A Nitric Oxide Signaling Pathway Controls CREB-Mediated Gene Expression in Neurons

A Nitric Oxide Signaling Pathway Controls CREB-Mediated Gene Expression in Neurons

Molecular Cell 21, 283–294, January 20, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.molcel.2005.12.006 A Nitric Oxide Signaling Pathway Controls CREB-Med...

772KB Sizes 1 Downloads 26 Views

Molecular Cell 21, 283–294, January 20, 2006 ª2006 Elsevier Inc.

DOI 10.1016/j.molcel.2005.12.006

A Nitric Oxide Signaling Pathway Controls CREB-Mediated Gene Expression in Neurons Antonella Riccio,1,2,6,* Rebecca S. Alvania,1,2,5 Bonnie E. Lonze,1,2,5 Narendrakumar Ramanan,1,2 Taeho Kim,1,2 Yunfei Huang,2 Ted M. Dawson,3,4 Solomon H. Snyder,2 and David D. Ginty1,2,* 1 Howard Hughes Medical Institute 2 Department of Neuroscience 3 Department of Neurology 4 Institute for Cell Engineering The Johns Hopkins University School of Medicine 725 North Wolfe Street Baltimore, Maryland 21201

Summary Prevailing views of neurotrophin action hold that the transcription factor CREB is constitutively bound to target genes with transcriptional activation occurring via CREB phosphorylation. However, we report that within several CRE-containing genes, CREB is not constitutively bound. Upon exposure of neurons to brainderived neurotrophic factor (BDNF), CREB becomes rapidly bound to DNA coincident with phosphorylation at its transcriptional regulatory site, Ser133. This inducible CREB-DNA binding is independent of CREB Ser133 phosphorylation and is not affected by inhibition of the ERK or PI3K signaling pathways. Instead, BDNF regulates CREB binding by initiating a nitric oxide-dependent signaling pathway that leads to S-nitrosylation of nuclear proteins that associate with CREB target genes. Pharmacological manipulation of neurons in vitro and analysis of mice lacking neuronal nitric oxide synthase (nNOS) suggest that NO mediates BDNF and activity-dependent expression of CREB target genes. Thus, in conjunction with CREB phosphorylation, the NO pathway controls CREB-DNA binding and CRE-mediated gene expression.

Introduction The neurotrophins comprise a family of closely related peptide growth factors that sculpt the vertebrate nervous system and regulate many cellular processes, including growth, survival, and plasticity (Huang and Reichardt, 2001). Over the past two decades, much insight into how neurotrophins mediate their diverse cellular effects has been gleaned; their receptors have been identified, and many intracellular effectors have been described (Huang and Reichardt, 2003; Segal, 2003). One nuclear effector of neurotrophin signaling pathways is CREB. Because many immediate early and delayed response genes contain CREB binding sites within their genomic regulatory regions, it has been proposed that *Correspondence: [email protected] (A.R.); [email protected] (D.D.G.) 5 These authors contributed equally to this work. 6 Present address: MRC Laboratory for Molecular Cell Biology and Department of Biology UCL, Gower Street, WC1E 6BT London, United Kingdom.

CREB and its closely related family members are key mediators of the general nuclear response to nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), and other neurotrophic growth factors (Lonze and Ginty, 2002). Neurotrophins and other stimuli exert their influence over CREB via multiple signaling cascades, including Ras/ERK, PI3K/Akt, Ca2+, nitric oxide, and p38MAPK signaling pathways (Mayr and Montminy, 2001; Shaywitz and Greenberg, 1999). Many of these signaling pathways ultimately lead to the activation of CREB kinases, which phosphorylate CREB on the transcriptional regulatory site Ser133. Upon phosphorylation of CREB Ser133, the transcriptional coactivator CBP, a histone acetyl transferase, is recruited to CREB where it promotes histone acetylation and assembly of the basal transcriptional complex (Vo and Goodman, 2001). The currently held view is that CREB is constitutively bound to CRE sequences within its target gene promoters and that initiation of transcription of these genes is controlled by CREB phosphorylation at Ser133 and possibly other sites (Gau et al., 2002; Kornhauser et al., 2002). Here, we show that in neurons, CREB DNA binding is tightly controlled by extracellular stimuli, including neurotrophins and excitatory synaptic transmission, and that an NO signaling pathway controls CREB binding. Thus, distinct signaling pathways mediate CREB DNA binding and phosphorylation of CREB’s transcriptional regulatory site, Ser133, to support CREB-dependent transcription in neurons. Results BDNF and Enhanced Excitatory Neurotransmission Induce CREB Binding to CREs Evidence that CREB is constitutively bound to CRE sequences stems chiefly from electrophoretic mobility shift assays (EMSAs) in which cell extracts are incubated in vitro with CRE oligonucleotides. To determine whether a constitutive interaction between CREB and its genomic DNA binding sites occurs in neurons, we employed a chromatin immunoprecipitation (ChIP) assay to assess CREB binding to the promoters of known CREB target genes: c-fos, nNOS, and VGF (Lonze and Ginty, 2002). Unexpectedly, in unstimulated cultures of cortical neurons, little or no CREB is bound to these promoters (Figures 1A and 1B). By contrast, exposure of neurons to BDNF, which signals through the tyrosine kinase receptor TrkB, elicits a rapid and robust induction of CREB binding to the promoters of c-fos, nNOS, VGF (Figures 1A and 1B), and to a lesser extent, zif-268 (A.R. and D.D.G., unpublished data). The antibody used for these ChIP experiments is specific for CREB because it detects a single band by immunoblot of cortical neuron extracts and it fails to precipitate the c-fos (Figure 1C) as well as nNOS and VGF promoters (A.R. and D.D.G., unpublished data) from extracts prepared from primary cortical neurons obtained from Creb null mice (Rudolph et al., 1998). Moreover, similar results were observed in ChIP experiments employing two

Molecular Cell 284

Figure 1. BDNF-Dependent CREB Binding to Endogenous CRE Sequences in Cortical Neurons (A) ChIP analysis of cortical neurons treated with BDNF (75 ng/ml) for the indicated times. ChIPs performed with anti-CREB antibody were followed by PCR amplification of the c-fos (top), nNOS (middle), and VGF (bottom) promoter regions containing their CREs. Immunoprecipitation input controls contain c-fos promoter sequences detectable by PCR (bottom) (n = 4). Although there is variability in the temporal pattern of CREB DNA binding, in all cases an increase in CREB binding occurred within 5 min of neuronal stimulation. (B) ChIP assays were performed with anti-CREB, and levels of immunoprecipitated c-fos, nNOS, and VGF promoters were measured by quantitative PCR. Data are represented as fold changes over unstimulated cortical neurons (n = 4–6). Shown are the means 6 SEM. (C) ChIP analysis of BDNF-treated cortical neurons derived from wild-type (left panels) or Creb null (right panels) mouse embryos. Neurons were stimulated with BDNF (75 ng/ml) for the indicated times. Immunoprecipitations were performed with anti-CREB (top) or anti-CBP (middle). PCR amplification of the c-fos promoter CRE indicates association of CREB and CBP with this DNA region (n = 2). (D) ChIP analysis of cortical neurons stimulated with BDNF (75 ng/ml) and immunoprecipitated with CREB (Upstate), CREB (Cell Signaling), SRF, or acetyl histone H4 antibodies followed by PCR amplification for c-fos (n = 3).

other CREB antibodies (Figure 1D). Another transcription factor that controls IEG expression, SRF, was constitutively bound to the c-fos promoter (Figure 1D). Also, with ChIP assays that employ anti-CBP, CBP occupancy of the c-fos promoter is reduced in neurons lacking CREB (Figure 1C), indicating that CBP association with the c-fos promoter depends in part upon CREB protein. Furthermore, and as expected, CREB is not phosphorylated on the transcriptional regulatory site

Ser133 under basal conditions, but after exposure of neurons to BDNF, a marked increase in phosphorylation of CREB Ser133 is observed (Figures 3B and 3D and Figure S1B available in the Supplemental Data with this article online), concomitant with the appearance of CREB DNA binding (Figures 1A and 1B and Figure S1A). Thus, upon growth factor stimulation of neurons, CREB binds to DNA and simultaneously becomes phosphorylated on Ser133. We propose that, together, these two

NO Regulates CREB DNA Binding in Neurons 285

BDNF-dependent events facilitate CBP recruitment and are required for the initiation of expression of CREB target genes. To ask whether stimulus-inducible CREB-CRE association is restricted to BDNF-treated cortical neurons or is a more general phenomenon, we performed ChIP analyses in multiple cell types under different stimulatory conditions, including primary fibroblast treated with EGF, PC12 cells treated with NGF, and HEK293 cells treated with forskolin. In all cases, we observe stimulus-inducible CREB-DNA binding (Figures 2B, 4A, and 4B and A.R. and D.D.G., unpublished data), although to varying degrees. Our findings are consistent with earlier observations that have noted stimulusdependent changes in occupancy of the CRE region of the tyrosine amino transferase gene in hepatoma cells (Weih et al., 1990). Interestingly, and consistent with recent observations (Impey et al., 2004), we found that CREB DNA binding activity also corresponds well with tissue specificity of gene activation. EGF stimulation of rat primary fibroblasts triggers CREB binding to the ubiquitously expressed c-fos gene but fails to induce its association with the promoters of the neuron-specific genes VGF and nNOS (Figure 2B), which are not expressed in fibroblasts. Discrepancies in the temporal characteristics of stimulus-induced CREB phosphorylation and target gene expression have been previously reported (for a review see Lonze and Ginty [2002]). However, the kinetics of transcriptional activation corresponds well with BDNFinduced CREB DNA binding. Within minutes of exposure of neurons to BDNF, CREB is associated with the c-fos promoter (Figures 1A, 1B, and 1C and Figure S1A) and c-fos expression is activated (Figure 6) (Kim et al., 2000). In contrast, CREB binding to the nNOS promoter peaks 90 min after BDNF stimulation (Figures 1A and 1B and Figure S1A), consistent with the relatively slow kinetics of expression of nNOS (Sasaki et al., 2000). CREB binding to the VGF promoter is also coincident with VGF expression (Hawley et al., 1992), with CREB binding to the VGF promoter increasing steadily and reaching a plateau after 30 min of BDNF treatment (Figures 1A and 1B and Figure S1A). Similarly, transcriptional activation by excitatory neurotransmission corresponds well with CREB DNA binding (Bading et al., 1993). Bicuculline, a GABA-A receptor antagonist that elicits prolonged CREB phosphorylation and CRE-mediated gene transcription in cultured neurons (Hardingham et al., 2002), effectively induces CREB binding to the c-fos (Figure 2A) and nNOS promoters (A.R. and D.D.G., unpublished data). In contrast, bath application of glutamate, which causes transient CREB phosphorylation and little to no transcription (Hardingham et al., 2002; Sala et al., 2000), fails to promote CREB DNA binding (Figure S1A, right panels). As reported previously, both bicuculline and bath-applied glutamate trigger phosphorylation of CREB Ser133 (Figure S1B) (Ginty et al., 1993; Hardingham et al., 2002). CREB DNA Binding Is Independent of Phosphorylation of CREB Ser133 We next asked whether a common BDNF-activated signal transduction pathway controls both phosphorylation of CREB Ser133 and CREB DNA binding activity and

Figure 2. CREB Binding to DNA Is Stimulus and Cell Type Specific (A) ChIP analysis of cortical neurons maintained in cultures for 20 days and stimulated with bicuculline (50 mM) for the indicated times. CREB immunoprecipitation and PCR for c-fos (n = 3). (B) ChIP analysis of primary fibroblasts treated with EGF (30 ng/ml) as indicated. CREB immunoprecipitation and PCR for c-fos (top), nNOS (middle), and VGF (bottom) promoters are shown (n = 3). Abbreviation: PI, preimmume antibody.

whether Ser133 phosphorylation is required for CREB DNA binding. In one set of experiments, pharmacological inhibitors of the Ras/ERK, PI3K/Akt, and calcium pathways were employed because these pathways mediate neurotrophin signaling events, including CREB Ser133 phosphorylation, and expression of at least some BDNF-responsive genes. Treatment of cortical neurons with the MEK inhibitor UO-126 has no effect on BDNF-induced binding of CREB to the c-fos promoter (Figures 3A and 3E), although the inhibitor blocks phosphorylation of CREB Ser133 (Figure 3B). The PI3 kinase inhibitor LY294002 also fails to block CREB DNA binding activity, but it completely inhibits activation of the PI3 kinase effector Akt (A.R. and D.D.G., unpublished data). In contrast, exposure of neurons to the calcium chelator BAPTA-AM attenuates both CREB DNA binding and histone acetylation (Figures 3C and 3E), suggesting that BDNF regulates the binding of CREB to its target promoters through one or more calcium signaling pathways. Consistent with independent regulation of CREB-DNA binding and CREB phosphorylation, pretreatment with BAPTA-AM does not affect BDNFinduced phosphorylation of CREB Ser133 (Figure 3D). To determine whether phosphorylation of CREB Ser133 influences CREB DNA binding, we performed ChIP analyses with cell lines expressing mutant forms of myc-tagged CREB. For these experiments, we used myc-tagged CREB (myc-CREB), myc-CREBm1, a wellcharacterized dominant-negative form of CREB that harbors a Ser133 to Ala mutation (Gonzalez and Montminy, 1989), and myc-CREBDIEDML, a constitutively active form of CREB in which seven residues surrounding Ser133 have been substituted with a motif that constitutively binds to CBP (Cardinaux et al., 2000). Treatment of HEK293 cells or PC12 cells expressing myc-CREB with

Molecular Cell 286

Figure 3. BDNF Induces CREB Phosphorylation and CREB DNA Binding through Independent Signaling Pathways (A) ChIP analysis of BDNF-stimulated cortical neurons in the presence (right panels) or the absence (left panels) of the MEK inhibitor UO-126. Neurons were pretreated with UO-126 (30 mM) for 1 hr prior to stimulation. CREB and acetyl histone H4 immunoprecipitation followed by PCR revealed association of CREB with the c-fos promoter (upper panels) under both conditions (n = 4). (B) Immunoblot analysis of phospho-CREB (top panels), phospho-ERK (middle panels), and CREB (bottom) in lysates from neurons subjected to the same treatment conditions described in (A) (n = 3). (C) ChIP analysis of BDNF-treated neurons in the presence (right lanes) or absence (left lanes) of the calcium chelator BAPTA-AM (50 mM). ChIPs were performed with anti-CREB, anti-CBP, and anti-acetyl histone H4. PCR analysis of the c-fos gene revealed an inhibition of CREB and CBP binding and histone H4 acetylation (n = 3). (D) Immunoblot analysis of phospho-CREB (top) and CREB (bottom) in lysates from neurons subjected to the same treatment conditions described in (C). (E) Real-time PCR quantitation of ChIP analyses of cortical neurons treated with BDNF (75 ng/ml) for the indicated times after pretreatment with either vehicle control, the MEK inhibitor U0-126 (30 mM), or BAPTA-AM (BAPTA, 50 mM). Data are represented as fold changes over unstimulated cortical neurons (n = 3, *p < 0.01). Shown are the averages 6 SEM.

forskolin or NGF, respectively, promotes binding of myc-CREB to the endogenous c-fos promoter (Figures 4A and 4B). Thus, as with endogenous CREB, myc-

CREB is inducibly bound to the c-fos promoter. As expected, myc-CREB-DLZ and myc-CREBm1-DLZ, which lack the CREB DNA binding and dimerization domain,

NO Regulates CREB DNA Binding in Neurons 287

Figure 4. Stimulus-Dependent Increase in CREB DNA Binding Occurs Independently of Phosphorylation of CREB Ser133 (A) Stimulus-inducible CREB-DNA binding occurs independently of Ser-133. ChIP and immunoblot analyses of PC12 cells and HEK293 cells (B) treated with NGF (50 ng/ml) or forskolin (10 mM), respectively, for the indicated times. Cells were transiently transfected with the myc-tagged CREB constructs indicated. ChIP assays were carried out with an anti-myc antibody to detect the DNA binding activity of the transfected myc-CREB fusion proteins (n = 4). Abbreviation: WB, immunoblot.

fail to associate with promoter DNA. Both, mycCREBm1 and myc-CREBDIEDML do, however, exhibit stimulus-dependent DNA binding activity (Figures 4A and 4B). Taken together, these results indicate that the association between CREB and target gene promoters requires intracellular calcium signaling and occurs independently of the state of phosphorylation of Ser133 and, presumably, its association with CBP. An NO Signaling Pathway Regulates CREB Binding and Gene Expression, but Not Phosphorylation of CREB Ser133, in Cortical Neurons Because stimulus-dependent CREB-DNA binding occurs independently of phosphorylation of CREB Ser133 and the Ras/ERK signaling pathway that controls it, we wondered whether a noncanonical neurotrophin signaling pathway controls CREB binding activity. Activities of certain bZIP transcription factors, including members of the AP-1 family, are controlled through regulation of their redox state (Abate et al., 1990; Goren et al., 2001). Moreover, NO and reactive oxygen species physiologically regulate neuronal growth, cell survival, and differentiation (Bedogni et al., 2003; Ciani et al., 2002; Peunova and Enikolopov, 1995; Poluha et al., 1997; Yermolaieva et al., 2000), processes also controlled by CREB target genes (Lonze et al., 2002; Mantamadiotis et al., 2002). To ask whether neurotrophins directly and rapidly activate NOS and an NO signaling pathway, cortical neurons were incubated with the NOS substrate [3H] arginine and stimulated with BDNF to monitor the production of [3H] citrulline (Bredt and Snyder, 1990), a product of the NOS-catalyzed reaction. BDNF stimulation of cortical neurons leads to a rapid increase of NOS activity, comparable to that observed after ionomycin treatment of HEK293 cells that stably overexpress nNOS (Figure 5A). BDNF activation of NOS is also observed in neurons treated with the MEK inhibitor U0-126. As expected, [3H] citrulline production was abolished by pretreatment of neurons with NG-Monomethyl-L-Arginine (L-NMMA), an NOS inhibitor. Thus, BDNF activates NOS activity in cortical neurons. To determine whether an NO signaling pathway could influence CREB association with CREs in a manner independent of Ser133 phosphorylation, cortical neurons were treated with either NO donors or inhibitors (Figures

5B–5F). Indeed, exposure of neurons to either of two different inhibitors of NOS, NG-Nitro-L-Arginine-Methyl Ester (L-NAME) or L-NMMA, blocks BDNF-induced association of CREB with the c-fos and nNOS promoters (Figures 5B and 5D and A.R. and D.D.G., unpublished data). L-NAME was without effect on SRF binding to the c-fos promoter (A.R. and D.D.G., unpublished data). Conversely, treatment of cortical neurons with the NO donors N-ethyl-2 ethanamine-12 (NOC-12) and S-nitroso-N-acetylpenicillamine (SNAP) elicits an increase of CREB binding to its target gene promoters in the absence of BDNF stimulation (Figures 5E and 5D). Interestingly, NO also appears to alter the acetylation state of histones at the promoter. The NOS inhibitor LNAME causes a decrease in BDNF induction of histone acetylation (Figure 5B). This inhibition of acetylation of histones surrounding the c-fos and nNOS promoters is not due to an indiscriminate effect of L-NAME on general acetylase activity, because the overall level of histone H4 acetylation in treated cells is unchanged (Figure 5C). Similarly, a reduction in histone acetylation is observed in response to BDNF in primary cortical cultures obtained from nNOS mutant mice (Huang et al., 1993) (Figure 5G). These findings regarding CREB DNA binding and modulation of promoter acetylation state may explain observations in PC12 cells that CREB-dependent transcription is amplified by NO donors (Peunova and Enikolopov, 1993). Although a previous report suggested that NO signaling modulates CREB Ser133 phosphorylation (Ciani et al., 2002), we found that NO signaling is neither necessary nor sufficient for BDNF induction of CREB Ser133 phosphorylation in cultured cortical neurons (Figures 5C and 5F). Importantly, BDNF induced robust CREB Ser133 phosphorylation in cortical neurons obtained from nNOS mutant mice (Figure 5H). The basis for the discrepancy is not known but may be due to different populations of neurons and different stimuli used in Ciani et al. and the present study. To determine whether NO signaling affects BDNF-induced gene expression in cortical neurons, Northern blot analysis was used to examine the expression of the BDNF-inducible, CREB target gene c-fos. NO donors sodium nitroprusside (SNP) and SNAP augment BDNF-induced expression of c-fos, although NO treatment alone is insufficient for c-fos expression (Figures

Molecular Cell 288

Figure 5. NO Signaling Regulates CREB-DNA Binding (A) BDNF induces a rapid increase of NOS activity. Cortical neurons were either pretreated with the NOS inhibitor L-NMMA (500 mM) or UO-126 (30 mM) for 1 hour prior to stimulation or left untreated. Cells were stimulated with PBS or BDNF (75 ng/ml) for 30 min, as indicated. HEK 293 cells overexpressing nNOS were stimulated with ionomycin (5 mM). NOS activity was determined by assessing the conversion of [3H] arginine to [3H] citrulline (n = 3, *p < 0.05, **p < 0.001). Shown are the means 6 SEM. (B) ChIP analysis of cortical neurons stimulated with BDNF as indicated either in the presence (right panels) or absence (left panels) of L-NAME (2 mM). CREB and acetyl histone H4 immunoprecipitation was followed by PCR amplification of the c-fos (top panels) and nNOS (bottom panels) CREs (n = 5). (C) Immunoblot analysis of phospho-CREB, acetyl histone H4, and tubulin in lysates from neurons subjected to the same treatment conditions described in (B) (n = 3).

NO Regulates CREB DNA Binding in Neurons 289

6A and 6B). Conversely, when neurons are pretreated with the NOS inhibitor L-NMMA, BDNF-dependent expression of c-fos is attenuated (Figure 6C). The residual amount of expression of c-fos in L-NMMA-treated cultures is presumably due to other signaling events that control transcription via the c-fos SRE and other cis-regulatory elements. Thus, independent BDNF signaling pathways, the Ras/ERK signaling pathway, and a calcium/nNOS/NO signaling pathway converge on CREB to mediate distinct events that underlie expression of CREB target genes. Impaired Expression of CREB-Dependent Genes in Neurons of Mice Lacking nNOS We next asked whether NO regulates expression of CREB target genes in vivo. For these experiments, nNOS mutant mice and wild-type (wt) controls were exposed to novel enriched environmental (NEE) conditions, which are well known to activate expression of CREB target genes in neurons of the hippocampus and somatosensory cortex (Pinaud et al., 2001; Williams et al., 2001). NEE also elicits a variety of plastic responses, including increased dendritic arborization, synaptic density, neurogenesis, and memory functions in rodents, and has been associated with a lower incidence of neurodegenerative diseases in humans (reviewed in van Praag et al. [2000]). The mechanisms underlying these changes have not been fully elucidated, but enhanced synaptic activity and increased neurotrophin signaling are implicated (Mattson et al., 2001; van Praag et al., 2000). In situ hybridization analysis of brain sections revealed robust induction of c-fos and zif-268 in both cortical and hippocampal neurons of wt mice exposed to NEE (Figure 6D, left panels, and Figure S2, left panels); however, this induction is dramatically impaired in mice lacking nNOS (Figure 6D, right panels, and Figure S2, right panels). These findings indicate that NO signaling is required in vivo for expression of CREB target genes. BDNF Induces S-Nitrosylation of Nuclear Proteins Associated with CRE Containing Promoters NO may influence CREB DNA binding and gene expression through modulation of guanylate cyclase and cGMP signaling (Ahern et al., 2002; Arnold et al., 1977) or, alternatively, through direct S-nitrosylation of proteins (Jaffrey et al., 2001; Stamler et al., 2001) associated with the CREB complex. We therefore asked whether either of these processes mediates NO actions on CREB signaling. To determine whether NO influences CREB function through a cGMP signaling pathway, we em-

ployed the potent and selective PKG inhibitor DT3. Pretreatment of cortical neurons with DT3 was without effect on either BDNF induction of CREB DNA binding to the c-fos promoter (Figure 7A) or phosphorylation of CREB Ser133 (Figure 7B). These observations, together with the previous finding that S-nitrosylation directly controls the activity of the transcription factor NF-kB (Marshall and Stamler, 2001; Reynaert et al., 2004), prompted us to test the idea that NO controls CREB activity through S-nitrosylation events. Interestingly, BDNF treatment leads to S-nitrosylation of proteins associated with the c-fos, nNOS, and VGF promoters (Figure 7C and Figure S3). Pretreatment of cortical neurons with the NOS inhibitors L-NAME or L-NMMA abolishes S-nitrosylation of proteins associated with CRE sequences (Figure 7C). To identify nuclear proteins that are modified by NO, we next investigated whether CREB itself is nitrosylated upon BDNF stimulation. Nitrocysteine immunoprecipitation followed by immunoblot analysis have so far failed to reveal any evidence for S-nitrosylation of CREB. We did find that BDNF signaling led to S-nitrosylation of histone H3 (A.R. and D.D.G., unpublished data). The observation that NO signaling modulates histone H3 and H4 acetylation (Figures 5B, 5E, and 5G), together with the more general finding that acetylation of histones can influence chromatin structure and transcription factor binding to chromatin (Cheung et al., 2000), prompted us to explore whether NO signaling influences CREB DNA binding through the modulation of acetylation of histones associated with CRE-containing promoters. During transcriptional activation, histones and other chromatin proteins become acetylated by histone acetyltransferases, such as CBP and P/CAF, whereas the reverse reaction, the removal of acetyl groups from lysine residues, is catalyzed by histone deacetylases (HDACs). We therefore asked whether BDNF and NO signaling modulate the activity of one or more histone deacetylases that may, in turn, alter acetylation of histones associated with CREB target genes. HDAC2 is expressed in developing cortical neurons (Figure 7E), and ChIP experiments show that it is associated with CREB target promoters under resting conditions (Figure 7D). Binding of HDAC2 to CREB-regulated gene promoters is modulated by a BDNF-activated NO signaling pathway (Figure 7D). Upon exposure of cortical neurons to BDNF, HDAC2 rapidly dissociates from the c-fos promoter, coincident with induction of histone acetylation and CREB association. HDAC2 dissociation is blocked both by L-NAME and L-NMMA, and conversely, NO donors trigger the dissociation of HDAC2 from the

(D) Real-time PCR quantitation of ChIP analyses of cortical neurons treated with BDNF (75 ng/ml) for the indicated times after pretreatment with either vehicle control, L-NAME (2 mM), or NO donors (SNAP and NOC-12, 100 mM each). Data are represented as fold changes over unstimulated cortical neurons (n = 4, p < 0.01). Shown are the means 6 SEM. (E) ChIP analysis of cortical neurons stimulated with BDNF (75 ng/ml) as indicated in the presence of or absence of NO donors, N-Ethyl-2 ethanamine-12 (NOC-12, 100 mM), and S-Nitroso-N-acetylpenicillamine (SNAP, 100 mM). CREB and acetyl histone H4 immunoprecipitations were followed by PCR for the c-fos promoter (n = 3). (F) Phospho-CREB and total CREB immunoblot analysis of cortical neuron lysates that had been treated as described in (E) (n = 3). The small change in CREB phosphorylation in SNAP and NOC-12-treated cells was not consistently observed. (G) ChIP analysis of BDNF-treated cortical neurons derived from wt (left lanes) or nNOS null (right lanes) mouse embryos. Neurons were stimulated with BDNF (75 ng/ml) for the indicated times. Immunoprecipitations were performed with an acetyl histone H4 antibody followed by real-time PCR for the c-fos promoter (n = 2). (H) Phospho-CREB and CREB immunoblot analysis of cortical neuron obtained from wt (left panels) or nNOS null mice (right panels) treated with BDNF (75 ng/ml) for the indicated times (n = 2).

Molecular Cell 290

Figure 6. NO Influences the Expression of CREB Target Genes (A and B) Northern blot analysis of neurons treated simultaneously with BDNF and the NO donor SNAP. Neurons were stimulated with the indicated concentrations of SNAP in the presence or absence of BDNF (1 ng/ml) for 45 min (n = 3). Treatment with increasing concentrations of SNP (B) yielded similar results (n = 3). (C) Northern blot analysis of cortical neurons stimulated with BDNF for the indicated times either in the presence or absence of L-NMMA (500 mM). Cells were pretreated with L-NMMA or PBS for 30 min and stimulated with BDNF (75 ng/ml) for the times indicated (n = 4). Neurons pretreated with L-NAME (2 mM) yielded similar results. (D) Expression of c-fos in brain of mice lacking nNOS. In situ hybridization of sagittal sections of cortex obtained from 10–12-week-old mice either maintained in standard cages (top panels) or exposed to novel enriched environment (NEE) (bottom panels) for 45 min (N = 3). Exposure to NEE greatly enhances c-fos expression in cortical neurons. Note that the absence of nNOS results in a decrease in NEE-induced c-fos expression.

promoter in the absence of BDNF (Figure 7D). These findings suggest that in unstimulated neurons HDAC2, by deacetylating histones and possibly other substrates, prevents CREB from binding to its target genes.

Conceivably, upon BDNF stimulation, NO signaling triggers the release of HDAC2 from chromatin, facilitating histone acetylation and CREB binding to CRE-containing promoter sequences. In support of this notion,

NO Regulates CREB DNA Binding in Neurons 291

Figure 7. Nitrosylation of Proteins Associated with the c-fos CRE (A) CREB DNA binding is independent of cGMP/PKG signaling. Neurons were pretreated with DT3 (100 nM) for 60 min and treated with BDNF (75 ng/ml) for the indicated times. Immunoprecipitation of CREB and acetyl histone H4 were followed by PCR of the c-fos promoter (n = 3). (B) Immunoblot analysis of phospho-CREB, and CREB in lysates from neurons subjected to the same treatment conditions described in (A) (n = 3). (C) ChIP analysis of cortical neurons stimulated with BDNF (75 ng/ml) as indicated (lanes 1–4) or pretreated with either NOS inhibitors (L-NAME, 2 mM or L-NMMA, 500 mM) or NO donors (SNAP and NOC-12, 100 mM each) (lanes 5–7). Antinitro-cysteine immunoprecipitation of cell lysates were followed by PCR analysis of the c-fos promoter (n = 3). (D) NO regulates the binding of HDAC2 to the c-fos NO regulates the binding of HDAC2 to the c-fos promoter. Cortical neurons were stimulated with BDNF in the absence (lanes 1–4) or presence (lanes 5 and 6) of NOS inhibitors (L-NAME, 2 mM and L-NMMA, 500 mM). HDAC2 was immunoprecipitated, and PCR for c-fos was performed. NO donors were added to unstimulated cells (lane 7) (n = 2). (E) Expression of HDAC2 in cortical neurons treated with BDNF (75 ng/ml) for the indicated times in the absence (lanes 1–4) or presence (lane 5) of L-NAME (2 mM) (n = 2). (F) Inhibition of histone deacetylases induces CREB DNA binding. ChIP analysis of cortical neurons treated with trichostatin A at the indicated concentrations and immunoprecipitated with CREB and acetyl-histone H4 antibodies is shown. PCR of c-fos promoter is shown (n = 3).

treatment of cortical neurons with the deacetylase inhibitors trichostatin A (TSA, Figure 7F) or sodium butyrate (A.R. and D.D.G., unpublished data) leads to a rapid and robust association between CREB and c-fos promoter. Discussion In prevailing models of CREB activation, CREB is constitutively bound to CREs within target gene promoters with activation initiated by phosphorylation of Ser133, which in turn, recruits CBP to the promoter. By contrast,

we found that in developing cortical neurons, CREB binding to many endogenous promoter sequences is stimulus inducible and involves a BDNF/NO signaling pathway. Upon exposure of neurons to BDNF or other extracellular stimuli, CREB becomes bound to DNA coincident with phosphorylation at Ser133. Interestingly, inducible CREB-DNA binding is independent of CREB Ser133 phosphorylation and the signaling pathways that control it. Rather, BDNF regulates CREB binding through a nitric oxide signaling pathway that leads to Snitrosylation of nuclear proteins, at least some of which associate with CREB target gene promoters. Our results

Molecular Cell 292

support a model in which parallel signaling pathways, a Ras/ERK/CREB Ser133 phosphorylation pathway and an NO pathway that controls CREB-DNA binding, converge to mediate CRE-dependent gene expression in neurons. Previous work had suggested that CREB is constitutively bound to promoter sequences and that phosphorylation of its transcriptional regulatory site Ser133 is the predominant trigger of transcriptional activation (reviewed in Lonze and Ginty [2002]). Our findings from experiments employing cortical neurons and ChIP assays indicate that for many CREB target genes extracellular stimuli promote induction of CREB DNA binding. We found that in cortical neurons, extracellular stimuli, including BDNF and enhanced synaptic activity, trigger an increase in binding of CREB to the promoters of c-fos, nNOS, VGF, and zif-268. The ChIP assay used to determine the state of CREB DNA binding relies on the ability of antibodies to recognize CREB protein bound to genomic DNA after crosslinking. In principle, an increase in coprecipitation of CREB and genomic DNA fragments could be a reflection of increased CREB occupancy to the DNA binding site of interest or, alternatively, an alteration of the CREB complex and ‘‘unmasking’’ of the epitope recognized by the CREB antibody used for immunoprecipitation. Our findings favor the former. First, we have used three distinct CREB antibodies to show that BDNF treatment of cortical neurons leads to an increase in CREB DNA binding activity. The 1418 CREB antibody, used for the majority of experiments in this study, was raised against the N-terminal 205 amino acids of CREB (Ginty et al., 1993). A second antibody recognizes amino acids 5–24, and the third antibody is directed against the full-length CREB. ChIP experiments employing each of these antibodies reveal stimulus-dependent precipitation of CRE-containing promoter sequences. Second, an anti-myc antibody efficiently immunoprecipitates the myc-CREB-genomic DNA complex in a stimulus-dependent manner. Because the myc epitope tag is not part of the native CREB protein, this finding provides strong support for the idea that the observed stimulus-dependent changes in ChIP are a reflection of changes in CREB DNA binding as opposed to the inability of the antibody to recognize a masked epitope in unstimulated cells. Taken together, our findings indicate that in cortical neurons, CREB binding to cis-acting regulatory elements in target genes is regulated by several extracellular stimuli. What is the nature of the BDNF/TrkB/NOS signaling pathway? We propose that BDNF and other extracellular stimuli control NOS activation and NO-mediated CREB DNA binding via a calcium/calmodulin-NOS signaling pathway. NOS is a well-known target of calcium/ calmodulin (Bredt and Snyder, 1990), and our finding that chelation of intracellular calcium using BAPTA attenuates BDNF induction of CREB DNA binding points to a key role for calcium ions in this process. One possibility is that BDNF activates PLC-g1, a mediator of TrkB signaling (Finkbeiner et al., 1997; Huang and Reichardt, 2003; Minichiello et al., 2002), and that this in turn leads to production of IP3, the release of intracellular pools of calcium, and activation of nNOS. Consistent with this idea, we found that the Ras/MEK and PI3K pathways, which mediate many of the effects of BDNF and other

neurotrophins (Huang and Reichardt, 2003), are not necessary for BDNF’s effects on CREB DNA binding. Although the details of how BDNF activates NOS remain to be fully resolved, our findings do indicate that distinct TrkB signaling pathways converge at the level of CREB to support transcriptional initiation. Our findings demonstrate a function of NO signaling in the control of CREB signaling and neuronal gene expression. It is noteworthy that CREB has been implicated in many neuronal processes, including axon and dendrite growth, survival, and plasticity (reviewed in Lonze and Ginty [2002]). The high levels of nNOS expression in many populations of developing neurons fits with extensive evidence for a prominent role of NO in neuronal growth and survival (Bredt and Snyder, 1994). In developing neurons, this modulation occurs, at least in part, at the level of transcription of CREB target genes. Experimental Procedures Mouse Lines Creb heterozygote mice were provided by Gu¨nther Schu¨tz, German Cancer Research Center, Heidelberg, Germany. nNOS homozygous null animals or wt control animals were intercrossed to generate null and control litters, respectively. Embryos were harvested at E16.5 for preparation of primary cortical cultures. Cell Culture For HEK293 cell transfections, cells were incubated in serum-free DMEM containing 5–10 mg of DNA and 10 ml of Lipofectamine 2000 (Invitrogen), and incubated for 3 hr before replacing with serumcontaining media. Experiments were performed 1–2 days after transfection. For PC12 cell transfections, cells were incubated in serum-free DMEM containing 10–15 mg of DNA and 40 ml of Lipofectamine 2000, and incubated overnight before replacing with serumcontaining media described above. Prior to ChIP experiments, cells were starved in medium containing FBS (0.5%) overnight. Dissociated cultures of cortical neurons were prepared as described (Threadgill et al., 1997) and grown in MEM containing FBS (10%) and HS (5%). Prior to ChIP experiments, cells were placed in low serum medium overnight (5% FBS). Cells were treated for the times indicated with BDNF, bicuculline (50 mM, Tocris), or glutamate (50 mM, Sigma). Primary cultures of fibroblasts were derived from E15.5 mouse embryos. Skin was removed from embryos, minced, and subjected to five sequential digestions in Trypsin/EDTA (Invitrogen). Undigested tissue was removed, and suspended cells were pelleted by centrifugation, rinsed, and resuspended in the culture medium, DMEM containing FBS (10%). ChIPs ChIP assays were performed largely as described (Weinmann and Farnham, 2002; Wells and Farnham, 2002). For all cell types, w3–5 3 106 cells were used per ChIP. Briefly, medium was removed from treated cells and replaced with PBS containing formaldehyde (1%). Cells were rinsed twice in PBS and harvested in harvesting buffer containing Tris-HCl (pH 9.4, 100 mM) and DTT (1 mM). Cells were collected by centrifugation and rinsed with PBS, and pellets were resuspended in lysis buffer containing SDS (0.1%), Triton X-100 (0.5%), Tris-HCl (pH 8.1, 20 mM), and NaCl (150 mM). Samples were sonicated with six 20 s pulses with a 10 s interpulse interval. Cell debris was removed by centrifugation, and supernatants were precleared by incubation with protein-A Sepharose beads (Amersham Biosciences) for 1 hr at 4ºC. Beads were collected by centrifugation, and supernatants were subjected to immunoprecipitation. A fraction of the supernatant was used for immunoprecipitation input control. The volume of each tube was adjusted to 500 ml with lysis buffer, and 5–10 mg of rabbit polyclonal antibody was added overnight at 4ºC. The following polyclonal antibodies were used: anti CREB 1418 (Ginty et al., 1993), anti-CREB (Upstate), anti-CREB (Cell Signaling), anti-SRF and anti-CBP (Santa Cruz), anti-acetyl histone H3 (Upstate), anti-acetyl histone H4 (Upstate), anti- HDAC2

NO Regulates CREB DNA Binding in Neurons 293

(Santa Cruz), control IgG (preimmune), anti-myc (Cell Signaling), and anti-nitrocysteine (A.G. Scientific). Immune complexes were collected by incubation with protein-A Sepharose beads for 1 hr at 4ºC. Beads were collected and subjected to a series of seven sequential washes. Two washes each were performed in lysis buffer and then in washing buffer containing SDS (0.1%), Triton X-100 (0.5%), EDTA (pH 8.0, 2 mM), Tris-HCl (pH 8.1, 20 mM), and NaCl (150 mM). One wash was performed in lithium buffer containing LiCl (0.25 M), NP-40 (1%), deoxycholate (1%), EDTA (pH 8.0, 1 mM), and Tris-HCl (pH 8.1, 10 mM). Two final washes were preformed in 13 TE (pH 8.1). Immune complexes were eluted from beads by vortexing in elution buffer containing SDS (1%) and NaHCO3 (pH 8.0, 0.1 M). NaCl was added (final concentration 0.33 M), and crosslinking was reversed by incubation overnight at 65ºC. DNA fragments were purified by using the QIAquick PCR purification kit (Qiagen). For PCR, specific sets of primers were designed that flank CRE regions within the upstream regulatory regions of the indicated genes. PCR conditions and cycle numbers were determined empirically for the different templates and primer pairs. Primers amplified fragments ranging in size from 200 to 400 bp. Primer sequences and PCR conditions are available on request.

Quantitative PCR PCR reactions (25 mL) contained 12.5 ml of PCR Sybr Green mix (NEB) and 0.3 mM primers. All reactions were performed in duplicate with an Opticon 2 System (MJ Research, Cambridge, MA), and each experiment included a standard curve, a preimmune control, and a no template control. Standard templates consisted of gel-purified PCR products of c-fos, nNOS, and VGF amplicons of known concentration, and each standard curve consisted of eight serial dilutions of template. At the end of 46 cycles of amplification, a dissociation curve was performed in which Sybr Green was measured at 1ºC interval between 50ºC and 100ºC. Melting temperatures for c-fos, nNOS, and VGF were 83ºC, 85ºC, and 92ºC, respectively. Results were normalized by total input DNA and expressed as fold changes over unstimulated control.

Immunoblotting Immunoblotting was performed with whole-cell extracts and antibodies against CREB (Cell Signaling), phospho-CREB (Upstate), phospho-ERKs (Cell Signaling), tubulin (Sigma), acetyl histone H4 (Upstate), HDAC2 (Zymed), and myc (Cell Signaling). HRP-conjugated anti-mouse IgG or anti-rabbit IgG secondary antibodies (Amersham Biosciences) were used, and detection was performed by using SuperSignal (Pierce) or ECL plus (Amersham Biosciences) substrates.

Supplemental Data Supplemental Data include three figures and can be found with this article online at http://www.molecule.org/cgi/content/full/21/2/283/ DC1/. Acknowledgments We are grateful to members of the Ginty laboratory, Seth Blackshaw and Joe Hurt for helpful discussions, and we thank Rejji Kuruvilla and Giovanni Pani for insightful comments on this manuscript. We thank Xi Chen and Catia Andreassi for help with the immunoblot analysis. Creb mutant mice were provided by Gu¨nther Schu¨tz and the Deutsches Krebsforschungszentrum. This work was supported by National Institutes of Health grants NS34814 (D.D.G.), NS37090, and 00266 (T.M.D.), a Conte Center for neuroscience research grant, and the Medical Scientist Training Program (B.E.L.). T.M.D. is the Leonard and Madlyn Professor of Neurodegenerative Diseases. D.D.G. is an Investigator of the Howard Hughes Medical Institute. Received: July 21, 2005 Revised: October 14, 2005 Accepted: December 2, 2005 Published: January 19, 2006 References Abate, C., Patel, L., Rauscher, F.J., 3rd, and Curran, T. (1990). Redox regulation of fos and jun DNA-binding activity in vitro. Science 249, 1157–1161. Ahern, G.P., Klyachko, V.A., and Jackson, M.B. (2002). cGMP and S-nitrosylation: two routes for modulation of neuronal excitability by NO. Trends Neurosci. 25, 510–517. Arnold, W.P., Mittal, C.K., Katsuki, S., and Murad, F. (1977). Nitric oxide activates guanylate cyclase and increases guanosine 30 :50 -cyclic monophosphate levels in various tissue preparations. Proc. Natl. Acad. Sci. USA 74, 3203–3207. Bading, H., Ginty, D.D., and Greenberg, M.E. (1993). Regulation of gene expression in hippocampal neurons by distinct calcium signaling pathway. Science 260, 181–186. Bedogni, B., Pani, G., Colavitti, R., Riccio, A., Borrello, S., Murphy, M., Smith, R., Eboli, M.L., and Galeotti, T. (2003). Redox regulation of cAMP-responsive element-binding protein and induction of manganous superoxide dismutase in nerve growth factor-dependent cell survival. J. Biol. Chem. 278, 16510–16519. Bredt, D.S., and Snyder, S.H. (1990). Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc. Natl. Acad. Sci. USA 87, 682–685.

Generation of CREB Constructs All epitope-tagged constructs were created with a vector encoding the N-terminal c-Myc epitope (pLP-CMV-Myc, Clontech). CREB-WT, CREB-DLZ, CREBm1, and CREBDIEDML full-length cDNAs were amplified by specific primers, and the PCR product was cloned into the pLP-CMV-Myc vector polylinker.

Bredt, D.S., and Snyder, S.H. (1994). Transient nitric oxide synthase neurons in embryonic cerebral cortical plate, sensory ganglia, and olfactory epithelium. Neuron 13, 301–313.

NOS Activity Cortical neurons were grown in 10 cm dishes for 5–7 days and labeled with [3H] arginine (10 mCi per plate) for 1 hr. Cells were treated as indicated and then rinsed with ACSF, harvested in 3 ml methanol, dried, and resuspended in buffer containing 20 mM Tris (pH 7.4). Fractions of the supernatants (500 ml) were loaded onto chromatography columns (Dowex 50WX8-400, Sigma in 20 mM Tris [pH 7.4]), and [3H] citrulline was eluted and quantified (Bredt and Snyder, 1990).

Cheung, P., Allis, C.D., and Sassone-Corsi, P. (2000). Signaling to chromatin through histone modifications. Cell 103, 263–271.

In Situ Hybridization Mice lacking nNOS and age-matched wt animals were maintained in standard cages for 10–12 weeks and exposed to NEE conditions for 45 min. Animals were sacrificed, and 16 micron sections were processed for in situ hybridization as described (Pasterkamp et al., 1998).

Gau, D., Lemberger, T., von Gall, C., Kretz, O., Le Minh, N., Gass, P., Schmid, W., Schibler, U., Korf, H.W., and Schutz, G. (2002). Phosphorylation of CREB Ser142 regulates light-induced phase shifts of the circadian clock. Neuron 34, 245–253.

Cardinaux, J.R., Notis, J.C., Zhang, Q., Vo, N., Craig, J.C., Fass, D.M., Brennan, R.G., and Goodman, R.H. (2000). Recruitment of CREB binding protein is sufficient for CREB-mediated gene activation. Mol. Cell. Biol. 20, 1546–1552.

Ciani, E., Guidi, S., Bartesaghi, R., and Contestabile, A. (2002). Nitric oxide regulates cGMP-dependent cAMP-responsive element binding protein phosphorylation and Bcl-2 expression in cerebellar neurons: implication for a survival role of nitric oxide. J. Neurochem. 82, 1282–1289. Finkbeiner, S., Tavazoie, S.F., Maloratsky, A., Jacobs, K.M., Harris, K.M., and Greenberg, M.E. (1997). CREB: a major mediator of neuronal neurotrophin responses. Neuron 19, 1031–1047.

Ginty, D.D., Kornhauser, J.M., Thompson, M.A., Bading, H., Mayo, K.E., Takahashi, J.S., and Greenberg, M.E. (1993). Regulation of

Molecular Cell 294

CREB phosphorylation in the suprachiasmatic nucleus by light and a circadian clock. Science 260, 238–241. Gonzalez, G.A., and Montminy, M.R. (1989). Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at Serine 133. Cell 59, 675–680. Goren, I., Tavor, E., Goldblum, A., and Honigman, A. (2001). Two cysteine residues in the DNA-binding domain of CREB control binding to CRE and CREB-mediated gene expression. J. Mol. Biol. 313, 695–709. Hardingham, G.E., Fukunaga, Y., and Bading, H. (2002). Extrasynaptic NMDARs oppose synaptic NMDARs by triggering CREB shut-off and cell death pathways. Nat. Neurosci. 5, 405–414. Hawley, R.J., Scheibe, R.J., and Wagner, J.A. (1992). Nerve growth factor induces the expression of the VGF gene through a cAMP response element. J. Neurosci. 12, 2573–2581. Huang, E.J., and Reichardt, L.F. (2001). Neurotrophins: roles in neuronal development and function. Annu. Rev. Neurosci. 24, 667–736. Huang, E.J., and Reichardt, L.F. (2003). Trk receptors: roles in neuronal signal transduction. Annu. Rev. Biochem. 72, 609–642. Huang, P.L., Dawson, T.M., Bredt, D.S., Snyder, S.H., and Fishman, M.C. (1993). Targeted disruption of the neuronal nitric oxide synthase gene. Cell 75, 1273–1286. Impey, S., McCorkle, S.R., Cha-Molstad, H., Dwyer, J.M., Yochum, G.S., Boss, J.M., McWeeney, S., Dunn, J.J., Mandel, G., and Goodman, R.H. (2004). Defining the CREB regulon: a genome-wide analysis of transcription factor regulatory regions. Cell 119, 1041–1054. Jaffrey, S.R., Erdjument-Bromage, H., Ferris, C.D., Tempst, P., and Snyder, S.H. (2001). Protein S-nitrosylation: a physiological signal for neuronal nitric oxide. Nat. Cell Biol. 3, 193–197. Kim, H.S., Lee, S.J., Kim, D.S., and Cho, H.J. (2000). Effects of brainderived neurotrophic factor and neurotrophin-3 on expression of mRNAs encoding c-Fos, neuropeptides and glutamic acid decarboxylase in cultured spinal neurons. Neuroreport 11, 3873–3876. Kornhauser, J.M., Cowan, C.W., Shaywitz, A.J., Dolmetsch, R.E., Griffith, E.C., Hu, L.S., Haddad, C., Xia, Z., and Greenberg, M.E. (2002). CREB transcriptional activity in neurons is regulated by multiple, calcium-specific phosphorylation events. Neuron 34, 221–233. Lonze, B.E., and Ginty, D.D. (2002). Function and regulation of CREB family transcription factors in the nervous system. Neuron 35, 605– 623.

Pinaud, R., Penner, M.R., Robertson, H.A., and Currie, R.W. (2001). Upregulation of the immediate early gene arc in the brains of rats exposed to environmental enrichment: implications for molecular plasticity. Brain Res. Mol. Brain Res. 91, 50–56. Poluha, W., Schonhoff, C.M., Harrington, K.S., Lachyankar, M.B., Crosbie, N.E., Bulseco, D.A., and Ross, A.H. (1997). A novel, nerve growth factor-activated pathway involving nitric oxide, p53, and p21WAF1 regulates neuronal differentiation of PC12 cells. J. Biol. Chem. 272, 24002–24007. Reynaert, N.L., Ckless, K., Korn, S.H., Vos, N., Guala, A.S., Wouters, E.F., van der Vliet, A., and Janssen-Heininger, Y.M. (2004). Nitric oxide represses inhibitory kappaB kinase through S-nitrosylation. Proc. Natl. Acad. Sci. USA 101, 8945–8950. Rudolph, D., Tafuri, A., Gass, P., Hammerling, G.J., Arnold, B., and Schutz, G. (1998). Impaired fetal T cell development and perinatal lethality in mice lacking the cAMP response element binding protein. Proc. Natl. Acad. Sci. USA 95, 4481–4486. Sala, C., Rudolph-Correia, S., and Sheng, M. (2000). Developmentally regulated NMDA receptor-dependent dephosphorylation of cAMP response element-binding protein (CREB) in hippocampal neurons. J. Neurosci. 20, 3529–3536. Sasaki, M., Gonzalez-Zulueta, M., Huang, H., Herring, W.J., Ahn, S., Ginty, D.D., Dawson, V.L., and Dawson, T.M. (2000). Dynamic regulation of neuronal NO synthase transcription by calcium influx through a CREB family transcription factor-dependent mechanism. Proc. Natl. Acad. Sci. USA 97, 8617–8622. Segal, R.A. (2003). Selectivity in neurotrophin signaling: theme and variations. Annu. Rev. Neurosci. 26, 299–330. Shaywitz, A.J., and Greenberg, M.E. (1999). CREB: a stimulusinduced transcription factor activated by a diverse array of extracellular signals. Annu. Rev. Biochem. 68, 821–861. Stamler, J.S., Lamas, S., and Fang, F.C. (2001). Nitrosylation. the prototypic redox-based signaling mechanism. Cell 106, 675–683. Threadgill, R., Bobb, K., and Ghosh, A. (1997). Regulation of dendritic growth and remodeling by Rho, Rac, and Cdc42. Neuron 19, 625–634. van Praag, H., Kempermann, G., and Gage, F.H. (2000). Neural consequences of environmental enrichment. Nat. Rev. Neurosci. 1, 191– 198. Vo, N., and Goodman, R.H. (2001). CREB-binding protein and p300 in transcriptional regulation. J. Biol. Chem. 276, 13505–13508.

Lonze, B.E., Riccio, A., Cohen, S., and Ginty, D.D. (2002). Apoptosis, axonal growth defects, and degeneration of peripheral neurons in mice lacking CREB. Neuron 34, 371–385.

Weih, F., Stewart, A.F., Boshart, M., Nitsch, D., and Schutz, G. (1990). In vivo monitoring of a cAMP-stimulated DNA-binding activity. Genes Dev. 4, 1437–1449.

Mantamadiotis, T., Lemberger, T., Bleckmann, S.C., Kern, H., Kretz, O., Martin Villalba, A., Tronche, F., Kellendonk, C., Gau, D., Kapfhammer, J., et al. (2002). Disruption of CREB function in brain leads to neurodegeneration. Nat. Genet. 31, 47–54.

Weinmann, A.S., and Farnham, P.J. (2002). Identification of unknown target genes of human transcription factors using chromatin immunoprecipitation. Methods 26, 37–47.

Marshall, H.E., and Stamler, J.S. (2001). Inhibition of NF-kappa B by S-nitrosylation. Biochemistry 40, 1688–1693.

Wells, J., and Farnham, P.J. (2002). Characterizing transcription factor binding sites using formaldehyde crosslinking and immunoprecipitation. Methods 26, 48–56.

Mattson, M.P., Duan, W., Lee, J., and Guo, Z. (2001). Suppression of brain aging and neurodegenerative disorders by dietary restriction and environmental enrichment: molecular mechanisms. Mech. Ageing Dev. 122, 757–778.

Williams, B.M., Luo, Y., Ward, C., Redd, K., Gibson, R., Kuczaj, S.A., and McCoy, J.G. (2001). Environmental enrichment: effects on spatial memory and hippocampal CREB immunoreactivity. Physiol. Behav. 73, 649–658.

Mayr, B., and Montminy, M. (2001). Transcriptional regulation by the phosphorylation-dependent factor CREB. Nat. Rev. Mol. Cell Biol. 2, 599–609.

Yermolaieva, O., Brot, N., Weissbach, H., Heinemann, S.H., and Hoshi, T. (2000). Reactive oxygen species and nitric oxide mediate plasticity of neuronal calcium signaling. Proc. Natl. Acad. Sci. USA 97, 448–453.

Minichiello, L., Calella, A.M., Medina, D.L., Bonhoeffer, T., Klein, R., and Korte, M. (2002). Mechanism of TrkB-mediated hippocampal long-term potentiation. Neuron 36, 121–137. Pasterkamp, R.J., De Winter, F., Holtmaat, A.J., and Verhaagen, J. (1998). Evidence for a role of the chemorepellent semaphorin III and its receptor neuropilin-1 in the regeneration of primary olfactory axons. J. Neurosci. 18, 9962–9976. Peunova, N., and Enikolopov, G. (1993). Amplification of calcium-induced gene transcription by nitric oxide in neuronal cells. Nature 364, 450–453. Peunova, N., and Enikolopov, G. (1995). Nitric oxide triggers a switch to growth arrest during differentiation of neuronal cells. Nature 375, 68–73.