Arx2/y; Pou3f4-Creþ (mutant embryos) and
39 000 transcripts and variants from 34 000 murine genes
(Affymetrix MOE430-2) resulted in the identification of 84
genes with at least a 2-fold change in expression within the
subpallium and with a false discovery rate (FDR) of ,5%
(Fig. 2A). Fifty-seven genes were increased and 27 genes
decreased in conditional Arx-null subpallium when compared
with control subpallium (Supplementary Material, Table S1).
Hierarchical centroid-linkage and unsupervised clustering
grouped differentially expressed genes according to relative
variation in gene expression patterns (Fig. 2B).
The microarray data set was validated by performing quan-
titative polymerase chain reaction (qPCR) on three indepen-
dentpairs ofsubpallia from
Pou3f4-Creþ) and control (Arxþ/Y; Pou3f4Creþ) embryos.
A subset of 10 genes identified as differentially expressed,
along with three genes that did not change significantly,
were selected for testing. All 10 genes designated as differen-
tially expressed by microarray analysis changed at least 2-fold
by qPCR, whereas the three control genes that failed to change
on the microarray also showed no difference in expression by
qPCR (Fig. 2D). Log2 expression observed by microarray
analysis was found to positively and significantly correlate
to the log2 expression observed by qPCR (Pearson r2¼
0.7031, P , 0.001) (Fig. 2E).
Functional classification of Arx target genes
To gain insight into the functions of identified Arx target
genes, we used two methods to identify Gene Ontology
(GO) biological processes (31) represented in the population
of differentially expressed genes. WebGestalt (32), consider-
ing only the differentially expressed genes relative to all
expressed genes, and ErmineJ Gene Set Resampling (GSR)
(33), which is a non-threshold based method that looks for
enrichment across the entire ranked expression set, were
used to gain insight into the functions of identified Arx
target genes. These analyses identified 38 enriched biological
processes (FDR-corrected P-value ,0.01) in the differentially
expressed genes relative to all expressed genes (Fig. 2C). The
identified genes included those implicated in cell/neuronal
migration (P ¼ 4.3 ? 1025, P ¼ 4.3 ? 1022), axonogenesis
(P ¼ 1.3 ? 1023, 1.2 ? 1029), neuron morphogenesis during
differentiation (1.6 ? 1023, 4.0 ? 1023) and regulation of
Human Molecular Genetics, 2008, Vol. 17, No. 233741
transcription (P ¼ 4.8 ? 1023; negative 1.1 ? 1022and posi-
tive 3.6 ? 1022), with P-values reported for WebGestalt and
GSR, respectively (Fig. 2C and Supplementary Material,
Tables S2 and S3).
Genes differentially expressed in Arx mutants show
spatial expression changes in the subpallium
To further validate the microarray results, we performed
in situ hybridization or immunohistochemistry for 10 of the
differentially expressed genes (Gbx2, Uncx, Magel2, Lmo1, Kitl,
Ndn, Shox2, Nts, Calb1 and Foxp1) on E14.5 brains from
mutant and control embryos. We observed Gbx2 expression
to be restricted to the basal forebrain in sections from the tele-
ncephalon of control mice. In sections from mutant mice, we
found a dorsal expansion of Gbx2 from the basal forebrain into
the medial GE (Fig. 3A). No expression of Uncx was observed
in the telencephalon of control mice; however, in the absence
of Arx, ectopic expression of Uncx was identified in the
anterior portion of the lateral GE (Fig. 3B). In control sections,
Magel2 expression was restricted to the septum and anterior
hypothalamus, while expression expanded laterally to fill
most of the medial ganglion eminence and striatum in
mutant brains (Fig. 3C). Magel2 expression was additionally
observed in the intermediate zone of the pallium in the
mutant telencephalon. Lmo1 expression within the control
telencephalon primarily localized to the VZ of the GEs. In
contrast, Lmo1 expression was greatly expanded ventrola-
terally to encompass nearly the entire volume of the GEs in
mutant mouse brains (Fig. 3D). Lmo1 expression was addition-
ally found in the medial cortex of both control and mutant
mice. In control mice, Kitl was strongly expressed within the
septum, the boundary between the lateral GE and striatum,
in the lateral cortex at the entry point of interneurons into
Figure 1. (A) Genomic organization of mouse Arx and details of the targeting construct. Arrowheads denote the location of primers used for genotyping. (B)
Brain section from a E14.5 Pou3f4-Cre?B6.129S4-Gt(ROSA)26Sortm1Sor/J stained with X-gal reveals the ubiquitous expression of the Pou3f4-Cre transgene
throughout the telencephalon. (C) PCR of genomic DNA from the hindbrains of E14.5 Arxþ/Y; Pou3f4-Creþ and Arx2/Y; Pou3f4-Creþ mice with primer
sets specific to Arx, Pou3f4 and Sry (left). Western blotting of whole cell lysate from E14.5 subpallia of Arxþ/Y; Pou3f4-Creþ and Arx2/Y; Pou3f4-Creþ
mice, using an antibody specific to the C-terminus of Arx (Arx-C), reveals the absence of Arx in conditionally mutant Arx mice (right). (D) Gross morphology
of prosencephalon from E18.5 Arxþ/Y; Pou3f4-Creþ (left) and Arx2/Y; Pou3f4-Creþ (right) mice. Total brain size appeared smaller and olfactory bulbs were
nearly absent in the conditional mutant (KO). (E) Lack of Arx protein within the subpallium of conditional knockouts. Telencephalon sections from E14.5 were
used for immunohistochemistry using the same anti-Arx antibody used for western blots. (F) Immunohistochemistry with an antibody specific for Calb1 on
sagittal sections of E18.5 brain revealed the presence of Calb1þ interneurons throughout the cortical plate of Arxþ/Y; Pou3f4-Creþ mice (left), while
Calb1þ interneurons were confined to the ventricular and subventricular zones of Arx2/Y; Pou3f4-Creþ mice. MZ, marginal zone; CP, cortical plate; WM,
white matter. Inset represents 4? magnification at the location represented by the black square.
3742Human Molecular Genetics, 2008, Vol. 17, No. 23
the pallium and within the pallial intermediate zone (Fig. 3E).
The mutant telencephalon showed similar Kitl expression;
however, within the subpallium, Kitl expression expanded
medially through the medial GE. In addition, we noted a stron-
ger expression of Kitl in the lateral cerebral cortex of mutant
mice when compared with control mice. Ndn was expressed
within the septum, basal forebrain, striatum and the ventricular
and intermediate zones of both control and mutant telencepha-
lon (Fig. 3F). However, within the mutant subpallium, the Ndn
domain expanded into the medial and lateral GEs. Shox2
exhibited no expression within the forebrain of control mice
(Fig. 3G), although expression was evident in the dorsal thala-
mus (data not shown). Within the mutant forebrain, a wide
array of Shox2 expression was detected in the septum and
basal forebrain. Nts was observed in a discrete region in the
ventrolateral cortex in control mice; however, within mutant
mice Nts expression expanded medially into the LGE
(Fig. 3H). Calb1 was observed throughout the subpallium
and in streams of presumptive interneurons within the
cortex of control and mutant mice (Fig. 3I). We observed a
Figure 2. (A) Volcano plot of log2fold-change versus 2log10FDR-corrected P-value for all probes deemed present in at least four of the eight microarrays
analyzed. Green lines denote the selected 2-fold change cutoff, while the red line denotes the 0.05 selected FDR-corrected P-value cutoff. Points in the
upper left and upper right quadrants represent probes that met the requirements to be called differentially expressed. (B) Hierarchical clustering of the 84
genes deemed as differentially expressed was performed using a correlation matrix and centroid linkage. Red represents increased expression, while green rep-
resents decreased expression. (C) Bar chart representing the 2log(p-values) for the top 15 overrepresented Gene Ontology (GO) Biological Processes as deter-
mined by WebGestalt (top) and ErmineJ Gene Set Resampling (bottom) algorithms. (D) qPCR on E14.5 subpallia from control and conditionally mutant mice for
genes deemed differentially expressed (Cadps2, Ndn, Rasgef1b, Kitl, Ebf3, Shox2, Sst, Magel2, Pde4dip and Calb1) and three control genes that were not differ-
entially expressed (Shroom2, Lgi1 and Ets1). Data were compared with Actb expression using the DDCT method and presented as the log2fold ratio
difference between mutant and control mice. (E) Microarray expression data (log2) and qPCR data (log2) were found to be positively correlated (P , 0.001,
Human Molecular Genetics, 2008, Vol. 17, No. 233743
significantly stronger expression of Calb1 in mutant subpal-
lium when compared with control subpallium. Finally,
Foxp1 was expressed in the striatum, basal forebrain, cortical
plate and cortical VZ of both control and mutant telencephalon
(Fig. 3J). However, within the mutant striatum and basal
forebrain we found a decrease in Foxp1 expression.
Identification of Arx target genes involved in interneuronal
Expression of the homeobox transcription factors Dlx1 and
Dlx2 in GE cells is necessary for tangential migration of inter-
neurons from the basal forebrain to the neocortex (34). Dlx1
and Dlx2 are thought to promote interneuron differentiation
by repressing oligodendroglial cell fate (35) and promote
interneuron migration by repressing neurite outgrowth (36).
Furthermore, early expression of Dlx genes matches the
expression of interneuron obligate genes, such as Gad1 and
Gad2, and is sufficient (although not necessary) for their
expression (37–40). These data suggest that activation of the
Dlx transcriptional pathway is shared by most, if not all, tele-
ncephalic interneurons. Given that Dlx1/2 expression appears
to overlap that of Arx and be sufficient, and within some
brain regions necessary, for Arx expression, it has been
suggested that Arx itself may be a transcriptional target of
As the loss of Arx results in a pleiotrophic phenotype (2)
and the genes identified in our screen could underlie any of
the various observed phenotypes, we sought to determine
whether there exists an identifiable gene signature shared
between genes significantly altered in Arx conditional
mutant and Dlx mutant mice. This signature would represent
those genes downstream of Arx that likely play a role in inter-
neuron differentiation and/or migration. We reasoned that
genes defining such a signature would (i) be enriched within
the GE-derived interneuron population and GABAergic pro-
jection neurons and, given Arx’s putative position downstream
of Dlx1/2 (41), (ii) show expression changes in the absence of
Dlx1/2 similar to those we observed in the absence of Arx.
To identify this gene signature, we took advantage of the
availability of two previously published data sets. The first
data set was derived from FAC-sorting GFP-positive and
-negative cells from the subpallium and pallium of E14.5
Dlx5/6Cre-IRES-eGFPembryos, which were then used in micro-
array analyses to compare the expression profiles between the
different populations of sorted cells (42). In the second data
set, Rubenstein and coworkers (36) compared the expression
profiles between E14.5 subpallium-derived cells from either
heterozygous or homozygous Dlx1/2 double mutant mice.
This data set was similar, but not identical, to that previously
published. We used gene set enrichment analysis (GSEA)
(43), a computational method for assessing whether a pre-
defined gene set is statistically enriched in one biological
state when compared with another (43), to determine (i)
whether and which Arx targets genes have differential
expression in the subpallium in the absence of Dlx1/2, (ii)
whether and which Arx target genes exhibit enriched
expression in interneurons, and (iii) of those Arx target
genes expressed in interneurons, which gene sets overlap
with those expressed in cells migrating into cortical regions
and which overlap with those expressed in cells migrating
into subcortical regions.
In the first analysis, we sought to determine whether the
gene sets that corresponded to genes that increased and
decreased in the absence of Arx were enriched in microarray
experiments comparing the gene expression changes in sub-
pallia from Dlx1/2 heterozygous and homozygous mice. We
observed that the gene sets both downregulated (NES ¼
1.102; P , 0.0001; FDR ¼ 0.105) (Fig. 4A) and upregulated
(NES ¼ 1.301; P , 0.0001; FDR ¼ 0.299) (Fig. 4B) in the
absence of Arx are enriched in the population of genes differ-
entially expressed in the subpallium in the absence of Dlx1/2.
Of the genes that decreased in the subpallium in the absence of
Arx, 12/26 also decreased in the subpallium of Dlx1/2 mutant
mice, while 2/26 increased in the subpallium of Dlx1/2 mutant
mice (Pde1c and Vcan). The remaining 12/26 genes decreased
in the subpallium of the Arx mutant but not the Dlx1/2 mutant.
Of those genes that increased in the subpallium in the absence
Figure 3. In situ hybridization on E14.5 mouse brain sections using probes directed to (A) Gbx2, (B) Uncx, (C) Magel2, (D) Lmo1, (E) Kitl, (F) Ndn, (G) Shox2
and (H) Nts. Immunohistochemistry on E14.5 mouse brain sections using (I) anti-Calb1 and (J) anti-Foxp1 antibodies. For each panel, Arxflox/Y; Pou3f4-Cre2 is
presented on the left, while Arx2/Y; Pou3f4-Creþ is presented on the right.
3744Human Molecular Genetics, 2008, Vol. 17, No. 23
ofArx, 20/52 also increased in the subpallium of Dlx1/2 mutant
mice. Interestingly, while Dlx1/2 appeared to enhance
expression of Calb1, Etv1, Rasgef1b, Phlda1, Lrrtm1, Lmo4,
Slc35b4 and Pcyt1b, Arx appeared to repress expression of
these genes. Alternatively, it is possible that we instead
observed an accumulation of a particular migratory cell sub-
population(s) that failed to exit the subpallium in the absence
of Arx. The remaining 24/52 genes increased in the subpallium
of the Arx mutant mouse but not the Dlx1/2 mutant mouse. Col-
lectively, these data indicate Arx is not simply a downstream
target for Dlx genes in subpallial cell development, but also
regulates a GRN that is independent of its role downstream
of Dlx, i.e. gene expression signatures were identified in the
subpallium of Arx mutant mice that were shared with those
of Dlx1/2 mutant mice; however, there also existed a gene
expression signature specific to the Arx mutant mice.
We next asked which Arx target genes were enriched in
Dlx5/6-derived versus non-Dlx5/6-derived cell populations.
Using GSEA we tested whether gene sets corresponding to
genes that increased or decreased in the absence of Arx were
Figure 4. GSEA revealed a shared GRN within the subpallium of Arx and Dlx1/2 mutants. (A) Graph denotes the Absolute Enrichment (NES ¼ 1.102; P ,
0.0001) of the gene set containing those genes whose expression decreased in the subpallium in the absence of Arx within the Rubenstein et al. data set represent-
ing those genes whose expression changed in the absence of Dlx1/2 (left). Green boxes contain those genes deemed as statistically enriched (right). The genes
contained within the top box are those genes whose expression decreased in the subpallium in both Arx and Dlx1/2 mutant mice. The lower box contains genes
whose expression increased in the subpallium of the Arx mutant mouse, but increased in the subpallium of Dlx1/2 mutant mice. (B) Graph denotes the Absolute
Enrichment (NES ¼ 1.301, P , 0.0001) of the gene set containing those genes whose expression increased in the subpallium in the absence of Arx within the
Rubenstein et al. data set (left). Green boxes contain those genes deemed as enriched (right). The genes contained within the top box are those genes whose
expression increase in the subpallium of Arx mice but that decrease in the subpallium of the Dlx1/2 mutant mouse. The lower box contains genes whose
expression increased in the subpallium of both Arx and Dlx1/2 mutant mice.
Human Molecular Genetics, 2008, Vol. 17, No. 233745
enriched within microarray experiments comparing gene
expression profiles from Dlx5/6-GFP-negative cells of the
pallium (primarily principal projection neurons and neuronal
progenitors of the VZ) to the combined populations of Dlx5/
6-GFP-positive cells of the pallium (interneurons) and subpal-
We observed that both genes whose expression decreased
(NES ¼ 1.333; P ¼ 0.034; FDR ¼ 0.002) and those whose
expressionincreased(NES ¼ 1.365;
0.069) in the absence of Arx were enriched in that population
of genes belonging to the Dlx5/6-GFP-positive cells of the
pallium and subpallium. Among those genes whose expression
increased in the absence of Arx, a set of 21/52 genes was
enriched in Dlx5/6-derived cells (Fig. 5B), while 12/26 of the
genes whose expression decreased in the absence of Arx were
also enriched in Dlx5/6-derived cells (Fig. 5A). Five genes
that increased in the subpallium in the absence ofArx were nor-
mally expressed in non-Dlx5/6-derived cells (Ebf3, Rasgef1b,
Lrrtm1, Nrp1 and Lmo1) (Fig. 5B), while five genes that
decreased in the subpallium in the absence of Arx were nor-
mally expressed in non-Dlx5/6-derived cells (Cntn2, Prkcd,
Frmd6, Rbl1 and Crabp1) (Fig. 5A). Among those genes
enriched in Dlx5/6-derived cells and that exhibited decreased
expression in the subpallium of Arx mutant mice, 7/12 were
also observed to decrease in the Dlx1/2 mutant subpallium
(Cxcr4, Maf, Arx, Zfp503, Has3, Ebf1 and Sox8) (Fig. 5C).
Among those genes enriched in Dlx5/6-derived cells and
that were observed to increase expression in the subpallium
of Arx mutant mice, 10/21 were also observed to increase in
the subpallium of Dlx1/2 mutant mice (2310045A20Rik,
Gbx2, Hap1, Lbxcor1, Magel2, Nap1l5, Olfm3, Plcxd3, Sst
and Stk33) (Fig. 5D). Two genes (Ebf3 and Nrp1) were
enriched in non-Dlx5/6-derived cells and increased expression
in the subpallium of both Dlx1/2 and Arx mutant mice. Inter-
estingly, 5/21 genes enriched in Dlx5/6-derived cells that
exhibited increased expression in the subpallium of Arx
mutant mice were decreased in the subpallium of Dlx1/2
mutant mice (Calb1, Etv1, Pcyt1b and Phlda1).
We next asked whether Arx target genes were preferentially
expressed in Dlx5/6-derived cells that migrated out of the sub-
pallium or those that migrate within the subpallium. To do
this, we looked for enrichment of gene sets corresponding to
genes that increased or decreased in the absence of Arx
within those genes differentially expressed in microarray
pallium relative to Dlx5/6-GFPþ cells within the subpallium.
Neurons that migrate from the LGE give rise to interneurons
of the olfactory bulb and the medium spiny neurons of the
striatum, while neurons that migrate from the MGE become
Calb1þ, Sstþ and Pvalþ interneurons of the neocortex, and
to a lesser extent of the anterior olfactory nucleus, piriform
cortex, medium spiny neurons of the striatum, multiple
nuclei of the amygdala and the CA1 and caudal aspect of
the CA3 of the hippocampus (44–50). Cells that arise in the
interganglionic region contribute to interneurons of the preop-
tic and hypothalamic regions (51). CGE-derived interneurons
populate the cortex (primarily those of layer 5 and those of
the Calb2þ and Npyþ subclasses), multiple nuclei of the
amygdala, CA1 of the hippocampus, and medium spiny
neurons of the striatum (45,52).
P ¼ 0.003;FDR ¼
Ebf1, Zfp503, Sox8 and Oprm1 were enriched in Dlx5/6-GFPþ
cells that remain in the subpallium (interneurons and striatal
GABAergic projection neurons), while Maf, Cxcr4, Mef2c and
Pde4dip were enriched in Dlx5/6-GFPþ cells contributing to
the pallium (interneurons) (Fig. 6J). In situ hybridization
images of E14.5 wild-type brains retrieved from the GenePaint
database (53) confirmed the predicted subpallial expression of
Sox8 (Fig. 6M), Zfp503 (Fig. 6N) and Ebf1 (Fig. 6O) and the
pallial expression of Cxcr4 (Fig. 6K) and Mef2c (Fig. 6L).
Of the genes that exhibited increased expression in the subpal-
lium of Arx mutant mice and enrichment in Dlx5/6-derived
cells, Gbx2, Stk33, Lbxcor1, Sorcs3, Nts, Olfm3, Magel2 and
Hap1 were enriched in Dlx5/6-GFPþ cells that remain in the
subpallium, while Cadps2, Kitl, Sst, Phlda1, AI427515,
Enpp5, Calb1 and Pcyt1b were enriched in Dlx5/6-GFPþ
cells contributing to the pallium (Fig. 6A). In situ hybridization
images confirm the predicted pallial expression of Kitl
(Fig. 6B), Sst (Fig. 6C) and Phlda1 (Fig. 6D) and the subpallial
expression of Hap1 (Fig. 6F), Magel2 (Fig. 6G), Lbxcor1
(Fig. 6H) and Gbx2 (Fig. 6I). A pattern emerged where genes
whose expression is normally confined to the subpallium and
that decreased in the absence of Arx localized to the striatum
in wild-type mice (Fig. 6M–O), while those that increased in
the absence of Arx tended to localize to the basal forebrain
and septum (Fig. 6F–I).
Homeobox transcription factor binding sites are enriched
in the promoter regions of genes differentially expressed
in the absence of Arx
In an attempt to determine which of the modulated genes may
be direct targets of Arx, we used oPOSSUM (54), a program
that determines the overrepresentation of transcription factor
binding sites (TFBSs), as determined by sliding position
weight matrices obtained from the JASPER database along
aligned and conserved sequences from given regions upstream
ofadefined transcriptional start site (TSS), withina set ofcoex-
pressed genes when compared with a precompiled background
set, to search for enrichment of conserved (TFBSs) 25 kb
upstream of the TSS of genes whose expression was modulated
in the absence of Arx. We reasoned that, given that transcrip-
tion factors within families often share similar binding sites
(55–57), it remained a formal possibility that, although the
binding site of Arx remains to be discovered, it might be
highly similar to that of transcription factors that have
already been annotated. We considered a site enriched if the
Z-score was .10 and the Fisher P-value cutoff was ,0.01,
as these conditions result in an average false positive rate of
,15% (54). Additionally, we required that there be at least
five target gene hits. When all differentially expressed genes
were considered, enrichment of the following TFBSs were
observed: Nkx2-5 (36.25% target gene hits, Z-score¼40.55,
P ¼ 4.70 ? 10212);
Z-score¼35.68, P ¼ 4.36 ? 1026); Sox5 (30% target gene
hits, Z-score¼13.41, P ¼ 1.16 ? 1025) and Sry (15% target
gene hits, Z-score¼10.69, P ¼ 9.42 ? 1025) (Fig. 7A). When
only the genes whose expression increased in the subpallium
in the absence of Arx are considered, enrichment of Nkx2.5
3746Human Molecular Genetics, 2008, Vol. 17, No. 23
(40.74% target gene hits, Z-score¼37.44, P ¼ 7.02 ? 1029),
Prrx2 (62.96% target gene hits, Z-score¼40.78, P ¼ 4.63 ?
1025) and Sox5 (27.78% target gene hits, Z-score¼10.13,
P ¼ 2.69 ? 1024) motifs is observed (Fig. 7B). If we only con-
siderthose genes whoseexpression decreased inthesubpallium
in the absence of Arx, enrichment is observed of the Sox5
(34.62% target gene hits, Z-score¼11.32, P ¼ 2.15 ? 1023),
Nkx2.5 (26.92% target gene hits, Z-score¼12.13, P ¼
6.28 ? 1023)
Z-score¼17.15, P ¼ 7.43 ? 1023) binding sites (Fig. 7C).
andPrrx2 (61.54% targetgene hits,
Figure 5. GSEA revealed that the gene sets corresponding to genes differentially expressed in the subpallium of Arx mutant mice are enriched in those genes
differentially expressed between Dlx5/6-derived and non-Dlx5/6-derived cells. (A) Graph denotes the Absolute Enrichment (NES ¼ 1.333, P ¼ 0.034) of the
gene set containing those genes whose expression decreased in the subpallium in the absence of Arx within a subset of the Marsh et al. data set (genes expressed
in GFPþ telencephalic cells versus those expressed in GFP- cortical cells FAC-sorted from Dlx5/6Cre-IRES-eGFPmice) (left). Green boxes contain those genes
deemed as enriched (right). The genes contained within the top box are those genes whose expression decreased in the subpallium Arx mutant mice and were
enriched in GFPþ interneurons. The lower box contains genes whose expression decreased in the subpallium of the Arx mutant mouse and were enriched in
GFP- non-interneuron cell populations. (B) Graph denotes the Absolute Enrichment (NES ¼ 1.365, P ¼ 0.003) of the gene set containing those genes whose
expression increased in the subpallium in the absence of Arx within a subset of the Marsh et al. data set (genes expressed in GFPþ telencephalic cells
versus those expressed in GFP2 cortical cells FAC sorted from Dlx5/6Cre-IRES-eGFPmice) (left). Green boxes contain those genes deemed as enriched
(right). The genes contained within the top box are those genes whose expression increased in the subpallium Arx mutant mice and were enriched in GFPþ
interneurons. The lower box contains genes whose expression increased in the subpallium of the Arx mutant mouse and were enriched in GFP2 non-interneuron
cell populations. (C) Venn diagram illustrates the overlap between genes whose expression was observed to decrease in the subpallium of conditional Arx mutant
and Dlx1/2 mutant mice (red) and those genes whose expression was observed to decrease in the subpallium in the absence of Arx and were enriched in Dlx5/
6-derived cells (blue). (D) Venn diagram illustrates the overlap between genes whose expression was observed to increase in the subpallium of conditional Arx
mutant and Dlx1/2 mutant mice (red) and those genes whose expression was observed to increase in the subpallium in the absence of Arx and were enriched in
Dlx5/6-derived cells (blue).
Human Molecular Genetics, 2008, Vol. 17, No. 23 3747
Analysis of the weblogos (58) corresponding to the position
weight matrices for the three enriched binding sites across
the three sets of genes revealed a similarity between the
enriched binding sites (Fig. 7D–F). Both Nkx2.5 and Prrx2
are homeodomain transcription factors, and the reverse comp-
lement of the Prrx2 site (TAATT) is contained within the
Nkx2.5 site. Prrx2 is a paired-boxed, homeodomain transcrip-
tion factor of the aristaless family as is Arx (59).
Genes containing E2F family TFBSs are enriched
in subpallium of wild-type mice when compared
with the subpallium of Arx mutants
In addition to using oPOSSUM to identify enriched TFBS in
those genes that were identified as differentially expressed,
we also used GSEA to look for enrichment, within the
ranked list of all gene expression values from the Arx array
experiments, of precomputed gene sets that share a cis-
regulatory motif contained within 4000 bp regions centered
at the predicted TSS that is conserved across the human,
mouse, rat and dog genomes (60). We identified 20 gene
sets enriched (P , 0.05; FDR , 25%) in the wild-type pheno-
type and 2 gene sets enriched in the mutant phenotype
(Fig. 8A). The two gene sets enriched in the mutant phenotype
correspond to highly conserved motifs that correspond to an
unknown transcription factor. To our surprise, 17/20 gene
sets enriched in the wild-type phenotype corresponded to con-
served motifs that represented binding sites of E2F family
member proteins. We next clustered the leading-edge genes
corresponding to the 20 gene sets enriched in the wild-type
Figure 6. (A) Graph denotes the Absolute Enrichment (NES ¼ 1.541, P , 0.001) of the gene set containing those genes whose expression increased in the sub-
pallium in the absence of Arx and were enriched in interneurons within a subset of the Marsh et al. data set (genes expressed in GFPþ subpallial cells versus
those expressed in GFPþ cortical cells FAC sorted from Dlx5/6Cre-IRES-eGFPmice) (left). Green boxes contain those genes deemed as enriched (right). The genes
contained within the top box are those genes whose expression increased in the subpallium of Arx mutant mice and were enriched in GFPþ interneurons within
the cortex. The lower box contains genes whose expression increased in the subpallium of the Arx mutant mouse and were enriched in GFPþ neurons that
remain within the subpallium. To validate the predicted expression data, in situ hybridization images of E14.5 embryonic brain were downloaded from the Gen-
ePaint web site (53) for (B) Kitl, (C) Sst, (D) Phlda1, (E) AI427515, (F) Hap1, (G) Magel2, (H) Lbxcor1 and (I) Gbx2. (J) Graph denotes the Absolute Enrich-
ment (NES ¼ 1.605, P , 0.001) of the gene set containing those genes whose expression decreased in the subpallium in the absence of Arx and were enriched in
Dlx5/6-derived cells within a subset of the Marsh et al. data set (genes expressed in GFPþ subpallial cells versus those expressed in GFPþ cortical cells FAC-
sorted from Dlx5/6Cre-IRES-eGFPmice) (left). Green boxes contain those genes deemed as enriched (right). The genes contained within the top box are those genes
whose expression decreased in the subpallium of Arx mutant mice and were enriched in GFPþ interneurons within the cortex. The lower box contains genes
whose expression decreased in the subpallium of the Arx mutant mouse and were enriched in GFPþ neurons that remain within the subpallium. To validate the
predicted expression data, in situ hybridization images of E14.5 embryonic brain were downloaded from the GenePaint web site for (K) Cxcr4, (L) Mef2c, (M)
Sox8, (N) Zfp503 and (O) Ebf1.
3748 Human Molecular Genetics, 2008, Vol. 17, No. 23
phenotype (Fig. 8B), which lead to the discovery of a cluster
of leading-edge genes shared between the 17 enriched motifs
corresponding to binding sites for E2F family proteins
(Fig. 8C). These results are highly suggestive that loss of
Arx function results in a loss of E2F-family proteins ability
to bind their target genes.
Ebf3, Lmo1 and Shox2 are direct transcriptional
targets of Arx
We first tested whether Arx could be recruited to the putative
enhancers of Arx target genes by performing chromatin
immunoprecipitation (ChIP) assays on Neuro2a neuroblas-
toma cells transfected with a V5-tagged Arx construct and
using an anti-V5 antibody (or anti-myc as a control) to
immunoprecipitate Arx-chromatin complexes. After the ChIP
procedure, immunoprecipitated chromatin was tested for the
presence of Ebf3 22064 to 21854 (Refseqs NM_010096,
NM_001113414, NM_001113415), Lmo1 2126 to 2295
(Aceview splice variant Lmo1.b; 25326 to 25495 relative
to Refseq NM_057173) and Shox2 2850 to 2748 (Refseq
NM_013665) putative enhancer regions by qPCR with specific
primer pairs spanning the identified phastCon regions contain-
ing the predicted Arx consensus binding sites (Fig. 9A).
Primer pairs designed to bind within the promoter region of
murine Gapdh were used as a control. Significant enrichment
was found for the Ebf3, Lmo1 and Shox2 putative enhancer
fragments when an anti-V5 antibody was used but not when
an anti-myc antibody was used (Fig. 9B). Furthermore, no
enrichment was observed at the Gapdh promoter. These data
indicate Arx can bind to the putative enhancers of Ebf3,
Lmo1 and Shox2.
Given that expression of Ebf3, Lmo1 and Shox2 were all
increased in the absence of Arx, we next asked whether Arx
binds and transcriptionally represses the Ebf3, Lmo1 and
Shox2 enhancers. We generated luciferase reporter constructs
by cloning the regions identified by ChIP assays upstream of
TK-luciferase. As a positive control, we generated a luciferase
reporter with four copies of the predicted Arx binding site,
TAATTA, and as a negative control we used the empty
TK-luciferase-containing vector. Reporter constructs were
transiently transfected into Neuro2a cells along with Arx or
an empty expression plasmid. Although Arx could minimally
repress the empty vector (22.2%), the TAATTA?4, Ebf3,
Shox2 and Lmo1 vectors were all significantly repressed by
Arx (72.2, 66.8, 56.8 and 43.4% repression, respectively).
These data suggest that TAATTA is likely the binding site
for Arx, and the regions we identified by ChIP were sufficient
to bind Arx and functionally repress transcription.
To determine whether Arx is capable of binding to the S8/
Prrx2 motif, EMSA was performed.32P-labeled oligonucleo-
tide (double-stranded 30mers) spanning the TAATTA found
Figure 7. (A) JASPAR TFBSs significantly enriched in the region 25000 bp upstream of the TSS of all differentially expressed genes relative to those regions
upstream of all genes present in at least four of the eight arrays where P , 0.01 and Z-score .10. (B) TFBSs significantly enriched in the regions 25000 bp
upstream of the TSS of genes whose expression increased within the subpallium in the absence of Arx. (C) TFBSs significantly enriched in the regions 25000 bp
upstream of the TSS of genes whose expression decreased within the subpallium in the absence of Arx. Weblogos of the position weight matrices corresponding
to the JASPAR (D) Nkx2.5, (E) Prrx2 and (F) Sox5 binding sites found to be enriched.
Human Molecular Genetics, 2008, Vol. 17, No. 23 3749
in the Ebf3 (21862 to 21892), Lmo1 (2217 to 2247) and
Shox2 (2780 to 2810) putative enhancers were probed
with purified GST-Arx. We observed that GST-Arx bound
binding was attenuated by competition with a 50-fold molar
excess of unlabeled template (Fig. 9D, lane 3). Furthermore,
the addition of an anti-Arx antibody to the binding reaction
supershifted the complex (Fig. 9D, lane 4). We observed
32P-labeled Shox2 30mer (Fig. 9D, lane 2), and this
that GST-Arx bound to the Lmo1 template (Fig. 9E, lane 2),
and this binding was attenuated by competition with a
50-fold molar excess of unlabeled template (Fig. 9E, lane 3).
GST-Arx binding of Lmo1, however, was not competed by
50-fold molar excess of an unlabeled template where
TAATTA was mutated to TCCTTA (Fig. 9E, lane 4).
Excess unlabeled template where a TG dinucleotide upstream
of the TAATTA was mutated to CC readily competed for
Figure 8. (A) GSEA of microarray data obtained from analysis of control versus Arx mutant subpallia, using a gene set including putative TF and microRNA
binding sites conserved within human, mouse, rat and dog (60), revealed TF and microRNA binding site motifs significantly enriched (P , 0.05 and FDR ,
0.25) in control (positive NES values) or conditional mutant (negative NES values) subpallia. (B) Clustering of significant gene sets and genes within those gene
sets enriched in control subpallia reveals a cluster of genes shared among 18 of the enriched gene sets. All 18 of these gene sets correspond to genes with various
known E2F family binding sites within 4000 bp centered on the TSS of the genes. (C) List of genes contained within the cluster.
3750 Human Molecular Genetics, 2008, Vol. 17, No. 23
binding of GST-Arx (Fig. 9E, lane 5), suggesting that the
TAATTA site within the oligonucleotide was necessary for
binding of GST-Arx. The addition of an anti-Arx antibody
to the binding reaction again supershifted the complex
(Fig. 9E, lane 6). Similar results were obtained with the
Ebf3 template (data not shown). These data, along with the
other data presented, indicate Ebf3, Lmo1 and Shox2 are
direct transcriptional targets of Arx whose transcription is
repressed by Arx.
Mutations in ARX have been identified in multiple human
neurodevelopmental disorders, yet its mechanistic role as a
transcription factor remains obscure. To gain insight into
ARX’s role in brain development we sought to define the
downstream targets of ARX in the developing brain. Based
on the clinical phenotypes observed in patients with ARX
mutations, we were specifically interested in evaluating the
role of ARX in cerebral cortical interneurons (61). Using
microarray assays on tissue derived from the subpallium of
mice in which Arx has been conditionally abrogated, we ident-
ified 57 genes significantly over expressed in the mutant mice
and 27 genes whose expression was downregulated when
compared with litter mate controls. These data were validated
by quantitative PCR and in situ hybridization for a subset of
genes. Based on our results we conclude Arx functions as
both a transcriptional activator and as a transcriptional
Although we observed both the activation and repression of
distinct sets of genes in the absence of Arx, our data indicate
Arx functions primarily as a transcriptional repressor in the
ventral forebrain. We base this on the fact that a greater
number of genes were over-expressed in the absence of Arx
when compared with those showing a decrease in expression.
A greater magnitude in change of gene expression was also
found for the over-expressed group of genes. Finally a
greater enrichment of putative Arx binding sites was present
in those genes that are overexpressed in the absence of Arx.
Arx was previously shown to regulate Pax4 during pancreas
development through binding to a conserved sequence
upstream of Pax4 and repress its expression (62). The ident-
ified sequence, however, lacks a canonical homeobox
binding site, suggesting that either (1) Arx can bind directly
to sequences other than canonical homeobox binding sites or
(2) Arx forms heterodimeric transcriptional complexes with
non-homeobox or homeobox transcription factors that recog-
nize a variant of the canonical homeobox binding site. The
Drosophila homolog of Arx, al, can bind clawless (mamma-
lian homologs are Tlx1, Tlx2 and Tlx3) and Chip (mammalian
Figure 9. (A) Schematics from the UCSC Genome Browser (135) demonstrate the presence of conserved S8/Prrx2 binding sites upstream of the Ebf3, Shox2 and
Lmo1 genes. Genes are oriented from right to left. The region chosen for ChIP for each of the genes is annotated and overlaps a PhastCon Vertebrate Conserved
Element (132). Below the PhastCon, in gray, the EMSA probe, which is located within the ChIP region, is annotated. Below the annotated EMSA probe, the
nucleotide conservation is demonstrated for several mammals with the S8/Prrx2 binding site designated by a red box. (B) Quantitative ChIP assays demonstrate
that anti-V5, but not anti-myc control, antibody immunoprecipitates the putative enhancer regions upstream of Ebf3, Shox2 and Lmo1, but not the negative
control Gapdh promoter in Neuro2a cells transfected to express V5-tagged Arx. Data are presented as percentage of input chromatin (+SEM). (C) Luciferase
reporter assays demonstrate that Arx, in Neuro2a cells, represses heterlogous expression of an engineered site containing four copies of the putative Arx binding
site, as well as, the putative enhancers for Ebf3, Shox2 and Lmo1, cloned upstream of TK-luciferase, to a greater extent than that observed for the empty vector
control. Luciferase data were normalized to Renilla expression. Data for the Arx-transfected data points are presented as percentage activation (+SEM) relative
to empty vector-transfected cells. (D) EMSA assays demonstrate that GST-Arx can bind a 30 bp oligonucleotide containing the putative Arx binding site and
corresponding to a region upstream of Shox2 (black arrowhead; lane 2), and that this binding is readily competed with cold competitor (lane 3). The complex was
supershifted in the presence of an anti-GST antibody (open arrowhead; lane 4). (E) EMSA assays demonstrate that GST-Arx can bind a 30 bp oligonucleotide
containing the putative Arx binding site and corresponding to a region upstream of Lmo1 (black arrowhead; lane 2), and that this binding is readily competed
with cold competitor (lane 3). The binding is not readily competed, however, with cold competitor where the TAATTA site has been mutated to TCCTTA
(lane 4), but is readily competed with cold competitor where a dinucleotide mutation is made outside of the putative Arx binding site (lane 5). The complex
was supershifted in the presence of an anti-GST antibody (open arrowhead; lane 6).
Human Molecular Genetics, 2008, Vol. 17, No. 233751
homologs are Ldb1 and Ldb2) (63,64). Tlx1, Tlx2 and Tlx3 are
homeobox transcription factors that bind TAAGTG (65).
Additionally, Arx has been suggested to interact with Nr2f1
and/or Nr2f2 (66) which bind the nuclear receptor binding
murine muscle, Arx appears to exists within a transcriptional
complex that includes Myod1 and Mef2c during myocyte
differentiation and mediates the activation of downstream
targets (26). Thus, while Pax4 was previously the only
known target for Arx, several lines of evidence support Arx
functioning within the telencephalon as a transcriptional acti-
vator and/or repressor. Our data indicate Arx binds, at least
within the subpallium, to ATTAAT in its role as a transcrip-
tional repressor. The fact that mutations in this sequence com-
peted out binding in our EMSA assays support this conclusion.
GSEA analyses were utilized to search for genes belonging
to specific sets of subpallial-derived neurons. The Arx-
dependent GRN identified in the current study was compared
with the expression profiles of Dlx5/6-derived and non-Dlx5/
6-derived cell populations within the telencephalon previously
elucidated by our laboratory (42) and the Dlx1/2 GRN pre-
viously identified by the Rubenstein laboratory (NCBI GEO
GSE2161). First, we found statistically significant overlap
between the Arx and Dlx1/2 GRNs, which is in agreement
with previous data, suggesting that Arx is downstream of
Dlx1/2 in the subpallium (41), and we have identified the indi-
vidual genes that are shared between these two networks.
However, we have also identified genes belonging to the
Arx GRN whose expression appears not to be affected by
loss of Dlx1/2. In fact, we find a small subset of genes that
appear to be activated in the absence of Arx but are repressed
in the absence of Dlx1/2. These data suggest that the Arx and
Dlx appear to be in both direct pathways as well as parallel or
We were able to determine that subsets of genes altered in
the absence of Arx were enriched in Dlx5/6-derived versus
non-Dlx5/6-derived expression profiles. Those genes shared
in the two data sets, define a set of genes localized to subpal-
lial destined interneurons and GABAergic projection neurons
versus interneurons that migrate into the pallium. These ana-
lyses allow for the prediction of which elements of the Arx
GRN underlie specific aspects of the Arx phenotype. For
example Maf, Cxcr4, Mef2c and Pde4dip normally localized
to pallial interneurons, and the expression of these genes
was lost in the absence of Arx, suggesting that the activation
of these genes by Arx may be necessary for either the
migration of interneurons into the pallium or differentiation
of interneuron subtypes that are pallial-specific. Cxcr4, for
example, is expressed on the surface of interneurons that
migrate into the cortex and appears to be required for
normal interneuron positioning from ventral–lateral
In contrast, genes such as Hap1, Magel2, Olfm3, Nts,
Sorcs3, Lbxcor1, Stk33 and Gbx2 normally localized to
septal and basal forebrain neurons and were ectopically
expressed in the GE in the absence of Arx. This suggests
that Arx may be required to repress this set of genes to
allow for maintenance of specific subpallial interneuron sub-
types or, for example, the cholinergic neurons which normally
arise from this region (47) and are lost in Arx knockout (28).
Our GSEA analyses also indicate that a subpopulation of
genes whose expression decreases in the absence of Arx nor-
mally localized to the striatum. Zfp503 (74,75), Sox8 (76),
Ebf1 (77,78) and Oprm1 (79) have all been shown to be
enriched in, or involved in development, of the striatum.
Ebf1 expression in the mantle of the LGE is required for main-
tenance of Cdh8 and Crabp1 mantle expression (77). We
observed Ebf1, Cdh8 and Crabp1 to all be decreased in the
subpallium of Arx conditional mutants. Ebf1 expression was
previously observed to be normal in the striatum of P0
Arx2/Ymice, but is lost in the globus pallidus at P0 (28). At
earlier time points, Ebf1 was excluded from its normal
expression pattern in the LGE corridor, and Ebf1- cells accu-
mulated in the SVZ (28). Colombo et al. (28) also observed an
ectopic expression of Etv1 throughout the globus pallidus and
accumulation of Calb1 positive cells in the MGE. These find-
ings are consistent with the increase in Etv1 and Calb1
expression and decrease in Ebf1, Cdh8 and Crabp1 expression
observed in the subpallium of our Arx conditional mutant mice
in the current study. Furthermore, Foxp1, whose expression
decreased in the absence of Arx, is expressed in the post-
mitotic neurons of the striatum and is specific to the
Ppp1r1b- and Gria2/3-expressing medium spiny projection
neurons, rather than the interneuron population (80–82). Simi-
Ppp1r1b-expressing medium spiny projection neuron popu-
lation, but appears to be turned on earlier than Foxp1
(75,77,78,83). Collectively, these data suggest that Arx may
be required to activate genes that play a role in the specifica-
tion of striatal GABAergic projection neurons.
GSEA enrichment analysis of TFBSs suggests that the
expression of a large cohort of E2F target genes is decreased
in the subpallium of Arx conditional mutant mice. This
implies that Arx may play a role in controlling cell cycle
exit of neural progenitors within the subpallium. The change
in expression of a majority of these E2F target genes fell
below our 2-fold cutoff, suggesting that these genes them-
selves are not likely direct targets of Arx; however, the
expression of the E2F family members was not altered
greater than our cutoff in the conditional Arx mutant.
However, two genes known to be involved in E2f1-signaling
were altered—Ndn and Rbl1, also known as p107. Ndn
binds to the transactivational domain of E2f1 and E2f4, and
expression of Ndn in neuroblastoma results in the induction
of neuronal markers (84). Furthermore, Ndn binds Dlx2 and
Dlx5 via Maged1 to promote the differentiation of Calb1þ
interneurons, and Ndn-deficient mice exhibit impaired inter-
neuron differentiation (86). Rbl1, normally expressed in the
VZ of developing brain, is rapidly downregulated at the
onset of neural differentiation (87–89). Further support for
this possibility comes from data indicating Rbl1 also regulates
the expansion of stem cell populations in brain and regulates
expression of Notch1 (90). Alternatively, it remains a formal
possibility that Arx could directly bind Rb, Rbl1 or Rbl2, as
has been observed, at least in vitro, for other paired-type
homeodomain proteins, including Prrx1, Pax3 and Vsx2 (91).
Finally, disruptions in interneuron development have been
implicated or linked to a number of neurodevelopmental
disorders (reviewed in 92–94). Mutations in Arx that cause
are expressed inthe
3752Human Molecular Genetics, 2008, Vol. 17, No. 23
neurodevelopmental disorders, such as epilepsy, mental retar-
dation, dystonia and autism spectrum disorder, may have no
morphological defects, favoring a primary interneuronopathy
(61). Thus, knowing the genes regulated by Arx in inter-
neurons provides an important resource for identifying
additional human disease genes and potentially better targets
for therapy. Evidence that our analyses has likely identified
genes relevant to human disorders comes from the recognition
that many of the genes whose expression was altered in the
subpallium in the absence of Arx are known to be involved,
or have been linked to, other CNS disorders associated with
mental retardation or autistic-like phenotypes. These include
MAGEL2 and NDN (95–97), CADPS2 (98), SOX8 (99,100),
GBX2 (101), AFF2 (102,103), PCDH19 (104) and FGF13
(105). Two other genes, LRRTM1 and HTR7, have been pre-
(106,107). It is likely that many of the other genes identified
in this study are also important in human disease. Further
studies designed to understand their function will unquestion-
ably bring new insights into our understanding of these
complex human disorders.
MATERIALS AND METHODS
All experiments were approved by and carried out in accord-
ance with the Animal Care and Use Committee of the Chil-
dren’s Hospital of Philadelphia.
Generation of floxed Arx animals
taining Arx (ATCC, Manassas, VA). A fragment containing
exons 1–3 was subcloned into pBS-SKþ. An frt-neoR-frt cas-
sette was obtained from Dr Susan Dymecki (Harvard Medical
School), and a loxP site was ligated into the 30end of this cas-
sette. The resulting frt-neoR-frt-loxP cassette was ligated into
into intron 1, using a unique Nsi1 site. Finally, diphtheria toxin
(DTA) was cloned onto the 50end of the targeting vector. The
resulting construct recombines to generate an allele of Arx
with loxP sites flanking exon 2. The floxed construct was elec-
troporated into 129 Sv6(T4) ES cells, as previously described
(108), and 576 NeoR colonies were isolated (data not shown).
Southern blots on 450 of these clones identified 6 recombinant
clones (data not shown). Two independent ES cell lines were
selected and injected into C57BL/6 blastocysts. Chimeric
mice were obtained, confirmed by Southern blot, and mated to
C57BL/6 mice. To prevent unwanted ectopic gene expression
of adjacent genes, triggered by the neoRpromoter, we
mice with a floxed Arx allele to those expressing Flp1
(B6;SJL-Tg(ACTFLPe)9205Dym/J; The Jackson Laboratory,
Pou3f4Creþ; Arxþ/Y(30) males to generate conditional Arx2/
Y; Pou3f4-Creþ mutant male mice. Hindbrain genomic DNA
from animals was genotyped by PCR using the following
GAATG-30; ArxLoxPR 50-TGGAGCGGGGACAGGGGTG
TCTACTCA-30to amplify the Arx locus; Brn4Fw 50-CAATG
ACTATC-30; SryFw 50-CAGAAATGAACTACTGCATCC
C-30and SryRev 50-AACTTGTGCCTCTCACCACG-30. The
Arx LoxPF, ArxLoxPR and ArxRePR were run as a multiplex
PCR. The combination of primers ArxLoxPF and ArxLoxPR
yields a 594 bp product for the wild-type allele. The combi-
nation of primers ArxLoxPF and ArxLoxPR amplifies a
637 bp product from the floxed allele. If recombination
occurs, then ArxLoxPF and ArxRePR amplify a 402 bp
Pou3f4-Cre mice were mated with B6.129S4-Gt(ROSA)26
Sortm1Sor/J mice (The Jackson Laboratory). Brains from
E14.5 embryos were harvested, fixed for 2 h at 4ºC, dropped in
30% sucrose overnight and then snap frozen and cut at 50 mm
on a cryostat. Sections were stained as free-floating sections
in 1 mg/ml 5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside
(X-gal) in 5 mM K3[Fe(CN)6], 5 mM K4[Fe(CN)6].3H2O, 2 mM
MgCl2in PBS overnight at room temperature. Sections were
then rinsed with several washes of PBS and mounted to Super-
GEs from E14.5 Arx2/Y; Pou3f4Creþ or Arxþ/Y; Pou3f4Creþ
brains were lysed in 10 mM Tris–HCl, pH 7.4; 150 mM NaCl;
0.1% Triton X-100, with complete EDTA-free protease inhibi-
tor mixture (Roche Diagnostics), passed six times through a
27-G needle to shear nucleic acids and spun at 14 000g for
20 min. The supernatant was then diluted with 4? LDS
sample buffer and 10? reducing agent (Invitrogen). For
whole cell extracts, 20 ml of lysate was used. Lysates were
separated using NuPAGE 10% Bis–Tris gels, transferred to
PVDF membranes, and after blocking were incubated with a
C-terminal anti-Arx antibody (1:1000; #sc-48843; Santa
Cruz Biotechnology, Santa Cruz, CA). Following incubation
with secondary antibody, blots were developed using ECL
(GE Healthcare) chemiluminescent reagent and exposed to
film. Blots were then stripped using Restore (Pierce) and
reprobed with anti-Gapdh (1:5000; #MAB374; Millipore) anti-
body as a protein loading control.
Extraction and amplification of cRNA
Each biological replicate consisted of a pair of GEs from a
single animal. A total of four animals from each genotype
(Arxþ/Y; Pou3f4Creþ or Arx2/Y; Pou3f4Creþ) were taken
from three separate litters. For microarray experiments, we
chose to use Arxþ/Y; Pou3f4Creþ, rather than Arxflox/Y;
Pou3f4-, mice as controls to eliminate any gene expression
changes that may result from any toxicity that may result
from Cre activity alone (50,109–113). After sacrificing, all
tissue was rapidly dissected from E14.5 GEs (MGE and LGE
Human Molecular Genetics, 2008, Vol. 17, No. 23 3753
using Trizol (Invitrogen), followed by purification using
RNeasy mini columns (Qiagen), and then labeled according
to the Affymetrix GeneChip. Expression Analysis Technical
Manual (114). Briefly, total RNA was converted to first-strand
cDNA using Superscript II reverse transcriptase, primed by a
poly(T) oligomer that incorporates the T7 promoter. Second-
strand cDNA synthesis was followed by in vitro transcription
for linear amplification of each transcript and incorporation of
biotinylated CTP and UTP. The cRNA products were then
fragmented and hybridized to eight GeneChip. Mouse
Genome 430 2.0 microarrays (Affymetrix, Santa Clara, CA),
each containing 45 101 probes. Arrays were washed and
scanned according to the manufacturer’s recommended
Microarray data analysis
Statistical analyses were performed using open source R soft-
ware packages available as part of the Bioconductor project
and summarization of gene expression data by GC Robust
Multi-array Averaging (116) using the affy package (117),
MAS5.0 presence/absence calls were used to filter out all
probe sets lacking a present call in at least four of the eight
samples. Differential gene expression, defined by the microar-
rays, from Arx2/Y; Pou3f4Creþ and Arxþ/Y; Pou3f4Creþ
tissues was then determined by computing empirical Bayes
moderated t-statistics with the limma package (118,119).
The moderated t-statistic P-values were adjusted for multiple
testing by Benjamini and Hochberg’s method to control
FDR (120). Differentially expressed genes were defined as
those having a FDR-corrected P-value ,0.05 and a log2
fold change in expression ?1 or ? 2 1. For clustering, log2-
transformed expression data for differentially expressed genes
were gene-centered (mean) and normalized. Hierarchical clus-
tering was performed using a correlation matrix and centroid
linkage using the Cluster 3.0 algorithm (121) and visualized
using the Java TreeView program (122). Microarray data gen-
erated for this study has been deposited to and is freely avail-
able from NCBI Gene Expression Omnibus (GEO) as record
(115). After normalization
Quantitative real-time polymerase chain reaction
Total RNA was extracted from the GEs of Arxþ/Y;
Pou3f4Creþ (n ¼ 5) and Arx2/Y; Pou3f4Creþ mice (n ¼ 5)
as described earlier and DNase-treated using DNA-free
(Ambion, Foster City, CA). Approximately 500 ng of total
RNA was reverse transcribed with SuperScript II reverse tran-
scriptase with random primers (Invitrogen, Carlsbad, CA), and
10 ng of cDNA was used for real-time PCR. Real-time PCR
was carried out using Taqman Gene Expression Assays
(Applied Biosystems, Foster City, CA). Assay IDs are avail-
able upon request. Each sample was run in triplicate, along
with probes for Actb on the same plate, on a Stratagene
MX3005P real-time PCR machine (La Jolla, CA) following
the manufacturer’s recommended protocol. Ten probes sets
corresponding to genes that changed on the arrays were
chosen (Pde4dip, Magel2, Cadps2, Ndn, Rasgef1b, Kitl,
Calb1, Ebf3, Shox2 and Sst) along with three negative
control genes (Shroom2, Ets1 and Cntnap4) that did not
change on the array. Specificity of the amplification was
checked by melting-curve analysis. Relative levels of mRNA
expression were calculated according to the DDCT method
(123), normalized by comparison to Actb mRNA expression.
GO term enrichment analysis
GO term enrichment analysis was performed using the online
version of WebGestalt (32). Enrichment was assessed by using
the hypergeometric test to compare the frequency of GO Bio-
logical Process categories represented in the non-redundant
list of Gene IDs corresponding to putative Arx target genes
(n ¼ 76) versus the global frequency of GO categories in the
reference set containing non-redundant Gene IDs correspond-
ing to all genes present in at least four out of the eight arrays.
We chose a P-value cutoff of P , 0.01 and required that all
categories tested contain at least five genes. Additionally, we
used ErmineJ (33) to assess the overrepresentation of GO Bio-
logical Process categories, which allows testing across the
entire population of genes without requiring a threshold to
divide genes into ‘selected’ and ‘nonselected’ genes. Raw
P-values corresponding to all 45 038 probes, as determined
by Bayes moderated t-statistics, were entered into the
program. The GSR method was used with the following par-
ameters: maximum gene set size ¼ 100; minimum gene set
size ¼ 5; with the mean of replicates, 100 000 iterations and
full resampling. Significant GO terms were identified using a
Benjamini–Hochberg FDR of ,0.05.
Gene set enrichment analysis
GSEA analysis was performed, using the GSEA version 2.0
program (124). Two sets of microarray data were downloaded
(as .CEL files) from the NCBI Gene Expression Omnibus
(GEO) (125) and GC-RMA normalized as described earlier.
The first data set, generated by the Rubenstein laboratory
(GSE2161), contains data in which cRNA, derived from
E14.5 subpallium dissected from either Dlx1/2 heterozygous
(n ¼ 3) or Dlx1/2 double mutant (n ¼ 3) mice, was hybridized
to Affymetrix MOE430-2 microarrays. A similar, companion
data set (GSE8311) was previously described (36). The
second data set (GSE5817), generated by our own laboratory
(42), contains a data set from microarrays hybridized to
cRNA derived from GFPþ pallial cells (n ¼ 6), GFP- pallial
cells (n ¼ 5) and GFPþ subpallial cells (n ¼ 6) FAC sorted
from the dissected brains of E14.5 Dlx5/6Cre-IRES-eGFPmice.
The GSEA program first collapsed the 45 037 probe sets rep-
resented by the microarrays to 21 890 individual genes, and
the GC-RMA normalized expression values for those genes
were ranked with respect to the phenotype (Dlx1/2 heterozy-
gote versus Dlx1/2 KO; GFP-negative pallial cells versus
GFP-positive pallial and subpallial cells; or GFP-positive
pallial cells versus GFP-positive subpallial cells) by using an
absolute value (rather than positive or negative) SNR (126).
With this approach, the final position in the ranked gene list
depended only on the strength of the gene in discriminating
between phenotypes rather than specific up- or down-
regulation in a given phenotype. Represented members of
3754 Human Molecular Genetics, 2008, Vol. 17, No. 23
the Arx target gene sets, divided into one gene set comprising
those genes whose expression increased in the absence of Arx
and a second gene set comprising those genes whose
expression decreased in the absence of Arx, were then
located within the ranked gene list, and the proximity of the
Arx target gene sets to the most differentially expressed genes
according to phenotype (i.e. those with the highest absolute
SNR value) was measured with a weighted Kolmogorov–
Smirnov statistic [ES, enrichment score (43)], with a higher
score corresponding to a higher proximity. The observed ES
ES scores after permutation of phenotype.
Identification of enriched TFBSs
The gene set including putative Arx target genes was analyzed
for enrichment of TFBSs using the oPOSSUM 2.0 program
(54). For each transcript, genes were mapped to EnsEMBL
gen. identifiers and submitted to Human SSA. The top 10%
of conserved regions (minimum conservation of 70%) within
5000 bp upstream and matrix match threshold 95% was
scanned for TFBSs in the JASPAR database (127) using a
position weight matrices algorithm. Enriched TFBSs were
identified using Fisher’s exact test at P , 0.01 by comparing
the putative Arx target genes with a random set of 1000
genes as background. Additionally, we performed GSEA, as
described earlier, on the entire Arx microarray data set,
testing for enrichment of the C3 motif gene set included
with the GSEA package (60). Briefly, this gene set includes
174 candidate TFBS motifs and 106 putative miRNA
binding motifs located within a 4 kb window centered at the
annotated TSS or the 30-UTR, respectively, and conserved in
human, mouse, rat and dog genomes.
In situ hybridization
dehydefixed cryostat sections
described (128). Briefly, cDNA for genes of interest was
obtained by PCR from an E10.5 whole-embryo mouse cDNA
library and cloned into the pBluescript-SK(2) plasmid (Strata-
gene). Primers used for PCR amplification are listed in Sup-
plementary Material, Table S5. Digioxigenin (DIG)-labeled
cRNA probes were generated by invitro transcription and hybri-
dized to tissue overnight at 708C. Hybridization signals were
detected with alkaline phosphatase-conjugated, anti-DIG Fab
fragments (1:2000; Roche Applied Science), followed by color
development with NBT and BCIP (Roche Applied Science).
In situ hybridizations, corresponding to E14.5 embryos and as
presented in Figure 6, were obtained from http://www.genepaint.
org, a digital atlas of gene expression patterns in the mouse
(53): Kitl (Set ID# HD11), Sst (Set ID# MH430), Phlda1 (Set
ID# EH2968), AI427515 (Set ID# EB1326), Hap1 (Set ID#
MH834), Magel2 (Set ID# EH1944), Lbxcor1 (Set ID#
EH934), Gbx2 (Set ID# MH867), Cxcr4 (Set ID# EN1117),
Mef2c (Set ID# MH765), Sox8 (Set ID# EN1112), Zfp503 (Set
ID# MH819) and Ebf1 (Set ID# MH1369).
Immunohistochemistry was performed as previously described
(129). Briefly, sections were washed in 0.1 M tris buffer
saline (TBS), endogenous peroxidases were quenched and
non-specific binding was blocked with 5% normal horse
serum and 0.3% Triton-X (Sigma) in TBS. The sections
were incubated overnight with the primary antibodies that
were detected using a secondary biotin-conjugated antibody
(1:2000; Jackson Immunoresearch, West Grove, PA) applied
for 1 h at room temperature. The sections were then incubated
with the ABC reagent (Vector Laboratories, Burlington, CA,
USA) for 1 h at room temperature, washed and developed
with 3,30-diaminobenzidine (DAB; Vector Laboratories). The
sections were mounted, dehydrated in graded ethanols,
scope. Primary antibodies included anti-Calb1 (1:2000, CB38;
Swant, Bellinzona, Switzerland), anti-Arx (1:100, #sc-48843;
Santa Cruz Biotechnology, Santa Cruz CA), anti-Foxp1
(1:400; Dr Edward E. Morrisey of the Department of Cell and
Developmental Biology, University of Pennsylvania) (130).
Quantitative ChIP assays
Chromatin immunoprecipitation assays were performed using
the ChIP Assay Kit (Millipore) as described by the manufac-
turer with minor modifications. Neuro2a cells were transfected
with pCDNA3.1-Arx-V5 (131) using FuGENE6 (Roche)
according to the manufacturer’s suggested protocol. After
48 h, transfected Neuro2a cells (2 ? 107) were dissociated
and fixed in 1% formaldehyde at room temperature for
10 min. Cells were lysed in SDS lysis buffer (ChIP Assay
Kit, Upstate USA, Charlottesville, VA), homogenized and
sonicated six times for 30 s using a Fisher Scientific Model
100 Sonic Dismembrator (Fisher) at power settings 3, 3, 4, 5,
6 and then 6 so as to fragment the DNA (‘input DNA’). Sub-
sequently, immunoprecipitation of chromatin equivalents of
2 ? 106cells was carried out at 48C overnight using 5 mg of
either anti-V5 (Invitrogen) or anti-myc (Sigma) antibodies,
and immune complexes pulled down using protein A-agarose
beads blocked with salmon sperm DNA (Upstate). Following
washes according to manufacturer instructions (Upstate), com-
plexes were eluted in 1% SDS/100 mM NaHCO3, and cross-
links were reversed by adding NaCl to 200 mM and heating to
658C for 14 h. DNA was extracted and subjected to quantitative
PCR using primers flanking the phastCon (132) site containing
the conserved Arx binding sites identified byour insilico analy-
sis. Amplicons were measured by SYBR Green fluorescence
(Qiagen) in25 ml reactions. Reactions were performed in tripli-
cate. The amount of product was determined relative to a stan-
dard curve of input chromatin. The primers used were as
follows: Shox2-Fw: 50-TCCAGTTCCCCAGTGTTTTACTAA
GT-30and Shox2-Rev: 50-GCTCTTGGCCATTAATCCAGGA
TT-30; Lmo1-Fw: 50- TAAGCTAATGGCGGGCACCT-30and
Lmo1-Rev: 50-CTCGCTCTCACCAGAGTGCA-30; and Ebf3-
Fw: 50-CCGTAATGGATTTTGAGATGGGA-30and Ebf3-
Rev: 50-TGAATTGGTGGTGTGTGTGC-30; Gapdh-Fw: 50- T
ACTCGCGGCTTTACGGG-30and Gapdh-Rev: 50- TGGAAC
AGGGAGGAGCAGAGAGCA-30. For statistical analyses,
student’s t-test was used.
Human Molecular Genetics, 2008, Vol. 17, No. 233755
Luciferase reporter assays
Neuro2a mouse neuroblastoma cells (0.8 ? 1025) were trans-
fected 24 h after plating with 200 ng of luciferase reporter
plasmid DNA, 50 ng of pCDNA3.1-Arx-V5-His (131) or
pCDNA3.1-V5-His (Invitrogen) and 50 ng of pRL-TK-Renilla
luciferase plasmid DNA (Promega) using FuGENE 6 (Roche
transfection, cell lysis and measurement of firefly and
Renilla luciferase activity was performed using the Dual-Glo
Luciferase Assay System (Promega) according to the manu-
facturer’s instructions using a Veritus Microplate Lumino-
meter (Turner BioSystems, Sunnyvale, CA). Transfections
were performed in quadruplicate, and three independent
experiments were performed. The firefly luciferase activity
was normalized according to the corresponding Renilla luci-
ferase activity, and luciferase activity was reported as mean
(+SEM) relative to pCDNA3.1-V5-His/luciferase transfec-
tion. Luciferase reporter constructs were generated using
primer sequences identical to those used for ChIP, modified
to include a HindIII on the 50end of the sense primer and a
BamHI site on the 30end of the antisense primer. These
primers were used to PCR amplify product from E14.5
mouse genomic DNA, and the amplified products were cut
with BamHI and HindIII and cloned into the BamHI and
HindIII sites of PPREX3-TK-Luc (Addgene plasmid 1015)
(133), which contains the thymidine kinase promoter upstream
of luciferase, replacing the PPAR response element.
Expression and purification of Arx (286–430)
in Escherichia coli
Sequences corresponding to amino acids 286–430 of Arx
were amplified by PCR using primers 286F-E (50-cgGAATT
Cga GAGGGCGGGGAGCTGTCGCC-30) and 430R-Xh (50-
30), digested with EcoRI and XhoI and then cloned into the
EcoRI and XhoI sites of pGEX-5X-3 (Amersham Bio-
sciences). GST-Arx (286–430) protein was expressed in
TOP10 competent cells (Invitrogen), and its expression was
induced by addition of 0.3 mM IPTG to late logarithmic cul-
tures (OD ¼ 0.5) for 3 h at 308 C. Cells were then harvested,
resuspended in PBS (phosphate buffer saline pH 7.4) contain-
ing 0.5% triton X-100 and disrupted by sonication in the
presence of protease inhibitor-cocktail (Sigma). After centri-
fugation, the supernatants were applied to glutathione–
Sepharose beads (Amersham Bioscience), and the beads
were washed with PBS. The GST fusion protein was finally
eluted with 30 mM glutathione in 100 mM Tris–Cl, pH 7.4.
Electrophoretic mobility-shift assay
as previously described with slight modifications (134). Oligo-
nucleotides (5 pmol) were labeled by T4 polynucleotide kinase
in the presence of [g-32P]ATP (6000 Ci/mmol; Amersham
Bioscience). Unincorporated [g-32P]ATP was removed by
column filtration using MicroSpin G-25 columns (Amersham
Biosciences). The purified probe was annealed in Buffer H
(Roche) by heating for 1 min and cooling to room temperature.
with 0.1 pmol of labeled-probe in total 20 ml of binding buffer
[20 mM HEPES pH 7.4, 50 mM KCl, 1 mM MgCl2, 1 mM DTT,
5% glycerol, 1 mg poly(dI-dC) (Roche)] for 30 min on ice. In
competition experiments, 10 pmol cold probes were added. In
antibody supershift experiment, 2 ml of anti-GST antibody
(Amersham Biosciences) was added. The mixture was loaded
on 5% polyacrylamide gel and electrophoresized (constant
33 mA) at 48C in 1? TBE. Gels were dried and visualized on
a phosphorimager (Molecular Dynamics). Oligonucleotides
used were as follows: Ebf3: 50-GCGATTTTCCCGATTAA
AATTAATCGGGAAAATCG-30; Ebf3mut1: 50-GCGATTTT
-30. Lmo1: 50-GTAATGAATTGATTTAATTAACAGGGGA
TTAACAGGGGAGTCTGA -30and 50-GTCAGACTCCCCT
GTTAAGGAAATCAATTCATTA-30; Lmo1mut2: 50-GTAA
This work was supported by NIH NS46616 (J.A.G.) and the
MRDDRCat the Children’s
Supplementary Material is available at HMG Online.
We would like to thank Dr E. Bryan Crenshaw III, The Chil-
dren’s Hospital of Philadelphia, for providing the Pou3f4-Cre
mice. We thank Dr Edward E. Morrisey, University of
Pennsylvania School of Medicine for the kind gift of anti-
Foxp1 antibody; Dr Susan Dymecki, Harvard Medical
School for kindly providing the frt-neoR-frt-containing
plasmid; Dr Bruce Spiegelman, Harvard Medical School, for
kindly providing the PPREX3-TK-Luc plasmid; Dr Gail
Martin, University of California, San Francisco for kindly pro-
viding the pGbx2-HA-FL plasmid; and Dr John Cobb, Univer-
sity of Calgary, for kindly providing the pGEM4-Shox2cds
plasmid. We also thank Dr Steven S. Scherer, University of
Pennsylvania, for kindly providing the Neuro2a neuroblas-
toma cell line. The microarray work was completed through
the National Institute of Health Neuroscience Microarray Con-
sortium (http://arrayconsortium.tgen.org) with the assistance
of Sheila Westman at the W.M. Keck Foundation Biotechnol-
ogy Resource Laboratory at Yale University. Finally, we
3756 Human Molecular Genetics, 2008, Vol. 17, No. 23
would like to thank Mr Jeremy C. Minarcik for assisting us
with animal husbandry.
Conflict of Interest statement. None declared.
1. Kato, M., Das, S., Petras, K., Kitamura, K., Morohashi, K., Abuelo,
D.N., Barr, M., Bonneau, D., Brady, A.F., Carpenter, N.J. et al. (2004)
Mutations of ARX are associated with striking pleiotropy and consistent
genotype-phenotype correlation. Hum. Mutat., 23, 147–159.
2. Kitamura, K., Yanazawa, M., Sugiyama, N., Miura, H., Iizuka-Kogo, A.,
Kusaka, M., Omichi, K., Suzuki, R., Kato-Fukui, Y., Kamiirisa, K. et al.
(2002) Mutation of ARX causes abnormal development of forebrain and
testes in mice and X-linked lissencephaly with abnormal genitalia in
humans. Nat. Genet., 32, 359–369.
3. Bhat, S.S., Rogers, R.C., Holden, K.R. and Srivastava, A.K. (2005) A
novel in-frame deletion in ARX is associated with lissencephaly with
absent corpus callosum and hypoplastic genitalia. Am. J. Med. Genet. A,
4. Hahn, A., Gross, C., Uyanik, G., Hehr, U., Hu ¨gens-Penzel, M., Alzen, G.
and Neubauer, B.A. (2004) X-linked lissencephaly with abnormal
genitalia associated with renal phosphate wasting. Neuropediatrics, 35,
5. Uyanik, G., Aigner, L., Martin, P., Gross, C., Neumann, D.,
Marschner-Scha ¨fer, H., Hehr, U. and Winkler, J. (2003) ARX mutations
in X-linked lissencephaly with abnormal genitalia. Neurology, 61,
6. Guerrini, R., Moro, F., Kato, M., Barkovich, A.J., Shiihara, T., McShane,
M.A., Hurst, J., Loi, M., Tohyama, J., Norci, V. et al. (2007) Expansion
of the first PolyA tract of ARX causes infantile spasms and status
dystonicus. Neurology, 69, 427–433.
7. Kato, M., Saitoh, S., Kamei, A., Shiraishi, H., Ueda, Y., Akasaka, M.,
Tohyama, J., Akasaka, N. and Hayasaka, K. (2007) A longer polyalanine
expansion mutation in the ARX gene causes early infantile epileptic
encephalopathy with suppression-burst pattern (Ohtahara syndrome).
Am. J. Hum. Genet., 81, 361–366.
8. Strømme, P., Mangelsdorf, M.E., Shaw, M.A., Lower, K.M., Lewis,
S.M., Bruyere, H., Lu ¨tcherath, V., Gedeon, A.K., Wallace, R.H.,
Scheffer, I.E. et al. (2002) Mutations in the human ortholog of Aristaless
cause X-linked mental retardation and epilepsy. Nat. Genet., 30,
9. Kato, M., Das, S., Petras, K., Sawaishi, Y. and Dobyns, W.B. (2003)
Polyalanine expansion of ARX associated with cryptogenic West
syndrome. Neurology, 61, 267–276.
10. Wohlrab, G., Uyanik, G., Gross, C., Hehr, U., Winkler, J., Schmitt, B.
and Boltshauser, E. (2005) Familial West syndrome and dystonia caused
by an Aristaless related homeobox gene mutation. Eur. J. Pediatr., 164,
11. Scheffer, I.E., Wallace, R.H., Phillips, F.L., Hewson, P., Reardon, K.,
Parasivam, G., Stromme, P., Berkovic, S.F., Gecz, J. and Mulley, J.C.
(2002) X-linked myoclonic epilepsy with spasticity and intellectual
disability: mutation in the homeobox gene ARX. Neurology, 59,
12. Frints, S.G., Froyen, G., Marynen, P., Willekens, D., Legius, E. and
Fryns, J.P. (2002) Re-evaluation of MRX36 family after discovery of an
ARX gene mutation reveals mild neurological features of Partington
syndrome. Am. J. Med. Genet., 112, 427–428.
13. Bienvenu, T., Poirier, K., Friocourt, G., Bahi, N., Beaumont, D.,
Fauchereau, F., Ben Jeema, L., Zemni, R., Vinet, M.C., Francis, F. et al.
(2002) ARX, a novel Prd-class-homeobox gene highly expressed in the
telencephalon, is mutated in X-linked mental retardation. Hum. Mol.
Genet., 11, 981–991.
14. Poirier, K., Lacombe, D., Gilbert-Dussardier, B., Raynaud, M.,
Desportes, V., de Brouwer, A.P., Moraine, C., Fryns, J.P., Ropers, H.H.,
Beldjord, C. et al. (2006) Screening of ARX in mental retardation
families: Consequences for the strategy of molecular diagnosis.
Neurogenetics, 7, 39–46.
15. Grønskov, K., Hjalgrim, H., Nielsen, I.M. and Brøndum-Nielsen, K.
(2004) Screening of the ARX gene in 682 retarded males. Eur. J. Hum.
Genet., 12, 701–705.
16. Partington, M.W., Turner, G., Boyle, J. and Ge ´cz, J. (2004) Three new
families with X-linked mental retardation caused by the
428-451dup(24 bp) mutation in ARX. Clin. Genet., 66, 39–45.
17. Stepp, M.L., Cason, A.L., Finnis, M., Mangelsdorf, M., Holinski-Feder,
E., Macgregor, D., MacMillan, A., Holden, J.J., Gecz, J., Stevenson, R.E.
et al. (2005) XLMR in MRX families 29, 32, 33 and 38 results from the
dup24 mutation in the ARX (Aristaless related homeobox) gene. BMC
Med. Genet., 6, 16.
18. Van Esch, H., Zanni, G., Holvoet, M., Borghgraef, M., Chelly, J., Fryns,
J.P. and Devriendt, K. (2005) X-linked mental retardation, short stature,
microcephaly and hypogonadism maps to Xp22.1-p21.3 in a Belgian
family. Eur. J. Med. Genet., 48, 145–152.
19. Miura, H., Yanazawa, M., Kato, K. and Kitamura, K. (1997) Expression
of a novel aristaless related homeobox gene ‘Arx’ in the vertebrate
telencephalon, diencephalon and floor plate. Mech. Dev., 65, 99–109.
20. Sherr, E.H. (2003) The ARX story (epilepsy, mental retardation, autism,
and cerebral malformations): one gene leads to many phenotypes. Curr.
Opin. Pediatr., 15, 567–571.
21. Colombo, E., Galli, R., Cossu, G., Ge ´cz, J. and Broccoli, V. (2004)
Mouse orthologue of ARX, a gene mutated in several X-linked forms of
mental retardation and epilepsy, is a marker of adult neural stem cells
and forebrain GABAergic neurons. Dev. Dyn., 231, 631–639.
22. Kitamura, K., Miura, H., Yanazawa, M., Miyashita, T. and Kato, K.
(1997) Expression patterns of Brx1 (Rieg gene), Sonic hedgehog,
Nkx2.2, Dlx1 and Arx during zona limitans intrathalamica and
embryonic ventral lateral geniculate nuclear formation. Mech. Dev., 67,
23. Poirier, K., Van Esch, H., Friocourt, G., Saillour, Y., Bahi, N., Backer,
S., Souil, E., Castelnau-Ptakhine, L., Beldjord, C., Francis, F. et al.
(2004) Neuroanatomical distribution of ARX in brain and its localisation
in GABAergic neurons. Brain Res. Mol. Brain Res., 122, 35–46.
24. Yoshihara, S., Omichi, K., Yanazawa, M., Kitamura, K. and Yoshihara,
Y. (2005) Arx homeobox gene is essential for development of mouse
olfactory system. Development, 132, 751–762.
25. Visel, A., Carson, J., Oldekamp, J., Warnecke, M., Jakubcakova, V.,
Zhou, X., Shaw, C.A., Alvarez-Bolado, G. and Eichele, G. (2007)
Regulatory pathway analysis by high-throughput in situ hybridization.
PLoS Genet, 3, 1867–1883.
26. Biressi, S., Messina, G., Collombat, P., Tagliafico, E., Monteverde, S.,
Benedetti, L., Cusella De Angelis, M.G., Mansouri, A., Ferrari, S.,
Tajbakhsh, S. et al. (2008) The homeobox gene Arx is a novel positive
regulator of embryonic myogenesis. Cell Death Differ, 15, 94–104.
27. Collombat, P., Mansouri, A., Hecksher-Sorensen, J., Serup, P., Krull, J.,
Gradwohl, G. and Gruss, P. (2003) Opposing actions of Arx and Pax4 in
endocrine pancreas development. Genes Dev., 17, 2591–2603.
28. Colombo, E., Collombat, P., Colasante, G., Bianchi, M., Long, J.,
Mansouri, A., Rubenstein, J.L. and Broccoli, V. (2007) Inactivation of
Arx, the murine ortholog of the X-linked lissencephaly with ambiguous
genitalia gene, leads to severe disorganization of the ventral
telencephalon with impaired neuronal migration and differentiation.
J. Neurosci., 27, 4786–4798.
29. Ahn, K., Mishina, Y., Hanks, M.C., Behringer, R.R. and Crenshaw, E.B.,
III (2001) BMPR-IA signaling is required for the formation of the apical
ectodermal ridge and dorsal-ventral patterning of the limb. Development,
30. Heydemann, A., Nguyen, L.C. and Crenshaw, E.B., III (2001)
Regulatory regions from the Brn4 promoter direct LACZ expression to
the developing forebrain and neural tube. Brain Res. Dev. Brain Res.,
31. Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H., Cherry,
J.M., Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T. et al. (2000)
Gene ontology: tool for the unification of biology. The Gene Ontology
Consortium. Nat. Genet., 25, 25–29.
32. Zhang, B., Kirov, S. and Snoddy, J. (2005) WebGestalt: an integrated
system for exploring gene sets in various biological contexts. Nucleic
Acids Res., 33, W741–W748.
33. Lee, H.K., Braynen, W., Keshav, K. and Pavlidis, P. (2005) ErmineJ:
tool for functional analysis of gene expression data sets. BMC
Bioinformatics, 6, 269.
34. Anderson, S.A., Eisenstat, D.D., Shi, L. and Rubenstein, J.L. (1997)
Interneuron migration from basal forebrain to neocortex: dependence on
Dlx genes. Science, 278, 474–476.
Human Molecular Genetics, 2008, Vol. 17, No. 233757
35. Petryniak, M.A., Potter, G.B., Rowitch, D.H. and Rubenstein, J.L. (2007)
Dlx1 and Dlx2 control neuronal versus oligodendroglial cell fate
acquisition in the developing forebrain. Neuron, 55, 417–433.
36. Cobos, I., Borello, U. and Rubenstein, J.L. (2007) Dlx transcription
factors promote migration through repression of axon and dendrite
growth. Neuron, 54, 873–888.
37. Anderson, S., Mione, M., Yun, K. and Rubenstein, J.L. (1999)
Differential origins of neocortical projection and local circuit neurons:
role of Dlx genes in neocortical interneuronogenesis. Cereb. Cortex, 9,
38. Stu ¨hmer, T., Anderson, S.A., Ekker, M. and Rubenstein, J.L. (2002)
Ectopic expression of the Dlx genes induces glutamic acid decarboxylase
and Dlx expression. Development, 129, 245–252.
39. Stu ¨hmer, T., Puelles, L., Ekker, M. and Rubenstein, J.L. (2002)
Expression from a Dlx gene enhancer marks adult mouse cortical
GABAergic neurons. Cereb. Cortex, 12, 75–85.
40. Anderson, S.A., Qiu, M., Bulfone, A., Eisenstat, D.D., Meneses, J.,
Pedersen, R. and Rubenstein, J.L. (1997) Mutations of the homeobox
genes Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and
differentiation of late born striatal neurons. Neuron, 19, 27–37.
41. Cobos, I., Broccoli, V. and Rubenstein, J.L. (2005) The vertebrate
ortholog of Aristaless is regulated by Dlx genes in the developing
forebrain. J. Comp. Neurol., 483, 292–303.
42. Marsh, E.D., Minarcik, J., Campbell, K., Brooks-Kayal, A.R. and
Golden, J.A. (2008) FACS-array gene expression analysis during early
development of mouse telencephalic interneurons. Dev. Neurobiol., 68,
43. Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L.,
Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S.
et al. (2005) Gene set enrichment analysis: a knowledge-based approach
for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci.
USA, 102, 15545–15550.
44. Anderson, S.A., Marı ´n, O., Horn, C., Jennings, K. and Rubenstein, J.L.
(2001) Distinct cortical migrations from the medial and lateral
ganglionic eminences. Development, 128, 353–363.
45. Fogarty, M., Grist, M., Gelman, D., Marı ´n, O., Pachnis, V. and Kessaris,
N. (2007) Spatial genetic patterning of the embryonic neuroepithelium
generates GABAergic interneuron diversity in the adult cortex.
J. Neurosci., 27, 10935–10946.
46. Jimenez, D., Lopez-Mascaraque, L.M., Valverde, F. and De Carlos, J.A.
(2002) Tangential migration in neocortical development. Dev. Biol., 244,
47. Marin, O., Anderson, S.A. and Rubenstein, J.L. (2000) Origin and
molecular specification of striatal interneurons. J. Neurosci., 20,
48. Pleasure, S.J., Anderson, S., Hevner, R., Bagri, A., Marin, O.,
Lowenstein, D.H. and Rubenstein, J.L. (2000) Cell migration from the
ganglionic eminences is required for the development of hippocampal
GABAergic interneurons. Neuron, 28, 727–740.
49. Wichterle, H., Turnbull, D.H., Nery, S., Fishell, G. and Alvarez-Buylla,
A. (2001) In utero fate mapping reveals distinct migratory pathways and
fates of neurons born in the mammalian basal forebrain. Development,
50. Schmidt, E.E., Taylor, D.S., Prigge, J.R., Barnett, S. and Capecchi, M.R.
(2000) Illegitimate Cre-dependent chromosome rearrangements in
transgenic mouse spermatids. Proc. Natl Acad. Sci. USA, 97,
51. Tripodi, M., Filosa, A., Armentano, M. and Studer, M. (2004) The
COUP-TF nuclear receptors regulate cell migration in the mammalian
basal forebrain. Development, 131, 6119–6129.
52. Nery, S., Fishell, G. and Corbin, J.G. (2002) The caudal ganglionic
eminence is a source of distinct cortical and subcortical cell populations.
Nat. Neurosci., 5, 1279–1287.
53. Visel, A., Thaller, C. and Eichele, G. (2004) GenePaint.org: an atlas of
gene expression patterns in the mouse embryo. Nucleic Acids Res., 32,
54. Ho Sui, S.J., Mortimer, J.R., Arenillas, D.J., Brumm, J., Walsh, C.J.,
Kennedy, B.P. and Wasserman, W.W. (2005) oPOSSUM: identification
of over-represented transcription factor binding sites in co-expressed
genes. Nucleic Acids Res., 33, 3154–3164.
55. Schones, D.E., Sumazin, P. and Zhang, M.Q. (2005) Similarity of
position frequency matrices for transcription factor binding sites.
Bioinformatics, 21, 307–313.
56. Luscombe, N.M., Austin, S.E., Berman, H.M. and Thornton, J.M. (2000)
An overview of the structures of protein-DNA complexes. Genome Biol.,
57. Sandelin, A. and Wasserman, W.W. (2004) Constrained binding site
diversity within families of transcription factors enhances pattern
discovery bioinformatics. J. Mol. Biol., 338, 207–215.
58. Crooks, G.E., Hon, G., Chandonia, J.M. and Brenner, S.E. (2004)
WebLogo: a sequence logo generator. Genome Res., 14, 1188–1190.
59. Meijlink, F., Beverdam, A., Brouwer, A., Oosterveen, T.C. and Berge,
D.T. (1999) Vertebrate aristaless-related genes. Int. J. Dev. Biol., 43,
60. Xie, X., Lu, J., Kulbokas, E.J., Golub, T.R., Mootha, V., Lindblad-Toh,
K., Lander, E.S. and Kellis, M. (2005) Systematic discovery of
regulatory motifs in human promoters and 30-UTRs by comparison of
several mammals. Nature, 434, 338–345.
61. Kato, M. and Dobyns, W.B. (2005) X-linked lissencephaly with
abnormal genitalia as a tangential migration disorder causing intractable
epilepsy: proposal for a new term, ‘interneuronopathy’. J. Child Neurol.,
62. Collombat, P., Hecksher-Sørensen, J., Broccoli, V., Krull, J., Ponte, I.,
Mundiger, T., Smith, J., Gruss, P., Serup, P. and Mansouri, A. (2005)
The simultaneous loss of Arx and Pax4 genes promotes a
somatostatin-producing cell fate specification at the expense of the
alpha- and beta-cell lineages in the mouse endocrine pancreas.
Development, 132, 2969–2980.
63. Kojima, T., Tsuji, T. and Saigo, K. (2005) A concerted action of a
paired-type homeobox gene, aristaless, and a homolog of Hox11/tlx
homeobox gene, clawless, is essential for the distal tip development of
the Drosophila leg. Dev. Biol., 279, 434–445.
64. Pueyo, J.I. and Couso, J.P. (2004) Chip-mediated partnerships of the
homeodomain proteins Bar and Aristaless with the LIM-HOM proteins
Apterous and Lim1 regulate distal leg development. Development, 131,
65. Tang, S. and Breitman, M.L. (1995) The optimal binding sequence of the
Hox11 protein contains a predicted recognition core motif. Nucleic Acids
Res., 23, 1928–1935.
66. Melkman, T. and Sengupta, P. (2005) Regulation of chemosensory and
GABAergic motor neuron development by the C. elegans Aristaless/Arx
homolog alr-1. Development, 132, 1935–1949.
67. Cooney, A.J., Tsai, S.Y., O’Malley, B.W. and Tsai, M.J. (1992) Chicken
ovalbumin upstream promoter transcription factor (COUP-TF) dimers
bind to different GGTCA response elements, allowing COUP-TF to
repress hormonal induction of the vitamin D3, thyroid hormone, and
retinoic acid receptors. Mol. Cell Biol., 12, 4153–4163.
68. Nakshatri, H. and Chambon, P. (1994) The directly repeated RG(G/
T)TCA motifs of the rat and mouse cellular retinol-binding protein II
genes are promiscuous binding sites for RAR, RXR, HNF-4 and ARP-1
homo- and heterodimers. J. Biol. Chem., 269, 890–902.
69. Li, G., Adesnik, H., Li, J., Long, J., Nicoll, R.A., Rubenstein, J.L. and
Pleasure, S.J. (2008) Regional distribution of cortical interneurons and
development of inhibitory tone are regulated by Cxcl12/Cxcr4 signaling.
J. Neurosci., 28, 1085–1098.
70. Lo ´pez-Bendito, G., Sa ´nchez-Alcan ˜iz, J.A., Pla, R., Borrell, V., Pico ´, E.,
Valdeolmillos, M. and Marı ´n, O. (2008) Chemokine signaling controls
intracortical migration and final distribution of GABAergic interneurons.
J. Neurosci., 28, 1613–1624.
71. Liapi, A., Pritchett, J., Jones, O., Fujii, N., Parnavelas, J.G. and
Nadarajah, B. (2008) Stromal-derived factor 1 signalling regulates radial
and tangential migration in the developing cerebral cortex. Dev.
Neurosci., 30, 117–131.
72. Tiveron, M.C., Rossel, M., Moepps, B., Zhang, Y.L., Seidenfaden, R.,
Favor, J., Konig, N. and Cremer, H. (2006) Molecular interaction
between projection neuron precursors and invading interneurons via
stromal-derived factor 1 (CXCL12)/CXCR4 signaling in the cortical
subventricular zone/intermediate zone. J. Neurosci., 26, 13273–13278.
73. Stumm, R.K., Zhou, C., Ara, T., Lazarini, F., Dubois-Dalcq, M.,
Nagasawa, T., Hollt, V. and Schulz, S. (2003) CXCR4 regulates
interneuron migration in the developing neocortex. J. Neurosci., 23,
74. Arlotta, P., Molyneaux, B.J., Jabaudon, D., Yoshida, Y. and Macklis,
J.D. (2008) Ctip2 controls the differentiation of medium spiny neurons
and the establishment of the cellular architecture of the striatum.
J. Neurosci., 28, 622–632.
3758 Human Molecular Genetics, 2008, Vol. 17, No. 23
75. Chang, C.W., Tsai, C.W., Wang, H.F., Tsai, H.C., Chen, H.Y., Tsai,
T.F., Takahashi, H., Li, H.Y., Fann, M.J., Yang, C.W. et al. (2004)
Identification of a developmentally regulated striatum-enriched
zinc-finger gene, Nolz-1, in the mammalian brain. Proc. Natl Acad. Sci.
USA, 101, 2613–2618.
76. Sock, E., Schmidt, K., Hermanns-Borgmeyer, I., Bosl, M.R. and Wegner,
M. (2001) Idiopathic weight reduction in mice deficient in the
high-mobility-group transcription factor Sox8. Mol. Cell Biol., 21,
77. Garel, S., Marin, F., Grosschedl, R. and Charnay, P. (1999) Ebf1 controls
early cell differentiation in the embryonic striatum. Development, 126,
78. Lobo, M.K., Karsten, S.L., Gray, M., Geschwind, D.H. and Yang, X.W.
(2006) FACS-array profiling of striatal projection neuron subtypes in
juvenile and adult mouse brains. Nat. Neurosci., 9, 443–452.
79. Moon Edley, S. and Herkenham, M. (1984) Comparative development of
striatal opiate receptors and dopamine revealed by autoradiography and
histofluorescence. Brain Res., 305, 27–42.
80. Ferland, R.J., Cherry, T.J., Preware, P.O., Morrisey, E.E. and Walsh,
C.A. (2003) Characterization of Foxp2 and Foxp1 mRNA and protein in
the developing and mature brain. J. Comp. Neurol., 460, 266–279.
81. Tamura, S., Morikawa, Y., Iwanishi, H., Hisaoka, T. and Senba, E.
(2003) Expression pattern of the winged-helix/forkhead transcription
factor Foxp1 in the developing central nervous system. Gene Expr.
Patterns, 3, 193–197.
82. Tamura, S., Morikawa, Y., Iwanishi, H., Hisaoka, T. and Senba, E.
(2004) Foxp1 gene expression in projection neurons of the mouse
striatum. Neuroscience, 124, 261–267.
83. Lobo, M.K., Yeh, C. and Yang, X.W. (2008) Pivotal role of early B-cell
factor 1 in development of striatonigral medium spiny neurons in the
matrix compartment. J. Neurosci. Res., 86, 2134–2146.
84. Kobayashi, M., Taniura, H. and Yoshikawa, K. (2002) Ectopic
expression of necdin induces differentiation of mouse neuroblastoma
cells. J. Biol. Chem., 277, 42128–42135.
85. Taniura, H., Taniguchi, N., Hara, M. and Yoshikawa, K. (1998) Necdin,
a postmitotic neuron-specific growth suppressor, interacts with viral
transforming proteins and cellular transcription factor E2F1. J. Biol.
Chem., 273, 720–728.
86. Kuwajima, T., Nishimura, I. and Yoshikawa, K. (2006) Necdin promotes
GABAergic neuron differentiation in cooperation with Dlx
homeodomain proteins. J. Neurosci., 26, 5383–5392.
87. Callaghan, D.A., Dong, L., Callaghan, S.M., Hou, Y.X., Dagnino, L. and
Slack, R.S. (1999) Neural precursor cells differentiating in the absence of
Rb exhibit delayed terminal mitosis and deregulated E2F 1 and 3
activity. Dev. Biol., 207, 257–270.
88. Ferguson, K.L. and Slack, R.S. (2001) The Rb pathway in neurogenesis.
Neuroreport, 12, A55–A62.
89. Gill, R.M., Slack, R., Kiess, M. and Hamel, P.A. (1998) Regulation of
expression and activity of distinct pRB, E2F, D-type cyclin, and CKI
family members during terminal differentiation of P19 cells. Exp. Cell
Res., 244, 157–170.
90. Vanderluit, J.L., Ferguson, K.L., Nikoletopoulou, V., Parker, M.,
Ruzhynsky, V., Alexson, T., McNamara, S.M., Park, D.S., Rudnicki, M.
and Slack, R.S. (2004) p107 regulates neural precursor cells in the
mammalian brain. J. Cell Biol., 166, 853–863.
91. Wiggan, O., Taniguchi-Sidle, A. and Hamel, P.A. (1998) Interaction of
the pRB-family proteins with factors containing paired-like
homeodomains. Oncogene, 16, 227–236.
92. Benes, F.M. and Berretta, S. (2001) GABAergic interneurons:
implications for understanding schizophrenia and bipolar disorder.
Neuropsychopharmacology, 25, 1–27.
93. Di Cristo, G. (2007) Development of cortical GABAergic circuits and its
implications for neurodevelopmental disorders. Clin. Genet., 72, 1–8.
94. Uhlhaas, P.J. and Singer, W. (2006) Neural synchrony in brain disorders:
relevance for cognitive dysfunctions and pathophysiology. Neuron, 52,
95. Boccaccio, I., Glatt-Deeley, H., Watrin, F., Roeckel, N., Lalande, M. and
Muscatelli, F. (1999) The human MAGEL2 gene and its mouse
homologue are paternally expressed and mapped to the Prader-Willi
region. Hum. Mol. Genet., 8, 2497–2505.
96. Jay, P., Rougeulle, C., Massacrier, A., Moncla, A., Mattei, M.G.,
Malzac, P., Roeckel, N., Taviaux, S., Lefranc, J.L., Cau, P. et al. (1997)
The human necdin gene, NDN, is maternally imprinted and located in the
Prader-Willi syndrome chromosomal region. Nat. Genet., 17, 357–361.
97. MacDonald, H.R. and Wevrick, R. (1997) The necdin gene is deleted in
Prader-Willi syndrome and is imprinted in human and mouse. Hum. Mol.
Genet., 6, 1873–1878.
98. Sadakata, T., Washida, M., Iwayama, Y., Shoji, S., Sato, Y., Ohkura, T.,
Katoh-Semba, R., Nakajima, M., Sekine, Y., Tanaka, M. et al. (2007)
Autistic-like phenotypes in Cadps2-knockout mice and aberrant
CADPS2 splicing in autistic patients. J. Clin. Invest., 117, 931–943.
99. Gibson, W.T., Harvard, C., Qiao, Y., Somerville, M.J., Lewis, M.E. and
Rajcan-Separovic, E. (2008) Phenotype-genotype characterization of
alpha-thalassemia mental retardation syndrome due to isolated
monosomy of 16p13.3. Am. J. Med. Genet. A, 146, 225–232.
100. Pfeifer, D., Poulat, F., Holinski-Feder, E., Kooy, F. and Scherer, G.
(2000) The SOX8 gene is located within 700 kb of the tip of
chromosome 16p and is deleted in a patient with ATR-16 syndrome.
Genomics, 63, 108–116.
101. Stodgell, C.J., Wanek, N., O’Bara, M., Hyman, S.L., Bryson, S.E. and
Rodier, P.M. (2003) Association of HOXD1 and GBX2 allelic variants
with autisms spectrum disorders. Am. J. Hum. Genet., 71, 2003.
102. Gecz, J., Gedeon, A.K., Sutherland, G.R. and Mulley, J.C. (1996)
Identification of the gene FMR2, associated with FRAXE mental
retardation. Nat. Genet., 13, 105–108.
103. Gu, Y., Shen, Y., Gibbs, R.A. and Nelson, D.L. (1996) Identification of
FMR2, a novel gene associated with the FRAXE CCG repeat and CpG
island. Nat. Genet., 13, 109–113.
104. Dibbens, L.M., Tarpey, P.S., Hynes, K., Bayly, M.A., Scheffer, I.E.,
Smith, R., Bomar, J., Sutton, E., Vandeleur, L., Shoubridge, C. et al.
(2008) X-linked protocadherin 19 mutations cause female-limited
epilepsy and cognitive impairment. Nat. Genet., 40, 776–781.
105. Gecz, J., Baker, E., Donnelly, A., Ming, J.E., McDonald-McGinn, D.M.,
Spinner, N.B., Zackai, E.H., Sutherland, G.R. and Mulley, J.C. (1999)
Fibroblast growth factor homologous factor 2 (FHF2): gene structure,
expression and mapping to the Borjeson-Forssman-Lehmann syndrome
region in Xq26 delineated by a duplication breakpoint in a BFLS-like
patient. Hum. Genet., 104, 56–63.
106. Francks, C., Maegawa, S., Lauren, J., Abrahams, B.S., Velayos-Baeza,
A., Medland, S.E., Colella, S., Groszer, M., McAuley, E.Z., Caffrey,
T.M. et al. (2007) LRRTM1 on chromosome 2p12 is a maternally
suppressed gene that is associated paternally with handedness and
schizophrenia. Mol. Psychiatry, 12, 1129–1139, 1057.
107. Ikeda, M., Iwata, N., Kitajima, T., Suzuki, T., Yamanouchi, Y.,
Kinoshita, Y. and Ozaki, N. (2006) Positive association of the serotonin
5-HT7 receptor gene with schizophrenia in a Japanese population.
Neuropsychopharmacology, 31, 866–871.
108. Tompers, D.M. and Labosky, P.A. (2004) Electroporation of murine
embryonic stem cells: a step-by-step guide. Stem Cells, 22, 243–249.
109. Naiche, L.A. and Papaioannou, V.E. (2007) Cre activity causes
widespread apoptosis and lethal anemia during embryonic development.
Genesis, 45, 768–775.
110. Buerger, A., Rozhitskaya, O., Sherwood, M.C., Dorfman, A.L., Bisping,
E., Abel, E.D., Pu, W.T., Izumo, S. and Jay, P.Y. (2006) Dilated
cardiomyopathy resulting from high-level myocardial expression of
Cre-recombinase. J. Card. Fail., 12, 392–398.
111. Lee, J.Y., Ristow, M., Lin, X., White, M.F., Magnuson, M.A. and
Hennighausen, L. (2006) RIP-Cre revisited, evidence for impairments of
pancreatic beta-cell function. J. Biol. Chem., 281, 2649–2653.
112. Schmidt-Supprian, M. and Rajewsky, K. (2007) Vagaries of conditional
gene targeting. Nat. Immunol., 8, 665–668.
113. Forni, P.E., Scuoppo, C., Imayoshi, I., Taulli, R., Dastru, W., Sala, V.,
Betz, U.A., Muzzi, P., Martinuzzi, D., Vercelli, A.E. et al. (2006) High
levels of Cre expression in neuronal progenitors cause defects in brain
development leading to microencephaly and hydrocephaly. J Neurosci.,
114. Affymetrix (2004) Affymetrix, Santa Clara, CA.
115. Gentleman, R.C., Carey, V.J., Bates, D.M., Bolstad, B., Dettling, M.,
Dudoit, S., Ellis, B., Gautier, L., Ge, Y., Gentry, J. et al. (2004)
Bioconductor: open software development for computational biology
and bioinformatics. Genome Biol., 5, R80.
116. Wu, Z., Irizarry, R.A., Gentleman, R., Murillo, F.M. and Spencer, F.
(2004) A model based background adjustment for oligonucleotide
expression arrays. J. Am. Stat. Assoc., 99, 909–917.
Human Molecular Genetics, 2008, Vol. 17, No. 233759
117. Gautier, L., Cope, L., Bolstad, B.M. and Irizarry, R.A. (2004) affy— Download full-text
analysis of Affymetrix GeneChip data at the probe level. Bioinformatics,
118. Smyth, G.K. (2004) Linear models and empirical bayes methods for
assessing differential expression in microarray experiments. Stat. Appl.
Genet. Mol. Biol., 3, Article3.
119. Smyth, G.K. (2005) In Gentleman, R., Carey, V.J., Dudoit, S., Irizarry,
R. and Huber, W. (eds), Bioinformatics and Computational Biology
Using R and Bioconductor, Springer, New York, pp. 397–420.
120. Benjamini, Y. and Hochberg, Y. (1995) Controlling the false discovery
rate: a practical and powerful approach to multiple testing. J R Statist Soc
B, 57, 289–300.
121. de Hoon, M.J., Imoto, S., Nolan, J. and Miyano, S. (2004) Open source
clustering software. Bioinformatics, 20, 1453–1454.
122. Saldanha, A.J. (2004) Java Treeview—extensible visualization of
microarray data. Bioinformatics, 20, 3246–3248.
123. Livak, K.J. and Schmittgen, T.D. (2001) Analysis of relative gene
expression data using real-time quantitative PCR and the 2(-Delta Delta
C(T)) Method. Methods, 25, 402–408.
124. Subramanian, A., Kuehn, H., Gould, J., Tamayo, P. and Mesirov, J.P.
(2007) GSEA-P: a desktop application for Gene Set Enrichment
Analysis. Bioinformatics, 23, 3251–3253.
125. Barrett, T., Troup, D.B., Wilhite, S.E., Ledoux, P., Rudnev, D.,
Evangelista, C., Kim, I.F., Soboleva, A., Tomashevsky, M. and Edgar, R.
(2007) NCBI GEO: mining tens of millions of expression profiles—
database and tools update. Nucleic Acids Res., 35, D760–D765.
126. Saxena, V., Orgill, D. and Kohane, I. (2006) Absolute enrichment: gene
set enrichment analysis for homeostatic systems. Nucleic Acids Res.,
127. Bryne, J.C., Valen, E., Tang, M.H., Marstrand, T., Winther, O., da
Piedade, I., Krogh, A., Lenhard, B. and Sandelin, A. (2008) JASPAR, the
open access database of transcription factor-binding profiles: new
content and tools in the 2008 update. Nucleic Acids Res., 36,
128. Cho, G., Lim, Y., Zand, D. and Golden, J.A. (2007) Sizn1 is a novel
protein that functions as a transcriptional coactivator of BMP signalling.
Mol. Cell Biol., 28, 1565–1572.
129. Marklund, N., Fulp, C.T., Shimizu, S., Puri, R., McMillan, A.,
Strittmatter, S.M. and McIntosh, T.K. (2006) Selective temporal and
regional alterations of Nogo-A and small proline-rich repeat protein 1A
(SPRR1A) but not Nogo-66 receptor (NgR) occur following traumatic
brain injury in the rat. Exp. Neurol., 197, 70–83.
130. Lu, M.M., Li, S., Yang, H. and Morrisey, E.E. (2002) Foxp4: a novel
member of the Foxp subfamily of winged-helix genes co-expressed with
Foxp1 and Foxp2 in pulmonary and gut tissues. Mech Dev., 119 (Suppl.
131. Nasrallah, I.M., Minarcik, J.C. and Golden, J.A. (2004) A polyalanine
tract expansion in Arx forms intranuclear inclusions and results in
increased cell death. J. Cell Biol., 167, 411–416.
132. Siepel, A., Bejerano, G., Pedersen, J.S., Hinrichs, A.S., Hou, M.,
Rosenbloom, K., Clawson, H., Spieth, J., Hillier, L.W., Richards, S. et al.
(2005) Evolutionarily conserved elements in vertebrate, insect, worm,
and yeast genomes. Genome Res., 15, 1034–1050.
133. Kim, J.B., Wright, H.M., Wright, M. and Spiegelman, B.M. (1998)
ADD1/SREBP1 activates PPARgamma through the production of
endogenous ligand. Proc. Natl Acad. Sci. USA, 95, 4333–4337.
134. Cho, G., Kim, J., Rho, H.M. and Jung, G. (1995) Structure-function
analysis of the DNA binding domain of Saccharomyces cerevisiae
ABF1. Nucleic Acids Res., 23, 2980–2987.
135. Kent, W.J., Sugnet, C.W., Furey, T.S., Roskin, K.M., Pringle, T.H.,
Zahler, A.M. and Haussler, D. (2002) The human genome browser at
UCSC. Genome Res., 12, 996–1006.
3760 Human Molecular Genetics, 2008, Vol. 17, No. 23