Article

In vivo transcriptional profile analysis reveals RNA splicing and chromatin remodeling as prominent processes for adult neurogenesis

The Rockefeller University, New York, New York, United States
Molecular and Cellular Neuroscience (Impact Factor: 3.84). 02/2006; 31(1):131-48. DOI: 10.1016/j.mcn.2005.10.005
Source: PubMed

ABSTRACT

Neural stem cells and neurogenesis persist in the adult mammalian brain subventricular zone (SVZ). Cells born in the rodent SVZ migrate to the olfactory bulb (Ob) where they differentiate into interneurons. To determine the gene expression and functional profile of SVZ neurogenesis, we performed three complementary sets of transcriptional analysis experiments using Affymetrix GeneChips: (1) comparison of adult mouse SVZ and Ob gene expression profiles with those of the striatum, cerebral cortex, and hippocampus; (2) profiling of SVZ stem cells and ependyma isolated by fluorescent-activated cell sorting (FACS); and (3) analysis of gene expression changes during in vivo SVZ regeneration after anti-mitotic treatment. Gene Ontology (GO) analysis of data from these three separate approaches showed that in adult SVZ neurogenesis, RNA splicing and chromatin remodeling are biological processes as statistically significant as cell proliferation, transcription, and neurogenesis. In non-neurogenic brain regions, RNA splicing and chromatin remodeling were not prominent processes. Fourteen mRNA splicing factors including Sf3b1, Sfrs2, Lsm4, and Khdrbs1/Sam68 were detected along with 9 chromatin remodeling genes including Mll, Bmi1, Smarcad1, Baf53a, and Hat1. We validated the transcriptional profile data with Northern blot analysis and in situ hybridization. The data greatly expand the catalogue of cell cycle components, transcription factors, and migration genes for adult SVZ neurogenesis and reveal RNA splicing and chromatin remodeling as prominent biological processes for these germinal cells.

Full-text

Available from: Benedicte Menn
In vivo transcriptional profile analysis reveals RNA splicing and
chromatin remodeling as prominent processes for adult neurogenesis
Daniel A. Lim,
a,
*
Mayte Sua´rez-Farin˜as,
b
Felix Naef,
b
Coleen R. Hacker,
c
Benedicte Menn,
a
Hirohide Takebayashi,
a
Marcelo Magnasco,
b
Nila Patil,
c
and Arturo Alvarez-Buylla
a,
*
a
Department of Neurological Surgery and Developmental and Stem Cell Biology Program, University of California,
San Francisco, CA 94143, USA
b
The Rockefeller University, 1230 York Ave., New York, NY 10021, USA
c
Perlegen Sciences Inc., 3380 Central Expressway, Santa Clara, CA 95051, USA
Received 10 May 2005; revised 21 August 2005; accepted 4 October 2005
Available online 5 December 2005
Neural stem cells and neurogenesis persist in the adult mammalian
brain subventricular zone (SVZ). Cells born in the rodent SVZ
migrate to the olfactory bulb (Ob) where they differentiate into
interneurons. To determine the gene expression and functional
profile of SVZ neurogenesis, we performed three complementary
sets of transcriptional analysis experiments using Affymetrix
GeneChips: (1) comparison of adult mouse SVZ and Ob gene
expression profiles with those of the striatum, cerebral cortex, and
hippocampus; (2) profiling of SVZ stem cells and ependyma isolated
by fluorescent-activated cell sorting (FACS); and (3) analysis of gene
expression changes during in vivo SVZ regeneration after anti-
mitotic treatment. Gene Ontology (GO) analysis of data from these
three separate approaches showed that in adult SVZ neurogenesis,
RNA splicing and chromatin remodeling are biological processes as
statistically significant as cell proliferation, transcription, and
neurogenesis. In non-neurogenic brain regions, RNA splicing and
chromatin remodeling were not prominent processes. Fourteen
mRNA splicing factors including Sf3b1, Sfrs2, Lsm4, and Khdrbs1/
Sam68 were detected along with 9 chromatin remodeling genes
including Mll, Bmi1, Smarcad1, Baf53a, and Hat1. We validated the
transcriptional profile data with Northern blot analysis and in situ
hybridization. The data greatly expand the catalogue of cell cycle
components, transcription factors, and migration genes for adult
SVZ neurogenesis and reveal RNA splicing and chromatin remodel-
ing as prominent biological processes for these germinal cells.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Subventricular zone (SVZ); Neurogenesis; Stem cell; Adult
brain; Microarray; Transcription; Transcriptional profile; Chromatin
remodeling; RNA splicing
Introduction
The adult brain harbors neurogenic stem cells within the
subventricular zone (SVZ) of the lateral ventricle wall (Garcia-
Verdugo et al., 1998; Gage, 2000). In neonatal (Luskin, 1993) and
adult mice (Lois and Alvarez-Buylla, 1994; Doetsch and Alvarez-
Buylla, 1996; Jankovski and Sotelo, 1996; Thomas et al., 1996),
cells born in the SVZ migrate a long distance to the olfactory bulb
(Ob) where they differentiate into interneurons. SVZ astrocytes
(type B cells) are neural stem cells (Doetsch et al., 1999a,b; Laywell
et al., 2000) and give rise to rapidly dividing, immature-appearing
cells (type C cells) that generate migratory neuroblasts (type A
cells) (Lois and Alvarez-Buylla, 1994; Peretto et al., 1997; Luskin,
1998; Doetsch et al., 1999a,b). See Figs. 1B, C. SVZ ependymal
cells are themselves not neurogenic (Chiasson et al., 1999; Laywell
et al., 2000; Capela and Temple, 2002) but may be important for
generating the SVZ neurogenic niche (Lim et al., 2000; Goldman,
2003; Peretto et al., 2004). Although the SVZ cellular architecture
(Gates et al., 1995; Jankovski and Sotelo, 1996; Doetsch et al.,
1997; Peretto et al., 1997), stem cell identity (Chiasson et al., 1999;
Doetsch et al., 1999a,b; Laywell et al., 2000; Rietze et al., 2001;
Capela and Temple, 2002; Imura et al., 2003), and neurogenic
lineage (Doetsch et al., 1999a,b) have been defined, the genetic
program for adult SVZ neurogenesis is poorly understood.
The transcriptional changes of differentiating neocortical (East-
erday et al., 2003; Karsten et al., 2003) and postnatal SVZ-derived
1044-7431/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.mcn.2005.10.005
Abbreviations: SVZ, subventricular zone; Ob, olfactory bulb; ObC,
olfactory bulb core; Ctx, cortex; St, striatum; Hp, hippocampus; ds cDNA,
double-stranded complementary DNA; cRNA, complementary (antisense)
RNA; FACS, fluorescent-activated cell sorting; ECM, extracellular matrix;
RMS, rostral migratory stream; EGF, epidermal growth factor; FGF,
fibroblast growth factor.
* Corresponding authors.
E-mail addresses: limd@neurosurg.ucsf.edu (D.A. Lim),
abuylla@itsa.ucsf.edu (A. Alvarez-Buylla).
Available online on ScienceDirect (www.sciencedirect.com).
www.elsevier.com/locate/ymcne
Mol. Cell. Neurosci. 31 (2006) 131 148
Page 1
(Gurok et al., 2004) neurospheres have also been studied in vitro.
Many neurospheres are derived from transit-amplifying cells (type
C cells) (Doetsch et al., 2002), and their exposure to growth factors
(EGF or FGF) in vitro deregulates the normal genetic control of
cell differentiation that occurs in vivo (Gabay et al., 2003; Santa-
Olalla et al., 2003; Hack et al., 2004); it is likely that the expression
profiles of neurospheres and endogenous SVZ precursors differ.
Furthermore, in vivo, SVZ neurogenesis involves a long-distance,
directional migration to the Ob while neurospheres do not appear
to have a similar migration capacity. Therefore, transcriptional
analysis of in vivo SVZ neurogenesis is required to identify genes
and biological processes involved in this continual generation of
neurons for the Ob.
Using high-density oligonucleotide (GeneChip) arrays, we
undertook three c omplementary approaches to determine the
transcriptional profile of in vivo SVZ neurogenesis. We first
compared gene expression differences of the SVZ-Ob system with
that of three other brain regions. We then utilized FACS methods to
compare the transcriptional profiles of type B cells the
neurogenic stem cell and the non-neurogenic ependyma. Finally,
we analyzed the transcriptional changes of the SVZ as it
regenerated type C and A cells from a population of type B cells.
Data integrated from these three approaches identified genes,
signaling pathways, and biological processes related to SVZ
neurogenesis. In addition to expanding the catalogue of cell cycle
components, transcription factors, and genes for migration, we
identified RNA splicing and chromatin remodeling as prominent
processes for adult neurogenesis. The importance of RNA splicing
and chromatin remodeling has not been described for SVZ neuro-
genesis, and we here provide evidence that these processes are as
upregulated as the expected processes of cell cycle, transcription,
and neurogenesis. We focused our Results and Discussion below
only on a subset of genes with special attention to RNA splicing and
chromatin remodeling, however, both the raw chip image data and
other data analyses are available (Supplementary data and at http://
asterion.rockefeller.edu/mayte/Neurogenesis) for future compara-
tive expression profile analyses with other developmental, adult, or
tumor cell populations.
Results
Brain region transcriptional profile analysis identified genes with
increased expression in the SVZ-ObC neurogenic system
We analyzed the transcriptional profiles of the SVZ, Ob core
(ObC), and three other brain regions indicated in Fig. 1A. The ObC
dissection excluded the mitral and periglomerular layers, providing a
RNA sample primarily representing migratory type A cel ls,
maturing neuroblasts, and mature granule cells. The hippocampus
(Hp) dissection included the non-neurogenic CA1 CA3 regions as
well the dentate gyrus. The striatum (St) was the region directly
underlying the SVZ dissection. The cortex (Ctx) did not include the
corpus callosum. Biotin-labeled complementary RNAs (cRNAs)
derived from each brain region were analyzed on GeneChip Mu11k
expression arrays, which contain more than 13,000 probe sets
analyzing the expression of over 11,000 unique genes. Each brain
region was analyzed independently twice; the data among the
duplicates were consistent (Supplementary data S1).
To focus our analysis on those genes more likely to be involved
in SVZ neurogenesis, we filtered the data (see Experimental
methods) for those genes that are (1) increased in the SVZ—the
SVZ profile, (2) increased in the ObC—the ObC profile, and (3)
increased in both the SVZ and ObC—the SO profile (Fig. 1D). The
SVZ profile (Supplementary data S2) contained 65 unique genes
(71 probe sets) with increased expression in the SVZ as compared
to all other regions (ObC, Hp, Ctx, St). The ObC profile
(Supplementary data S3) included 168 genes (209 probe sets),
and the SO profile (Supplementary data S4) contained 60 genes (80
probe sets). Genes in the SVZ, SO, ObC profiles are shown
Fig. 1. (A) Brain regions dissected for the brain region transcriptional
analysis. Dissected areas are shown in yellow. The SVZ contains three
populations of neurogenic precursors—type B, C, and A cells. (B) The
lineage of SVZ neurogenesis. Type B cells (blue) are SVZ astrocytes that self-
renew and give rise to a rapidly dividing population of immature-appearing
cells—type C cells (green). The transit-amplifying type C cells then become
type A cells (red), the neuroblasts that migrate into the ObC. (C) Architecture
of the SVZ. The ventricle is to the left. Ciliated ependymal cells (gray) line
the ventricle wall. Some type B cells (blue) make contact with the ventricle
lumen (arrow). Both type C (green) and A cells (red) are in direct contact with
the type B cells. In this panel, type A cells are migrating toward the ObC in a
direction perpendicular to the page. (D) Sagittal view of the mouse brain.
Within the SVZ, there is an extensive network of type A cells migrating
tangentially toward the ObC. This network of pathways coalesces at the
anterior of the SVZ to form the rostral migratory stream (curved arrow). The
rostral migratory stream enters the ObC where type A cells then migrate
radially and disperse (red dots) throughout the Ob. The major biological
processes that occur in the SVZ alone (SVZ profile), SVZ and ObC (SO
profile), and ObC alone (ObC profile) are listed to the left, middle, and right,
respectively; the cell types of the SVZ and ObC profiles are in bold.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148132
Page 2
clustered in a color matrix in Fig. 3A. Genes that had decreased
expression in the SVZ, SO, and ObC can be found in
Supplementary data S12.
To assess the sensitivity of the SVZ, ObC, and SO profiles, we
surveyed the literature to identify those genes that would be
expected to be detected in our analysis. Of the genes represented
on the Mu11K GeneChip set, we identified 15 that are highly
expressed in the SVZ and/or ObC relative to the other brain regions
(Hp, Ctx, St). Of these 15 genes, our analysis detected 11 (73%)
with a profile matching the published in situ hybridization or
immunohistochemical data (Table 1). Six other genes previously
described to be expressed highly in the SVZ and/or ObC were not
represented on the Mu11K GeneChip set.
To validate the array data with another measure of transcript
levels, we analyzed 8 genes by Northern blot: Ccnd2, Hmgb2,
Mia, Pdyn, Dlx1, 2310021G01Rik, Sox11, and Col6a1. For all of
the genes tested, the Northern blot data paralleled the pattern of
expression observed on GeneChip analysis (Fig. 2).
Gene Ontology analysis identifies RNA splicing and chromatin
regulation as prominent biological events in the SVZ and ObC
brain regions
To translate the gene expression data into functional profiles, we
used Gene Ontology (GO) analysis. GO provides an organized
vocabulary of terms that can be used to describe a gene product’s
attributes (www.geneontology.org). GO terms are organized into
three categories (biological process, cellular component, and
molecular function) in structures called directed acyclic graphs;
these structures differ from hierarchies in that a Fchild_ (more
specialized term) can have several Fparents_ (less specialized term).
To analyze the GO terms of the SVZ, SO, and ObC profiles, we
used Onto-Express (Khatri et al., 2002, 2004). For each GO term,
Onto-Express computes its significance ( P value), allowing one to
distinguish prominent biological processes from non-significant
events. A complete list of GO terms for the SVZ, SO, and ObC
profiles with associated P values is in Supplementary data S5, and
the parent child relationship of the GO terms can be browsed with
Onto-Express (see Experimental methods). The functional profiles
of SVZ, SO, and ObC gene expression are shown in the pie charts of
Figs. 3B–D.
The SVZ is the primary site where type B and C cells are
maintained and proliferate. Compared to the other brain regions in
our analysis, the SVZ is the most proliferative. As expected, the
biological processes of proliferation and cell cycle were prominent
in the SVZ profile (Fig. 3B). From the SVZ, type A cells tangentially
migrate into the ObC where they then turn to migrate radially into the
granule cell layer. Within the granule cell layer, the type A cells
undergo terminal differentiation and integrate into local circuits (Fig.
1D). There is also a continual turnover of young neurons in the Ob
involving apoptosis (Najbauer and Leon, 1995; Fiske and Brunjes,
2001; Pet reanu and Alvarez-Buylla, 2002). The ObC profile
therefore should reflect these later stages of SVZ-Ob neurogenesis
as well as granule cell turnover. Indeed, significant biological
processes were development, neurogenesis, and cell differentiation;
other highly significant GO terms included CNS and brain
development, negative regulation of cell proliferation, axon o-
genesis, and apoptosis/programmed cell death (Fig. 3D). Therefore,
GO analysis described many of the major known and expected
biological processes that occur in the SVZ and ObC regions.
The SO profile represents gene expression common to both the
SVZ and ObC. As expected, SO profile terms related to cell growth,
transcription, protein metabolism, and development (Fig. 3C). The
most prominent biological process in the SO profile, however, was
RNA splicing (Fig. 3C); terms related to RNA splicing appeared 9
times in this analysis (in Supplementary data S5), all with very high
significance ( P values are in the figure). GO terms related to
chromatin regulation terms appeared 7 times, including terms from
all three GO categories (Fig. 3C, and in Supplementary data S5). The
SVZ profile also was significant for Festablishment and/or mainte-
nance of chromatin architecture _ as well as components of chromatin
Table 1
Correlation between published expression data and GeneChip brain region
profile analysis
Published expressionGene Profile
SVZ ObC Reference
Dlx1 SO + + Lois, 1996
Dlx2 SO + + Doetsch et al., 2002
Sox2 SO + + Ferri et al., 2004
Pbx1 ObC + ++ Redmond et al., 1996
Mash1 –++Parras et al., 2004
Er81/Etv1 –++Stenman et al., 2003
Vim SVZ + Doetsch et al., 1997
Mki67 SVZ + Zhu et al., 2003
Rrm1 SVZ + Zhu et al., 2003
Notch1 –+?Stump et al., 2002
Wnt5a ObC + Shimogori et al., 2004
Thra SVZ + Lemkine et al., 2005
Nog ObC + ++ Peretto et al., 2004
Nestin –+ Doetsch et al., 1997
Cd24 SO + + Calaora et al., 1996
Genes expressed in SVZ and/or ObC but not on Mu11K chip Olig2, Emx2,
Slit, Ng2, Dcx, Gli1.
Fig. 2. Northern hybridizations substantiate array data. Northern blot
analysis was performed on the cRNA samples (left). Corresponding
GeneChip array data for the brain region analysis (duplicate data shown,
indicated as 1 or 2 under the brackets) is shown as a color matrix (right,
red—increased expression, green—decreased expression). Eight out of
8 genes tested by Northern blot had good agreement with the array data.
Fold change scale (log
2
) for the color matrix is shown at the bottom right.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 133
Page 3
Fig. 3. The brain region transcriptional profiles. (A) Color matrix of the SVZ, SO, and ObC profiles. Genes are ordered along the vertical axis using hierarchical
clustering. Duplicate profiles of the brain regions are presented on the horizontal axis. The color and color intensity of each cell in the matrix relate to the
expression ratio of each gene. Red indicates a positive ratio (expression greater than the mean of the other brain regions), green indicates a negative ratio, and
black indicates a ratio of 1. A color scale (log
2
) indicating the magnitude of the expression ratios is shown in the bottom. (B D) GO analysis pie charts for the
brain region profiles. The entire pie represents all GO terms in the analysis. Pie slices are proportional to the number of genes (in parentheses) related to a
particular GO Fparent_ term (legend for color code is in the inset to the right of each panel). GO terms that are Fchildren_ of a parent term are listed next to the
pie chart with an indicating line. Further parent child relationship of the GO tree structure is indicated by indentation with hyphen. All listed GO terms are
statistically significant, and color of the type indicates the GO category (see legend at the lower right of the figure).
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148134
Page 4
and nucleosomes (Fig. 3B). Thus, the data suggest that both RNA
splicing and chromatin regulation are important biological processes
for SVZ neurogenesis.
To determine the relative prominence of RNA splicing and
chromatin remodeling for SVZ neurogenesis in comparison to non-
neurogenic brain regions, we performed GO analysis on the sets of
genes that were increased in the Ctx (Ctx profile), St (St profile), and
Hp (Hp profile) (probe set lists in Supplementary data S11, GO term
lists in S5). No terms related to RNA splicing were statistically
significant in the Ctx, St, or Hp profiles. In the Ctx profile, the term
Fchromatin remodeling_ was associated with 2 genes and a P value
of 0.02; however, the parent term of Festablishment and/or
maintenance of chromatin architecture_ was not statistically
significant ( P = 0.37). No GO t erms r elated to chromatin
remodeling were significant in the St or Hp profiles. Thus, RNA
splicing and chromatin remodeling were much more prominent in
the SVZ and SO profiles than in the Ctx, St, and Hp.
In the Supplementary text, we identify and discuss the genes
detected in our SVZ-Ob analysis related to cell cycle, transcription,
migration, and apoptosis. The majority of those genes has not been
previously described for adult SVZ-Ob neurogenesis, and thus, the
data present a wealth of gene candidates for future study. In this
manuscript, we focus on RNA splicing and chromatin remodeling
because they are important biological processes but not well
described for the adult SVZ and Ob.
Using the GO analysis and a review of the literature, we
identified genes related to RNA splicing and chromatin remodel-
ing in the SVZ, SO, and ObC expression profiles. The SO profile
contained RNA splicing factors Sf3b1, Sfrs2, Lsm4, Snrpg,
Snrpd2 , Hnrpa2b1, Hnrpd, Hnrpm, Hnrpdl, Hnrph1,and
Khdrbs1/Sam68, and the ObC profile contained Snrpb (Table
2). Chromatin-remodeling genes Mll, Hat1, Hmgb3, and Baf53a
were detected in the SO profile, Hmgb2 and H2afx were in the
SVZ profile, and the ObC profile contained Bmi1 and Smarcad1
(Table 2).
Gene expression comparison of the type B SVZ stem cell and the
non-neurogenic ependyma reveals chromatin regulation as a
prominent process in type B cells
Neurogenic SVZ cells are closely associated with the non-
proliferative ependymal cells that line the walls of the lateral
ventricle (see Fig. 1C). The SVZ and SO profiles therefore
contained the gene expression of non-neurogenic ependyma. We
used fluorescent-activated cell sorting (FACS) to separate the type
B cells and ependyma and compared their gene expression profiles.
Table 2
Chromatin-remodeling and RNA splicing genes in the brain region profiles
Highlighted cells indicate the profile to which each probe set/gene belongs (e.g., Baf53a has cells in both the SVZ and ObC columns highlighted, indicating the
SO profile).
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 135
Page 5
To isolate type B cells, we used antibodies to GFAP (Doetsch et al.,
1999a,b). Immunocytochemistry for this i ntracellular antigen
requires permeabilization of the cell membrane. We developed
methods to isolate RNA from cells permeabilized by a non-ionic
detergent (Tween-20) and confirmed that the RNAs are stable
through the immunostaining protocol (Figs. 4H, I, K). GFAP+ cells
were generally round or elliptical and not ciliated (Figs. 4C, D).
We used CD24 antibodies to purify ependymal cells (Capela and
Temple, 2002). CD24 staining was also performed with Tween-20
so that any changes in the gene expression profile associated with
this agent would be comparable to those observed in the GFAP+
population. To a lesser degree, CD24 antibodies also stain SVZ Type
A cells (Calaora et al., 1996); however, our dissociation protocol and
Tween-20 treatment eliminated the CD24 epitope from the surface
of type A cells. CD24 antibody staining strongly labeled multi-
ciliated ependymal cells (Figs. 4A, B); CD24+ non-ciliated cells
were not observed.
SVZ cells immunostained for CD24 and GFAP were sorted by
FACS (Figs. 4E, F). Total RNA from type B and ependymal cell
populations was isolated, and mRNAs were amplified as schema-
tized in Fig. 4G and described in Experimental methods. The
amplification procedure preserved the appropriate mRNA size
distribution as well as differential expression of GFAP and CD24
(Figs. 4I, J). The cRNAs produced for GeneChip analysis were
also of an appropriate size distribution, and GAPDH Northern blot
analysis shows a single band of expected size, indicating that the
amplification procedure did not produce degraded transcripts (Fig.
4K). Scatter plots comparing expression profiles of duplicate
samples show good reproducibility (see Supplementary data S6).
Differential expression of 1324 probe sets (1282 unique genes)
was detected between GFAP+ and CD24+ cells. 54% of the genes
had increased expression in GFAP+ cells, and 46% were increased
in the CD24+ cells. To confirm the FACS cell separation and
cDNA amplification, we examined the data for expected differen-
tial gene expression. Cd24 itself was strongly increased (146-fold)
in the CD24+ population, paralleling the RT-PCR result of Fig. 4J.
In the SVZ, Sox2 is expressed highest in the ependyma (Ferri et al.,
2004), and the FACS data reported Sox2 expression as 3.8-fold
higher in the ependymal cells relative to the type B cells. Spa17 is
a component of cilia (Grizzi et al., 2004), and it was expressed 11-
fold higher in the ciliated CD24+ ependymal cells. The probe set
for Gfap did not show differential expression, however, the Gfap
mRNA was differentially represented in the representative cDNA
libraries as shown by RT-PCR (Fig. 4J). A small fraction of the
probe sets on the Mu11K arrays assess transcript levels poorly (N.
Patil, personal communication), and it is possible that the probe set
for Gfap is problematic. NOG (Noggin) protein has been
previously shown to be highly expressed in ependymal cells
(Lim et al., 2000; Peretto et al., 2004) and SVZ astrocytes (Peretto
et al., 2004), however, we did not find elevated expression of Nog
in either the SVZ profile or CD24+ cells. There may be a mismatch
between transcription and translation for the Nog gene, resulting in
a pattern of low mRNA transcript levels but high Noggin protein
concentrations in the SVZ and ependymal cells. It is also possible
that differential expression for any gene is not detected due to a
loss of transcript during FACS or cDNA amplification.
GO analysis showed that type B cells are significant for cell
proliferation and cell cycle, while ependymal cells are significant for
cell cycle arrest (Table 3). These data are consistent with the finding
that ependymal cells do not divide in vivo (Doetsch et al., 1999a,b;
Capela and Temple, 2002; Spassky et al., 2005). The process of
neurogenesis was also significant in type B cells and not in
ependyma, supporting the data that ependyma are non-neurogenic
(Chiasson et al., 1999; Capela and Temple, 2002). Like the SVZ and
SO profiles, establishment and/or m aintenance of chromatin
architecture was prominent in type B cells along with histone
acetyltransferase activity. FmRNA metabolism,_FmRNA proc-
essing,_ and Fnuclear mRNA splicing via spliceosome_ were not
significant GO terms in either cell population. Ependymal cells have
a basal apical orientation, and the GO term for F apical plasma
membrane_ was significant in these cells along with peroxidase
activity. A complete listing of GO terms for the FACS data is in
Supplementary data S7.
There were 82 probe sets (78 unique genes) at the intersection of
the FACS data and the brain region profile data. Fold-change values
for genes at this int ersection are indicated in the tables of
Supplementary data S2 4. Cell cycle related genes Ccnd2, Cdca7,
Mki67, Rrm2, and Mcm7 were increased in type B cells; no cell
cycle genes were statistically significantly elevated in the CD24+
population. Of the 10 RNA splicing genes in the SO profiles, only
Snrpg was differentially expressed (1.7-fold increased in CD24+
cells). Of the chromatin-remodeling genes, Mll, H2afx, and Hmgb3
Fig. 4. FACS analysis of SVZ cells. (A D) Immunostaining of dissociated SVZ cells. (A) DIC image and (B) immunofluorescent image of a multiciliated CD24-
positive ependymal cell. Arrow indicates ependymal cilia. Panels C and D show respective DIC and immunofluorescence images of a GFAP-positive SVZ cell. (E,
F) FACS of immunostained SVZ cells. (E) SVZ cells stained only with secondary antibodies. Cross-bars shown isolate >99% of the non-specific signal in the
lower left quadrant. (F) SVZ cells stained for CD24 and GFAP. Rectangle R1 indicates the collection gate for the GFAP+, CD24 population. R2 indicates the
collection gate for the GFAP, CD24+ cells. (G) Schematic of cDNA amplification procedure. Briefly, mRNA is reverse transcribed from an oligo-dT primer
containing a T7 RNA polymerase promoter sequence. A specific oligonucleotide (SMARTIII oligo) containing a stretch of dG nucleotides is included in the
reaction, and the ‘strand-switching’ activity of the reverse transcriptase copies the SMARTIII sequence to the end of the cDNA. With primers to the SMARTIII
and oligo-dT T7 promoter sequences, two rounds of long-distance PCR (LD-PCR) are used to amplify the cDNA. For hybridization, cRNAs are produced from the
3V T7 promoter. See Experimental methods for details. (H) Cellular RNAs are stable through the immunostaining protocol. 1 10
6
SVZ cells were double
immunostained for GFAP and CD24. Omission of 0.1% Tween-20 results in no GFAP staining. 1% Tween-20, RNasin, and DTT were added to the staining
solutions where indicated (+). After staining, cells were incubated at 4-C for an additional 1.5 h. Total cellular RNAwas then extracted and analyzed on agarose gel.
No RNA degradation was detected in any staining protocol. Note that if SVZ cells are freeze thawed and incubated at 37-C, all of the 28S and 18S RNAs are
degraded (right lane). (I) Analysis of ds cDNA libraries from FACS SVZ cells. A portion of the ds cDNAs after the first round of LD-PCR was used as the template
in a second round of control LD-PCR reactions in which aliquots were taken after 6, 8, 10, and 12 cycles. The cDNA aliquots were analyzed on agarose gels (left
panel). The size distribution of the amplified cDNAs was not biased toward smaller products by the LD-PCR. Southern blot signal for GAPDH was a single band,
indicating that the initial mRNA was not heavily degraded. The linear range of amplification was determined by both the GAPDH signal intensity and visual
inspection of the ethidium bromide stained cDNA population. (J) Semi-quantitative RT-PCR confirms the separation of SVZ cells by FACS. The GFAP message
was more than 10-fold enriched in the cDNAs prepared from the GFAP+, CD24 cell population (R1) as compared to the GFAP, CD24+ population (R2).
Conversely, the CD24 message was more than 20-fold enriched in the cDNAs from the R2 population in comparison with that of the R1 population. (K) Agarose
gel and Northern analysis of cRNAs from FACS-derived ds cDNAs. Size distributions were as expected for brain tissue and GAPDH messages did not show signs
of mRNA degradation.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148136
Page 6
were increased in type B cells by 3.7, 9.4, and 1.6-fold, respectively
(Table 2). Therefore, some chromatin-remodeling genes may begin
expression in the stem cell population of the SVZ and continue into
the ObC. Discussion of some of the other notable gene expression
differences between type B cells and ependyma is in the
Supplementary text.
Analysis of SVZ gene expression changes during SVZ regeneration
also identifies RNA splicing and chromosome organization as
prominent biological processes
We next analyzed gene expression changes during in vivo
regeneration of the SVZ germinal zone. Osmotic pump infusion of
the anti-mitotic cytosine arabinoside (AraC) onto the surface of the
brain eliminates type A and C cells, leaving behind only type B cells
and ependyma. After AraC pump removal, the SVZ regenerates with
remarkable fidelity. First, type B cells begin dividing. Between 2 to 4
days after pump removal, type C cells emerge, and after that, type A
cells form. Within 10 days, the entire network of migrating
neuroblasts with clusters of B and C cells is reconstituted (Doetsch
et al., 1999a,b). See Fig. 5A for illustration of SVZ regeneration.
We profiled gene expression at 1, 3, and 10 days (A1, A3, A10)
after AraC pump removal. To control for the effects of surgery, we
analyzed gene expression of saline infusion at 1 day (S1) and 10
days (S10) after pump removal. We also in parallel analyzed SVZ
from unmanipulated animals.
First, we identified genes whose expression was significantly
regulated ( P < 0.05) in at least one comparison to untreated SVZ
(total of 1758 probe sets). SVZ dissections include a small amount of
underlying striatal tissue; to focus our analysis on genes expressed
strongly in the SVZ, we filtered the AraC data with the list of genes
(985 probe sets) that were determined to be increased in the SVZ as
compared to the underlying striatum ( P < 0.05) in the brain region
experiment. The 229 probe sets at the intersection of these two lists
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 137
Page 7
were then analyzed with Principle Component Analysis (PCA) to
allow us to separate the gene expression changes of SVZ
regeneration from that of surgery and saline infusion (see
Experimental methods for details of the filters and PCA). The gene
expression pattern of the 59 probe sets (57 unique genes) most
related to SVZ regeneration is shown clustered in a color matrix (Fig.
5B), and a list of these genes is in Supplementary data S8.
The 59 probe sets shown share a similar expression pattern
representing the initial destruction and later regeneration of the
SVZ. At A1, gene expression is decreased relative to S1 (A1 < S1).
Between A1 and A10, gene expression returns to near normal
levels (A10 ; S10) or even Fsupranormal _ levels (A10 > S10);
these Fsupranormal_ levels may be due to the robust surge of
neurogenesis after AraC treatment, producing chains of type A
cells more dense than in saline controls (Doetsch et al., 1999a,b;
Doetsch and Alvarez-Buylla, 1996).
We applied GO analysis to the genes regulated during SVZ
regeneration. Similar to the SO profile, terms related to mRNA
splicing were the most significant (Fig. 5C). GO terms related to
regulation of cell cycle, proliferation, enzyme regulation, and
chromosome organization, and chromatin/nucleosome structure
were also significant (Fig. 5C, and Supplementary data S9 contains
a list of all GO terms for SVZ regeneration). Of the 59 probe sets in
this analysis, 16 (29%) were also found in the SVZ or SO profiles
(Table 4). The probability of having such an intersection at random
is approximately 10
50
, with the expected number of genes in the
random intersection being 0.7. Of these 16 genes, 4 had increased
expression in the FACS GFAP+ population (Table 4); the
probability of this intersection by chance is smaller than 10
10
.
In situ hybridization (ISH) validates gene expression data
The SVZ, SO, and ObC expression profiles suggested genes
that may be important for SVZ-Ob neurogenesis. Because these
profiles are derived from filters based on expression levels relative
to an artificial mean (see Experimental methods), they are not
intended to indicate the absolute presence or absence of gene
expression in the brain regions analyzed. For instance, a gene in the
ObC profile should be expressed at a level statistically higher than
the calculated average of all brain regions; however, an ObC
profile gene may not necessarily be expressed exclusively in the
ObC. To better understand how the expression profile data predicts
in vivo expression patterns, we performed ISH for some of the
genes.
Dlx5 and Mrg1/Meis2 were found in the ObC profile, and ISH
demonstrated that both Dlx5 and Mrg1/Meis2 are expressed in both
the ObC and the SVZ (Figs. 6A, B, E, F). To provide a comparison
to an SO profile gene, we performed ISH for Dlx2 in parallel (Figs.
6C, D). As assessed by ISH, ObC profile genes Dlx5 and Mrg1/
Meis2 both were more intensely expressed in the ObC as compared
to the SVZ; in comparison, the SO profile gene Dlx2 was
expressed higher in the SVZ than in the ObC. Therefore, ObC
profile genes may be expressed in SVZ, but the ObC/SVZ
expression ratio is higher than that of SO profile genes. The
GeneChip data also predict that Mrg/Meis2 expression levels in the
SVZ and St should be similar, and the ISH data are consistent with
this prediction. Thus, the GeneChip data provide a reasonable
estimation of relative gene expression levels as assessed by ISH.
We next used ISH to examine the gene expression of the RNA
splicing genes Sfrs2, Sf3b1, Lsm4, and Khdrbs1/Sam68 and
chromatin remodeling genes Mll and Smarcad1 (Fig. 6). Sfrs2 is
clearly expressed in the SVZ and ObC. A low level of Lsm4
expression was detected in the ObC, however, ISH was not evident
outside of that region; it is likely that the ISH detection threshold
for this gene was low, and we confirmed Lsm4 expression in both
the SVZ and ObC with RT-PCR (data not shown). Sf3b1 and
Khdrbs1/Sam68 were both clearly expressed in the SVZ and ObC
at levels higher than the other brain regions. The chromatin-
remodeling gene Mll was expressed at moderate levels in all brain
regions; however, it was detected in the SVZ and at relatively
higher levels in the ObC. Similarly, SWI/SNF family member
Smarcad1 was expressed moderately in all brain regions; however,
its expression was very prominent in the SVZ and ObC.
Discussion
We used Affymetrix GeneChips in three different approaches to
identify gene sets associated with in vivo SVZ neurogenesis. We
first obtained the gene expression profiles of five adult mouse brain
regions and filtered for genes that had increased expression in the
germinal SVZ and/or Ob target of neuronal differentiation. GO
analysis identified RNA splicing and chromatin remodeling as
prominent biological processes in the neurogenic SVZ and Ob
brain regions. Using FACS and cDNA amplification, we then
compared the expression profiles of two SVZ cell populations
important for neurogenesis: the SVZ astrocytes which function as
the stem cells (Doetsch et al., 1999a,b), and the ependymal cells
which contribute to the creation of a neurogenic niche (reviewed in
Goldman, 2003; Alvarez-Buylla and Lim, 2004); SVZ astrocytes
were significant for the processes of cell proliferation, neuro-
genesis, and chromatin remodeling. For a more dynamic portrait of
SVZ neurogenesis, we analyzed the transcriptional profiles during
SVZ regeneration which proceeds sequentially from B to C to A
cells (Doetsch et al., 1999a,b); GO analysis of the SVZ
Table 3
GO term differences between type B cells (GFAP+) and ependyma
(CD24+)
Highlighting indicates statistical significance of the listed GO term (e.g.,
Fcell cycle arrest_ is significant in the CD24+ cells and not the GFAP+ cells.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148138
Page 8
regeneration data also found RNA splicing and chromosome
organization as prominent biological processes.
These three approaches have distinct advantages and dis-
advantages. The brain region comparison yields the cleanest
expression data, but it represents the average expression profile of
entire regions and may reveal components beyond those related
to neurogenesis. The cell-type comparison is a more direct
analysis of the neurogenic transcriptional profile, but the extra
amplification required for chip hybridization results in noisier
data. The regeneration analysis is a fairly direct test for genes that
are dynamically regulated during neurogenesis, yet the invasive-
ness of the procedure complicates anal ysis. Because the
expression data derived from these three approaches differ in
quality and nature, we analyzed the GeneChip array data of the
three experiments separately. For the brain region and cell-
specific transcriptional profile analyses, we used the t test to
determine differential gene expression; for the SVZ regeneration
experiment, we used PCA to separate the gene expression due to
SVZ regeneration from that of surgery and saline infusion (see
Experimental methods, Data analysis for details of these
methods). Each experimental approach provided us with a
different view of the transcriptional profile for SVZ neurogenesis,
and the transcriptional profiles from all three approaches were
unified by GO analysis, which gave us an overview of the
biological processes involved.
Supporting our experimental approaches, we found that some of
our expression data matched previously known regional and cell-
specific expression patterns, and Northern blot analysis and ISH
validated other data. A large number of genes identified in this study
have not been previously described to be present in the SVZ or Ob
and are available in the Supplementary data. In the Results section,
we presented data mostly for the RNA splicing and chromatin
remodeling genes, however, taken together, the data appeared to fit
into a biological ‘story’ of SVZ neurogenesis, progressing through
cell cycle, transcriptional regulation, RNA processing, migration,
and apoptosis (see Fig. 7 and Supplementary text).
Recent progress in the description of stem cell gene expression
has be en made by comparing gene profiles of embryonic,
hematopoietic, and neural stem cells grown as neurospheres
(Ivanova et al., 2002; Ramalho-Santos et al., 2002). These analyses
identified sets of genes that may be important for basic stem cell
properties such as self-renewal; however, the process of neuro-
genesis was not specifically addressed. Prior gene expression studies
of neurogenesis have been performed with neurospheres in vitro.
Fig. 5. Transcriptional profile of SVZ regeneration after AraC treatment. (A) Schematic of AraC infusion and associated changes in SVZ cellular composition
after AraC pump removal. At 1 day, only ependyma (gray) and type B cells (blue) remain. At 3 days, type C (green) cells return. At 10 days, all SVZ cell types
including type A cells (red) have been regenerated. (B) Transcriptional profile of SVZ regeneration. The columns labeled A1, A3, and A10 represent the
timepoints after AraC infusion. Columns S1 and S10 are the timepoints after control saline infusion. The SVZ column is the gene expression of unmanipulated
controls. Genes are ordered along the vertical axis using hierarchical clustering. The color and color intensity of each cell in the matrix relate to the expression
ratio of each gene. Red indicates a positive ratio (expression greater than the mean of the other brain regions), green indicates a negative ratio, and black
indicates a ratio of 1. A color scale (log
2
) indicating the magnitude of the expression ratios is at the bottom of the panel. (C) GO analysis pie chart for SVZ
regeneration. The entire pie represents all GO terms in the analysis. Pie slices are proportional to the number of genes (in parentheses) related to a particular GO
Fparent_ term (legend for color code is in the inset to the right of each panel). GO terms that are Fchildren_ of a parent term are listed next to the pie chart with an
indicating line. Further parent child relationship of the GO tree structure is indicated by indentation with hyphen. All listed GO terms are statistically
significant, and color of the type indicates the GO category (see legend at the lower right of the figure).
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 139
Page 9
Neurospheres are spherical clusters of cells propagated in vitro from
single cells by addition of EGF and/or FGF. Neurospheres can
generate neurons, astrocytes, and oligodendrocytes (Reynolds and
Weiss, 1992; Morshead et al., 1994; Gritti et al., 1996; Kukekov et
al., 1999; Caldwell et al., 2001). For the transcriptional profile
studies, neurospheres were obtained from embryonic and early
postnatal cortex (not SVZ) (Geschwind et al., 2001; Easterday et al.,
2003; Karsten et al., 2003), embryonic striatum (contains SVZ)
(Zhou et al., 2001; Wen et al., 2002), or postnatal SVZ (Gurok et al.,
2004); the adult SVZ differs in gene expression and cellular
composition from that of embryonic and postnatal SVZ as well as
developing cortex (Tramontin et al., 2003). Also, the high levels of
exogenous growth factors (EGF or F GF) used to propagate
neurospheres deregulates normal gene expression (Gabay et al.,
2003; Hack et al., 2004), likely leading to significant alterations in
their transcriptional profiles. Notwithstanding these differences,
there were genes and biological processes overlapping between our
in vivo analysis and the in vitro neurosphere studies: certain cell
cycle genes (Ccnd2, Mcm3, Mcm7, S100a6, Mdk, Pcna, Gadd45b),
cytoskeletal/migration genes (Tubb3, Tagln, Racgap1), Hmgb2,
Fyn, and Rbp1 were common to our analysis and one or more of the
neurosphere gene expression studies (Geschwind et al., 2001;
Easterday et al., 2003; Karsten et al., 2003; Gurok et al., 2004). In
addition to identifying these genes, our study provided spatial (brain
region and SVZ cell type) and/or temporal (during regeneration)
expression information. The raw data sets and complete gene lists
are available in the Supplementary data, allowing further analysis of
the similarities and differences between mouse in vitro neurospheres
and in vivo SVZ neurogenesis. Such analyses along with compar-
isons to human neurosphere transcriptional profiles (Wright et al.,
2003) may allow us to narrow down the list of genes that may be
important for neural stem cell function.
The GFAP+ and CD24+ transcriptional profiles allowed us to
assign a subset of genes to either the neurogenic type B cells or the
non-neurogenic ependyma. It is possible that the GFAP+ cells in the
SVZ are intrinsically different from GFAP+ astrocytes in non-
germinal regions. It will be interesting to compare the SVZ GFAP+
transcriptional profile to those of astrocytes without stem cell
properties; the differences revealed by such an analysis may reveal
the molecular basis of the stem cell properties unique to SVZ
astrocytes. There is very little information about the gene expression
of ependymal cells. These important epithelial cells are born in the
embryo (Spassky et al., 2005) and play essential roles in brain
cerebrospinal fluid circulation and homeostasis. Ependymal cell also
contribute to the neurogenic niche (Lim et al., 2000; Goldman, 2003;
Peretto et al., 2004). Our transcriptional profile of the CD24+ cells
provides a gene expression database for ependymal cells and should
serve as an important resource for further molecular analysis of these
cells (see Supplementary text). The gene expression profile of
isolated type A cells has also been studied (Pennartz et al., 2004);
therefore, to date, transcriptional profiles of type B, ependymal, and
type A cells are available, and together they should assist
investigators in the formation of hypotheses about gene function
in the SVZ.
RNA splicing in SVZ neurogenesis
It has been proposed that RNA splicing is vital for
generating the complexity of the nervous system (Grabowski
and Black, 2001; Black and Grabowski, 2003). Alternative
splicing of the same gene can induce dramatic changes in neural
developmental; for instance, distinct splice isoforms of Numb
direct either proliferation or differentiation (Verdi et al., 1999).
RNA splicing can regulate cell fate, transcription factor activity,
axon guidance, neuro tran smitt er recepto r and ion ch anne l
function, and apoptosis; because all of these processes occur
in the SVZ throughout adult life, the SVZ may be an ideal
system in which to study RNA splicing function in neural
Table 4
Intersection with SVZ regeneration data
Highlighted cells indicate the profile to which each probe set/gene belongs (e.g., Ccnd2 has its cell in SVZ column highlighted, indicating the SVZ profile).
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148140
Page 10
development. In this study, we identified 11 genes for RNA
splicing that may be important for adult SVZ neurogenesis. The
SO profile contained Sf3b1 (splicing factor 3b, subunit 1), Sfrs2
(splicing factor, arginine/serine-rich 2, SC35), Lsm4 (LSM4
homologue, U6 small nuclear RNA associated), Snrpg (small
nuclear ribonucleoprotein, polypeptide G), Khdrbs1/Sam68 (KH
domain containing, RNA binding, signal transduction associated
1), and four members of the heterogeneous nuclear ribonucleo-
protein family—Hnrpa2b1, Hnrpm, Hnrph1, and Hnrpd. The
analysis of SVZ regeneration also recognized Sf3b1, Hnrpd, and
Lsm4; additionally, three other genes for RNA splicing were
identified in the regeneration experiment: Brunol4, Prpf8, and
Hnrpab (Supplementary data S8).
Sf3b1, Sfrs2, Prpf8, Lsm4, Snrpg, Hnrpa2b1, Hnrpm, Hnrph1,
Hnrpd, and Hnrpab are all components of the spliceosome complex
(reviewed in Jurica and Moore, 2003). The activity and specificity of
the spliceosome are regulated; for instance, changes in levels of
Hnrpab mediate mRNA splice site selection in developing
erythroblas ts (Hou e t al., 2002). The he terog eneous nuclea r
ribonucleoprotein (Hnrp) family members (e.g., Hnrpab) them-
selves are regul ated by methylation at arginine (review ed in
McBride and Silver, 2001), and the arginine methyltransferase
Fig. 6. In situ hybridization (ISH) validates transcriptional profile expression data. ISH was performed for Dlx2 (A, B), Dlx5 (C, D), Meis2 (E, F), Sfrs2 (G, H),
Sf3b1 (I, J), Lsm4 (K, L), Khrdbs1/Sam68 (M, N), Mll (O, P), and Smarcad1 (R, S) on coronal adult brain sections. The dotted line in panel A shows the
boundary between the corpus callosum (CC) and the Ctx, and the SVZ is indicated by arrows. The ventricle is to the left. Scale bars = 100 Am (A, C, E, G, I, K,
M, O, R), 500 A m (B, D, F, H, J, L, N, P, S).
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 141
Page 11
Hmrt1l2 (Scott et al., 1998) was in the SO profile, suggesting its
interaction with the Hnrps. Brunol4 belongs to the bruno/elav
family of RNA binding proteins that regulate mRNA processing
(Good et al., 2000); the human homologue of Brunol4 promotes
specific exon exclusion in developing muscle (Ladd et al., 2001).
Perhaps most intriguingly, Khdrbs1/Sam68 is a prototype splice
site regulator whose activity is modified by extracellular signal-
regulated kinase (ERK) transduction (Matter et al., 2002); as such,
Khdrbs1/Sam68 may link t he SVZ precursor RNA splicing
machinery to changes in the extracellular environment. Khdrbs1/
Sam68, like the Hnrp family members, is also regulated by arginine
methylation (Bedford et al., 2000). Fyn is a kinase found in the ObC
profile, and FYN phosphorylation of KHDRBS1/SAM68 changes
its subcellular localization, interaction with the spliceosome
components, and splice site selection (Hartmann et al., 1999); the
increased expression of Fyn in the ObC could induce Khdrbs1/
Sam68 to change mRNA splicing regulation in type A cells leading
to their cell cycle exit, change to radial migration, and integration
into local circuits.
Neuroblasts born in the SVZ have different destinations in the
Ob. Some end up in the granule cell layer, while others migrate
farther into the periglomerular layer. Granule cell and periglomer-
ular interneurons have different synaptic organization as well as
neurotransmitter phenotypes. If these two types of Ob interneurons
are derived from the same SVZ neural stem cell (this is currently
unclear), it is possible that alternative splicing may be critical for
determining the migratory path of the neuroblasts as well as the cell
fate choice. Recently, a genome-wide analysis of alternative
splicing determined by the Nova splicing factor has indicated that
RNA splicing may play important roles in synapse formation,
axonogenesis, neurite morphogenesis, and neurogenesis (Ule et al.,
2005). Eph/ephrin signaling plays a role in SVZ migration and
proliferation (Conover et al., 2000), and alternative splice forms of
certain Eph receptors can regulate cellular repulsion or adhesion
(Holmberg et al., 2000). Hence, alternative splicing of the same
sets of transcripts could account for the generation of different
destinations and phenotypes of SVZ-born neuroblasts.
Chromatin remodeling in SVZ neurogenesis
Chromatin remodeling can engage or maintain particular
genetic ‘programs’ and therefore likely plays a critical role in
both stem cell maintenance as well as daughter cell differenti-
ation (reviewed in Rasmussen, 2003; Cerny and Quesenberry,
2004; Ehrenhofer-Murray, 200 4). There also is increasing
evidence that chromatin remodeling is important for neural
development (reviewed in Hsieh and Gage, 2004). Bmi1,a
member of the Polycomb group of chromatin modifiers, is
important for self-renewal of embryonic and postnatal SVZ stem
cell regulation (Molofsky et al., 2003); in the adult SVZ, we
identified Bmi1 in the ObC profile. Polycomb group members
such as Bmi1 work in concert with trithorax group proteins to
regulate chromatin structure (Orlando, 2003); appropriately, Mll,
a member of the trithorax family, was expressed in the SO
profile. BMI1 physically interacts with and is antagonized by
MLL (Hanson et al., 1999; Xia et al., 2003).
Mll establishes and maintains specific gene expression patterns
through serial mitotic cell cycles (Yu et al., 1998; Milne et al.,
2002). The increased expression of Mll in the B cell population and
presence in the SO profile (Table 2) suggests that Mll expression
begins in B cells and continues through the lineage to type A cells.
Mll therefore potentially regulates global developmental transcrip-
tional patterns throughout the entire SVZ neurogenic lineage. Mll
regulates Dlx1, Dlx2, and Dlx5 (Ferrari et al., 2003), transcription
factors in the SO profile, and MLL fusion proteins regulate Pbx3
and Meis1 (ObC profile) (Zeisig et al., 2004). Additionally, using
transcriptional profile analysis, Schraets et al. identified potential
gene targets of Mll regulation (Schraets et al., 2003), and among
the top candidates are Col6a (SO profile), Fhl1 (Four-and-a-half
LIM domains 1, ObC profile), Nestin (neural precursor cell marker
expressed in SVZ (Gates et al., 1995; Doetsch et al., 1997)), and
Tenascin-C (SVZ stem cell niche ECM component (Garcion et al.,
2004)). Hence, we have not only identified Mll in the SVZ but also
9 genes that Mll may regulate.
H2afx (SVZ profile, regulated during regeneration) is a histone
H2A variant that is critical for chromatin remodeling and
inactivation of sex chromosomes in meiosis (Fernandez-Capetillo
et al., 2003). Methylation of histone arginine residues modifies
chromatin function (reviewed in Trievel, 2004), and the arginine
methyltransferase Hmrt1l2 (Scott et al., 1998) was found in the SO
profile. One of the best characterized histone modifications is
lysine acetylation (reviewed in Sterner and Berger, 2000 ), and
Hat1 (histone acetyltransferase 1) was in the SVZ profile. In
addition to modifying histones, Hat1 can acetylate high mobility
group proteins (HMGs), which were also present in our analysis.
Hmgb2 (SVZ profile) and Hmgb3 (SO profile, increased in type B
cells) are members of the high-mobility group B (HMGB) family,
which can activate or repress transcription by modifying DNA
histone complexes (Ge and Roeder, 1994; Shykind et al., 1995;
Thom as, 2001). Hmgb2 wasalsoidentifiedinneurospheres
(Karsten et al., 2003; Gurok et al., 2004). In primitive blood cell
precursors, enforced expression of Hmgb3 inhibits B cell and
myeloid lineages (Nemeth et al., 2003), and Hmgb3-deficient mice
have dysregulated lymphoid and myeloid cell development
(Nemeth et al., 2004).
SWI/SNF chromatin modifiers also regulate transc ription.
Smarcad1 (ObC profile) is a SWI/SNF component, and
Smarcad1-deficient mice have impaired fertility, skeletal dyspla-
sias, and growth retardation (Schoor et al., 1999). Arp (actin-
related protein) family members regulate SWI/SNF complexes
(reviewed in Olave et al., 2002), and Baf53a (ArpNa) was
identified in the SO profile. Intriguingly, Baf53a is brain specific
and expressed in developing neurons in vitro (Kuroda et al.,
2002). Among the 216 ‘stemness’ genes common to brain,
blood, and embryonic stem cells are two members of the SWI/
SNF family of chromatin modifiers (Ramalho-Santos et al.,
2002), further suggesting the importance of chromatin modifica-
tion for stem cell regulation.
Fig. 7. Schematic of genes, biological processes, and gene interactions for SVZ neurogenesis. Data from the SVZ, SO, ObC profiles, the FACS data, and the
SVZ regeneration analysis are integrated. This figure highlights 89 genes selected from the data; these genes are discussed in the Results section and
Supplementary text. Genes in the SVZ, SO, and ObC profiles are arranged over a yellow background in vertical columns. Genes increased in GFAP+ and
CD24+ cells are boldfaced in blue and black, respectively. Genes regulated during SVZ regeneration are circled. Known physical and genetic interactions are
indicated by dotted lines and red arrows, respectively. See the legend at the lower right.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148142
Page 12
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 143
Page 13
Concluding remarks
Any attempt to understand adult neurogenesis at the molecular
level needs to take into consideration large sets of genes acting in
parallel. This study provides data on genes that contribute to adult
neurogenesis. The data hint to the groups of genes involved in
proliferation, migration, and differentiation and reveal chromatin
remodeling and RNA splicing as important components of these
processes. This in vivo molecular description of SVZ neurogenesis
provides the launching point of future studies into the regulation of
this adult germinal zone. The challenge now is to understand the
contribution of individual genes in the context of the complexity
revealed by this study.
Experimental methods
Production of ds T7 cDNA from adult brain regions
Adult (2 3 months) CD-1 (Charles River Laboratories) mouse
brains were used for RNA isolation. SVZ was dissected as
previously described (Lim and Alvarez-Buylla, 1999), and Ctx
and St were obtained from the same coronal slice. ObC was
dissected from serial coronal slices of the Ob. Hp was isolated by
cutting the fimbria and blunt dissection. 10 mice were used for
each of the 2 experimental replicates. Dissected tissues were snap
frozen in 1.5-ml tubes with liquid N
2
. Tissues were disrupted in
RNeasy (Qiagen) lysis buffer with needle trituration and Qiash-
redder columns (Qiagen). DNase treated total RNA was isolated
with RNeasy mini-col umns (Qiagen). PolyA RNA was then
purified with magnetic oligo-dT beads (Dynal). For each brain
region, 1 Ag of polyA RNA was converted to ds T7cDNA with the
T7LD3V primer using standard Superscript II reverse transcriptase
and DNA polymerase protocols (Invitrogen).
FACS isolation of type B and ependymal cells and ds cDNA
production
Adult SVZ cells were dissociated, cleared of dead cells and
debris by 22% Percoll (Sigma) step gradient as previously
described (Lim et al., 2000), and passed through a 40-Am nylon
cell strainer (BD Biosciences). All immunostaining incubations
and washes were performed at 0 4-C with pre-chilled buffers.
Biotinylated mCD24 antibody (BD Biosciences, Pharmingen)
was used at 1:10 and rabbit GFAP antibody (DakoCytomation)
was used at 1:100. About 1 10
6
SVZ cells were resuspended
in 100 Al PBS containing both primary antibodies, 0.1% Tween-
20 (Sigma), and 100 200 units of RNasin (Promega) and
incubated for 15 min on ice. Cells were pelleted by gentle
centrifugation and washed in PBS three times. Cells were then
resuspended in 100 Al of PBS containing streptavidin-Cy2 at
1:100 and anti-rabbit F(ab)
2
at 1:25 (Jackson Immunoresearch),
0.1% Tween-20, and 100 200 units of RNasin and incubated
for 10 min on ice. Cells were again washed 3 times with PBS.
Omission of primary antibodies resulted in no staining.
Immunostained cells were isolated with the FACS Vantage
(BD Biosciences). For each of the 2 experimental replicates,
10,000 cells (from the SVZ of 25 mice) were collected directly
into RNeasy lysis buffer, and DNAse-treated RNA was isolated
with RNeasy columns. RNA was then converted to cDNA with
the T7LD3V primer using standard Superscript II reverse
transcriptase protocols. Using the cDNA as template, 20 cycles
of LD-PCR (BD Biosciences, Clontech) were performed. ds
T7cDNA from the LD-PCR reactions we re phenol/CHCl
3
extracted and spun through a Chromospin400 column (BD
Biosciences, Clontech). An aliquot of the ds T7cDNA was used
as template for a second LD-PCR, and aliquots were removed at
6, 8, 10, and 12 cycles; these were analyzed on agarose gels by
ethidium bromide staining and GAPDH Southern blot to
determine the linear range of amplification.
Analysis of regenerating SVZ
2% AraC in vehicle (saline 0.9%) or vehicle alone were infused
onto the surface of 2- to 3-month-old CD-1 mice for 6 days by
mini-osmotic pump (Alzet, Palo Alto, CA, Model 1007D) as
described (Doetsch et al., 1999a,b). At the end of infusion, osmotic
pumps were surgically removed from their suprascapular place-
ment; cannulas were left in place until after animals were
sacrificed. Only the SVZ from the side of cannula placement
(right side) was dissected. A total of 18 mice were used for this
experiment: 4 for the no-surgery control, 4 for A1, 3 for A3, 3 for
A10, 2 for S1, and 2 for S10. Total RNA from dissected SVZ tissue
was isolated as described above. 312 Ag of total RNA from
pooled SVZ tissue for each time point/condition was converted to
ds T7 cDNA using the above protocols, and equal amounts of
biotin-labeled cRNA were used for GeneChip hybridizations.
GeneChip probe production and hybridizations
Biotin-labeled cRNAs were produced from the ds T7cDNA
libraries and hybridized to Mu11K chips according to standard
protocols (Affymetrix, Santa Clara, CA). Chips were scanned on
a GeneArray scanner (Affymetrix). For each brain region, cRNAs
were prepared from independent ds cDNA libraries from different
dissection sessions. Likewise, for each FACS population, cRNAs
were generated from independent ds cDNA libraries prepared
from different dissection sessions and FACS runs.
Northern and Southern blots and PCR analysis
Northern and Southern blots were performed according to
standard protocols using ExpressHyb (BD Biosciences, Clontech)
or ULTRAhyb (Ambion). Probes for hybridizations were produced
by PCR cloning. All probes were sequenced to verify their identity.
Semi-quantitative PCR analysis for CD24 and GFAP was
performed as previously described (Lim et al., 2000).
In situ hybridization
For in situ hybridization (ISH) on cryosections, we used a
modification of methods previously described (Wilkinson, 1999).
After perfusion fixation of 2- to 3-month old mice, brain tissues
were fixed with 4% PFA, cryoprotected by 10% and 20% sucrose
PBS, embedded in OCT compound (Sakura Finetechnical Co.
Ltd., Tokyo, Japan), frozen, and sectioned at 18 Am thickness.
After ISH staining, the sections were counterstained by nuclear
fast red. The following mouse cDNA was used for making
digoxigenin labeled probes: Meis2 (GenBank accession num-
ber:BF472214), Dlx2 (GenBank:NM
_
010054,nt746-1355), Dlx5
(GenBank:AW046057), Mll1 (GenBank:BC044818), Smarcad1
(GenBank: BC042442), Sf3b1 (GenBank:BC037098), Sfrs2 (Gen-
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148144
Page 14
Bank:BC005493), Sam68 (GenBank:BC002051) and Lsm4 (Gen-
Bank:BC026747). Dlx2 cDNA was cloned by RT-PCR, and others
were obtained as EST clones.
Data analysis
Data analysis was performed with the R packages available at
the Bioconductor project site (www.bioconductor.org). We used
the GCRMA algorithm to obtain expression measures from the
fluorescent intensities of the individual probes. This algorithm
employs a statistical model that uses probe sequence information
for background adjustment (Naef and Magnasco, 2003; Wu and
Irizarry, in press) which proved to be more sensitive than other
preprocessing methods (see http://affycomp.biostat.jhsph.edu) in-
cluding the GeneChip software (MAS 5). The normalization step
utilizes a quantile normalization algorithm (Bolstad et al., 2003;
Bolstad, 2004) and probe sets were summarized using medianpol-
ish (Bolstad et al., 2003; Irizarry et al., 2003).
To identify genes of the SVZ, SO, and ObC profile, we
first filtered the data to exclude genes with low variability
across all brain region samples (standard deviation smaller than
0.15). Then the t test was used to determine the set of genes
differentially expressed in th e region und er question as
compared to the other brain regions. P values were adjusted
for multiple hypothesis test as suggested in Benjamini and
Hochberg (2001) and Dudoit and Shaffer (2003), using the
Benjamin and Hochbert procedure; the permutations procedure
was not used. We then filtered for genes with statistical
significance ( P < 0.05) and with difference greater than 0.5
(i.e., more than 1.42-fold change) to obtain 71, 80, and 209
filtered probe sets in the SVZ, SO and ObC profiles
respectively (65, 60, 168 UniGene identifiers). Similar proce-
dures were carried out in other comparisons: t tests were
applied to FACS data (to determine those genes that are
differentially expressed between CD24+ and GFAP+ cells, P <
0.05) and to determine those genes expressed higher in the
SVZ as compared to the St (to provide the list of genes that
was used to as a filter for the AraC data, see below).
The gene expression analysis of SVZ regeneration is confound-
ed by the changes induced by the surgery. Some of these effects
may be adequately controlled by comparison with the saline
control groups; however, the response to surgical lesions is variable
from animal to animal and may differ between saline and AraC
treated animals. For this reason, we pooled the RNA from the SVZ
for each of the different time points (see Analysis of Regenerating
SVZ, above). Since this pooled RNA was analyzed on a single chip
set, we used Principal Component Analysis (PCA) after filtering
the data. To filter the data, we considered the differences between
untreated SVZ and all other samples (three AraC and two saline
time points) within this experiment; we selected genes that show
differences in at least one comparison; the threshold of the t test
was based on the distribution of the differences for all genes, rather
than on a gene-by-gene basis. This set of 1764 probe sets was
filtered with the list of genes that are increased in the SVZ as
compared with St ( P < 0.05, in the brain region analysis) to
eliminate from analysis those genes that normally are expressed at
high levels in the striatum. In order to separate the gene expression
changes of SVZ regeneration from that of surgery and infusion of
saline vehicle, the 229 probe sets at the intersection of these two
lists were analyzed by applying PCA to the expression matrix of
the 229 probe sets and the 6 chips.
PCA, a widely used data mining technique (see, e.g., Jolliffe,
2003), creates new independent variables (the principal compo-
nents) as those linear combinations of the original variables that
capture as much of the variability of the original system as possible.
In other words, PCA models a cloud of points in high dimensional
space by finding the direction along which the cloud has the largest
spread (the first component), the perpendicular direction with the
second largest spread (the second component), and so on. We found
that the first three principal components were enough to explain
almost 90% of the variability among chips, thereby reducing our 6-
dimension space to a 3-dimension space. In this new space, the first
component was basically the overall expression of the genes. The
second component described the ‘recovery’ from surgery and saline
infusion, while the third component captured the gene expression
due to the regeneration of the SVZ cellular population (see
Supplementary data S10). We emphasize that these 3 new variables
are independent in the population considered, and so the recovery
from saline infusion and the SVZ regeneration are now independent
variables. The 229 probe sets were then listed by magnitude of the
third component so that those genes at the top represent those most
related to SVZ regeneration and not the effect of saline or surgery.
The expression array data for the top 25% of this list (59 probe sets,
56 unique genes, Supplementary data S8) was then clustered and is
shown in Fig. 4.
Clustering analysis was done using Gene Cluster 3.0 software
and Tree View 1.6 (Eisen et al., 1998) available at http://rana.lbl.gov/
EisenSoftware.htm. We used hierarchical clustering with Complete
Average Linkage method and Euclidean distance as similarity
matrix for the SVZ, SO, and ObC profile data, and with the Pearson
Correlation coefficient for the AraC data.
Analysis of GO annotations was done using the Onto-Express
(Khatri et al., 2002; Draghici et al., 2003; Khatri et al., 2004)a
web-based tool available at http://vortex.cs.wayne.edu/Proj -
ects.html. To find those GO terms that were over-represented in
the transcriptional profile in question (e.g., the SVZ, SO, or ObC
profiles), we compared the list of genes in the profile with the
entire set of genes in Mu11K A and B chips. Significance was
assessed by using the hypergeometric distribution, and P values
were corrected for multiple hypothesis, controlling fdr (false
discovery rate). Only nodes (in the ontology tree) with fdr <0.1
and >1 gene were considered. Supplementary data S11 contains
the probe set identifiers for the SVZ, SO, ObC, Ctx, St, Hp,
GFAP+, CD24+, and SVZ regeneration profiles as well as the
background Mu11kA and B chips; these probe set lists can be
used with the Onto-Express tool, allowing one to browse through
the GO terms organized in the tree structure.
Primer sequences
T7LD3V :ATTCTAGAGGCCGAGGCGGCCGACATG-
TAATACGACTCACTATAGGGCGTTTTTTTTTTTTTTTTT-
TTTTTTTTTTTTTVN (V = A,G,C; N = A,G,C,T)
SMART III: AAGCAGTGGTATCAACGCAGAGTGGCCAT-
TATGGCCGGG
5VPCR: AAGCAGTGGTATCAACGCAGAGTGGCCAT-
TATGG
3VPCR: ATTCTAGAGGCCGAGGCGGCCGACATGTAA-
TACGACTCACTATAGGGCG
Gapdh: CCCACTAACATCAAATGGGG, CTCACTTGT-
GGCCCAGGTAT
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 145
Page 15
Dlx1: TCCTGAATGGTCTTCTTCCG, CTGGGGTGGTAC-
GAAGATGG
2310021601Rik: AGATGATAGCTG AG CAGC GG , CTGG-
CAGAGAGGTTCAAAGC
Sox11: CAGGCACTTCTTCCCTTTTG, CAGCTCT-
GAGGTCTATGTCACC
Col6a1: CCCCATTGGACCTAAAGGAT, CAGCACGAA-
GAGGATGTCAA
Ccnd2: CCTCACGACTTCATTGAGCA, ATGCTGCTCTT-
GACGGAACT
Hmgb2: AGCTTGGGGAAGGAAGTCTC, AGCAAAACAG-
GAAGAAGGCA
Mia: AGCCCAGAGACCTCGTTCTT, ATC AATTTTGC-
CAGGTTTCG
Pdyn: GATCAGGTAGGGCATGAGGA, TTCTCT-
GGATTCTGGGATGG
Gfap: CTCAATGCTGGCTTCAAGGAGA, GACG-
CAGCGTCTGTGAGGTC
Cd24: ATGCAAAGGAG CCAAAA CTG , GTGACCATGC-
GAACAAAAGA
Acknowledgments
We thank Miguel Ramalho-Santos for the many helpful dis-
cussions and editorial comments. H.T. was supported by the
Mochida Memorial Foundation for Medical and Pharmaceutical
Research. This work was supported by NIH grant NS28478-12 to
AAB.
Appendix A. Supplementary data
Supplementary data associated with this article can be found in
the online version at doi:10.1016/j.mcn.2005.10.005.
References
Alvarez-Buylla, A., Lim, D.A., 2004. For the long run: maintaining
germinal niches in the adult brain. Neuron 41, 683 686.
Bedford, M.T., Frankel, A., Yaffe, M.B., Clarke, S., Leder, P., Richard,
S., 2000. Arginine methylation inhibits the binding of proline-rich
ligands to Src homology 3, but not WW, domains. J. Biol. Chem.
275, 16030 16036.
Benjamini, Y., Hochberg, Y.C., 2001. Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J. R. Stat. Soc. 57,
289 300.
Black, D.L., Grabowski, P.J., 2003. Alternative pre-mRNA splicing and
neuronal function. Prog. Mol. Subcell. Biol. 31, 187 216.
Bolstad, B.M., 2004. Low level analysis of high-density oligonucleotide
array data: background, normalization and summarization. Biostatistics.
University of California, Berkeley, pp. 156.
Bolstad, B.M., Irizarry, R.A., Astrand, M., Speed, T.P., 2003. A comparison
of normalization methods for high density oligonucleotide array data
based on variance and bias. Bioinformatics 19, 185 193.
Calaora, V., Chazal, G., Nielsen, P.J., Rougon, G., Moreau, H., 1996.
mCD24 expression in the developing mouse brain and in zones of
secondary neurogenesis in the adult. Neuroscience 73, 581 594.
Caldwell, M.A., He, X., Wilkie, N., Pollack, S., Marshall, G., Wafford, K.A.,
Svendsen, C.N., 2001. Growth factors regulate the survival and fate of
cells derived from human neurospheres. Nat. Biotechnol. 19, 475 479.
Capela, A., Temple, S., 2002. LeX/ssea-1 is expressed by adult mouse CNS
stem cells, identifying them as nonependymal. Neuron 35, 865 875.
Cerny, J., Quesenberry, P.J., 2004. Chromatin remodeling and stem cell
theory of relativity. J. Cell. Physiol. 201, 1 16.
Chiasson, B.J., Tropepe, V., Morshead, C.M., Van der Kooy, D., 1999.
Adult mammalian forebrain ependymal and subependymal cells
demonstrate proliferative potential, but only subependymal cells have
neural stem cell characteristics. J. Neurosci. 19, 4462 4471.
Conover, J.C., Doetsch, F., Garcia-Verdugo, J.M., Gale, N.W., Yancopou-
los, G.D., Alvarez-Buylla, A., 2000. Disruption of Eph/ephrin signaling
affects migration and proliferation in the adult subventricular zone. Nat.
Neurosci. 3, 1091 1097.
Doetsch, F., Alvarez-Buylla, A., 1996. Network of tangential pathways for
neuronal migration in adult mammalian brain. Proc. Natl. Acad. Sci. U.
S. A. 93, 14895 14900.
Doetsch, F., Garcia-Verdugo, J.M., Alvarez-Buylla, A., 1997. Cellular
composition and three-dimensional organization of the subventric-
ular germinal zone in the adult mammalian brain. J. Neurosci. 17,
5046 5061.
Doetsch, F., Caille, I., Lim, D.A., Garcia-Verdugo, J.M., Alvarez-Buylla,
A., 1999a. Subventricular zone astrocytes are neural stem cells in the
adult mammalian brain. Cell 97, 703 716.
Doetsch, F., Garcia-Verdugo, J.M., Alvarez-Buylla, A., 1999b. Regenera-
tion of a germinal layer in the adult mammalian brain. Proc. Natl. Acad.
Sci. U. S. A. 96, 11619 11624.
Doetsch, F., Petreanu, L., Caille, I., Garcia-Verdugo, J.M., Alvarez-Buylla,
A., 2002. EGF converts transit-amplifying neurogenic precursors in the
adult brain into multipotent stem cells. Neuron 36, 1021 1034.
Draghici, S., Khatri, P., Bhavsar, P., Shah, A., Krawetz, S.A., Tainsky,
M.A., 2003. Onto-Tools, the toolkit of the modern biologist: Onto-
Express, Onto-Compare, Onto-Design and Onto-Translate. Nucleic
Acids Res. 31, 3775 3781.
Dudoit, S., Shaffer, J., 2003. Multiple hypothesis testing in microarray
experiments. Stat. Sci. 18, 71 103.
Easterday, M.C., Dougherty, J.D., Jackson, R.L., Ou, J., Nakano, I., Paucar,
A.A., Roobini, B., Dianati, M., Irvin, D.K., Weissman, I.L., Terskikh,
A.V., Geschwind, D.H., Kornblum, H.I., 2003. Neural progenitor genes.
Germinal zone expression and analysis of genetic overlap in stem cell
populations. Dev. Biol. 264, 309 322.
Ehrenhofer-Murray, A.E., 2004. Chromatin dynamics at DNA replication,
transcription and repair. Eur. J. Biochem. 271, 2335 2349.
Eisen, M.B., Spellman, P.T., Brown, P.O., Botstein, D., 1998. Cluster
analysis and display of genome-wide expression patterns. Proc. Natl.
Acad. Sci. U. S. A. 95, 14863 14868.
Fernandez-Capetillo, O., Mahadevaiah, S.K., Celeste, A., Romanienko, P.J.,
Camerini-Otero, R.D., Bonner, W.M., Manova, K., Burgoyne, P.,
Nussenzweig, A., 2003. H2AX is required for chromatin remodeling
and inactivation of sex chromosomes in male mouse meiosis. Dev. Cell
4, 497 508.
Ferrari, N., Palmisano, G.L., Paleari, L., Basso, G., Mangioni, M., Fidanza,
V., Albini, A., Croce, C.M., Levi, G., Brigati, C., 2003. DLX genes as
targets of ALL-1: DLX 2,3,4 down-regulation in t(4;11) acute
lymphoblastic leukemias. J. Leukocyte Biol. 74, 302 305.
Ferri, A.L., Cavallaro, M., Braida, D., Di Cristofano, A., Canta, A.,
Vezzani, A., Ottolenghi, S., Pandolfi, P.P., Sala, M., DeBiasi, S.,
Nicolis, S.K., 2004. Sox2 deficiency causes neurodegeneration and
impaired neurogenesis in the adult mouse brain. Development 131,
3805 3819.
Fiske, B.K., Brunjes, P.C., 2001. Cell death in the developing and sensory-
deprived rat olfactory bulb. J. Comp. Neurol. 431, 311 319.
Gabay, L., Lowell, S., Rubin, L.L., Anderson, D.J., 2003. Deregulation of
dorsoventral patterning by FGF confers trilineage differentiation
capacity on CNS stem cells in vitro. Neuron 40, 485 499.
Gage, F.H., 2000. Mammalian neural stem cells. Science 287, 1433 1438.
Garcia-Verdugo, J.M., Doetsch, F., Wichterle, H., Lim, D.A., Alvarez-
Buylla, A., 1998. Architecture and cell types of the adult subventricular
zone: in search of the stem cells. J. Neurobiol. 36, 234 248.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148146
Page 16
Garcion, E., Halilagic, A., Faissner, A., ffrench-Constant, C., 2004.
Generation of an environmental niche for neural stem cell development
by the extracellular matrix molecule tenascin C. Development 131,
3423 3432.
Gates, M.A., Thomas, L.B., Howard, E.M., Laywell, E.D., Sajin, B.,
Faissner, A., Gotz, B., Silver, J., Steindler, D.A., 1995. Cell and
molecular analysis of the developing and adult mouse subventricular
zone of the cerebral hemispheres. J. Comp. Neurol. 361, 249 266.
Ge, H., Roeder, R.G., 1994. The high mobility group protein HMG1 can
reversibly inhibit class II gene transcription by interaction with the
TATA-binding protein. J. Biol. Chem. 269, 17136 17140.
Geschwind, D.H., Ou, J., Easterday, M.C., Dougherty, J.D., Jackson, R.L.,
Chen, Z., Antoine, H., Terskikh, A., Weissman, I.L., Nelson, S.F.,
Kornblum, H.I., 2001. A genetic analysis of neural progenitor
differentiation. Neuron 29, 325 339.
Goldman, S., 2003. Glia as neural progenitor cells. Trends Neurosci. 26,
590 596.
Good, P.J., Chen, Q., Warner, S.J., Herring, D.C., 2000. A family of human
RNA-binding proteins related to the Drosophila Bruno translational
regulator. J. Biol. Chem. 275, 28583 28592.
Grabowski, P.J., Black, D.L., 2001. Alternative RNA splicing in the
nervous system. Prog. Neurobiol. 65, 289 308.
Gritti, A., Parati, E.A., Cova, L., Frolichsthal, P., Galli, R., Wanke, E.,
Faravelli, L., Morassutti, D.J., Roisen, F., Nickel, D.D., Vescovi, A.L.,
1996. Multipotential stem cells from the adult mouse brain proliferate
and self-renew in response to basic fibroblast growth factor. J. Neurosci.
16, 1091 1100.
Grizzi, F., Chiriva-Internati, M., Franceschini, B., Bumm, K., Colombo, P.,
Ciccarelli, M., Donetti, E., Gagliano, N., Hermonat, P.L., Bright, R.K.,
Gioia, M., Dioguardi, N., Kast, W.M., 2004. Sperm protein 17 is
expressed in human somatic ciliated epithelia. J. Histochem. Cytochem.
52, 549 554.
Gurok, U., Steinhoff, C., Lipkowitz, B., Ropers, H.H., Scharff, C., Nuber,
U.A., 2004. Gene expression changes in the course of neural progenitor
cell differentiation. J. Neurosci. 24, 5982 6002.
Hack, M.A., Sugimori, M., Lundberg, C., Nakafuku, M., Gotz, M., 2004.
Regionalization and fate specification in neurospheres: the role of Olig2
and Pax6. Mol. Cell. Neurosci. 25, 664 678.
Hanson, R.D., Hess, J.L., Yu, B.D., Ernst, P., van Lohuizen, M., Berns, A.,
van der Lugt, N.M., Shashikant, C.S., Ruddle, F.H., Seto, M.,
Korsmeyer, S.J., 1999. Mammalian Trithorax and polycomb-group
homologues are antagonistic regulators of homeotic development. Proc.
Natl. Acad. Sci. U. S. A. 96, 14372 14377.
Hartmann, A.M., Nayler, O., Schwaiger, F.W., Obermeier, A., Stamm, S.,
1999. The interaction and colocalization of Sam68 with the splicing-
associated factor YT521-B in nuclear dots is regulated by the Src family
kinase p59(fyn). Mol. Biol. Cell 10, 3909 3926.
Holmberg, J., Clarke, D.L., Frisen, J., 2000. Regulation of repulsion
versus adhesion by different splice forms of an Eph receptor. Nature
408, 203 206.
Hou, V.C., Lersch, R., Gee, S.L., Ponthier, J.L., Lo, A.J., Wu, M., Turck,
C.W., Koury, M., Krainer, A.R., Mayeda, A., Conboy, J.G., 2002.
Decrease in hnRNP A/B expression during erythropoiesis mediates a
pre-mRNA splicing switch. EMBO J. 21, 6195 6204.
Hsieh, J., Gage, F.H., 2004. Epigenetic control of neural stem cell fate.
Curr. Opin. Genet. Dev. 14, 461 469.
Imura, T., Kornblum, H.I., Sofroniew, M.V., 2003. The predominant
neural stem cell isolated from postnatal and adult forebrain but
not early embryonic forebrain expresses GFAP. J. Neurosci. 23,
2824 2832.
Irizarry, R.A., Hobbs, B., Collin, F., Beazer-Barclay, Y.D., Antonellis, K.J.,
Scherf, U., Speed, T.P., 2003. Exploration, normalization, and summa-
ries of high density oligonucleotide array probe level data. Biostatistics
4, 249 264.
Ivanova, N.B., Dimos, J.T., Schaniel, C., Hackney, J.A., Moore, K.A.,
Lemischka, I.R., 2002. A stem cell molecular signature. Science 298,
601 604.
Jankovski, A., Sotelo, C., 1996. Subventricular zone-olfactory bulb
migratory pathway in the adult mouse: Cellular composition and
specificity as determined by heterochronic and heterotopic transplanta-
tion. J. Comp. Neurol. 371, 376 396.
Jolliffe, I., 2003. Principle Component Analysis. Springler-Verlag, New
York.
Jurica, M.S., Moore, M.J., 2003. Pre-mRNA splicing: awash in a sea of
proteins. Mol. Cell 12, 5 14.
Karsten, S.L., Kudo, L.C., Jackson, R., Sabatti, C., Kornblum, H.I.,
Geschwind, D.H., 2003. Global analysis of gene expression in neural
progenitors reveals specific cell-cycle, signaling, and metabolic net-
works. Dev. Biol. 261, 165 182.
Khatri, P., Draghici, S., Ostermeier, G.C., Krawetz, S.A., 2002. Profiling
gene expression using onto-express. Genomics 79, 266 270.
Khatri, P., Bhavsar, P., Bawa, G., Draghici, S., 2004. Onto-Tools: an
ensemble of web-accessible, ontology-based tools for the functional
design and interpretation of high-throughput gene expression experi-
ments. Nucleic. Acids Res. 32, W449 W456.
Kukekov, V.G., Laywell, E.D., Suslov, O., Davies, K., Scheffler, B.,
Thomas, L.B., O’Brien, T.F., Kusakabe, M., Steindler, D.A., 1999.
Multipotent stem/progenitor cells with similar properties arise from
two neurogenic regions of adult human brain. Exp. Neurol. 156,
333 344.
Kuroda, Y., Oma, Y., Nishimori, K., Ohta, T., Harata, M., 2002. Brain-
specific expression of the nuclear actin-related protein ArpNalpha and
its involvement in mammalian SWI/SNF chromatin remodeling
complex. Biochem. Biophys. Res. Commun. 299, 328 334.
Ladd, A.N., Charlet, N., Cooper, T.A., 2001. The CELF family of RNA
binding proteins is implicated in cell-specific and developmentally
regulated alternative splicing. Mol. Cell. Biol. 21, 1285 1296.
Laywell, E.D., Rakic, P., Kukekov, V.G., Holland, E.C., Steindler, D.A.,
2000. Identification of a multipotent astrocytic stem cell in the
immature and adult mouse brain. Proc. Natl. Acad. Sci. U. S. A. 97,
13883 13888.
Lemkine, G.F., Raji, A., Alfama, G., Turque, N., Hassani, Z., Alegria-
Prevot, O., Samarut, J., Levi, G., Demeneix, B.A., 2005. Adult neural
stem cell cycling in vivo requires thyroid hormone and its alpha
receptor. FASEB J. 19, 863 865.
Lim, D.A., Alvarez-Buylla, A., 1999. Interaction between astrocytes and
adult subventricular zone precursors stimulates neurogenesis. Proc.
Natl. Acad. Sci. U. S. A. 96, 7526 7531.
Lim, D.A., Tramontin, A.D., Trevejo, J.M., Herrera, D.G., Garcia-Verdugo,
J.M., Alvarez-Buylla, A., 2000. Noggin antagonizes BMP signaling to
create a niche for adult neurogenesis. Neuron 28, 713 726.
Lois, C., 1996. Long Distance Neuronal Migration in the Adult Mammalian
Brain. Rockefeller, New York, pp. 160.
Lois, C., Alvarez-Buylla, A., 1994. Long-distance neuronal migration in the
adult mammalian brain. Science 264, 1145 1148.
Luskin, M.B., 1993. Restricted proliferation and migration of postnatally
generated neurons derived from the forebrain subventricular zone.
Neuron 11, 173 189.
Luskin, M.B., 1998. Neuroblasts of the postnatal mammalian forebrain:
their phenotype and fate. J. Neurobiol. 36, 221 233.
Matter, N., Herrlich, P., Konig, H., 2002. Signal-dependent regulation of
splicing via phosphorylation of Sam68. Nature 420, 691 695.
McBride, A.E., Silver, P.A., 2001. State of the arg: protein methylation at
arginine comes of age. Cell 106, 5 8.
Milne, T.A., Briggs, S.D., Brock, H.W., Martin, M.E., Gibbs, D., Allis,
C.D., Hess, J.L., 2002. MLL targets SET domain methyltransferase
activity to Hox gene promoters. Mol. Cell 10, 1107 1117.
Molofsky, A.V., Pardal, R., Iwashita, T., Park, I.K., Clarke, M.F., Morrison,
S.J., 2003. Bmi-1 dependence distinguishes neural stem cell self-
renewal from progenitor proliferation. Nature 425, 962 967.
Morshead, C.M., Reynolds, B.A., Craig, C.G., McBurney, M.W., Staines,
W.A., Morassutti, D., Weiss, S., Van der Kooy, D., 1994. Neural stem
cells in the adult mammalian forebrain: a relatively quiescent
subpopulation of subependymal cells. Neuron 13, 1071 1082.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148 147
Page 17
Naef, F., Magnasco, M.O., 2003. Solving the riddle of the bright
mismatches: labeling and effective binding in oligonucleotide arrays.
Phys. Rev., E Stat. Nonlinear Soft Matter Phys. 68, 011906.
Najbauer, J., Leon, M., 1995. Olfactory experience modulated apoptosis in
the developing olfactory bulb. Brain Res. 674, 245 251.
Nemeth, M.J., Curtis, D.J., Kirby, M.R., Garrett-Beal, L.J., Seidel, N.E.,
Cline, A.P., Bodine, D.M., 2003. Hmgb3: an HMG-box family member
expressed in primitive hematopoietic cells that inhibits myeloid and B-
cell differentiation. Blood 102, 1298 1306.
Nemeth, M.J., Cline, A.P., Anderson, S.M., Garrett-Beal, L.J., Bodine,
D.M., 2005. Hmgb3 deficiency deregulates proliferation and differen-
tiation of common lymphoid and myeloid progenitors. Blood 105,
627 634.
Olave, I.A., Reck-Peterson, S.L., Crabtree, G.R., 2002. Nuclear actin and
actin-related proteins in chromatin remodeling. Annu. Rev. Biochem.
71, 755 781.
Orlando, V., 2003. Polycomb, epigenomes, and control of cell identity. Cell
112, 599 606.
Parras, C.M., Galli, R., Britz, O., Soares, S., Galichet, C., Battiste, J.,
Johnson, J.E., Nakafuku, M., Vescovi, A., Guillemot, F., 2004. Mash1
specifies neurons and oligodendrocytes in the postnatal brain. EMBO J.
23, 4495 4505.
Pennartz, S., Belvindrah, R., Tomiuk, S., Zimmer, C., Hofmann, K.,
Conradt, M., Bosio, A., Cremer, H., 2004. Purification of neuronal
precursors from the adult mouse brain: comprehensive gene expression
analysis provides new insights into the control of cell migration,
differentiation, and homeostasis. Mol. Cell. Neurosci. 25, 692 706.
Peretto, P., Merighi, A., Fasolo, A., Bonfanti, L., 1997. Glial tubes in the
rostral migratory stream of the adult rat. Brain Res. Bull. 42, 9 21.
Peretto, P., Dati, C., De Marchis, S., Kim, H.H., Ukhanova, M., Fasolo, A.,
Margolis, F.L., 2004. Expression of the secreted factors noggin and
bone morphogenetic proteins in the subependymal layer and olfactory
bulb of the adult mouse brain. Neuroscience 128, 685 696.
Petreanu, L., Alvarez-Buylla, A., 2002. Maturation and death of adult-born
olfactory bulb granule neurons: role of olfaction. J. Neurosci. 22,
6106 6113.
Ramalho-Santos, M., Yoon, S., Matsuzaki, Y., Mulligan, R.C., Melton,
D.A., 2002. ‘Stemness’: transcriptional profiling of embryonic and
adult stem cells. Science 298, 597 600.
Rasmussen, T.P., 2003. Embryonic stem cell differentiation: a chromatin
perspective. Reprod. Biol. Endocrinol. 1, 100.
Redmond, L., Hockfield, S., Morabito, M.A., 1996. The divergent
homeobox gene PBX1 is expressed in the postnatal subven-
tricular zone and interneurons of the olfactory bulb. J. Neurosci.
16, 2972 2982.
Reynolds, B., Weiss, S., 1992. Generation of neurons and astrocytes from
isolated cells of the adult mammalian central nervous system. Science
255, 1707 1710.
Rietze, R.L., Valcanis, H., Brooker, G.F., Thomas, T., Voss, A.K., Bartlett,
P.F., 2001. Purification of a pluripotent neural stem cell from the adult
mouse brain. Nature 412, 736 739.
Santa-Olalla, J., Baizabal, J.M., Fregoso, M., del Carmen Cardenas, M.,
Covarrubias, L., 2003. The in vivo positional identity gene expression
code is not preserved in neural stem cells grown in culture. Eur. J.
Neurosci. 18, 1073 1084.
Schoor, M., Schuster-Gossler, K., Roopenian, D., Gossler, A., 1999.
Skeletal dysplasias, growth retardation, reduced postnatal survival,
and impaired fertility in mice lacking the SNF2/SWI2 family member
ETL1. Mech. Dev. 85, 73 83.
Schraets, D., Lehmann, T., Dingermann, T., Marschalek, R., 2003. MLL-
mediated transcriptional gene regulation investigated by gene expres-
sion profiling. Oncogene 22, 3655 3668.
Scott, H.S., Antonarakis, S.E., Lalioti, M.D., Rossier, C., Silver, P.A.,
Henry, M.F., 1998. Identification and characterization of two putative
human arginine methyltransferases (HRMT1L1 and HRMT1L2).
Genomics 48, 330 340.
Shimogori, T., VanSant, J., Paik, E., Grove, E.A., 2004. Members of the
Wnt, Fz, and Frp gene families expressed in postnatal mouse cerebral
cortex. J. Comp. Neurol. 473, 496 510.
Shykind, B.M., Kim, J., Sharp, P.A., 1995. Activation of the TFIID-TFIIA
complex with HMG-2. Genes Dev. 9, 1354 1365.
Spassky, N., Merkle, F.T., Flames, N., Tramontin, A.D., Garcia-Verdugo,
J.M., Alvarez-Buylla, A., 2005. Adult ependymal cells are postmitotic
and are derived from radial glial cells during embryogenesis. J. Neurosci.
25, 10 18.
Stenman, J., Toresson, H., Campbell, K., 2003. Identification of two distinct
progenitor populations in the lateral ganglionic eminence: implications
for striatal and olfactory bulb neurogenesis. J. Neurosci. 23, 167 174.
Sterner, D.E., Berger, S.L., 2000. Acetylation of histones and transcription-
related factors. Microbiol. Mol. Biol. Rev. 64, 435 459.
Stump, G., Durrer, A., Klein, A.L., Lutolf, S., Suter, U., Taylor, V., 2002.
Notch1 and its ligands Delta-like and Jagged are expressed and active in
distinct cell populations in the postnatal mouse brain. Mech. Dev. 114,
153 159.
Thomas, J.O., 2001. HMG1 and 2: architectural DNA-binding proteins.
Biochem. Soc. Trans. 29, 395 401.
Thomas, L.B., Gates, M.A., Steindler, D.A., 1996. Young neurons from the
adult subependymal zone proliferate and migrate along an astrocyte,
extracellular matrix-rich pathway. Glia 17, 1 14.
Tramontin, A.D., Garcia-Verdugo, J.M., Lim, D.A., Alvarez-Buylla, A.,
2003. Postnatal development of radial glia and the ventricular zone
(VZ): a continuum of the neural stem cell compartment. Cereb. Cortex
13, 580 587.
Trievel, R.C., 2004. Structure and function of histone methyltransferases.
Crit. Rev. Eukaryotic Gene Expression 14, 147 169.
Ule, J., Ule, A., Spencer, J., Williams, A., Hu, J.S., Cline, M., Wang, H.,
Clark, T., Fraser, C., Ruggiu, M., Zeeberg, B.R., Kane, D., Weinstein,
J.N., Blume, J., Darnell, R.B., 2005. Nova regulates brain-specific
splicing to shape the synapse. Nat. Genet. 37, 844 852.
Verdi, J.M., Bashirullah, A., Goldhawk, D.E., Kubu, C.J., Jamali, M.,
Meakin, S.O., Lipshitz, H.D., 1999. Distinct human NUMB isoforms
regulate differentiation vs. proliferation in the neuronal lineage. Proc.
Natl. Acad. Sci. U. S. A. 96, 10472 10476.
Wen, T., Gu, P., Minning, T.A., Wu, Q., Liu, M., Chen, F., Liu, H., Huang,
H., 2002. Microarray analysis of neural stem cell differentiation in the
striatum of the fetal rat. Cell. Mol. Neurobiol. 22, 407 416.
Wilkinson, D.G. (Ed.), 1999. In Situ Hybridization: A Practical Approach.
Oxford Univ. Press, Oxford.
Wright, L.S., Li, J., Caldwell, M.A., Wallace, K., Johnson, J.A., Svendsen,
C.N., 2003. Gene expression in human neural stem cells: effects of
leukemia inhibitory factor. J. Neurochem. 86, 179 195.
Wu, Z., Irizarry, R.A., Gentleman, R., Martinez-Murillo, F., Spencer, F.,
2004. A model based background adjustment for oligonucleotide
expression arrays. J. Am. Stat. Assoc. 99, 909 917.
Xia, Z.B., Anderson, M., Diaz, M.O., Zeleznik-Le, N.J., 2003. MLL
repression domain interacts with histone deacetylases, the polycomb
group proteins HPC2 and BMI-1, and the corepressor C-terminal-
binding protein. Proc. Natl. Acad. Sci. U. S. A. 100, 8342 8347.
Yu, B.D., Hanson, R.D., Hess, J.L., Horning, S.E., Korsmeyer, S.J., 1998.
MLL, a mammalian trithorax-group gene, functions as a transcriptional
maintenance factor in morphogenesis. Proc. Natl. Acad. Sci. U. S. A.
95, 10632 10636.
Zeisig, B.B., Milne, T., Garcia-Cuellar, M.P., Schreiner, S., Martin,
M.E., Fuchs, U., Borkhardt, A., Chanda, S.K., Walker, J., Soden,
R., Hess, J.L., Slany, R.K., 2004. Hoxa9 and Meis1 are key targets
for MLL-ENL-mediated cellular immortalization. Mol. Cell. Biol.
24, 617 628.
Zhou, F.C., Duguid, J.R., Edenberg, H.J., McClintick, J., Young, P., Nelson,
P., 2001. DNA microarray analysis of differential gene expression of 6-
year-old rat neural striatal progenitor cells during early differentiation.
Restor. Neurol. Neurosci. 18, 95 104.
Zhu, H., Wang, Z.Y., Hansson, H.A., 2003. Visualization of proliferating
cells in the adult mammalian brain with the aid of ribonucleotide
reductase. Brain Res. 977, 180 189.
D.A. Lim et al. / Mol. Cell. Neurosci. 31 (2006) 131 148148
Page 18
    • "In V-SVZ cells, Sox9 is targeted by miR-124, and Sox9 down-regulation is required for neurogenesis, suggesting that the posttranscriptional regulation of Sox9 by miR- 124 is an important aspect of adult neurogenesis. Interestingly, in embryonic brain NSCs, miR-124 (and miR-9) has been shown to be involved in the posttranscriptional regulation of Baf53a (Yoo et al. 2009), and expression of this BAF complex subunit is prominent in the V-SVZ transcriptome (Lim et al. 2006). "
    [Show abstract] [Hide abstract] ABSTRACT: A large population of neural stem/precursor cells (NSCs) persists in the ventricular-subventricular zone (V-SVZ) located in the walls of the lateral brain ventricles. V-SVZ NSCs produce large numbers of neuroblasts that migrate a long distance into the olfactory bulb (OB) where they differentiate into local circuit interneurons. Here, we review a broad range of discoveries that have emerged from studies of postnatal V-SVZ neurogenesis: the identification of NSCs as a subpopulation of astroglial cells, the neurogenic lineage, new mechanisms of neuronal migration, and molecular regulators of precursor cell proliferation and migration. It has also become evident that V-SVZ NSCs are regionally heterogeneous, with NSCs located in different regions of the ventricle wall generating distinct OB interneuron subtypes. Insights into the developmental origins and molecular mechanisms that underlie the regional specification of V-SVZ NSCs have also begun to emerge. Other recent studies have revealed new cell-intrinsic molecular mechanisms that enable lifelong neurogenesis in the V-SVZ. Finally, we discuss intriguing differences between the rodent V-SVZ and the corresponding human brain region. The rapidly expanding cellular and molecular knowledge of V-SVZ NSC biology provides key insights into postnatal neural development, the origin of brain tumors, and may inform the development regenerative therapies from cultured and endogenous human neural precursors.
    No preview · Article · Apr 2016 · Cold Spring Harbor perspectives in biology
  • Source
    • "Epigenetic regulatory pathways could play a key role by imparting stable states of transcriptional activity. It is now clear that epigenetic processes play critical roles in activitydependent regulation of gene expression, and are required for adult neurogenesis, synaptic plasticity, and memory formation, consolidation and extinction (Lim et al., 2006; Miller and Sweatt, 2007; Schor et al., 2009; Feng et al., 2010; Gräff et al., 2012; Cortés-Mendoza et al., 2013; Day et al., 2013). Dnmt1 and Dnmt3a mRNA, protein, and activity are reduced by neuronal membrane depolarization (Sharma et al., 2008), and contextual fear conditioning increases Dnmt3a and Dnmt3b expression in the hippocampus (Miller and Sweatt, 2007). "
    [Show abstract] [Hide abstract] ABSTRACT: Cytosine DNA methylation is a stable epigenetic modification with established roles in regulating transcription, imprinting, female X-chromosome inactivation, and silencing of transposons. Dynamic gain or loss of DNA methylation reshapes the genomic landscape of cells during early differentiation, and in post-mitotic mammalian brain cells these changes continue to accumulate throughout the phases of cortical maturation in childhood and adolescence. There is also evidence for dynamic changes in the methylation status of specific genomic loci during the encoding of new memories, and these epigenome dynamics could play a causal role in memory formation. However, the mechanisms that may dynamically regulate DNA methylation in neurons during memory formation and expression, and the function of such epigenomic changes in this context, are unclear. Here we discuss the possible roles of DNA methylation in encoding and retrieval of memory.
    Full-text · Article · Aug 2015 · Frontiers in Neuroscience
  • Source
    • "HMGB2 mRNA was previously reported to be present in the SVZ of P90 adult mice [26]. We verified expression of HMGB2 protein in SVZ progenitor cells in brain sections of young adult Nestin-GFP+ transgenic mice (Fig. S1A–C). "
    [Show abstract] [Hide abstract] ABSTRACT: Neural stem and progenitor cells (NSCs/NPCs) are distinct groups of cells found in the mammalian central nervous system (CNS). Previously we determined that members of the High Mobility Group (HMG) B family of chromatin structural proteins modulate NSC proliferation and self-renewal. Among them HMGB2 was found to be dynamically expressed in proliferating and differentiating NSCs, suggesting that it may regulate NSC maintenance. We report now that Hmgb2(-/-) mice exhibit SVZ hyperproliferation, increased numbers of SVZ NSCs, and a trend towards aberrant increases in newly born neurons in the olfactory bulb (OB) granule cell layer. Increases in the levels of the transcription factor p21 and the Neural cell adhesion molecule (NCAM), along with down-regulation of the transcription/pluripotency factor Oct4 in the Hmgb2-/- SVZ point to a possible pathway for this increased proliferation/differentiation. Our findings suggest that HMGB2 functions as a modulator of neurogenesis in young adult mice through regulation of NSC proliferation, and identify a potential target via which CNS repair could be amplified following trauma or disease-based neuronal degeneration.
    Full-text · Article · Dec 2013 · PLoS ONE
Show more