Snf2l Regulates Foxg1-Dependent
Progenitor Cell Expansion
in the Developing Brain
Darren J. Yip,1,3,6Chelsea P. Corcoran,1,3,6Matı ´as Alvarez-Saavedra,1,4,6Adriana DeMaria,1Stephen Rennick,1,3
Alan J. Mears,2,4Michael A. Rudnicki,1,4Claude Messier,5and David J. Picketts1,3,*
1Regenerative Medicine Program
Ottawa Hospital Research Institute, Ottawa, ON K1H 8L6, Canada
3Department of Biochemistry, Microbiology, and Immunology
4Department of Cellular and Molecular Medicine
5School of Psychology
University of Ottawa, Ottawa, ON K1H 8M5, Canada
6These authors contributed equally to this work
Balancing progenitor cell self-renewal and differenti-
ation is essential for brain development and is regu-
lated by the activity of chromatin remodeling com-
plexes. Nevertheless, linking chromatin changes to
specific pathways that control cortical histogenesis
remains a challenge. Here we identify a genetic inter-
action between the chromatin remodeler Snf2l and
Foxg1, a key regulator of neurogenesis. Snf2l mutant
mice exhibit forebrain hypercellularity arising from
increased Foxg1 expression, increased progen-
itor cell expansion, and delayed differentiation. We
demonstrate that Snf2l binds to the Foxg1 locus at
midneurogenesis and that the phenotype is rescued
by reducing Foxg1 dosage, thus revealing that
Snf2l and Foxg1 function antagonistically to regulate
The development of the mammalian forebrain is a tightly regu-
lated process that involves expansion of the neural progenitor
pool followed by waves of asymmetric division to generate an
array of specialized neuronal subtypes that comprise the six
layers of the cortex (Gupta et al., 2002). The decision of pre-
cursor cells to self renew or differentiate is regulated by extrinsic
factors and a cell intrinsic program largely mediated by neuro-
genic transcription factors (Go ¨tz and Huttner, 2005; Guillemot,
2007; Shen et al., 2006). The role of epigenetic factors in fore-
brain development is implicit based in part, on the rising number
of neurodevelopmental disorders caused by mutations in genes
encoding chromatin remodeling proteins (van Bokhoven and
Kramer, 2010). Further, the use of mouse models and neural
stem cell cultures have begun to elucidate the epigenetic mech-
anisms controlling neurogenesis (Be ´rube ´ et al., 2005; Fasano
et al., 2009; Lessard et al., 2007; Lim et al., 2009; Molofsky
et al., 2005; Molofsky et al., 2003). Nonetheless, the continued
deciphering of the interplay between epigenetic regulators
and neurogenic transcription factors is paramount to our under-
standing of forebrain development.
The mammalian ISWI chromatin remodeling proteins, Snf2h
and Snf2l, are components of multiple protein complexes with
diverse functions that include nucleosome assembly/spacing
during replication and transcriptional regulation (Dirscherl and
Krebs, 2004). In the developing brain, Snf2h expression is prev-
alent in progenitor cells whereas Snf2l expression increases
during terminal differentiation (Barak et al., 2003; Lazzaro and
Picketts, 2001). Other studies suggest these genes may have
distinct roles during neural development. Inactivation of Snf2h
results in proliferation defects and embryonic lethality in mice
whereas ectopic expression of Snf2l induces terminal differ-
entiation of cultured neuroblastoma cells (Barak et al., 2003;
Stopka and Skoultchi, 2003). To further assess the in vivo
requirement for Snf2l we used a conditional targeting approach
to impair remodeling activity by the removal of the ATP-binding
motif of the Snf2l gene. These Snf2l mutant mice exhibit
deregulated Foxg1 expression resulting in enhanced progenitor
expansion, delayed neurogenesis and hypercellularity in the
RESULTS AND DISCUSSION
Hypercellularity in Snf2l-Deficient Mice
We utilized a conditional gene targeting approach by inserting
loxP sites that flanked exon 6 (Figure 1A). Exon 6 encodes the
ATP-binding motif of the SNF2 domain that is critical for the
chromatin remodeling activity of Snf2l containing protein com-
plexes (Figure 1A). To assess the impact of widespread early
embryonic loss of Snf2l activity we bred Snf2lf/fmice to
GATA1-Cre animals, which exhibit ubiquitous and early embry-
onic Cre activity (Mao et al., 1999). Unlike the embryonic lethality
of Snf2h null mice (Stopka and Skoultchi, 2003), the Snf2lf/Y;
GATA1-Cre+/?animals were healthy, fertile, and born at normal
Mendelian ratios. As such, we bred the exon 6-deleted allele
independent of Cre recombinase (hereafter named Ex6DEL).
Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc. 871
The Ex6DEL male mice expressed a stable Snf2l transcript
lacking exon 6 and produced a corresponding protein product
reduced in size by 7 kDa (Figures 1B and S1A available online).
The mice displayed no overt phenotype and performed equally
well to wild-type (WT) littermates in several behavior tests
(Figures S1B–S1D). The only gross morphological difference
ure S1E). Because Snf2l can promote neuronal differentiation
in vitro, we reasoned that the loss of Snf2l activity might result
in unfettered proliferation and increased brain size. Indeed,
there was a 1.4-fold increase in the Ex6DEL brain:body mass
ratio (Figure 1C) that arose from an increased brain mass
(0.65 g ± 0.18 g; n = 29; 8–43 weeks of age) compared to WT
littermates (0.48 g ± 0.06 g; n = 26; p = 1.53 3 10?5), with no
difference in body weight (Ex6DEL, 36.5 g ± 6.8 g; WT,
37.9 g ± 7.6 g; p = 0.48). Homozygous Ex6DEL female mice
showed a similar increase in brain:body mass ratio, whereas
heterozygous Ex6DEL females had normal ratios, suggesting
Figure 1. Increased Brain Mass and Cell Density in Ex6DEL Mice
(A) Schematic of the Snf2l locus, targeted allele, and the Ex6DEL allele (LoxP sites, gray triangles; NeoR, neomycin resistance gene). Below, sequence
conservation of the 60 amino acids encoded by exon 6. Lys residue critical for ATPase activity is highlighted in red.
(B) Snf2l immunoblot of E15.5 cortical extracts from WT, HET, and Ex6DEL mice.
(C) Plot of brain weight to body weight ratios for Ex6DEL, WT, or HET mice. Brains from individual mice are represented by a dot with the mean indicated by
a horizontal line.
(D and E) Nissl stained P7 rostral coronal brain sections (D) with high magnification images underneath (boxed areas in D) and corresponding plot (E) of cortical
thickness from medial and medial-lateral regions.
(F) Cell counts from E15.5 cortical sections.
(G and H) Cell counts within a fixed brain volume from P7 (G) or Adult (H) brains. CP, cortical plate; II/III, IV, V/VI: cortical layers II and III, IV, and V and VI,
respectively; IZ, intermediate zone; MZ, medial zone; n.s., not significant; VZ, ventricular zone. For (E–H), n = 4. Bars correspond to the mean values, whereas
error bars indicate the SEM; an asterisk (*) represents a statistically significant change by a two-tailed paired Student t test (p % 0.05) as compared to the WT
sample unless otherwise indicated by brackets.
Snf2l Regulates IPC Expansion via Foxg1
872 Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc.
that the Ex6DEL allele is not functioning as a dominant-negative
or gain-of-function allele but is consistent with a loss-of-func-
tion phenotype (Figure 1C). A similar analysis of multiple organs
at a single age (11 weeks) showed that heart mass was the only
other organ increased in size (Table S1). We conclude that the
Ex6DEL mice have an increase in organ mass that primarily
affects the brain.
To characterize the brains we measured cortical thickness at
medial and medial-lateral positions in both rostral and caudal
sections of P7 brains. A statistically significant increase in thick-
ness throughout the medial region of the Ex6DEL cortex was
observed (Figures 1D and 1E and S1F). To determine if in-
creased cortical thickness correlated with cell number we per-
formed cell counts and observed hypercellularity within the
cortex at several ages (Figures 1F–1H). Specifically, increased
cell density was observed beginning embryonically at E15.5,
although at this age statistical significance was restricted to
the ventricular zone (VZ; Figure 1F). At P7, cell density was
increased in cortical layers II/III (Ex6DEL 504 ± 16 cells; WT
440 ± 22 cells; n = 4; p = 0.015), IV (Ex6DEL 364 ± 30 cells;
WT 238 ± 16 cells; n = 4; p = 0.003), and V/VI (Ex6DEL 423 ±
16 cells; WT 258 ± 30 cells; n = 4; p = 0.001) (Figure 1G). Simi-
larly, the Ex6DEL adult cortex also showed increased cell
density in the cortical layers (Figure 1H). We conclude that
Snf2l loss causes hypercellularity and an increased brain mass
without significantly altering the structural morphology of the
Figure 2. Ex6DEL Forebrains Have Altered
(A) E15.5 forebrain sections stained for terminally
differentiated neurons (Tuj) and mitotic progenitors
(PH3). All nuclei are counterstained with DAPI.
Boxed regions are enlarged below. Scale bar
represents 30 mm.
(B) Quantification of PH3+cells to total cells in the
VZ of E15.5 embryos (n = 3).
(C) E15.5 cortical sections stained for BrdU after
a 2 hr pulse (n = 4). Scale bar represents 100 mm.
(D) Normalized plot of the percentage of BrdU+
cells following a 2 hr BrdU pulse (n = 3).
(E and F) Tbr2 and BrdU double labeled E13.5
sections. Scale bar represents 100 mm. The boxed
regions are shown below at higher magnification
(scale bar represents 50 mm) and were represen-
tative of the images used to count the percentage
of double-labeled cells (n = 3) shown in (F).
(G) Schematic of IdU and BrdU injection times to
determine S-phase length. Plot of normalized
S-phase length (h) at E13.5 and E15.5 (n = 3). For
(B), (D), (F), and (G), bars represent the mean ±
SEM; *p < 0.05 by t test.
Increased Progenitor Self-Renewal
in Developing Brain of Ex6DEL Mice
Increased cell numbers can arise from
decreased cell death, changes in cell
cycle kinetics, or increased self-renewal.
We did not observe any differences in
the proportion of apoptotic cells by
TUNEL staining in WT or mutant mice
indicating that altered cell survival is not
the cause of the hypercellularity (Figures S1G and S1H). As an
initial measure of cell kinetics we stained for mitotic cells at
E15.5 using antibodies to phospho-histone H3 (PH3). We ob-
served a 2.9-fold increase in the proportion of mitotic cells in
the VZ of Ex6DEL mice compared to control animals (0.118 ±
7.9 3 10?5versus 0.040 ± 0.002 per 125 mm2, respectively; n =
3; p < 0.05) prompting us to investigate other cell cycle parame-
ters (Figures 2A and 2B). Next, we examined the proportion of
S-phase cells by pulse-labeling cells with 5-bromo-20-deoxyuri-
dine (BrdU) at E13.5, 15.5, and 17.5 (Figures 2C and 2D). We
observed a statistically significant increase in BrdU+cells in
the Ex6DEL neocortex at E13.5 (1.47-fold increase; n = 4; p =
0.0488) and at E15.5 (1.35-fold increase; n = 6; p = 0.0074). A
similar trend, albeit not significant, was also observed at E17.5
(1.26-fold increase; n = 3; p = 0.1503). An increased proportion
of cells in both S- and M-phase suggest that proliferation is
increased in the Ex6DEL animals.
Consistent with increased proliferation, we observed in-
creased numbers of proliferating apical (Pax6+/BrdU+) and in-
termediate/basal (Tbr2+/BrdU+) progenitor cells (IPCs) in the
Ex6DEL neocortex at E13.5 (Figures 2E and 2F and S2A). By
E15.5, the level of proliferating apical progenitors was normal,
whereas IPC numbers (Tbr2+) remained increased (Figure 2F).
As a more direct assessment of cell cycle kinetics we used
Idu/BrdU double labeling to measure S-phase length (Fukumitsu
et al., 2006). We observed that S-phase length was significantly
shorter in the Ex6DEL animals at E15.5 confirming that the
Snf2l Regulates IPC Expansion via Foxg1
Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc. 873
progenitors cycled at a faster rate (Figure 2G). A similar trend
was observed at E13.5, although significance was not attained.
To address the possibility that increased self-renewal also
contributed to the increased number of proliferating progenitors,
cells were pulse-labeled with BrdU at E12.5, or E16.5 and the
brains harvested 1 day later for staining with Ki67 antigen,
a protein expressed in all dividing cells. The number of BrdU+
Ki67+double-labeled cells represents the proportion of progen-
itors that have remained in cycle. We observed a significant
increase in the proportion of double-labeled cells in the mutant
animals at E13.5 (WT: 39.2 ± 3.2%; Ex6DEL: 50.8 ± 4.2%;
n = 9; p = 0.0056) suggesting that a greater fraction of progeni-
tors undergo self-renewal within a 24 hr period (Figures 3A
and 3B). A similar increase was observed in the mutants at
E17.5 (Figure 3B). Taken together, the data demonstrate that
increased hypercellularity observed in the Ex6DEL mice results
from a combination of an increased progenitor cell cycle rate
and enhanced self-renewal, primarily of IPCs.
Altered Timing of Neurogenesis and Increased Neuronal
The altered proliferation of Ex6DEL progenitors coupled with the
increased cortical thickness prompted us to examine the fore-
brain for changes in cortical lamination, the timing of neurogen-
esis, and the production of both deep (layer V, VI) and superficial
(layers II–IV) neurons. Staining with six different layer-specific
markers at several stages (E15.5, 17.5, 18.5) revealed that
each marker was expressed in its correct laminar position (see
Figures 3C and 3E, S2B and S2E, and S3B and S3C). Marker
positive cell counts at E18.5 revealed that both early-born
deep (Nurr1, CTIP2) and later-born superficial (Satb2, Cux1)
neurons were increased in number (Figures S2C and S3A and
S3D). Nonetheless, this was not true of all markers (e.g.,
Foxp1) indicating that specific fate changes are likely occurring
identities within layer V/VI are characterized by CTIP2 and Tbr1
levels. We observed a qualitative increase in the proportion of
Figure 3. Delayed Neurogenesis and Increased Neuronal Output
(A) Merged images for cycling progenitors pulsed with BrdU at E16.5 then harvested at E17.5 and stained for Ki67 (red) and BrdU (green). Scale bar represents
(B) Plot from (A) of the percentage of cells re-entering the cell cycle (BrdU+, Ki67+; n = 9).
(C) E18.5sectionsstainedforCTIP2andBrdUfollowing neuronalbirthdating atE12.5.Scalebar represents50mm.Boxedregion isenlargedonright andbrightly-
fluorescent BrdU+cells are circled. Scale bar represents 20 mm.
(D) The location of BrdU-densely+cells from (C) were quantified and normalized to WT levels (n = 3).
(E) E18.5 sections stained for SATB2 and Cux1, markers of superficial neuronal layers. Scale bars represent 50 mm. For (B) and (D), bars represent the mean ±
SEM; *p < 0.05 by t test.
Snf2l Regulates IPC Expansion via Foxg1
874 Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc.
CTIP2high/Tbr1lowexpressing cells at E15.5 within layers V/VI
(Figure S2E) that could be suggestive of a fate change (McKenna
et al., 2011). Similarly, at E17.5 the number of CTIP2+cells in
layer V remained high and we observed a reduction in Foxp1+
cells (Figure S3B), which may be reflective of altered sensory
motor projections (Su ¨rmeli et al., 2011). Taken together, these
observations highlight that the Ex6DEL phenotype is complex
with specific fate changes likely accompanying the increased
neuronal output within the cortical layers.
The cortical plate is established by the successive birth of
neurons that migrate to different layers. Layer I neurons are
born at ?E11.5 followed around E12.5 by the production of layer
VI and layer V neurons, respectively. Thus, a delay in cell cycle
exit would be predicted to result in a shift in the timing of this
sequence such that the production of ‘‘early’’ neurons will occur
at later stages. To examine whether additional progenitor cell
division(s) altered the timing of neurogenesis we pulse-labeled
progenitors with BrdU at E12.5 then double labeled cells at
E18.5 for BrdU and CTIP2, a marker of layer V–VI neurons.
Brightly fluorescent BrdU-labeled cells (BrdU staining >75% of
nucleus) are representative of cells born at E12.5. We observed
a significant reduction in the number of CTIP2/BrdU-double
positive cells in layer V of the mutant cortex (Figures 3C and
3D). Indeed, the majority of the brightly fluorescent cells in the
Ex6DEL cortex resided within layers 1 and VI suggesting that
the neurons ‘‘born’’ at E12.5 in the mutants were adopting fates
that are characteristic of an earlier stage. Similar results were
obtained for Nurr1 (Figure S2D). Thus, aberrant Snf2l function
disrupts progenitor cell cycle kinetics, self-renewal decisions,
and alters the timing of neurogenesis that collectively, increases
overall cell number in the developing Ex6DEL neocortex.
Foxg1 Is Misregulated in Ex6DEL Mice
Gene profiling at E15.5 revealed a significant increase in the
expression of Foxg1 in the mutant cortex that was confirmed
by qPCR (Table S2 and Figure 4A). Foxg1/Brain factor-1 is a
forkhead homeodomain transcription factor that controls NSC
self-renewal, IPC expansion and the timing of neurogenesis
(Fasano et al., 2009; Shen et al., 2006; Siegenthaler and Miller,
2005; Siegenthaler et al., 2008). Mice deficient in Foxg1 have
a severe reduction in the size of the cerebral hemispheres as
neural progenitors undergo premature differentiation, exhaust
the progenitor pool at the expense of late-born neurons, and
undergo lateral to medial repatterning of Cajal-Retzius neurons
(Dou et al., 1999; Hanashima et al., 2004; Muzio and Mallamaci,
2005; Xuan et al., 1995). Conversely, NSCs transduced with
Foxg1-expressinglentiviralvectors increasedtheneural progen-
itor pool, delayed neurogenesis, and increased neuronal output
(Brancaccio et al., 2010). Thus, we postulated that increased
Foxg1 expression represented an excellent mechanism to
explain the Ex6DEL phenotype. Consistent with this hypothesis,
Foxg1 protein levels were increased in the mutant animals
compared to WT mice (Figures 4B and S4A–S4C). Notably, the
SVZ is comprised of a mixture of high and low Foxg1 expressing
cells and the expression persists at a high level in the cortical
plate (Figures 4B and S4B and S4C).
Recent studies have shown that Foxg1 mediates the repres-
sion of Cdkn1a (p21Cip1) to inhibit cell cycle exit and promote
IPC expansion (Siegenthaler and Miller, 2005; Siegenthaler
et al., 2008). Consistent with the increase in Foxg1 we observed
ure 4A). We also assessed the expression of other cdk inhibitors
and observed reduced expression of Cdnk1b (p27Kip1) and
Cdnk2a (p16INK4a) but not Cdnk1c (p57Kip2). We conclude
that the Ex6DEL mice have altered Foxg1expression thataffects
downstream target gene expression.
We reasoned that Snf2l complexes are recruited to the Foxg1
locus to limit the expression level of the gene and promote
terminal differentiation. To assess whether Snf2l is enriched at
the Foxg1 gene we performed chromatin immunoprecipitation
(ChIP) assays. Chromatin isolated from cortices dissected from
E15.5 embryos was incubated with a Snf2l antibody previously
used to identify Snf2l target genes, or with sheep IgG as a nega-
tive control (Barak et al., 2003; Lazzaro et al., 2006). Eleven
amplicons (R1–R11) were designed for qPCR that span the
Foxg1 gene (Figure 4C). The results of these experiments (Fig-
ure 4D) showed that Snf2l binding was specifically enriched at
the region that encompasses the P1 promoter (R8, p % 0.012),
which drives expression of the more abundant class 1 Foxg1
transcript (Li et al., 1996). Luciferase reporter assays confirmed
the importance of a 1.2 kb fragment for gene expression.
Although this fragment encompassed the R8 region, addition
of Snf2l or Snf2h in the assay system did not have any effect
indicating that the functionality may be context-dependent
(Figures S4D–S4F). Nonetheless, we conclude that Snf2l binds
to the Foxg1gene, consistent with a direct rolefor Snf2l-contain-
ing complexes in the regulation of Foxg1 expression. These
results further suggest that misregulation of Foxg1 underlies
the phenotype of Ex6DEL mice.
Reduction of Foxg1 Rescues the Ex6DEL Phenotype
If Snf2l-containing chromatin remodeling complexes are re-
quired to regulate Foxg1 expression then we predicted that
genetic reduction of Foxg1 dosage should rescue the phenotype
of the Ex6DEL mice. Ex6DEL mice were bred to Foxg1 heterozy-
gous mice and the resulting Ex6DEL; Foxg1+/?animals were
examined at E15.5 for signs of genetic rescue. In this regard,
we examined the proportion of mitotic cells and the fraction of
IPCs. As shown in Figures 4E and S4G, WT and Foxg1 heterozy-
gous mice had no significant difference in the proportion of PH3+
cells (0.034 ± 0.003 versus 0.033 ± 0.004 per fixed area). The
Ex6DEL mice had a 3-fold increase in PH3+cells (0.107±
0.006) but this was rescued in the Ex6DEL; Foxg1+/?animals
(0.039 ± 0.002). Similarly, only the Ex6DEL mice showed a signif-
icant difference in the percentage of Tbr2/BrdU-double-positive
cells in the VZ/SVZ of E15.5 cortical sections when compared
to WT or Foxg1 heterozygous littermates (Figures 4F and 4G).
Remarkably, the increase in double positive cells was com-
that the increased IPC numbers was dependent on Foxg1
expression levels (Figures 4F and 4G). Taken together, these
data demonstrate that Snf2l functions in the same genetic
pathway as Foxg1 to control the balance between progenitor
proliferation and differentiation.
Modification of chromatin structure has emerged as a funda-
mental process controlling brain development through disease
gene identification and a growing number of in vitro studies
examining NSC proliferation and differentiation (Hamby et al.,
Snf2l Regulates IPC Expansion via Foxg1
Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc. 875
2008; van Bokhoven and Kramer, 2010). Inthisstudy, weidentify
a genetic link between Foxg1 and the chromatin remodeling
progenitor self-renewal and differentiation. We predict that Snf2l
remodeling at the Foxg1 gene results in its transcriptional re-
pression and the subsequent de-repression of p21, thereby
promoting the timely exit and terminal neuronal differentiation
of progenitors that ultimately, controls brain size (Figure 4H).
role for ISWI in transcriptional repression often in association
with the Sin3 repressive complex (Burgio et al., 2008; Fazzio
et al., 2001). Other studies highlight a more global role for ISWI
in chromatin compaction (Corona et al., 2007; Li et al., 2010).
The Snf2l protein is part of two different complexes, CERF
(Banting et al., 2005) and NURF (Barak et al., 2003), but which
one regulates Foxg1 expression is unclear. The NURF complex
is largely considered an activator of gene expression, whereas
the function of the CERF complex remains unknown (Lazzaro
et al., 2006; Wysocka et al., 2006). Because the Ex6DEL cortical
phenotype is mild, the possibility exists that Snf2h-containing or
other chromatin remodeling complexes may be providing some
compensation. Indeed, future studies are required to delineate
Figure 4. Foxg1 Is Misregulated in Ex6DEL Mice
(A) qRT-PCR analysis of Foxg1, cdk inhibitors (Cdnk1a, 1b,1c, 2a), and cyclins (Ccnb2, Ccnd1) normalized to WT levels (n = 3; bars represent the mean ± SEM;
*p < 0.05 by t test).
(B) Foxg1 immunohistochemistry of E17.5 cortical sections. Scale bar represents 50 mm. Boxed region within SVZ is enlarged on right. Scale bar represents
(C) Schematic diagram of the Foxg1 locus showing the location of qPCR primer pairs R1-R11 used for ChIP analysis. P1 and P2 represent alternative promoters
driving expression of class 1 and 2 Foxg1 transcripts, respectively.
(D) ChIP analysis depicting the binding of Snf2l relative to % input across the Foxg1 locus. There was a significant level of Snf2l binding over IgG control at the R8
primer set (p < 0.012 by t test). Each bar represents the mean ± SD from three biological replicates (each replicate represents a pool of E15.5 dissected cortices
from 10 embryos) and two qPCR reactions from each replicate.
(E) Proportion of mitotic (PH3+) cells in WT or Ex6DEL mice with normal (Foxg1+/+) or reduced (Foxg1+/?) Foxg1 dosage (n = 3; bars represent the mean ± SEM;
*p < 0.05 by t test; PH3 stained images are shown in Figure S3G).
(F) E15.5 cortical sections from indicated genotype stained for Tbr2 (red) and BrdU (green) to identify cycling IPCs. Scale bar represents 50 mm.
(G) The proportion of Tbr2, BrdU-double labeled cells was rescued on the Foxg1+/?background (n = 3; bars represent the mean ± SEM; *p < 0.05 by t test).
(H) Model suggesting that Snf2l represses Foxg1 expression allowing for p21 derepression and the promotion of terminal differentiation.
Snf2l Regulates IPC Expansion via Foxg1
876 Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc.
the specific roles of the ISWI-containing complexes in neural
progenitor regulation. Nonetheless, an increased cortical cell
density is most likely compensated by a reduction in another
component of the tissue (e.g., elaboration of neuronal projec-
tions). Additional studies will be required to define such changes
although they may be related to corticofugal and sensory-motor
projections as reflected by altered CTIP2- and Foxp1-positive
cell numbers in the Ex6DEL animals.
Generation and Maintenance of Ex6DEL Mice
A 4.84 kb KpnI/XmnI genomic fragment encompassing exon 6 of the mSnf2l
gene was used to generate the targeting vector. A single loxP site was intro-
duced distal to exon 6 (Pst1-EcoR1) and a Neo cassette flanked by loxP sites
was inserted proximally (Xba1-Pst1) leaving short and long homologous arms
of 1.6 kb and 2.6 kb, respectively. The targeting construct was electroporated
into J1 ES cells, and G418-resistant clones were selected as described previ-
ously (Li et al., 1992). Homologous recombinants identified by Southern blot
with probes located within exon 2 and exon 6, were utilized for blastocyst
injection to generate exon 6 floxed mice. The exon 6 floxed line was bred to
Gata1-Cre mice (Mao et al., 1999) to generate the Ex6DEL allele that was
then maintained (independentof
Ex6DEL;Foxg1+/?mice were generated by crossing Ex6DEL mice to Foxg1-
Cre knock-in animals (He ´bert and McConnell, 2000). For timed mating
purposes, the day of vaginal plug detection was counted as embryonic day
0.5 (E0.5). All experiments wereapproved by the University of Ottawa’s Animal
Care ethics committee adhering to the guidelines of the Canadian Council on
Cre)ona 129Sv background.
BrdU Labeling Experiments
Bromodeoxyuridine (BrdU) was administered by intraperitoneal injections
(0.1 mg/g body weight) to time-mated females. For pulse labeling, pups
were harvested 2 hr after a single injection. The percentage of BrdU+cells to
total VZ cells were counted and normalized to WT levels for all embryonic
time points. For cell cycle re-entry experiments, embryos pulse-labeled with
BrdU at E12.5 or E16.5 were harvested 1 day later and stained for BrdU and
the cell cycle marker Ki67. Cell cycle re-entry was determined as the propor-
tion of cells that were BrdU+and Ki67+to the total number of BrdU+cells. For
neuronal birthdating experiments, BrdU was injected at E12.5 and offspring
harvested at E18.5 for processing. BrdU+cells were scored as densely labeled
if BrdU comprised >75% of the nucleus. To determine S-phase length, iodo-
deoxyuridine (IdU; 300 mg) was injected intraperitoneally into pregnant mares
at gestational days E13.5 and 15.5. This was followed, 1.5 hr (Ti) later, by
a similar injection of BrdU (300 mg). The female was euthanized and the
embryos removed 30 min after the second injection. These two separate
analogs were differentially detected by mouse anti-BrdU/IdU (1:100; BD
Biosciences) and rat anti-BrdU/CldU (1:200; Abcam) antibodies. Cells positive
for IdU but not BrdU have exited S phase (Lc) but double-labeled cells remain
in S phase (Sc). Length of S phase was calculated as: Ts= Ti3 Sc/Lcas
described elsewhere (Martynoga et al., 2005).
Coronal brain sections (10 mm) were mounted on SuperFrost Slides (Fisher
Scientific, ON) and frozen at ?80?C until use. Sections were washed five times
in PBST (PBS with 0.1% Triton X-100), blocked (1 hr, room temperature) in
10% horse serum/PBST, and incubated (overnight, 4?C) in primary antibodies.
The following primary antibodies were used: rat anti-BrdU (1:400); mouse anti-
SATB2 (1:10); rat anti-CTIP2 (1:500); rabbit anti-FoxP1(1:400); rabbit anti-Ki67
(1:250; NCL-Ki67-P, Novocastro); rabbit anti-Phosphohistone H3 antibody
(1:100; 06-570, Upstate); rabbit anti-TBR2 (1:250; ab23345, AbCam); rabbit
anti-Nurr1 (1:500, Santa Cruz, CA); mouse anti-b III tubulin (Tuj; 1:400 dilution;
01409, Stem Cell Technologies); and mouse anti-NeuN (1:500; Millipore). The
following day, sections were washed five times in PBST and incubated (2 hr,
RT) with DyLight488, DyLight594, or DyLight649-conjugated secondary anti-
bodies (1:1,000, Jackson ImmunoResearch, PA). All sections were counter-
stained with the nuclear marker DAPI (Invitrogen). Coverslips were mounted
with Dako Fluorescence Mounting Medium (Dako Canada, ON).
Image Acquisition and Processing
For IF experiments, tissuesectionswereexamined and images captured using
a Zeiss 510 laser scanning confocal microscope with UV (405 nm), argon
(488 nm), helium/neon (546 nm), and helium/neon (633 nm) lasers. All images
were acquired as 8 mm Z stacks (in 2-mm intervals) and analyzed as projections
using the LSM 510 Image Browser software (Zeiss). For cell density analysis,
a 40-mm section was used to acquire 24 mm Z stacks (in 2-mm intervals) that
were used for 3D reconstruction and cell counts.For counting of marker+cells,
and inadjacent ‘‘noprimary’’controlsandthemeanvalue wassubtracted from
the tissue under examination, thus generating a normalized immunoreactive
positive intensity value that was used as the baseline to score a positive cell.
The ChIP assay was performed as described in the protocol from the Millipore
ChIPAssayKit(product17-295) withsomemodifications.Cerebral cortex was
dissected from E15.5 Snf2lf/fmice. The tissue was mechanically dissociated
and crosslinked in 1% paraformaldehyde for 1 hr on ice. Tissue from approx.
10 embryos (4 3 107cells) was used for each experiment. Cells were lysed
(50 mM Tris, 1 mM EDTA, 1% SDS, Roche Complete Mini protease inhibitor
cocktail) on ice and then sonicated to generate fragments of 200–500 bp in
length. Sepharose G slurry (25 mL; GE Biosciences) with 1 mg of either Sheep
anti-Snf2l or Sheep IgG (Sigma) was used for IP. Quantitative PCR analysis
was run on a MX3000P instrument (Stratagene) using Absolute QPCR SYBR
Green Mix (Thermo Scientific). The cycling conditions were: one cycle at
95?C for 10 min and 40 cycles of 95?C for 30 s, 60?C for 30 s, 72?C for 30 s.
Percent input of target DNA in the IP samples was calculated off a curve
derived from serial dilutions of input chromatin.
For all data sets a minimum of three biological replicates (mice, embryos, or
brains) were analyzed (n = 3). For cell counts, the mean cell number was deter-
most instances the data was normalized to WT. Unless indicated otherwise,
histograms represent the mean ± the standard error of the mean (SEM). An
asterisk (*) represents a statistically significant change by a two-tailed Student
t test (p % 0.05) as comparedto the WT sample, unless indicated differently by
the use of brackets.
The microarray data are available in the Gene Expression Omnibus (GEO) data-
Supplemental Information includes four figures and two tables and can be
found with this article online at doi:10.1016/j.devcel.2012.01.020.
We would like to thank K. Yan and J. Coulombe for technical assistance and
Drs. R. Kothary, L. Megeney, and V. Wallace for helpful discussions and
comments on the manuscript. The HSFCSR provided partial financial support
toward the purchase of the confocal microscope used in thisstudy.D.J.Y. was
funded by an OGSST award. C.P.C. was an Ontario Graduate Scholarship
recipient. D.J.P. was funded by CIHR operating grants.
Received: February 15, 2011
Revised: January 11, 2012
Accepted: January 26, 2012
Published online: April 16, 2012
Snf2l Regulates IPC Expansion via Foxg1
Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc. 877
Banting, G.S., Barak, O., Ames, T.M., Burnham, A.C., Kardel, M.D., Cooch,
N.S., Davidson, C.E., Godbout, R., McDermid, H.E., and Shiekhattar, R.
(2005). CECR2, a protein involved in neurulation, forms a novel chromatin
remodeling complex with SNF2L. Hum. Mol. Genet. 14, 513–524.
Barak, O., Lazzaro, M.A., Lane, W.S., Speicher, D.W., Picketts, D.J., and
Shiekhattar, R. (2003). Isolation of human NURF: a regulator of Engrailed
gene expression. EMBO J. 22, 6089–6100.
Be ´rube ´, N.G., Mangelsdorf, M., Jagla, M., Vanderluit, J., Garrick, D., Gibbons,
R.J., Higgs, D.R., Slack, R.S., and Picketts, D.J. (2005). The chromatin-
remodeling protein ATRX is critical for neuronal survival during corticogenesis.
J. Clin. Invest. 115, 258–267.
Brancaccio, M., Pivetta, C., Granzotto, M., Filippis, C., and Mallamaci, A.
(2010). Emx2 and Foxg1 inhibit gliogenesis and promote neuronogenesis.
Stem Cells 28, 1206–1218.
Burgio, G., La Rocca, G., Sala, A., Arancio, W., Di Gesu `, D., Collesano, M.,
Sperling, A.S., Armstrong, J.A., van Heeringen, S.J., Logie, C., et al. (2008).
Genetic identification of a network of factors that functionally interact with
the nucleosome remodeling ATPase ISWI. PLoS Genet. 4, e1000089.
Corona, D.F., Siriaco, G., Armstrong, J.A., Snarskaya, N., McClymont, S.A.,
Scott, M.P., and Tamkun, J.W. (2007). ISWI regulates higher-order chromatin
structure and histone H1 assembly in vivo. PLoS Biol. 5, e232.
Dirscherl, S.S., and Krebs, J.E. (2004). Functional diversity of ISWI complexes.
Biochem. Cell Biol. 82, 482–489.
Dou, C.L., Li, S., and Lai, E. (1999). Dual role of brain factor-1 in regulating
growth and patterning of the cerebral hemispheres. Cereb. Cortex 9, 543–550.
Fasano, C.A., Phoenix, T.N., Kokovay, E., Lowry, N., Elkabetz, Y., Dimos, J.T.,
Lemischka, I.R., Studer, L., and Temple, S. (2009). Bmi-1 cooperates with
Foxg1 to maintain neural stem cell self-renewal in the forebrain. Genes Dev.
Fazzio, T.G., Kooperberg, C., Goldmark, J.P., Neal, C., Basom, R., Delrow, J.,
and Tsukiyama, T. (2001). Widespread collaboration of Isw2 and Sin3-Rpd3
chromatin remodeling complexes in transcriptional repression. Mol. Cell.
Biol. 21, 6450–6460.
Fukumitsu, H., Ohtsuka, M., Murai, R., Nakamura, H., Itoh, K., and Furukawa,
S. (2006). Brain-derived neurotrophic factor participates in determination of
neuronal laminar fate in the developing mouse cerebral cortex. J. Neurosci.
Go ¨tz, M., and Huttner, W.B. (2005). The cell biology of neurogenesis. Nat. Rev.
Mol. Cell Biol. 6, 777–788.
Guillemot, F. (2007). Spatial and temporal specification of neural fates by
transcription factor codes. Development 134, 3771–3780.
Gupta, A.,Tsai, L.H., and Wynshaw-Boris,A. (2002). Life is ajourney:a genetic
look at neocortical development. Nat. Rev. Genet. 3, 342–355.
Hamby, M.E., Coskun, V., and Sun, Y.E. (2008). Transcriptional regulation of
neuronal differentiation: the epigenetic layer of complexity. Biochim.
Biophys. Acta 1779, 432–437.
Hanashima, C., Li, S.C., Shen, L., Lai, E., and Fishell, G. (2004). Foxg1
suppresses early cortical cell fate. Science 303, 56–59.
He ´bert, J.M., and McConnell, S.K. (2000). Targeting of cre to the Foxg1 (BF-1)
locus mediates loxP recombination in the telencephalon and other developing
head structures. Dev. Biol. 222, 296–306.
Lazzaro, M.A., Pe ´pin, D., Pescador, N., Murphy, B.D., Vanderhyden, B.C., and
Picketts, D.J. (2006). The imitation switch protein SNF2L regulates steroido-
genic acute regulatory protein expression during terminal differentiation of
ovarian granulosa cells. Mol. Endocrinol. 20, 2406–2417.
Lazzaro, M.A., and Picketts, D.J. (2001). Cloning and characterization of the
murine Imitation Switch (ISWI) genes: differential expression patterns suggest
distinct developmental roles for Snf2h and Snf2l. J. Neurochem. 77, 1145–
Lessard, J., Wu, J.I., Ranish, J.A., Wan, M., Winslow, M.M., Staahl, B.T., Wu,
H., Aebersold, R., Graef, I.A., and Crabtree, G.R. (2007). An essential switch
in subunit composition of a chromatin remodeling complex during neural
development. Neuron 55, 201–215.
Li, E., Bestor, T.H., and Jaenisch, R. (1992). Targeted mutation of the DNA
methyltransferase gene results in embryonic lethality. Cell 69, 915–926.
of the gene for transcription factor BF-1, an essential regulator of forebrain
development. Brain Res. Mol. Brain Res. 37, 96–104.
Li, M., Belozerov, V.E., and Cai, H.N. (2010). Modulation of chromatin
boundary activities by nucleosome-remodeling activities in Drosophila mela-
nogaster. Mol. Cell. Biol. 30, 1067–1076.
Lim, D.A., Huang, Y.C., Swigut, T., Mirick, A.L., Garcia-Verdugo, J.M.,
Wysocka, J., Ernst, P., and Alvarez-Buylla, A. (2009). Chromatin remodelling
factor Mll1 is essential for neurogenesis from postnatal neural stem cells.
Nature 458, 529–533.
toring Cre recombinase-mediated DNA excisions in mice. Proc. Natl. Acad.
Sci. USA 96, 5037–5042.
Martynoga, B., Morrison, H., Price, D.J., and Mason, J.O. (2005). Foxg1 is
required for specification of ventral telencephalon and region-specific regula-
tion of dorsal telencephalic precursor proliferation and apoptosis. Dev. Biol.
McKenna, W.L., Betancourt, J., Larkin, K.A., Abrams, B., Guo, C., Rubenstein,
J.L., and Chen, B. (2011). Tbr1 and Fezf2 regulate alternate corticofugal
neuronal identities during neocortical development. J. Neurosci. 31, 549–564.
Molofsky, A.V., He, S., Bydon, M., Morrison, S.J., and Pardal, R. (2005). Bmi-1
growth and survival by repressing the p16Ink4a and p19Arf senescence path-
ways. Genes Dev. 19, 1432–1437.
Molofsky, A.V., Pardal, R., Iwashita, T., Park, I.K., Clarke, M.F., and Morrison,
S.J. (2003). Bmi-1 dependence distinguishes neural stem cell self-renewal
from progenitor proliferation. Nature 425, 962–967.
Muzio, L., and Mallamaci, A. (2005). Foxg1 confines Cajal-Retzius neurono-
genesis and hippocampal morphogenesis to the dorsomedial pallium.
J. Neurosci. 25, 4435–4441.
Shen, Q., Wang, Y., Dimos, J.T., Fasano, C.A., Phoenix, T.N., Lemischka, I.R.,
Ivanova, N.B., Stifani, S., Morrisey, E.E., and Temple, S. (2006). The timing of
cortical neurogenesis is encoded within lineages of individual progenitor cells.
Nat. Neurosci. 9, 743–751.
Siegenthaler, J.A., and Miller, M.W. (2005). Transforming growth factor beta 1
promotes cell cycle exit through the cyclin-dependent kinase inhibitor p21 in
the developing cerebral cortex. J. Neurosci. 25, 8627–8636.
Siegenthaler, J.A., Tremper-Wells, B.A., and Miller, M.W. (2008). Foxg1
haploinsufficiency reduces the population of cortical intermediate progenitor
cells: effect of increased p21 expression. Cereb. Cortex 18, 1865–1875.
Stopka, T., and Skoultchi, A.I. (2003). The ISWI ATPase Snf2h is required for
early mouse development. Proc. Natl. Acad. Sci. USA 100, 14097–14102.
Su ¨rmeli, G., Akay, T., Ippolito, G.C., Tucker, P.W., and Jessell, T.M. (2011).
Patterns of spinal sensory-motor connectivity prescribed by a dorsoventral
positional template. Cell 147, 653–665.
van Bokhoven, H., and Kramer, J.M. (2010). Disruption of the epigenetic code:
an emerging mechanism in mental retardation. Neurobiol. Dis. 39, 3–12.
Wysocka, J., Swigut, T., Xiao, H., Milne, T.A., Kwon, S.Y., Landry, J., Kauer,
M., Tackett, A.J., Chait, B.T., Badenhorst, P., et al. (2006). A PHD finger of
NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling.
Nature 442, 86–90.
Xuan, S., Baptista, C.A., Balas, G., Tao, W., Soares, V.C., and Lai, E. (1995).
Winged helix transcription factor BF-1 is essential for the development of
the cerebral hemispheres. Neuron 14, 1141–1152.
Snf2l Regulates IPC Expansion via Foxg1
878 Developmental Cell 22, 871–878, April 17, 2012 ª2012 Elsevier Inc.