Limits...
A genome-wide screen of CREB occupancy identifies the RhoA inhibitors Par6C and Rnd3 as regulators of BDNF-induced synaptogenesis.

Lesiak A, Pelz C, Ando H, Zhu M, Davare M, Lambert TJ, Hansen KF, Obrietan K, Appleyard SM, Impey S, Wayman GA - PLoS ONE (2013)

Bottom Line: Interestingly, CREB occupied a cluster of non-canonical CRE motifs in the Rnd3 promoter region.Lastly, we show that BDNF-stimulated synaptogenesis requires the expression of Par6C and Rnd3, and that overexpression of either protein is sufficient to increase synaptogenesis.Thus, we propose that BDNF can regulate formation of functional synapses by increasing the expression of the RhoA inhibitors, Par6C and Rnd3.

View Article: PubMed Central - PubMed

Affiliation: Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Program in Neuroscience, Washington State University, Pullman, Washington, United States of America.

ABSTRACT
Neurotrophin-regulated gene expression is believed to play a key role in long-term changes in synaptic structure and the formation of dendritic spines. Brain-derived neurotrophic factor (BDNF) has been shown to induce increases in dendritic spine formation, and this process is thought to function in part by stimulating CREB-dependent transcriptional changes. To identify CREB-regulated genes linked to BDNF-induced synaptogenesis, we profiled transcriptional occupancy of CREB in hippocampal neurons. Interestingly, de novo motif analysis of hippocampal ChIP-Seq data identified a non-canonical CRE motif (TGGCG) that was enriched at CREB target regions and conferred CREB-responsiveness. Because cytoskeletal remodeling is an essential element of the formation of dendritic spines, within our screens we focused our attention on genes previously identified as inhibitors of RhoA GTPase. Bioinformatic analyses identified dozens of candidate CREB target genes known to regulate synaptic architecture and function. We showed that two of these, the RhoA inhibitors Par6C (Pard6A) and Rnd3 (RhoE), are BDNF-induced CREB-regulated genes. Interestingly, CREB occupied a cluster of non-canonical CRE motifs in the Rnd3 promoter region. Lastly, we show that BDNF-stimulated synaptogenesis requires the expression of Par6C and Rnd3, and that overexpression of either protein is sufficient to increase synaptogenesis. Thus, we propose that BDNF can regulate formation of functional synapses by increasing the expression of the RhoA inhibitors, Par6C and Rnd3. This study shows that genome-wide analyses of CREB target genes can facilitate the discovery of new regulators of synaptogenesis.

Show MeSH
The non-canonical CRE is associated with CREB occupancy and CREB-responsiveness of endogenous genes.(A and B) Histograms depict spatial accumulation of mouse hippocampal CREB ChIP-Seq peaks relative to all genomic occurrences of the canonical CRE (TGACG) and the non-canonical CRE (TGGCG). Randomized peaks over the same genomic extent are depicted in grey. (C-E) UCSC genome browser tracks show hippocampal CREB ChIP-Seq (CREB hip), embryonic cortical neuron CREB ChIP-Seq (CREB embryonic) data relative to RefSeq genes, non-canonical CRE motifs (red), and CpG islands (green). Rat CREB SACO data mapped to the mouse genome is also depicted with purple bars denoting SACO cluster extent (CREB SACO). The depicted loci did not contain the canonical CRE motif. (F) Chromatin from rat and mouse hippocampal neurons was immunoprecipitated with the indicated antibodies. Real-time PCR with primers directed against the ChIP-Seq peak (mouse) or the orthologous rat locus (rat) were used to assess CREB occupancy (n = 3; SEM). Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p). (G) All RefSeq genes whose annotated transcriptional start is within 1 kb of a hippocampal neuron ChiP-Seq peak were selected for gene ontology analyses. Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p).
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3675129&req=5

pone-0064658-g002: The non-canonical CRE is associated with CREB occupancy and CREB-responsiveness of endogenous genes.(A and B) Histograms depict spatial accumulation of mouse hippocampal CREB ChIP-Seq peaks relative to all genomic occurrences of the canonical CRE (TGACG) and the non-canonical CRE (TGGCG). Randomized peaks over the same genomic extent are depicted in grey. (C-E) UCSC genome browser tracks show hippocampal CREB ChIP-Seq (CREB hip), embryonic cortical neuron CREB ChIP-Seq (CREB embryonic) data relative to RefSeq genes, non-canonical CRE motifs (red), and CpG islands (green). Rat CREB SACO data mapped to the mouse genome is also depicted with purple bars denoting SACO cluster extent (CREB SACO). The depicted loci did not contain the canonical CRE motif. (F) Chromatin from rat and mouse hippocampal neurons was immunoprecipitated with the indicated antibodies. Real-time PCR with primers directed against the ChIP-Seq peak (mouse) or the orthologous rat locus (rat) were used to assess CREB occupancy (n = 3; SEM). Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p). (G) All RefSeq genes whose annotated transcriptional start is within 1 kb of a hippocampal neuron ChiP-Seq peak were selected for gene ontology analyses. Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p).

Mentions: If the non-canonical CRE contributes to recruitment of CREB it should show similar spatial correlation with CREB occupancy as the canonical motif. A histogram of CREB occupancy at all genomic positions of the non-canonical motif shows the same narrow spatial localization as the known motif (Figure 2A and B). Moreover, enrichment of ChIP-Seq peak counts at both motifs was highly significant (Wilcoxon rank-sum: p<1×10−11). Similar results were seen with both embryonic ChIP-Seq data [43], and PC12 SACO data (data not shown). To test whether this was due to promoter or GC bias we selected ChIP-Seq peaks that were greater than 5 kb distal to annotated promoters and performed a similar analysis. Enrichment for both motifs in this smaller set of non-GC-rich regions was also very significant (Wilcoxon rank-sum: p<1×10−6) (data not shown). To test whether this motif is correlated with functional regulation of gene expression, we selected ChIP-Seq peaks adjacent to proximal promoters that contained the non-canonical, but not the canonical motif (1000 bp window). In this data the distribution of the non-canonical motif at individual genes showed tight correlation with hippocampal ChIP-Seq peaks, and with ChIP data from other studies (Figure 2C-E). ChIP-Seq PCR analyses confirmed significant CREB occupancy at these regions (Figure 2F). These data suggest that the non-canonical motif can recruit CREB to target genes in vivo.


A genome-wide screen of CREB occupancy identifies the RhoA inhibitors Par6C and Rnd3 as regulators of BDNF-induced synaptogenesis.

Lesiak A, Pelz C, Ando H, Zhu M, Davare M, Lambert TJ, Hansen KF, Obrietan K, Appleyard SM, Impey S, Wayman GA - PLoS ONE (2013)

The non-canonical CRE is associated with CREB occupancy and CREB-responsiveness of endogenous genes.(A and B) Histograms depict spatial accumulation of mouse hippocampal CREB ChIP-Seq peaks relative to all genomic occurrences of the canonical CRE (TGACG) and the non-canonical CRE (TGGCG). Randomized peaks over the same genomic extent are depicted in grey. (C-E) UCSC genome browser tracks show hippocampal CREB ChIP-Seq (CREB hip), embryonic cortical neuron CREB ChIP-Seq (CREB embryonic) data relative to RefSeq genes, non-canonical CRE motifs (red), and CpG islands (green). Rat CREB SACO data mapped to the mouse genome is also depicted with purple bars denoting SACO cluster extent (CREB SACO). The depicted loci did not contain the canonical CRE motif. (F) Chromatin from rat and mouse hippocampal neurons was immunoprecipitated with the indicated antibodies. Real-time PCR with primers directed against the ChIP-Seq peak (mouse) or the orthologous rat locus (rat) were used to assess CREB occupancy (n = 3; SEM). Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p). (G) All RefSeq genes whose annotated transcriptional start is within 1 kb of a hippocampal neuron ChiP-Seq peak were selected for gene ontology analyses. Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p).
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3675129&req=5

pone-0064658-g002: The non-canonical CRE is associated with CREB occupancy and CREB-responsiveness of endogenous genes.(A and B) Histograms depict spatial accumulation of mouse hippocampal CREB ChIP-Seq peaks relative to all genomic occurrences of the canonical CRE (TGACG) and the non-canonical CRE (TGGCG). Randomized peaks over the same genomic extent are depicted in grey. (C-E) UCSC genome browser tracks show hippocampal CREB ChIP-Seq (CREB hip), embryonic cortical neuron CREB ChIP-Seq (CREB embryonic) data relative to RefSeq genes, non-canonical CRE motifs (red), and CpG islands (green). Rat CREB SACO data mapped to the mouse genome is also depicted with purple bars denoting SACO cluster extent (CREB SACO). The depicted loci did not contain the canonical CRE motif. (F) Chromatin from rat and mouse hippocampal neurons was immunoprecipitated with the indicated antibodies. Real-time PCR with primers directed against the ChIP-Seq peak (mouse) or the orthologous rat locus (rat) were used to assess CREB occupancy (n = 3; SEM). Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p). (G) All RefSeq genes whose annotated transcriptional start is within 1 kb of a hippocampal neuron ChiP-Seq peak were selected for gene ontology analyses. Significance was assessed using the Storey-adjusted Fisher exact test (FDR-adjusted p).
Mentions: If the non-canonical CRE contributes to recruitment of CREB it should show similar spatial correlation with CREB occupancy as the canonical motif. A histogram of CREB occupancy at all genomic positions of the non-canonical motif shows the same narrow spatial localization as the known motif (Figure 2A and B). Moreover, enrichment of ChIP-Seq peak counts at both motifs was highly significant (Wilcoxon rank-sum: p<1×10−11). Similar results were seen with both embryonic ChIP-Seq data [43], and PC12 SACO data (data not shown). To test whether this was due to promoter or GC bias we selected ChIP-Seq peaks that were greater than 5 kb distal to annotated promoters and performed a similar analysis. Enrichment for both motifs in this smaller set of non-GC-rich regions was also very significant (Wilcoxon rank-sum: p<1×10−6) (data not shown). To test whether this motif is correlated with functional regulation of gene expression, we selected ChIP-Seq peaks adjacent to proximal promoters that contained the non-canonical, but not the canonical motif (1000 bp window). In this data the distribution of the non-canonical motif at individual genes showed tight correlation with hippocampal ChIP-Seq peaks, and with ChIP data from other studies (Figure 2C-E). ChIP-Seq PCR analyses confirmed significant CREB occupancy at these regions (Figure 2F). These data suggest that the non-canonical motif can recruit CREB to target genes in vivo.

Bottom Line: Interestingly, CREB occupied a cluster of non-canonical CRE motifs in the Rnd3 promoter region.Lastly, we show that BDNF-stimulated synaptogenesis requires the expression of Par6C and Rnd3, and that overexpression of either protein is sufficient to increase synaptogenesis.Thus, we propose that BDNF can regulate formation of functional synapses by increasing the expression of the RhoA inhibitors, Par6C and Rnd3.

View Article: PubMed Central - PubMed

Affiliation: Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Program in Neuroscience, Washington State University, Pullman, Washington, United States of America.

ABSTRACT
Neurotrophin-regulated gene expression is believed to play a key role in long-term changes in synaptic structure and the formation of dendritic spines. Brain-derived neurotrophic factor (BDNF) has been shown to induce increases in dendritic spine formation, and this process is thought to function in part by stimulating CREB-dependent transcriptional changes. To identify CREB-regulated genes linked to BDNF-induced synaptogenesis, we profiled transcriptional occupancy of CREB in hippocampal neurons. Interestingly, de novo motif analysis of hippocampal ChIP-Seq data identified a non-canonical CRE motif (TGGCG) that was enriched at CREB target regions and conferred CREB-responsiveness. Because cytoskeletal remodeling is an essential element of the formation of dendritic spines, within our screens we focused our attention on genes previously identified as inhibitors of RhoA GTPase. Bioinformatic analyses identified dozens of candidate CREB target genes known to regulate synaptic architecture and function. We showed that two of these, the RhoA inhibitors Par6C (Pard6A) and Rnd3 (RhoE), are BDNF-induced CREB-regulated genes. Interestingly, CREB occupied a cluster of non-canonical CRE motifs in the Rnd3 promoter region. Lastly, we show that BDNF-stimulated synaptogenesis requires the expression of Par6C and Rnd3, and that overexpression of either protein is sufficient to increase synaptogenesis. Thus, we propose that BDNF can regulate formation of functional synapses by increasing the expression of the RhoA inhibitors, Par6C and Rnd3. This study shows that genome-wide analyses of CREB target genes can facilitate the discovery of new regulators of synaptogenesis.

Show MeSH