The repressor element-1 silencing transcription factor/neuron-restrictive silencer factor (REST/NRSF) is a master regulator of neuronal gene expression. REST functions as a modular scaffold for dynamic recruitment of epigenetic regulatory factors including its primary cofactor, the corepressor for element-1-silencing transcription factor (CoREST), to genomic loci that contain the repressor element-1 (RE1) binding motif. While REST was initially believed to silence RE1 containing neuronal genes in neural stem cells (NSCs) and non-neuronal cells, emerging evidence shows an increasingly complex cell type- and developmental stage-specific repertoire of REST target genes and functions that include regulation of neuronal lineage maturation and plasticity.
In this study, we utilized chromatin immunoprecipitation on chip (ChIP-chip) analysis to examine REST and CoREST functions during NSC-mediated specification of cholinergic neurons (CHOLNs), GABAergic neurons (GABANs), glutamatergic neurons (GLUTNs), and medium spiny projection neurons (MSNs). We identified largely distinct but overlapping profiles of REST and CoREST target genes during neuronal subtype specification including a disproportionately high percentage that are exclusive to each neuronal subtype.
Our findings demonstrate that the differential deployment of REST and CoREST is an important regulatory mechanism that mediates neuronal subtype specification by modulating specific gene networks responsible for inducing and maintaining neuronal subtype identity. Our observations also implicate a broad array of factors in the generation of neuronal diversity including but not limited to those that mediate homeostasis, cell cycle dynamics, cell viability, stress responses and epigenetic regulation.
Citation: Abrajano JJ, Qureshi IA, Gokhan S, Zheng D, Bergman A, Mehler MF (2009) REST and CoREST Modulate Neuronal Subtype Specification, Maturation and Maintenance. PLoS ONE 4(12): e7936. https://doi.org/10.1371/journal.pone.0007936
Editor: Hitoshi Okazawa, Tokyo Medical and Dental University, Japan
Received: August 4, 2009; Accepted: October 18, 2009; Published: December 7, 2009
Copyright: © 2009 Abrajano et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: A. B. is supported by grants from the National Institutes of Health (AG028872, AG027734) and M.F.M. is supported by grants from the National Institutes of Health (NS38902, MH66290), as well as by the F.M. Kirby, the Rosanne H. Silbermann, the Alpern Family, the Mildred and Bernard H. Kayden and the Roslyn and Leslie Goldstein Foundations. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The repressor element-1 silencing transcription factor/neuron-restrictive silencer factor (REST/NRSF) is a transcriptional regulator with genome-wide effects important for orchestrating neuronal development . REST binds to repressor element-1 (RE1) consensus motifs  which are primarily located in promoter regions of genes responsible for fundamental mature neuronal traits including: ion channels, adhesion molecules, synaptic vesicle proteins, growth factors and hormones, axonal guidance and vesicle trafficking, and neurotransmitter receptors , , . While REST was initially believed to repress transcription of these neuronal genes in neural stem cells (NSCs) and in non-neuronal cells, recent evidence suggests a much broader role for REST, with context-specific and sometimes seemingly paradoxical functions in embryonic stem cells (ESCs), NSCs, mature neurons, and other cell types , , .
The differential roles played by REST depend on its ability to recruit a series of epigenetic and regulatory cofactors to its N- and C-terminal domains. These highly plastic macromolecular complexes often include another important transcriptional regulator, the corepressor for element-1-silencing transcription factor (CoREST), the primary cofactor of REST . Like REST, CoREST recruits additional epigenetic factors similarly associated with gene activation and repression, including methyl CpG binding proteins (e.g., MeCP2), histone deacetylases (e.g., HDAC1/2), histone modifying enzymes (e.g., LSD1, EHMT2/G9a, and SUV39H1), and components of SWI-SNF chromatin remodeling complexes (e.g., BAF57, BRG1, and BAF170) , , , .
Various studies have begun to characterize the specific roles played by REST and CoREST complexes during NSC-mediated neuronal lineage specification and maturation. For example, distinct REST and CoREST complexes were found to regulate target gene expression in post-mitotic neurons , . In these studies, for a subset of targets designated as class I neuronal genes, a maximum level of expression was observed with REST complex dissociation from the gene promoter. For class II genes, a submaximal level of expression was found when the REST complex dissociated from the gene promoter due to the presence of a separate CoREST complex at a different promoter site . In addition, other studies have also started to describe the specific roles played by cofactors in these complexes. During adult hippocampal neurogenesis, REST is converted from a transcriptional repressor into an activator by a small modulatory double stranded RNA (dsRNA) , . Moreover, a truncated isoform of REST, REST4, exerts a dominant-negative effect on REST and possibly derepresses or activates expression of RE-1 containing genes in neurons . Additional studies have identified cell type-specific profiles of REST target genes and suggested that many REST target genes and functions have yet to be discovered , , , .
In this study, we examined the roles of REST and CoREST during NSC-mediated neuronal subtype specification and maintenance. While these processes are critical for establishing functional diversity, homeostasis, and neural network connectivity and plasticity within the nervous system , the regulatory circuitry responsible for governing the elaboration of different neuronal subtypes remains poorly understood. We found specific REST and CoREST target gene profiles in neuronal subtypes, including cholinergic neurons (CHOLNs), GABAergic neurons (GABANs), glutamatergic neurons (GLUTNs), and medium spiny projection neurons (MSNs)], including those not previously associated with the establishment of neuronal subtype identity, such as particular epigenetic factors, cell cycle regulators and homeostatic modulators. These observations suggest that the complex gene networks underlying neuronal subtype specification and maintenance are selectively modulated by the differential deployment of REST and CoREST.
We examined the molecular mechanisms that underlie neuronal subtype specification by characterizing REST and CoREST protein expression and target gene profiles using ChIP-chip analysis in a developmental paradigm associated with the elaboration of mature neuronal subtypes. We identified REST and CoREST target genes that are unique to a specific neuronal subtype and also those found in multiple subtypes. In addition, we interrogated the potential roles of REST and CoREST in regulating gene expression during neuronal subtype specification by examining correlations between profiles of REST and CoREST promoter occupancy and corresponding target gene expression patterns during critical developmental transitions.
Identification of Dorsal and Ventral Forebrain Developmental and Mature Cellular Species by the Expression of Selective Neural Lineage Markers and Associated Transcription Factor Codes
Both dorsal and ventral NSCs are identified by their expression of the neuroepithelial marker, nestin in the absence of expression of other lineage markers associated with intermediate neural progenitors and neuronal or glial species (Figure S1). Radial glial cells derived from dorsal forebrain NSCs continued to express nestin while acquiring RC2  (Figure S1). N/OPs are defined by their expression of two basic helix-loop-helix transcription factors, Olig2 and Mash1  in addition to presence of the neuropeithelial marker, nestin (Figure S1). GLUTN derived from RG cells ,  acquired expression of the early neuronal linage marker, β tubulin III and exhibited complete overlap with the excitatory neurotransmitter, glutamate, while ventral forebrain derived neuronal species were identified by their co-expression of β tubulin III in concert with acquisition of distinct neurotransmitters (GABA for GABAN) or substrates and enzymes required for the processing and synthesis of neuronal subtype specific neurotransmitters (DARPP32, MSN and ChAT, CHOLN, respectively)  (Figure S1).
Determination of REST and CoREST Expression and Subcellular Localization
We characterized REST and CoREST protein expression and localization during the process of neuronal subtype specification using immunofluorescence microscopy and Western blot analysis (Figures S2, S3, S4). We found that both proteins displayed more nuclear expression in NSCs and the intermediate progenitors, RGs and N/OPs, while being differentially distributed to the nucleus and cytoplasm of all neuronal subtypes examined (Figures S2, S3, S4). These observations suggest that both REST and CoREST are present in all neuronal subtypes and therefore have the potential to regulate gene expression during neuronal subtype specification.
Genome-Wide Promoter Occupancy Profiles for REST and CoREST
While REST and CoREST expression and localization remain relatively constant, there is significant variation in the number of gene promoter sites occupied across the different neuronal subtypes (Figure 1A). We found a total of 2,123 REST and 2,384 CoREST target genes across all neuronal cell types examined. Amongst these genes, we found that 1,453 are “exclusive” targets of REST, 1,714 are exclusive targets of CoREST, and 670 are targets of both REST and CoREST (REST-CoREST). We designated these genes as “unique” targets if they were bound only within one of the neuronal subtypes examined. For CHOLNs, GABANs, GLUTNs, and MSNs, we identified 622, 587, 481, and 477 unique and exclusive REST target genes; and we found 600, 814, 266, and 967 unique and exclusive CoREST target genes. For CHOLNs, GABANs, GLUTNs, and MSNs, we further identified 54, 97, 36, and 58 unique REST-CoREST target genes. These observations show that profiles of REST, CoREST, and REST-CoREST target genes are largely unique and vary significantly in number across different neuronal subtypes, suggesting that REST and CoREST may perform diverse and differential functions during neuronal subtype specification.
(A) The number of exclusive REST, exclusive CoREST, and REST and CoREST (REST-CoREST) target genes uniquely present in individual neuronal subtypes as identified through chromatin immunoprecipitation on promoter chip (ChIP-chip) experiments. We identified 2,123 total REST target genes and 2,384 total CoREST target genes. (B) The percentages of REST, CoREST, or REST-CoREST target genes in individual neuronal subtypes that contain previously characterized repressor element-1 (RE1) motifs , .
We also compared the REST and CoREST target genes we identified across the four neuronal subtypes with a previously identified set of canonical and non-canonical RE1 sequence containing genes (Figure 1B) , . Interestingly, for the subgroups of genes that are REST, CoREST, and REST-CoREST targets, the percentages of RE1 containing genes range only from 30 to 40%. Our results also show that, in these neuronal subtypes, REST and CoREST principally target genes that have not previously been found to contain RE1 motifs. These observations suggest that REST and CoREST genomic interactions are highly complex and provide further support for the existence of a neuronal lineage-specific regulatory hierarchy for REST and CoREST that employs differential DNA binding motifs to encode functional information .
To begin characterizing the roles of REST and CoREST in modulating neuronal subtype specific gene networks, we compared the profiles of REST and CoREST target genes from all neuronal subtypes to those from a corresponding study we performed in a subset of immature neural and macroglial cell types (Figure 2). Unexpectedly, we observed that these composite profiles are quite distinct from each other. We found that only 13% of REST target genes in neuronal subtypes are also REST targets in the other cell types, 14% of CoREST target genes in neuronal subtypes are also CoREST targets in the other cell types, and 5% REST-CoREST target genes in neuronal subtypes are also REST-CoREST targets in the other cell types.
The distinct and overlapping profiles of REST, CoREST and REST-CoREST target genes present in neuronal subtypes and in all other neural cell types. We compared profiles of REST and CoREST target genes in neuronal subtypes [cholinergic neurons (CHOLNs), medium spiny projection neurons (MSNs), GABAergic neurons (GABANs), glutamatergic neurons (GLUTNs)] with other neural cell types [neural stem cells, neuronal/oligodendrocyte (OL) precursors, radial glia, astrocytes, OL precursors, OL progenitors, post-mitotic OLs, and myelinating OLs]. Note the relatively small number of REST and CoREST target genes found in common between these neural species.
Pathway Analysis of REST and CoREST Target Genes
To compare the functional roles of REST and CoREST, we analyzed profiles of REST and CoREST target genes in neuronal subtypes using Ingenuity Pathway Analysis and found a myriad of enriched pathways that may play roles in neuronal subtype elaboration (Table 1). For exclusive REST target genes, these pathways include hypoxia signaling. In contrast, for exclusive CoREST target genes, these pathways include CD28 signaling, chemokine signaling, glucocorticoid receptor signaling, and Huntington's disease signaling. Also, for exclusive REST target genes and exclusive CoREST target genes, these pathways include the protein ubiquitination and estrogen receptor signaling. Finally, for REST-CoREST target genes, these pathways include tight junction signaling, induction of apoptosis, protein ubiquitination, PTEN signaling, and neuregulin signaling. These results indicate that REST, CoREST, and REST-CoREST coordinately regulate a diverse array of neuronal homeostasis pathways involved in stress responses, cell cycle progression, and trophic factor signaling and suggest that REST and CoREST perform similar functional roles through different subsets of target genes.
Pathway Analysis of Target Genes in Individual Neuronal Subtypes
To further examine the distinct and overlapping roles for REST and CoREST within neuronal subtypes, we examined REST and CoREST target genes in four mature neuronal subtypes (CHOLNs, GABANs, GLUTNs, and MSNs) (Table 2). We found that REST target genes are involved in a diverse set of biological functions. Specifically, in CHOLNs, these pathways include IL-17 signaling and C21-steroid hormone metabolism. In GABANs, these pathways include p38 MAPK signaling and arachidonic acid metabolism. In contrast, CoREST target genes are involved in a more delimited set of biological functions. For example, in MSNs, these pathways include nitrogen metabolism and aminosugar metabolism. In CHOLNs these pathways include riboflavin metabolism and estrogen receptor signaling. Finally, in GABANs, these pathways include complement system and cell cycle: G2/M DNA damage checkpoint regulation.
Roles of REST and CoREST in Neuronal Subtype Identity, Maintenance, and Function
Cholinergic neurons (CHOLNs) are basal forebrain excitatory neurons that provide the major cholinergic input to the neocortex . Our results suggest that a subset of REST and CoREST target genes (Table S1) modulate aspects of CHOLN specification and maintenance, including those involved in gradient morphogen signaling (e.g., Bmp2), cellular processes (e.g., Smad2), transcriptional regulation (e.g., Lhx8), and cell-cell signaling (e.g., Cnr1). Specifically, bone morphogenetic protein (BMP) signaling plays an important role during various stages of neural development including the differentiation of cortical progenitors into distinct neuronal subtypes . While Bmp2 promotes CHOLN specific gene expression in the basal forebrain , Smad2 is essential for ventral forebrain patterning and has been shown to regulate excitability of CHOLNs , , . In addition, REST targets the LIM homeobox transcription factor Lhx8, which is selectively expressed in the embryonic medial ganglionic eminence (MGE), where it is required for CHOLN maintenance in the developing basal forebrain and may also play a role in establishing CHOLNs rather than GABANs , , . Finally, Cnr1 is important for CHOLN specification and long-range axonal patterning . Our cumulative evidence demonstrates that REST and CoREST modulate an array of target genes, including those with roles in CHOLN morphology and maintenance as well as other genes that actively suppress the elaboration of alternate neuronal subtypes.
We identified target genes with diverse roles in forebrain specification and development, including genes involved in transcriptional regulation (e.g., Dlx2, Mib1, and Bhlhb5), neuronal subtype specification (e.g., Neo1 and Gabrg1), and neuroblast proliferation and differentiation (e.g., Rarβ and Elavl2). Dlx2 is a homeobox factor that promotes the specification of GABANs . We also identified Mib1, an ubiquitin ligase necessary for efficient Notch signaling that specifies GABANs through interactions with the Dlx family of homeobox transcription factors . Further, we identified other factors with specific roles in GABAN differentiation, including Neo1, a netrin receptor that plays a role in the tangential migration of cortical GABANs , and Bhlhb5, a basic helix-loop-helix (bHLH) factor that is induced during the intermediate phase of the Ngn2-regulated cortical differentiation cascade that determines the decision to differentiate between GLUTNs and GABANs . Finally, besides specific roles in GABANs, we also found that REST and CoREST mediate neuronal homeostasis programs. Specifically, sequential Rarβ and Rarα signaling can lead to the proliferation and differentiation of forebrain progenitors in concert with Shh and FGF signaling pathways . Elavl2 is a neuronal RNA binding protein that is a target of Ngn2 and is important for controlling the switch from proliferation to neuronal differentiation through the post-transcriptional regulation of p21Cip1/Waf1 , . Our results suggest that REST and CoREST mediate factors directly involved in diverse aspects of GABAN subtype specification and terminal differentiation as well as other factors with roles in neuronal homeostasis programs. These findings are consistent with evolving evidence that homeostatic factors, such as cell cycle regulators, are critical in post-mitotic neuronal functions .
We found target genes critical for GLUTN specification and forebrain patterning, including those involved in neuronal morphogenesis and plasticity (e.g., Ctnnd2), transcriptional regulation (e.g., Tlx1), and cytokine signaling (e.g., Bmp2 and Lif). We identified Ctnnd2, a factor essential for the trafficking and proper localization of glutamate receptors (GluR2) in dendritic spines . Also, Tlx1 is a homeobox transcription factor that promotes the specification of GLUTNs and inhibits the acquisition of GABANs by inhibiting the effects of Lbx1, a GABAN differentiation factor . We also identified genes encoding factors directly involved in the specification of GLUTNs from RG  including BMP2, which promotes lineage restriction of RG into GLUTNs in concert with LIF , an interleukin-6 (IL6) family cytokine.
We further identified target genes with roles in forebrain development, including those involved in neuroblast proliferation (e.g., Pten1), lineage commitment (e.g., Gjc1), maturation (e.g., Dcp1a and Smad2), and synaptogenesis (e.g., Akt1). While Pten promotes endogenous NSC expansion and self-renewal , Gjc is selectively expressed in committed neuronal progenitor cells , , . We also identified mediators of TGF-β signaling, including BMP-related pathways, such as Dcp1a and Smad2, both of which are involved in neural patterning, specification and maturation. Finally, Akt1 is involved in aspects of prefrontal cortex architecture, including neuronal morphology, synaptic function, and axonal myelination . These cumulative results suggest that REST and CoREST regulate factors with specific roles in defining GLUTN specification and maintenance. Our results also suggest that neuronal maturation is intrinsically linked to neuronal subtype specification and that these fundamental developmental processes are, in large part, coordinately regulated by REST and CoREST.
Medium spiny projection neurons.
We identified target genes involved in promoting MSN morphogenesis (e.g., Cobl), maturation and subtype specific functions (e.g., Mash1, Nts and Grm5). Mash1 is a proneural bHLH member that promotes neural precursor cell cycle exit and differentiation . Nts is a neuropeptide expressed in MSNs that is important for striatal activity , and Grm5 is a metabotropic glutamate receptor densely expressed in the spines of MSNs . Intriguingly, cortical glutamate and substantia nigra dopamine (DA) afferents converge onto striatal MSN dendritic spines where they modulate motor and cognitive functions.
We further identified target genes that have roles in neuronal morphogenesis (e.g., Evl and Nefm) and migration (e.g., Plxnb1). Evl, an actin regulatory protein, is highly expressed in the developing cortical plate and regulates cortical neuronal positioning , , while Nefm is a neurofilament subunit that modulates cytoskeletal stability and determines cortical neuronal functions . Also, Plxnb1 is a semaphorin 4D receptor widely expressed in the developing neopallial cortex that regulates neuronal migration . Finally, we found gene targets involved in both pan-neuronal (e.g., Purb and Tbca) and neural (e.g., Egfr) differentiation. Purb, a nucleic acid binding protein similar to Pura, controls DNA replication and transcription, mediates dendritic translocation of ncRNAs and mRNAs, and assists in establishing the postsynaptic compartment in developing neurons . Tbca promotes proper folding of a pan-neuronal protein, β-tubulin . Also, EGFR-mediated signaling pathways appear to have diverse roles in neuronal subtype specification . Specifically, asymmetric distribution of EGFR contributes to forebrain development by creating progenitors with different apoptotic, proliferative, migratory, and differentiation responses to ligand , , , . These cumulative findings suggest that REST and CoREST regulate factors associated with MSN-specific morphology and differentiation. They also indicate that REST and CoREST differentially regulate the deployment of gene networks preferentially involved in mediating broader developmental and homeostatic programs during neuronal subtype specification.
Roles of REST and CoREST in Developmental Processes
We examined both common and unique gene targets for REST and CoREST within mature neuronal subtypes (Figures 3 and 4). Interestingly, the common REST and CoREST targets we identified in all four mature neuronal subtypes are directly or indirectly involved or related to EHMT2/G9a, a H3K9 histone methyltransferase (HMT) primarily associated with heterochromatic regions and recruited directly to the REST complex by CoREST , . Specifically, REST targets EHMT1, a G9a related HMT that mediates gene silencing through mono- and dimethylation of H3K9 , , . This HMT may, in fact, be an integrative factor critical for maintaining epigenetic states that link histone modifications with DNA methylation states previously shown to be mediated by components of the REST repressor complex , . Furthermore, CoREST targets the gene locus for Defb42, a G9a-regulated factor that is a member of the defensin family, which is implicated in mediating innate and acquired immunity and may be important for immune surveillance in the nervous system , .
We examined the distinct and overlapping profiles of REST target genes present in individual neuronal subtypes.
We examined the distinct and overlapping profiles of CoREST target genes present in individual neuronal subtypes.
We also identified a diverse array of REST and CoREST targets involved in epigenetic processes (Table S2), including genes that mediate DNA methylation states (e.g., Mbd6), histone modifications (e.g., Smcx, Jarid1d, Jmjd1a, and Lsd1), and the SWI/SNF family of chromatin remodeling enzymes (e.g., Smarca5, Smarcc1, Smarcc2, and Smarce1). A number of these factors have previously been associated with the REST complex and its functions, which range from maintenance of higher-order chromosomal organization and stability to DNA methylation, nucleosome dynamics, and chromatin remodeling , . Specifically, we identified Jarid1c, a H3K4me3 demethylase recruited by the REST complex that is essential for neuronal survival and dendritic development . Jarid1c mutations are associated with both X-linked MR and autism spectrum disorder , . Moreover, we identified components of dynamic combinatorial SWI/SNF nucleosome remodeling complexes, including subunits of the Brg1/Brm Associated Factors (BAF) complex (e.g., Smarce1 and Smarcc1) and imitation SWI (ISWI) members (e.g., Smarca5, Baz1b, and Smarca1). Interestingly, Smarce1 and Smarcc1 can heterodimerize and associate with SWI/SNF complexes in combination with Smarca4, Smarca2, and Smarcc2. Also, CoREST recruits Smarce1, which has neuron-specific isoforms that are differentially expressed during neurogenesis and are believed to play specific roles in establishing neuronal subtype identity . Similarly, ISWI members play important roles in brain development through effects on transcriptional regulation, heterochromatin replication, chromatin assembly, and chromatin higher-order structure . Furthermore, Snf2h forms a complex with Wstf and is implicated in neural progenitor cell proliferation, while Snf2l has gene-specific effects during neuronal terminal differentiation . These cumulative findings suggest that REST and CoREST mediate deployment of various classes of epigenetic factors in neuronal subtypes, thereby, suggesting important roles for these factors in regulating multi-layered epigenetic cascades during neuronal subtype specification.
Cell cycle, ubiquitination, and apoptosis.
A number of REST and CoREST target genes encode factors involved in cell cycle regulation (Table S3), including Mad2, an essential component of proper spindle checkpoint assembly that is regulated by ubiquitin ligase SCF(beta-TrCP)-dependent degradation of REST during the G2 phase of the cell cycle . We also identified key factors involved in the ubiquitin-proteosome pathway (Table S4), including Cul1, a scaffolding protein for SCF ubiquitin ligases implicated in the regulation of cell cycle progression, synaptic morphology and function, and neuronal apoptosis , . Furthermore, we uncovered a subset of other apoptotic factors (Table S5), including Casp1, a modulator of neuronal apoptosis which acts as part of a viability rheostat , . In concert with recent studies, our results indicate that REST and CoREST coordinate neuronal subtype specific gene expression programs by modulating factors involved in cell cycle regulation  and protein turnover  as well as those that orchestrate the sequential developmental elaboration of mature, post-mitotic neurons.
Neuronal identity factors.
We found that REST and CoREST modulate genes encoding neuronal identity factors that regulate various aspects of forebrain development (Table S6). For example, Jag1 is a Notch ligand that promotes seminal neural fate decisions , suggesting REST and CoREST modulate developmental signaling pathways important for neuronal subtype specification. Jag1 also promotes microtubule stability and axon integrity . In addition, Neo1 is a netrin receptor involved in axon pathfinding and regulation of neuronal differentiation . Our findings thereby suggest a role for REST and CoREST in modulating axonal processes and plasticity in neuronal subtypes, which is supported by previous evidence showing that misexpression of REST causes axon guidance errors .
We identified factors with roles in transcriptional regulation of forebrain patterning. For example, NeuroD4/Math3 is a dorsal bHLH transcription factor that cooperates with the proneural bHLH transcription factor, Neurog2, to orchestrate regional gene expression profiles in the developing cortex . In addition, Bmp5 is a gradient morphogen also involved in neuronal specification in the dorsal forebrain . In contrast, Dlx4 is a homeobox protein that exists in a bigene cluster with Dlx3, which is expressed in the ventral forebrain . These cumulative results suggest that REST and CoREST orchestrate neuronal subtype specification through effects on transcription factors with important roles in both dorsal and ventral regional patterning.
Association of Promoter Occupancy with Developmental Gene Expression Profiles
To assess the potential functional roles played by REST and CoREST in neuronal developmental gene expression programs, we also correlated profiles for REST and CoREST promoter occupancy with differential gene expression patterns. Specifically, we evaluated the developmental transitions of the four neuronal subtypes from their most proximate progenitors: different NSC species into mature neuronal subtypes (CHOLNs and MSNs), radial glia (RG) into GLUTNs, and bipotent neuronal-OL progenitors (N/OPs) into GABANs (Table 3).
We identified genes that displayed either gain or loss of REST or CoREST promoter occupancy and also exhibited differential expression between the two cell types comprising the developmental transition. Among these, we found subsets of genes with gain of REST or CoREST promoter occupancy during the transitions that were associated with both gene activation and repression. Conversely, we also identified subsets of genes with loss of REST or CoREST promoter occupancy during the transitions that were similarly associated with both gene activation and repression. For example, among genes where REST was bound in NSCs but not bound in MSNs, 75 genes were up regulated and 30 were down regulated in MSNs. Also, among genes where CoREST was bound in NSCs but not bound in CHOLNs, 138 genes were up regulated and 67 were down regulated in CHOLNs. In contrast, among genes where REST was not bound in N/OPs but bound in GABANs, 118 genes were up regulated and 160 were down regulated in GABANs. Among genes where CoREST was not bound in NSCs but bound in MSNs, 114 genes were up regulated and 139 were down regulated in MSNs. Our cumulative results suggest that REST and CoREST modulate NSC-mediated neuronal subtype specification through different profiles of REST and CoREST promoter occupancy that mediate a complex repertoire of gene expression programs involved in neuronal homeostasis, cell cycle progression, apoptosis, and neuronal subtype specific transcriptional processes.
The developmental fates of stem and progenitor cells within the nervous system are determined by a complex series of spatial and temporal extracellular cues which induce selective classes of cross-repressive patterning genes and combinatorial transcription factor codes that ultimately lead to cellular functional diversification. We interrogated these molecular mechanisms and found that REST and CoREST appear to play key roles in orchestrating neuronal subtype specific gene expression programs. By characterizing corresponding profiles of genome-wide REST and CoREST promoter occupancy and gene expression in a number of terminally differentiated neuronal subtypes (e.g., CHOLNs, GABANs, GLUTNs, and MSNs), we identified neuronal subtype specific REST and CoREST “regulons”  which seem to be important for promoting features associated with a particular neuronal subtype while concurrently repressing characteristics of alternative neural cell types. These observations support the conclusion that the highly selective REST and CoREST regulons that we uncovered mediate neuronal subtype specification.
Consistent with previous reports, the REST and CoREST target genes that we uncovered encode a broad array of factors, including those which define cell identity and connectivity (e.g., transcription factors, gradient morphogens and cytokines, cell surface molecules, neurotransmitter receptors, and synaptic vesicle proteins) and modulate homeostasis (e.g., metabolic factors, DNA repair enzymes, and cell cycle regulators) , , , . REST and CoREST targeted genes that encode members of key neural developmental signaling pathways, such as FGF, RA, EGF, Notch, BMP, SHH, and WNT. They also bound to the promoters of genes with important neuronal subtype specific roles in CHOLNs (e.g., Bmp2, Smad2, Lhx8, and Cnr1), GABANs (e.g., Bhlhb5, Rarβ, and Elavl2), GLUTNs (e.g., Tlx1, Bmp2, and Lif), and MSNs (e.g., Nts, Grm5, and Mash1). Together, these findings are consistent with emerging evidence demonstrating that REST and CoREST complexes can act with high degrees of context-specificity depending on developmental stage, cell type, and gene locus. While functions of REST have previously been examined in ESCs, certain populations of NSCs, and other cell types, this is the first study to link differential deployment of REST and CoREST to neuronal subtype specification, and it significantly expands our understanding of the mechanisms that mediate this critical process.
Recent reports have suggested that the properties of neuronal subtypes are sculpted by trans-acting terminal selector genes (TSGs), which modulate the expression of cis-regulatory terminal selector motif (TSM) associated genes . In this model, TSGs, such as transcription factors, specify neuronal subtype identity by modulating terminal differentiation gene expression through interactions with their cognate TSMs contained within neuronal subtype specific genes . Despite the recognition of this terminal regulatory logic, the molecular circuitry that explains how different neuronal subtype specific gene networks are dynamically constrained by other types of selector genes has yet to be elucidated.
Like TSGs, REST and CoREST are important nodes in the regulatory circuitry underlying the generation of neuronal diversity, though they have broader effects on neuronal identity and function by controlling a larger cohort of positively and negatively interacting epigenetic modulators, transcription factors, and neuronal differentiation genes. While REST and CoREST are distinct from TSGs, they seem to share a number of properties, and therefore, we suggest that they act as “facultative selector genes” (FSGs). This role for REST and CoREST in orchestrating neuronal subtype specification is supported by their regulation of RE1-associated neuronal target genes as well as additional molecular mechanisms. First, TSGs can function either alone or in synergistic combination with each other , and previous work as well as our current findings indicate that REST and CoREST can function both independently or in tandem as members of a combinatorial complex , , . Second, TSGs perform maintenance functions throughout the life of a neuron in order to maintain the post-mitotic, differentiated state . Our results indicate that both REST and CoREST are expressed in the nucleus of all neuronal subtypes examined, suggesting roles in the maintenance of terminally differentiated neuronal subtypes. Also, the function of TSGs is orchestrated by positive autoregulation (i.e. TSGs maintain their own expression), a process that may mediate REST function because of the presence of an RE1 in the promoter of the REST gene , . TSGs can also repress alternative cell fates by repressing selector genes for other cell types . Our data support a similar role for REST and CoREST as these factors modulate glial genes in neuronal subtypes (e.g., Mobp). Moreover, a number of REST and CoREST targets encode factors that repress alternative neuronal fates (e.g., Lhx8 in CHOLN, Bhlhb5 in GABAN, and Tlx1 in GLUTN), suggesting that REST and CoREST mediate the functional properties of specific neuronal subtypes in a myriad of ways. Finally, our results indicate that the majority of REST and CoREST target genes are unique for each neuronal subtype (Figures 3 and 4), suggesting that a REST and CoREST mediated generic neuronal differentiation program does not exist. Together, these observations imply that REST and CoREST function as FSGs that act upstream of TSGs and have a broad range of effects on neuronal gene networks in order to promote the acquisition and maintenance of neuronal subtype identity while repressing alternative cellular fates .
A more detailed understanding of the properties of REST and CoREST transcriptional networks and their dynamic functions is necessary to further characterize the regulatory logic underlying neuronal subtype specification and also to elucidate their potential roles in diseases with selective vulnerability of neuronal subtypes to neurodegeneration. For example, Huntington's disease (HD) is caused by a polyglutamine expansion repeat in Huntingtin (Htt) and characterized, in part, by abnormal REST trafficking and transcriptional dysregulation. We found that REST and CoREST specifically target genes encoding a number of Htt interacting proteins in distinct neuronal subtypes (CHOLNs, GABANs, and GLUTNs but not MSNs) including Hip2, which is a ubiquitin conjugating enzyme involved in polyglutamine induced cell death, Ift57 (Hippi), which regulates apoptosis, and Prpf40b, which is a spliceosomal factor also associated with methyl CpG binding protein 2 (Mecp2) and Rett syndrome. Neuronal subtype specific REST and CoREST targeting of these factors may be important for mediating the selective vulnerability of certain neuronal subpopulations in HD. Potential roles for REST and CoREST in modulating the specific profiles of neuronal loss in neurodegenerative diseases are further suggested by our finding that REST and CoREST targeted a number of Alzheimer's disease associated genes. These included Blmh, Lrp1, and ApoE, in CHOLN, and A2m, Appbp1, Abca2, Sorl1, and Icam5 in other neuronal subtypes. Moreover, REST and CoREST selectively targeted genes that, when mutated, cause a broader array of diseases characterized by varying patterns of neurodegeneration including spinocerebellar ataxias (e.g., Atxn7, Fgf14, Atxn2, and Tbp), spastic paraplegias (e.g., Spg7, Nipa1, Kif5a, and Spg20), Parkinson's disease (e.g., Park2 and Snca), frontotemporal dementia (e.g., Mapt), myoclonic and dopa-responsive dystonias (e.g., Drd2 and Gch1), pantothenate kinase associated neurodegeneration (e.g., Pank2), and hemochromataosis (e.g., Hfe).
In this study, we suggest that REST and CoREST are responsible for integrating the intrinsic mechanisms (e.g., transcription factor codes and epigenetic modifications) and extrinsic cues (e.g., growth factors, cytokines, and extracellular signaling pathways), which maintain the fidelity, diversity, and plasticity of neuronal subtypes. Uncovering the molecular codes that determine neuronal subtype identity and function will likely add to our understanding of normal CNS processes and contribute to the characterization of neurological diseases involving deregulation of specific regional neuronal subtypes.
Materials and Methods
Cultures preparations were generated as previously described with minor modifications , , , , , . Briefly, multipotent and more lineage restricted progenitor species derived from embryonic day 14.5 (E14.5) ventral forebrain regions of CD1 mice were plated and propagated in serum-free media (SFM) composed of DMEM/F12 (GIBCO) containing B27 and N2 (GIBCO) with the addition of specified factors for various time intervals and subsequently examined by immunofluorescence microscopy to define neural lineage profiles , , , , , , by Western blot analysis to detect REST and CoREST protein expression , , and by QChip and ChIP-chip to identify DNA binding sites for REST and CoREST as previously described , , . Primary neural stem cell (NSC) clones were generated by application of basic fibroblast growth factor (bFGF, 10 ng/ml) for 7 days in vitro (DIV) and then dissociated using 0.05% trypsin (GIBCO) for 15 minutes at 37°C. Individual cells were re-propagated in bFGF for an additional 2 DIV to form secondary NSC clones that were used for experiments and are referred to herein as NSCs. The expansion of secondary NSC clones was limited to 2 DIV to avoid inclusion of differentiated neural species from this culture condition. This culture paradigm eliminates intermediate neural progenitor species and other proliferating neural developmental cell types present in primary NSC clones. Lineage-restricted neuronal-OL progenitor (N/OP) clones were generated from dissociated primary NSC clones by application of bFGF and the N-terminal active form of Shh (N-Shh, 100 ng/ml) for 2 days in vitro (2 DIV) . GABAergic neurons (GABANs) were generated by application of BMP2 (10 ng/ml) to N/OPs for 2 DIV  propagated on poly-D-lysine (PDL) coated culture dishes with the addition of laminin (3 µg/ml, BD Biosciences). Cholinergic neurons (CHOLNs) were generated from E14.5 ventral forebrain-derived NSCs by minor modification of a previously published method . Briefly, NSCs were dissociated in 0.05% trypsin for 15 minutes at 37°C and re-propagated in Neurobasal medium (GIBCO) supplemented with N2 on PDL coated culture dishes with the addition of laminin (3 µg/ml), bFGF (10 ng/ml), N-Shh (100 ng/ml) and nerve growth factor (NGF, 200 ng/ml) for 2 DIV followed by application of N-Shh and NGF (same dose) for an additional 14 DIV. Medium spiny projection neurons (MSNs) were also generated from E14.5 ventral forebrain derived NSCs by modifying a previously published method . Dissociated NSCs were plated on PDL-coated cultures dishes containing Neurobasal, B27 and laminin (3 µg/ml) and were treated with bFGF (10 ng/ml), N-Shh (50 ng/ml) and brain derived neurotrophic factor (BDNF, 50 ng/ml) for 1 DIV followed by application of BDNF (100 ng/ml) for an additional 14 DIV. Media preparations containing NGF and BDNF were replenished every 3 DIV for CHOLNs and MSNs, respectively. Glutamatergic neurons (GLUTNs) were generated directly from radial glial cells (RG) that were obtained from dorsal forebrain derived primary NSC clones. Briefly, individual cells dissociated from dorsal forebrain derived NSC clones were propagated in SFM on PDL-coated dishes in SFM supplemented with laminin (3 µg/ml) in the presence of bFGF (10 ng/ml) and LIF (10 ng/ml) for 2 DIV to elaborate RG. GLUTN were subsequently generated by withdrawal of bFGF and Shh from RG for an additional 4 DIV.
For comparative analysis, OL precursor cells (OLpres) were obtained from N/OP clones by application of platelet-derived growth factor (PDGF-AA, 10 ng/ml) for 2 DIV. Sequential developmental stages of the OL lineage were obtained by propagation in SFM containing laminin (3 µg/ml, BD Biosciences) on poly-D-lysine (PDL) coated culture dishes. OL progenitors (OLpros) were obtained from OL precursor species by application of PDGF-AA (10 ng/ml) for an additional 2-4 DIV. Post-mitotic OLs (pmOLs) and myelinating OLs (myOLs) were obtained from OL progenitors by withdrawal of PDGF-AA for an additional 2 and 4 DIV, respectively . Astrocytes (ASs) were obtained by dissociation of NSC clones using 0.05% trypsin (Sigma) for 15 minutes at 37°C and re-propagation of individual cells in epidermal growth factor (EGF) for 7 DIV with subsequent addition of bone morphogenetic protein 2 (BMP2) for 5 DIV , . We followed institutional IACUC guidelines for experiments in which primary mouse tissue specimens were used.
Specific Antibody Preparations
All antibodies exhibited selective immunoreactivity for mouse cells and tissue sections, and each antibody exhibited a complete absence of alternate cross-reactivity. The following antibodies were utilized: CoREST, REST, and normal rabbit IgG (1∶100, Upstate, Temecula, CA, USA), the neuroepithelial marker (nestin, mIgG1, 1∶200, Pharmingen), N/OP markers (Olig2, goat IgG, 1∶300, R&D and Mash1, mIgG1, 1∶100, Pharmingen), neuronal marker (β Tubulin III, mIgG2b, 1∶700, Sigma), GABAN marker (GABA, rIgG, 1∶1000, Sigma), GLUTN marker (Glutamate, rIgG, 1∶1000, Sigma), CHOLN marker (ChAT, goat IgG, 1∶100, Millipore), MSN marker (DARPP32, rIgG, 1∶100, Santa Cruz). Isotype specific secondary antibodies were utilized at 1∶1500 dilution according to the required fluorophore combinations (Invitrogen). Secondary antibodies utilized for Western blot analysis were HRP conjugated (GE Healthcare).
Western Blot Analysis
Cell cultures were processed for Western blot analysis as described previously , , , . Briefly, cells were homogenized in nine volumes of buffer comprising 0.32 m sucrose, 50 mm Tris-HCl, pH 8.0, EDTA-free protease inhibitors cocktail (Roche) and 0.5 mm phenylmethylsulfonyl fluoride using a glass–Teflon homogenizer (10 strokes at 800 rpm) on ice and centrifuged at 900 g for 10 min and lysed in sodium dodecyl sulfate sample-loading buffer for Western blot analysis.
Growth Factor Preparations
To generate the various neural stem, progenitor and more differentiated neuronal subtype species the following growth factor preparations were utilized: recombinant bFGF (Collaborative Biomedical Products), N-Shh and recombinant mouse β-NGF (R&D Systems), human BDNF (BioVision), BMP2 (gift from Genetics Institute) and LIF (Chemicon). To generate the various complementary and comparative glial species the following additional growth factor preparations were utilized: recombinant EGF and PDGF-AA (R&D Systems).
Quantitative Chromatin Immunoprecipitation (QChIP)
Cultures were prepared essentially as previously described ,  and used in chromatin immunoprecipitations (ChIP) , , , with slight modifications. Between 1×106 and 3×106 cells were used per antibody. The following antibodies were utilized: CoREST, REST, and normal rabbit IgG (Upstate, Temecula, CA, USA). An additional control included the absence of antibody (input chromatin). Antibodies were first validated using a peptide competition assay. For 1×106 cells, 10 µg of antibody was used. Enrichment of fragments by ChIP was quantified by using 1 µl of ChIP product for real-time PCR using the SYBR® Green kit (Applied Biosystems, Foster City, CA, USA) in a 7000 Real Time PCR system® (Applied Biosystems, CA, USA). Validated CoREST and REST promoter binding sites were used as positive controls, while a non-target gene promoter was used as a negative control. ΔCT values were obtained using the ΔΔCT method . For validation of ChIP-chip results, primers, 50 to 150 base pairs in length, were designed to flank the binding peak within each of the respective promoter sequences.
To investigate the specific roles of CoREST and REST throughout these early neural developmental stages, we examined the differential binding profiles of CoREST and REST by determining its gene-specific targets through a series of ChIP-chip experiments. Chromatin immunoprecipitation (ChIP) is an in vivo technique that can be used to identify transcription factor binding sites. To determine that the CoREST and REST antibodies were specific for known targets, ChIP was first performed in an Oli-Neu cell line  with a CoREST antibody, REST antibody, control IgG, or no antibody (input). Samples were then analyzed by quantitative PCR (QPCR) using primers specific for both a known and previously validated CoREST and REST target gene, GluR2, and a negative control , . Chromatin immunoprecipitation targets were determine by utilizing a fold-enrichment greater than 1.5. We then performed a series of ChIP-chip experiments in order to build a comprehensive profile of CoREST and REST targets throughout early neural developmental stages. The mouse promoter array was based on MM8/mouse genome build 36 from February 2006. The design is based only on normal RefSeq genes (17,355 genes) and includes 2,000 base pairs upstream of the transcriptional start sites and 500 base pairs downstream. The probe size ranges from 50–75 base pairs and the spacing interval is 100 base pairs.
ChIP-Chip Data Analysis
Analysis was performed essentially as previously described . Enrichment was calculated for each probe by computing the log-ratio value for the ChIP immunoprecipitated product in comparison to the input chromatin. For all ChIP-chip experiments, in order to find promoter peaks, a maximum log-ratio value for a window consisting of three consecutive probes was determined for both experimental data and a random permutation of the data. A positive threshold was then established to determine the probability for real enrichment. This positive threshold was determined by examining signals of known CoREST and REST binding sites, GluR2, and calbindin, respectively , . A 90% positive threshold was used. The gene target lists were generated based on the intersection of genes across a minimum of two arrays for each experimental paradigm. We validated these ChIP-chip results in a representative sample of cell types with QChIP and found 83% and 94% correlation for CoREST and REST, respectively. These results indicate that the ChIP-chip technique and data analysis methods we used to characterize CoREST and REST target genes are effective approaches for identifying valid binding sites.
Functional Classification of Target Genes
Target genes were functionally annotated through the application of Ingenuity Pathways Analysis (Ingenuity® Systems, www.ingenuity.com).
Gene Expression Assays and Analysis
The gene expression array was based on MM8/mouse genome build 36 from February 2006 (NimbleGen, Madison, WI). The Robust Multi-Array Average (RMA) algorithm was used and data was also analyzed by the significant analysis of microarray methods (SAM). Each of the 11 different cell types was compared to neural stem cells. Three biological replicates were conducted for each cell type.
Identification of neural stem cells (NSCs) and their more lineage restricted progeny that give rise to selective dorsal and ventral neuronal subtypes characterized by the expression of selective neural lineage markers and developmental stage specific transcription factors. Immunofluorescence microscopy of expression profiles for neural lineage markers and transcription factors in NSCs derived from dorsal and ventral mouse forebrain, and progressively maturing intermediate neural progenitors and neuronal subtype species. The neuroepithelial marker, nestin (FITC) is expressed by NSCs, radial glia (RG), and neuronal-oligodendrocyte progenitors (N/OPs). RC2 (Cy5/TRITC) expression is limited to radial glial cells. The basic helix-loop-helix transcription factors, Mash1 (FITC) and Olig2 (TRITC) are only expressed by N/OPs (Nestin-Mash1-Olig2 co-expression studies cannot be performed due to antibody isotype incompatibility). Dorsal (glutamatergic-, GLUTN) and ventral (GABAergic-, GABAN; medium spiny projection-; MSN and cholinergic-, CHOLN) neurons were identified by complete overlap of the early neuronal lineage marker, beta-tubulin III with subtype specific neurotransmitters, or substrate/enzymes involved in the synthesis of those neurotransmitters (glutamate, GABA, DARPP32 and ChAT).
(3.53 MB TIF)
Expression and subcellular localization of REST in neural stem cells (NSCs) and more lineage restricted progeny that give rise to selective dorsal and ventral forebrain derived neuronal subtypes. Immunofluorescence microscopy of REST (TRITC) expression profiles in NSCs, lineage restricted intermediate neural progenitor species and their progeny composed of a selective subset of ventral and dorsal forebrain neuronal species. REST is expressed in the nucleus of all undifferentiated NSCs and intermediate neural progenitors including radial glia (RG) and neuronal-oligodendrocyte progenitors (N/OP). REST expression is noted in both nuclear and cytoplasmic compartments in mature dorsal (glutamatergic-; GLUTN) and ventral (GABAergic-, GABAN; medium spiny projection-, MSN; and cholinergic-, CHOLN) neuronal species. Antibodies to specific markers for NSCs, RGs, and CHOLN (FITC) were used to identify distinct stages of cellular maturation. Due to the absence of a specific lineage marker, N/OPs were labeled with the NSC marker, nestin (FITC), but selectively identified by the presence of two bHLH transcription factors, Olig2 and Mash1 (Figure S1), while GLUTN, GABAN and MSN were labeled with the early neuronal marker, beta-tubulin III (FITC) because of the isotype incompatibility between the REST antibody and neuronal subtype specific markers. The complete overlap of beta-tubulin III with these neuronal subtype specific markers is documented in Figure S1. Scale bars = 100 µm.
(3.76 MB TIF)
Expression and subcellular localization of CoREST in neural stem cells (NSCs) and more lineage restricted progeny that give rise to selective dorsal and ventral forebrain derived neuronal subtypes. Immunofluorescence microscopy of CoREST (TRITC) expression profiles in NSCs, lineage restricted intermediate neural progenitor species and more mature ventral and dorsal forebrain neuronal species. CoREST is expressed in the nucleus of all undifferentiated NSCs and intermediate neural progenitors including radial glia (RG) and neuronal-oligodendrocyte progenitors (N/OP) as well as dorsal (glutamatergic; GLUTN) and ventral (GABAergic-, GABAN; medium spiny projection-, MSN; and cholinergic-, CHOLN) forebrain derived neuronal species. Antibodies specific to NSCs, RG, and CHOLN (FITC) were used to identify selective cellular developmental stages. N/OPs were labeled with nestin (FITC) as well as Olig2/Mash1 (Figure S1), GLUTN, GABAN and MSN were co-labeled with the early neuronal marker, beta-tubulin III (FITC) to avoid isotype incompatibility of CoREST antibody with neuronal subtype specific markers (Figure S1). Scale bars = 100 µm.
(4.23 MB TIF)
Western blot analysis of REST and CoREST expression in NSCs, intermediate progenitors and mature neuronal subtypes. REST and CoREST are ubiquitously expressed in all cell types examined in our developmental paradigm.
(0.46 MB TIF)
Composite profiles of REST and CoREST target genes in individual neuronal subtypes.
(1.69 MB XLS)
Selective profiles of REST and CoREST target genes encoding epigenetic factors in individual neuronal subtypes.
(0.09 MB DOC)
Selective profiles of REST and CoREST target genes encoding cell cycle factors in individual neuronal subtypes.
(0.08 MB DOC)
Selective profiles of REST and CoREST target genes encoding ubiquitin-proteasome factors in individual neuronal subtypes.
(0.07 MB DOC)
Selective profiles of REST and CoREST target genes encoding apoptosis and cell viability factors in individual neuronal subtypes.
(0.05 MB DOC)
Conceived and designed the experiments: JJA IAQ SG MFM. Performed the experiments: JJA IAQ SG MFM. Analyzed the data: JJA IAQ SG DZ AB MFM. Contributed reagents/materials/analysis tools: JJA IAQ SG DZ AB MFM. Wrote the paper: JJA IAQ SG MFM.
- 1. Chong JA, Tapia-Ramirez J, Kim S, Toledo-Aral JJ, Zheng Y, et al. (1995) REST: a mammalian silencer protein that restricts sodium channel gene expression to neurons. Cell 80: 949–957.
- 2. Schoenherr CJ, Anderson DJ (1995) The neuron-restrictive silencer factor (NRSF): a coordinate repressor of multiple neuron-specific genes. Science 267: 1360–1363.
- 3. Chen ZF, Paquette AJ, Anderson DJ (1998) NRSF/REST is required in vivo for repression of multiple neuronal target genes during embryogenesis. Nat Genet 20: 136–142.
- 4. Sun YM, Greenway DJ, Johnson R, Street M, Belyaev ND, et al. (2005) Distinct profiles of REST interactions with its target genes at different stages of neuronal development. Mol Biol Cell 16: 5630–5638.
- 5. Brivanlou AH (1998) Should the master regulator Rest in peace? Nat Genet 20: 109–110.
- 6. Johnson R, Teh CH, Kunarso G, Wong KY, Srinivasan G, et al. (2008) REST regulates distinct transcriptional networks in embryonic and neural stem cells. PLoS Biol 6: e256.
- 7. Ooi L, Wood IC (2007) Chromatin crosstalk in development and disease: lessons from REST. Nat Rev Genet 8: 544–554.
- 8. Lakowski B, Roelens I, Jacob S (2006) CoREST-like complexes regulate chromatin modification and neuronal gene expression. J Mol Neurosci 29: 227–239.
- 9. Battaglioli E, Andres ME, Rose DW, Chenoweth JG, Rosenfeld MG, et al. (2002) REST repression of neuronal genes requires components of the hSWI.SNF complex. J Biol Chem 277: 41038–41045.
- 10. Ballas N, Mandel G (2005) The many faces of REST oversee epigenetic programming of neuronal genes. Curr Opin Neurobiol 15: 500–506.
- 11. Andres ME, Burger C, Peral-Rubio MJ, Battaglioli E, Anderson ME, et al. (1999) CoREST: a functional corepressor required for regulation of neural-specific gene expression. Proc Natl Acad Sci U S A 96: 9873–9878.
- 12. Ballas N, Grunseich C, Lu DD, Speh JC, Mandel G (2005) REST and its corepressors mediate plasticity of neuronal gene chromatin throughout neurogenesis. Cell 121: 645–657.
- 13. Greenway DJ, Street M, Jeffries A, Buckley NJ (2007) RE1 Silencing transcription factor maintains a repressive chromatin environment in embryonic hippocampal neural stem cells. Stem Cells 25: 354–363.
- 14. Kuwabara T, Hsieh J, Nakashima K, Taira K, Gage FH (2004) A small modulatory dsRNA specifies the fate of adult neural stem cells. Cell 116: 779–793.
- 15. Shimojo M, Paquette AJ, Anderson DJ, Hersh LB (1999) Protein kinase A regulates cholinergic gene expression in PC12 cells: REST4 silences the silencing activity of neuron-restrictive silencer factor/REST. Mol Cell Biol 19: 6788–6795.
- 16. Bruce AW, Donaldson IJ, Wood IC, Yerbury SA, Sadowski MI, et al. (2004) Genome-wide analysis of repressor element 1 silencing transcription factor/neuron-restrictive silencing factor (REST/NRSF) target genes. Proc Natl Acad Sci U S A 101: 10458–10463.
- 17. Mortazavi A, Leeper Thompson EC, Garcia ST, Myers RM, Wold B (2006) Comparative genomics modeling of the NRSF/REST repressor network: from single conserved sites to genome-wide repertoire. Genome Res 16: 1208–1221.
- 18. Otto SJ, McCorkle SR, Hover J, Conaco C, Han JJ, et al. (2007) A new binding motif for the transcriptional repressor REST uncovers large gene networks devoted to neuronal functions. J Neurosci 27: 6729–6739.
- 19. Bruce AW, Lopez-Contreras AJ, Flicek P, Down TA, Dhami P, et al. (2009) Functional diversity for REST (NRSF) is defined by in vivo binding affinity hierarchies at the DNA sequence level. Genome Res.
- 20. Gokhan S, Mehler MF (2001) Basic and clinical neuroscience applications of embryonic stem cells. Anat Rec 265: 142–156.
- 21. Li H, Grumet M (2007) BMP and LIF signaling coordinately regulate lineage restriction of radial glia in the developing forebrain. Glia 55: 24–35.
- 22. Yung SY, Gokhan S, Jurcsak J, Molero AE, Abrajano JJ, et al. (2002) Differential modulation of BMP signaling promotes the elaboration of cerebral cortical GABAergic neurons or oligodendrocytes from a common sonic hedgehog-responsive ventral forebrain progenitor species. Proc Natl Acad Sci U S A 99: 16273–16278.
- 23. Noctor SC, Flint AC, Weissman TA, Dammerman RS, Kriegstein AR (2001) Neurons derived from radial glial cells establish radial units in neocortex. Nature 409: 714–720.
- 24. Gotz M, Hartfuss E, Malatesta P (2002) Radial glial cells as neuronal precursors: a new perspective on the correlation of morphology and lineage restriction in the developing cerebral cortex of mice. Brain Res Bull 57: 777–788.
- 25. Snyder GL, Fisone G, Greengard P (1994) Phosphorylation of DARPP-32 is regulated by GABA in rat striatum and substantia nigra. J Neurochem 63: 1766–1771.
- 26. Pisani A, Bernardi G, Ding J, Surmeier DJ (2007) Re-emergence of striatal cholinergic interneurons in movement disorders. Trends Neurosci 30: 545–553.
- 27. Mehler MF, Mabie PC, Zhu G, Gokhan S, Kessler JA (2000) Developmental changes in progenitor cell responsiveness to bone morphogenetic proteins differentially modulate progressive CNS lineage fate. Dev Neurosci 22: 74–85.
- 28. White PM, Morrison SJ, Orimoto K, Kubu CJ, Verdi JM, et al. (2001) Neural crest stem cells undergo cell-intrinsic developmental changes in sensitivity to instructive differentiation signals. Neuron 29: 57–71.
- 29. Williams S, Souchelnytskyi S, Danik M (2002) TGFbeta2 mediates rapid inhibition of calcium influx in identified cholinergic basal forebrain neurons. Biochem Biophys Res Commun 290: 1321–1327.
- 30. Jin JZ, Gu S, McKinney P, Ding J (2006) Expression and functional analysis of Tgif during mouse midline development. Dev Dyn 235: 547–553.
- 31. Itoh S, Itoh F, Goumans MJ, Ten Dijke P (2000) Signaling of transforming growth factor-beta family members through Smad proteins. Eur J Biochem 267: 6954–6967.
- 32. Zhao Y, Marin O, Hermesz E, Powell A, Flames N, et al. (2003) The LIM-homeobox gene Lhx8 is required for the development of many cholinergic neurons in the mouse forebrain. Proc Natl Acad Sci U S A 100: 9005–9010.
- 33. Mori N, Schoenherr C, Vandenbergh DJ, Anderson DJ (1992) A common silencer element in the SCG10 and type II Na+ channel genes binds a factor present in nonneuronal cells but not in neuronal cells. Neuron 9: 45–54.
- 34. Mori T, Yuxing Z, Takaki H, Takeuchi M, Iseki K, et al. (2004) The LIM homeobox gene, L3/Lhx8, is necessary for proper development of basal forebrain cholinergic neurons. Eur J Neurosci 19: 3129–3141.
- 35. Harkany T, Hartig W, Berghuis P, Dobszay MB, Zilberter Y, et al. (2003) Complementary distribution of type 1 cannabinoid receptors and vesicular glutamate transporter 3 in basal forebrain suggests input-specific retrograde signalling by cholinergic neurons. Eur J Neurosci 18: 1979–1992.
- 36. Zhang H, Wang Y, Zhao Y, Yin Y, Xu Q, et al. (2008) Immortalized human neural progenitor cells from the ventral telencephalon with the potential to differentiate into GABAergic neurons. J Neurosci Res 86: 1217–1226.
- 37. Pitsouli C, Delidakis C (2005) The interplay between DSL proteins and ubiquitin ligases in Notch signaling. Development 132: 4041–4050.
- 38. Fitzgerald DP, Cole SJ, Hammond A, Seaman C, Cooper HM (2006) Characterization of neogenin-expressing neural progenitor populations and migrating neuroblasts in the embryonic mouse forebrain. Neuroscience 142: 703–716.
- 39. Mattar P, Langevin LM, Markham K, Klenin N, Shivji S, et al. (2008) Basic helix-loop-helix transcription factors cooperate to specify a cortical projection neuron identity. Mol Cell Biol 28: 1456–1469.
- 40. Goncalves MB, Agudo M, Connor S, McMahon S, Minger SL, et al. (2009) Sequential RARbeta and alpha signalling in vivo can induce adult forebrain neural progenitor cells to differentiate into neurons through Shh and FGF signalling pathways. Dev Biol 326: 305–313.
- 41. Gohlke JM, Armant O, Parham FM, Smith MV, Zimmer C, et al. (2008) Characterization of the proneural gene regulatory network during mouse telencephalon development. BMC Biol 6: 15.
- 42. Yano M, Okano HJ, Okano H (2005) Involvement of Hu and heterogeneous nuclear ribonucleoprotein K in neuronal differentiation through p21 mRNA post-transcriptional regulation. J Biol Chem 280: 12690–12699.
- 43. Frank CL, Tsai LH (2009) Alternative functions of core cell cycle regulators in neuronal migration, neuronal maturation, and synaptic plasticity. Neuron 62: 312–326.
- 44. Medina M, Marinescu RC, Overhauser J, Kosik KS (2000) Hemizygosity of delta-catenin (CTNND2) is associated with severe mental retardation in cri-du-chat syndrome. Genomics 63: 157–164.
- 45. Cheng L, Arata A, Mizuguchi R, Qian Y, Karunaratne A, et al. (2004) Tlx3 and Tlx1 are post-mitotic selector genes determining glutamatergic over GABAergic cell fates. Nat Neurosci 7: 510–517.
- 46. Hevner RF (2006) From radial glia to pyramidal-projection neuron: transcription factor cascades in cerebral cortex development. Mol Neurobiol 33: 33–50.
- 47. Gregorian C, Nakashima J, Le Belle J, Ohab J, Kim R, et al. (2009) Pten deletion in adult neural stem/progenitor cells enhances constitutive neurogenesis. J Neurosci 29: 1874–1886.
- 48. Cina C, Bechberger JF, Ozog MA, Naus CC (2007) Expression of connexins in embryonic mouse neocortical development. J Comp Neurol 504: 298–313.
- 49. Rozental R, Srinivas M, Gokhan S, Urban M, Dermietzel R, et al. (2000) Temporal expression of neuronal connexins during hippocampal ontogeny. Brain Res Brain Res Rev 32: 57–71.
- 50. Imbeault S, Gauvin LG, Toeg HD, Pettit A, Sorbara CD, et al. (2009) The extracellular matrix controls gap junction protein expression and function in postnatal hippocampal neural progenitor cells. BMC Neurosci 10: 13.
- 51. Lai WS, Xu B, Westphal KG, Paterlini M, Olivier B, et al. (2006) Akt1 deficiency affects neuronal morphology and predisposes to abnormalities in prefrontal cortex functioning. Proc Natl Acad Sci U S A 103: 16906–16911.
- 52. Alvarez-Rodriguez R, Pons S (2009) Expression of the proneural gene encoding Mash1 suppresses MYCN mitotic activity. J Cell Sci 122: 595–599.
- 53. Martorana A, Martella G, D'Angelo V, Fusco FR, Spadoni F, et al. (2006) Neurotensin effects on N-type calcium currents among rat pallidal neurons: an electrophysiological and immunohistochemical study. Synapse 60: 371–383.
- 54. Bonsi P, Platania P, Martella G, Madeo G, Vita D, et al. (2008) Distinct roles of group I mGlu receptors in striatal function. Neuropharmacology 55: 392–395.
- 55. Janssens K, De Kimpe L, Balsamo M, Vandoninck S, Vandenheede JR, et al. (2009) Characterization of EVL-I as a protein kinase D substrate. Cell Signal 21: 282–292.
- 56. Kwiatkowski AV, Rubinson DA, Dent EW, Edward van Veen J, Leslie JD, et al. (2007) Ena/VASP Is Required for neuritogenesis in the developing cortex. Neuron 56: 441–455.
- 57. Fiumelli H, Riederer IM, Martin JL, Riederer BM (2008) Phosphorylation of neurofilament subunit NF-M is regulated by activation of NMDA receptors and modulates cytoskeleton stability and neuronal shape. Cell Motil Cytoskeleton 65: 495–504.
- 58. Worzfeld T, Puschel AW, Offermanns S, Kuner R (2004) Plexin-B family members demonstrate non-redundant expression patterns in the developing mouse nervous system: an anatomical basis for morphogenetic effects of Sema4D during development. Eur J Neurosci 19: 2622–2632.
- 59. Ohashi S, Kobayashi S, Omori A, Ohara S, Omae A, et al. (2000) The single-stranded DNA- and RNA-binding proteins pur alpha and pur beta link BC1 RNA to microtubules through binding to the dendrite-targeting RNA motifs. J Neurochem 75: 1781–1790.
- 60. Kortazar D, Fanarraga ML, Carranza G, Bellido J, Villegas JC, et al. (2007) Role of cofactors B (TBCB) and E (TBCE) in tubulin heterodimer dissociation. Exp Cell Res 313: 425–436.
- 61. Sun Y, Goderie SK, Temple S (2005) Asymmetric distribution of EGFR receptor during mitosis generates diverse CNS progenitor cells. Neuron 45: 873–886.
- 62. Wagner B, Natarajan A, Grunaug S, Kroismayr R, Wagner EF, et al. (2006) Neuronal survival depends on EGFR signaling in cortical but not midbrain astrocytes. Embo J 25: 752–762.
- 63. Aguirre A, Rizvi TA, Ratner N, Gallo V (2005) Overexpression of the epidermal growth factor receptor confers migratory properties to nonmigratory postnatal neural progenitors. J Neurosci 25: 11092–11106.
- 64. Ivkovic S, Canoll P, Goldman JE (2008) Constitutive EGFR signaling in oligodendrocyte progenitors leads to diffuse hyperplasia in postnatal white matter. J Neurosci 28: 914–922.
- 65. Roopra A, Qazi R, Schoenike B, Daley TJ, Morrison JF (2004) Localized domains of G9a-mediated histone methylation are required for silencing of neuronal genes. Mol Cell 14: 727–738.
- 66. Kleefstra T, Brunner HG, Amiel J, Oudakker AR, Nillesen WM, et al. (2006) Loss-of-function mutations in euchromatin histone methyl transferase 1 (EHMT1) cause the 9q34 subtelomeric deletion syndrome. Am J Hum Genet 79: 370–377.
- 67. Hao HN, Zhao J, Lotoczky G, Grever WE, Lyman WD (2001) Induction of human beta-defensin-2 expression in human astrocytes by lipopolysaccharide and cytokines. J Neurochem 77: 1027–1035.
- 68. Tachibana M, Nozaki M, Takeda N, Shinkai Y (2007) Functional dynamics of H3K9 methylation during meiotic prophase progression. Embo J 26: 3346–3359.
- 69. Cedar H, Bergman Y (2009) Linking DNA methylation and histone modification: patterns and paradigms. Nat Rev Genet.
- 70. Tahiliani M, Mei P, Fang R, Leonor T, Rutenberg M, et al. (2007) The histone H3K4 demethylase SMCX links REST target genes to X-linked mental retardation. Nature 447: 601–605.
- 71. Abidi F, Holloway L, Moore CA, Weaver DD, Simensen RJ, et al. (2009) Novel human pathological mutations. Gene symbol: JARID1C. Disease: mental retardation, X-linked. Hum Genet 125: 344.
- 72. Adegbola A, Gao H, Sommer S, Browning M (2008) A novel mutation in JARID1C/SMCX in a patient with autism spectrum disorder (ASD). Am J Med Genet A 146A: 505–511.
- 73. Kazantseva A, Sepp M, Kazantseva J, Sadam H, Pruunsild P, et al. (2009) N-Terminally Truncated Baf57 Isoforms Contribute to the Diversity of Swi/Snf Complexes in Neurons. J Neurochem.
- 74. Wu JI, Lessard J, Crabtree GR (2009) Understanding the words of chromatin regulation. Cell 136: 200–206.
- 75. Pepin D, Vanderhyden BC, Picketts DJ, Murphy BD (2007) ISWI chromatin remodeling in ovarian somatic and germ cells: revenge of the NURFs. Trends Endocrinol Metab 18: 215–224.
- 76. Guardavaccaro D, Frescas D, Dorrello NV, Peschiaroli A, Multani AS, et al. (2008) Control of chromosome stability by the beta-TrCP-REST-Mad2 axis. Nature 452: 365–369.
- 77. Staropoli JF, Abeliovich A (2005) The ubiquitin-proteasome pathway is necessary for maintenance of the postmitotic status of neurons. J Mol Neurosci 27: 175–183.
- 78. Ang XL, Seeburg DP, Sheng M, Harper JW (2008) Regulation of postsynaptic RapGAP SPAR by Polo-like kinase 2 and the SCFbeta-TRCP ubiquitin ligase in hippocampal neurons. J Biol Chem 283: 29424–29432.
- 79. Pasinelli P, Borchelt DR, Houseweart MK, Cleveland DW, Brown RH Jr (1998) Caspase-1 is activated in neural cells and tissue with amyotrophic lateral sclerosis-associated mutations in copper-zinc superoxide dismutase. Proc Natl Acad Sci U S A 95: 15763–15768.
- 80. Ando Y, Liang Y, Ishigaki S, Niwa J, Jiang Y, et al. (2003) Caspase-1 and -3 mRNAs are differentially upregulated in motor neurons and glial cells in mutant SOD1 transgenic mouse spinal cord: a study using laser microdissection and real-time RT-PCR. Neurochem Res 28: 839–846.
- 81. Segref A, Hoppe T (2009) Think locally: control of ubiquitin-dependent protein degradation in neurons. EMBO Rep 10: 44–50.
- 82. Salero E, Hatten ME (2007) Differentiation of ES cells into cerebellar neurons. Proc Natl Acad Sci U S A 104: 2997–3002.
- 83. Ferrari-Toninelli G, Bonini SA, Bettinsoli P, Uberti D, Memo M (2008) Microtubule stabilizing effect of notch activation in primary cortical neurons. Neuroscience 154: 946–952.
- 84. Dugas JC, Tai YC, Speed TP, Ngai J, Barres BA (2006) Functional genomic analysis of oligodendrocyte differentiation. J Neurosci 26: 10967–10983.
- 85. Paquette AJ, Perez SE, Anderson DJ (2000) Constitutive expression of the neuron-restrictive silencer factor (NRSF)/REST in differentiating neurons disrupts neuronal gene expression and causes axon pathfinding errors in vivo. Proc Natl Acad Sci U S A 97: 12318–12323.
- 86. Lein PJ, Beck HN, Chandrasekaran V, Gallagher PJ, Chen HL, et al. (2002) Glia induce dendritic growth in cultured sympathetic neurons by modulating the balance between bone morphogenetic proteins (BMPs) and BMP antagonists. J Neurosci 22: 10377–10387.
- 87. Zhu H, Bendall AJ (2006) Dlx3 is expressed in the ventral forebrain of chicken embryos: implications for the evolution of the Dlx gene family. Int J Dev Biol 50: 71–75.
- 88. Turrigiano GG (2008) The self-tuning neuron: synaptic scaling of excitatory synapses. Cell 135: 422–435.
- 89. Zhang SC (2006) Neural subtype specification from embryonic stem cells. Brain Pathol 16: 132–142.
- 90. Hobert O (2008) Regulatory logic of neuronal diversity: terminal selector genes and selector motifs. Proc Natl Acad Sci U S A 105: 20067–20071.
- 91. Flavell SW, Greenberg ME (2008) Signaling mechanisms linking neuronal activity to gene expression and plasticity of the nervous system. Annu Rev Neurosci 31: 563–590.
- 92. Johnson DS, Mortazavi A, Myers RM, Wold B (2007) Genome-wide mapping of in vivo protein-DNA interactions. Science 316: 1497–1502.
- 93. Zheng D, Zhao K, Mehler MF (2009) Profiling RE1/REST-mediated histone modifications in the human genome. Genome Biol 10: R9.
- 94. Gokhan S, Marin-Husstege M, Yung SY, Fontanez D, Casaccia-Bonnefil P, et al. (2005) Combinatorial profiles of oligodendrocyte-selective classes of transcriptional regulators differentially modulate myelin basic protein gene expression. J Neurosci 25: 8311–8321.
- 95. Marmur R, Kessler JA, Zhu G, Gokhan S, Mehler MF (1998) Differentiation of oligodendroglial progenitors derived from cortical multipotent cells requires extrinsic signals including activation of gp130/LIFbeta receptors. J Neurosci 18: 9800–9811.
- 96. Marmur R, Mabie PC, Gokhan S, Song Q, Kessler JA, et al. (1998) Isolation and developmental characterization of cerebral cortical multipotent progenitors. Dev Biol 204: 577–591.
- 97. Zhu G, Mehler MF, Mabie PC, Kessler JA (2000) Developmental changes in neural progenitor cell lineage commitment do not depend on epidermal growth factor receptor signaling. J Neurosci Res 59: 312–320.
- 98. Zhu G, Mehler MF, Mabie PC, Kessler JA (1999) Developmental changes in progenitor cell responsiveness to cytokines. J Neurosci Res 56: 131–145.
- 99. Bao D, Pang Z, Morgan JI (2005) The structure and proteolytic processing of Cbln1 complexes. J Neurochem 95: 618–629.
- 100. Oberley MJ, Tsao J, Yau P, Farnham PJ (2004) High-throughput screening of chromatin immunoprecipitates using CpG-island microarrays. Methods Enzymol 376: 315–334.
- 101. Sandoval J, Rodriguez JL, Tur G, Serviddio G, Pereda J, et al. (2004) RNAPol-ChIP: a novel application of chromatin immunoprecipitation to the analysis of real-time gene transcription. Nucleic Acids Res 32: e88.
- 102. Weinmann AS, Farnham PJ (2002) Identification of unknown target genes of human transcription factors using chromatin immunoprecipitation. Methods 26: 37–47.
- 103. Reilly JO, Karavanova ID, Williams KP, Mahanthappa NK, Allendoerfer KL (2002) Cooperative effects of Sonic Hedgehog and NGF on basal forebrain cholinergic neurons. Mol Cell Neurosci 19: 88–96.
- 104. Ivkovic S, Ehrlich ME (1999) Expression of the striatal DARPP-32/ARPP-21 phenotype in GABAergic neurons requires neurotrophins in vivo and in vitro. J Neurosci 19: 5409–5419.
- 105. Kondo T, Raff M (2000) Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 289: 1754–1757.
- 106. Koller KJ, Coyle JT (1985) The characterization of the specific binding of [3H]-N-acetylaspartylglutamate to rat brain membranes. J Neurosci 5: 2882–2888.
- 107. Cohen RI, Marmur R, Norton WT, Mehler MF, Kessler JA (1996) Nerve growth factor and neurotrophin-3 differentially regulate the proliferation and survival of developing rat brain oligodendrocytes. J Neurosci 16: 6433–6442.
- 108. Chakrabarti SK, James JC, Mirmira RG (2002) Quantitative assessment of gene targeting in vitro and in vivo by the pancreatic transcription factor, Pdx1. Importance of chromatin structure in directing promoter binding. J Biol Chem 277: 13286–13293.
- 109. Jung M, Kramer E, Grzenkowski M, Tang K, Blakemore W, et al. (1995) Lines of murine oligodendroglial precursor cells immortalized by an activated neu tyrosine kinase show distinct degrees of interaction with axons in vitro and in vivo. Eur J Neurosci 7: 1245–1265.
- 110. Huang Y, Doherty JJ, Dingledine R (2002) Altered histone acetylation at glutamate receptor 2 and brain-derived neurotrophic factor genes is an early event triggered by status epilepticus. J Neurosci 22: 8422–8428.
- 111. Polo JM, Juszczynski P, Monti S, Cerchietti L, Ye K, et al. (2007) Transcriptional signature with differential expression of BCL6 target genes accurately identifies BCL6-dependent diffuse large B cell lymphomas. Proc Natl Acad Sci U S A 104: 3207–3212.