Introduction

Neurotransmission at synapses requires the conversion of an electrical signal into a chemical signal via neurotransmitter release. In the postsynaptic cell, neurotransmitter receptors convert back this chemical signal into an electrical signal. Correct transmission of the information emitted by the presynaptic neuron requires a highly specialized postsynaptic network. Among the proteins accumulated at the neuromuscular junction (NMJ), the nicotinic acetylcholine receptor (AChR) plays a central role as it binds acetylcholine, the neurotransmitter released by the nerve, and allows the transmission of the nerve influx to the muscle.

The AChR is a heteropentameric cationic channel composed of four subunits, ₂ in the embryo and ₂ in adult muscle (Sanes and Lichtman, 2001). The current model stipulates that AChR accumulation at the NMJ results from the combined action of several distinct mechanisms that all depend on the activation of the muscle tyrosine kinase receptor MuSK by the neural factor agrin (Sanes and Lichtman, 1999). The first consequence of MuSK activation is the physical clustering of AChR at the NMJ via the adaptator protein Rapsyn. Besides protein aggregation, the clustering of synaptic proteins is also paralleled by the compartmentalization of gene expression. Upon acetylcholine binding, AChR molecules aggregated at the NMJ trigger membrane depolarization and electrical activity in the muscle cell, thereby inhibiting the expression of the AchR , , and subunit genes (Sanes and Lichtman, 1999). Finally, the activation of MuSK, either directly or indirectly via neuregulin, locally triggers the MAPK and JNK intracellular signalling pathways that strongly activate the expression of the genes coding for the components of the NMJ (the synaptic genes) in the nuclei located directly beneath the NMJ (Schaeffer et al, 2001; Lacazette et al, 2003). The downstream nuclear target of these signalling pathways is the Ets-related transcription factor GABP, which binds the N box present in several synaptic genes, including AChR genes (Koike et al, 1995; Duclert et al, 1996; Fromm and Burden, 1998, 2001; Schaeffer et al, 1998; Gramolini et al, 1999; Briguet and Ruegg, 2000; de Kerchove et al, 2002). The motor neuron thus produces two anterograde signals that control AChR expression, agrin and acetylcholine, which respectively activate AChR gene expression in subsynaptic nuclei and inhibit AChR gene expression in extrasynaptic nuclei.

In the nucleus, DNA is embedded in chromatin, a repetitive nucleoproteic structure. Nucleosomes are formed by the wrapping of 146 DNA base pairs around a histone octamer and constitute the basic unit of chromatin. The N-terminal tails of histones are subject to numerous post-translational modifications such as phosphorylation, ubiquitination, ADP ribosylation, glycosylation, methylation and acetylation (Fischle et al, 2003; Margueron et al, 2005). These modifications can facilitate or inhibit transcription, DNA repair, DNA replication, DNA condensation and chromosome segregation. Histone H3 phosphorylation on serine 10 (PH3 S10) is well known to decorate chromosomes during mitosis (Gurley et al, 1978; Hsu et al, 2000; Murnion et al, 2001; Crosio et al, 2002), but this modification has also been reported to be associated with the acetylation and activation of the immediate-early gene promoter cJun (Sassone-Corsi et al, 1999; Thomson et al, 1999). Lysine acetylation on H3 and H4 histone tails is the best-characterized histone modification, and hyperacetylated histones are found in active domains of chromatin, whereas hypoacetylated histones are associated to the silent regions (Wolffe and Kurumizaka, 1998). Histone acetylation levels are regulated by two antagonistic enzymatic activities carried by two classes of transcription factors: histone acetyltransferases (HAT) and histone deacetylases (HDAC).

We have previously demonstrated that the inhibition of AChR expression in extrasynaptic nuclei involved the control of muscle chromatin acetylation by electrical activity via HDACs (Mejat et al, 2005). In subsynaptic nuclei, AChR genes expression is resistant to electrical activity. In the present study, we show that the specific gene expression profile of subsynaptic nuclei correlates with a particular state of chromatin. The observation of NMJ electron micrographs reveals different chromatin organizations in subsynaptic and extrasynaptic nuclei. This implies that the chromatin remodelling machinery is differentially regulated in muscle subsynaptic nuclei. In subsynaptic nuclei, chromatin appeared to be less condensed, a feature that could correlate with increased histone acetylation. We have therefore sought to determine if the particular subsynaptic chromatin organization correlated with increased histone acetylation, and if specific HATs and/or HDACs were involved in the regulation of AChR expression at the NMJ and thus participated in the specification of the unique gene expression program of subsynaptic nuclei.

Our results show that chromatin is hyperacetylated and hyperphosphorylated in subsynaptic nuclei, and that the HAT activity of the coactivator p300 is required for AChR genes expression at the NMJ. They further demonstrate that histone hyperacetylation on AChR promoters is induced by the neural factors agrin and neuregulin. Finally, our results show that histone acetylation on AChR genes is due to the loss of HDAC1 and the recruitment of the histone acetyl transferase p300 by GABP, whereas in the absence of activation GABP rather recruits the histone deacetylase HDAC1.

Results

The postsynaptic chromatin structure is different in subsynaptic nuclei

The specific morphology of subsynaptic nuclei was already noticed by Ranvier in 1875, who described them as fundamental nuclei. The observation of muscle nuclei by electron microscopy reveals that in subsynaptic nuclei, chromatin appears much less dense to electrons, thus suggesting that chromatin compaction is lower in these nuclei (Figure 1). Histone acetylation being a marker of relaxed chromatin, antibodies directed against hyperacetylated histone H3 were used to perform immunofluorescence experiments on isolated muscle fibres. The results showed a strong labelling of subsynaptic nuclei, whereas extrasynaptic nuclei were hardly detected (Figure 2A and B).

Phosphorylation of the serine 10 of H3 was reported to be associated with the acetylation of the neighbouring lysine on immediate-early genes (Cheung et al, 2000; Clayton et al, 2000; Lo et al, 2000). Using an antibody directed against histone H3 both phosphorylated on serine 10 and acetylated on lysine 14 (PAcH3 S10 K14), we indeed observed that histone H3 was strongly phosphoacetylated in subsynaptic nuclei, whereas these modifications were not detected in extrasynaptic nuclei (Figure 2C and D).

Synaptic gene activation in subsynaptic nuclei depends on MuSK and agrin, and ectopic expression of agrin is sufficient to induce ectopic postsynaptic specialization, with AChR clustering and synaptic gene activation. To determine whether histone hyperacetylation and hyperphosphorylation were also triggered by agrin, ectopic synapses were induced to form in the soleus muscle by injection of purified neural agrin. Two weeks after agrin injection, -bungarotoxin labelling showed ectopic AChR accumulation. Immunofluorescence experiments performed on these ectopic postsynaptic structures with anti-acetylated histone H3 and anti-phosphoacetylated histone H3 antibodies respectively demonstrated histone hyperacetylation and histone phosphoacetylation in the nuclei located beneath AChR clusters (Figure 3A and B).

Histone H3 acetylation and phosphoacetylation are triggered by neuregulin in cultured myotubes

C2C12 myotubes have been extensively used to dissect the mechanisms of synaptic genes activation in response to neural cues (Schaeffer et al, 2001; Mejat et al, 2003). However, although neural agrin induced activation of episomic AChR reporter constructs in C2C12 myotubes, as previously described, we could not see any significant activation of endogenous AChR subunit gene expression (data not shown). This prevented the use of agrin to observe histone acetylation at genomic loci in cultured myotubes. Conversely, neuregulin activates both episomic and endogenous AChR subunit gene expression in C2C12 myotubes (Mejat et al, 2003). In addition, similar to agrin, the pathways triggered by neuregulin to activate AChR gene expression converge on GABP (Schaeffer et al, 1998; Fromm and Burden, 2001; Lacazette et al, 2003). Moreover, in cultured myotubes, neuregulin and agrin activate AChR expression through a common pathway, neuregulin acting as agrin second messenger, as agrin-dependent AChR transcriptional activation is blocked by erbB dominant-negative mutants (Meier et al, 1998; Lacazette et al, 2003). Neuregulin was therefore used to study histone acetylation upon AChR gene activation in C2C12 myotubes.

In C2C12 cells, the activation of AChR expression is a biphasic process, with an early phase corresponding to the activation of the immediate-early genes cfos and cJun, and a late phase, in which AChR genes are activated (Si et al, 1999). To determine if AChR activation correlated with chromatin acetylation in cultured muscle cells, C2C12 myotubes were treated with recombinant neuregulin, and histone H3 phosphorylation and acetylation were evaluated by Western blot. Figure 4A shows that histone H3 acetylation gradually increased after neuregulin treatment. The analysis of histone H3 serine 10 phosphorylation revealed two peaks of phosphorylation; a first peak 15 min after neuregulin addition and a second peak 4–6 h later (Figure 4A). The two phases of histone modifications occurred concomitantly to the two waves of gene activation previously described by Si et al (1999), therefore suggesting that they could affect different sets of genes, that is, immediate-early genes and AChR genes. The rather high basal level of histone acetylation in C2C12 cells observed in Western blot could mask the presence of two acetylation peaks linked to neuregulin treatment. Chromatin immunoprecipitation (ChIP) was thus used to quantitate the levels of histone acetylation on the promoters of cJun and AChR subunit genes at different times after neuregulin addition to the culture medium. The results showed that on the cJun promoter, histone H3 acetylation transiently increased during the first half an hour, whereas on the AChR promoter histone H3 acetylation increased 4 h later and remained elevated thereafter (Figure 4B). The initial phase of histone acetylation therefore corresponds to the transient activation of immediate-early genes, and the second phase corresponds to the stable activation of AChR gene expression.

HDAC1 is recruited on AChR genes

We next sought to identify the histone-modifying activities involved in the neuronal control of postsynaptic chromatin acetylation. In a first attempt, synaptic and extrasynaptic zones of diaphragm muscles were microdissected and the expression levels of several HATs (p300, PCAF and GCN5) and HDACs were measured by quantitative RT–PCR. No differential expression could be detected. The amount of mRNA coding these factors was also quantified in neuregulin-treated or control myotubes, and neuregulin did not affect their expression (data not shown).

Previous studies on class I HDACs report that their regulation is rather post-translational than transcriptional (Zhou et al, 2000; Pflum et al, 2001; Galasinski et al, 2002; Qiu et al, 2006; Vashisht Gopal et al, 2006). To follow the evolution of class I HDACs protein levels during AChR activation, C2C12 myotubes were treated with neuregulin and analysed by immunofluorescence. In the case of HDAC1, neuregulin induced a marked decrease in the immunostaining of the protein, whereas the levels of HDAC2 and HDAC3 were not affected (Figure 5A; data not shown). This suggested that neuregulin could potentially induce a decrease in the amount of HDAC1 in the myotubes. In addition, the reduction of HDAC1 immunostaining was accompanied by a redistribution of the protein in the nucleus, with a dramatic decrease in the number of immunoreactive spots (Figure 5A).

To determine if the reduction of the immunostaining resulted from a reduction of the amount of HDAC1, or a conformational change, or from the masking of its epitope by a new partner, HDAC1 levels were evaluated by Western blot. Consistent with the immunofluorescence results, the amount of HDAC1 started to decrease 4 h after neuregulin addition (Figure 5B).

The reduction in the amount of HDAC1 took place 4 h after the addition of neuregulin, precisely when H3 acetylation increased on the AChR promoter, thus suggesting that HDAC1 could participate in the repression of this gene. To determine if HDAC1 was present on the AChR promoter, and if the activation of AChR expression correlated with a displacement of HDAC1, ChIP experiments were performed with an anti-HDAC1 antibody. The results indicated that HDAC1 was indeed present on the AChR subunit promoter, and that neuregulin induced a two-fold decrease in the amount of HDAC1 bound to the promoter (Figure 5C).

HDAC1 thus probably participates in the repression of chromatin acetylation and AChR gene expression in the absence of neural factors. However, removal of HDACs is usually not sufficient to obtain histone hyperacetylation, and HAT recruitment is also required.

p300 participates in the activation of AChR genes expression

The best-characterized transcription factor involved in the activation of synaptic genes at the NMJ is GABP (Schaeffer et al, 2001; Mejat et al, 2003). GABP also activates gene expression in many non-muscle cells. GABP has been shown to cooperate with the HAT p300 for the activation of IL16 gene in T lymphocytes (Bannert et al, 1999), and for the activation of the CD18 gene in myeloid cells (Bush et al, 2003; Resendes and Rosmarin, 2006). The recruitment of p300 on the AChR subunit gene promoter upon activation by neuregulin was therefore evaluated by ChIP. As shown in Figure 6A, p300 was indeed recruited to the AChR promoter in the presence of neuregulin.

In addition to its HAT activity, p300 possesses several activities and functions, and some of its coactivating functions have been shown to be HAT independent (Chan and La Thangue, 2001). Lysine–coenzymeA (LysCoA) has been shown previously to specifically inhibit the HAT activity of p300 (Lau et al, 2000; Polesskaya et al, 2001). However, the use of this inhibitor has been limited in vivo by the fact that it is not cell-permeant. To circumvent this problem, we took advantage of the susceptibility of skeletal muscle to in vivo electroporation. To determine whether the HAT activity of p300 was required for AChR gene expression, LysCoA was electroporated in Tibialis anterior muscles and the expression of the AChR subunit gene was measured by quantitative RT–PCR. The results were normalized to the effect of the inhibitor on -actin mRNA levels. Two other genes (cycloB and HPRT) were also used for normalization and produced similar results (data not shown). In the presence of LysCoA, the AChR subunit gene mRNA dropped to less than 30% of its original level in 24 (Figure 6B) and 48 h (data not shown), whereas it was not affected by the electroporation of a mixture of Lysine and coenzyme A. In skeletal muscle, the AChR subunit gene is expressed only in subsynaptic nuclei. The reduction of AChR mRNA levels thus resulted from the inhibition of the expression of the AChR subunit gene in subsynaptic nuclei. p300 HAT activity is thus crucial for AChR expression at the NMJ.

To confirm this result, LysCoA was also coelectroporated with a reporter plasmid coding for the luciferase placed under the control of the mouse AChR promoter. As shown in Figure 6C, the presence of LysCoA strongly inhibited the expression of luciferase at 24 and 48 h. Even with episomic constructs, the HAT activity of p300 is therefore required to activate transcription through the AChR promoter.

GABP selectively interacts with p300 or HDAC1

In the previous sections we have demonstrated that HDAC1 and p300 were differentially recruited on the AChR promoter in the presence or absence of neuregulin.

The transcription factor GABP is present both in synaptic and extrasynaptic nuclei, but nevertheless specifically activates the expression of synaptic genes in subsynaptic nuclei (Schaeffer et al, 2001; Mejat et al, 2003). We therefore examined the possibility that GABP could differentially recruit p300 or HDAC1, depending on the activation/repression context of synaptic genes. For this purpose, co-immunoprecipitation experiments were performed to detect interactions between GABP and p300, or HDAC1. The experiments were first performed with transfected tagged versions of p300 and HDAC1, and the results demonstrated that GABP could indeed interact with either p300 or HDAC1 (Figure 7A). To determine whether the interaction between GABP and p300, or HDAC1, correlated with the activation/repression status of synaptic genes, the co-immunoprecipitation experiments were performed with the endogenous proteins, in control and neuregulin-treated myotubes. The results show that GABP binds p300 more efficiently in the presence of neuregulin, whereas it binds more HDAC1 in the absence of neuregulin (Figure 7B).

Therefore, in conditions where synaptic genes are activated, GABP recruits the HAT p300, whereas in conditions where they are repressed, GABP recruits the histone deacetylase HDCA1.

Discussion

Altogether, our results demonstrate that chromatin is hyperphosphorylated and hyperacetylated in subsynaptic nuclei. This is consistent with our finding that in cultured cells the activation of the AChR subunit gene coincides with histone H3 hyperphosphorylation and hyperacetylation on its promoter. We further demonstrated that AChR activation and hyperacetylation correlate with the recruitment of the HAT p300 by GABP, whereas in the absence of activation, GABP recruits HDAC1. Consistently, in vivo electroporation of the p300 HAT activity-specific inhibitor LysCoA reduced the transcriptional activation of the AChR subunit gene. The histone acetyltransferase activity of p300 is therefore necessary for the activation of the AChR subunit gene expression at the NMJ.

GABP differentially recruits transcriptional coactivators or repressors on the AChR gene promoter

GABP has been shown to regulate the expression of several synaptic proteins, among which are AChR, utrophin and MuSK. Previous studies mainly focused on the upstream signals that triggered GABP to activate the expression of synaptic genes (for review, see Schaeffer et al, 2001; Mejat et al, 2003), but the molecular mechanisms by which GABP specifically activated transcription in response to neural cues remained elusive. Our results provide a first insight into such mechanisms, as they demonstrate that upon transcriptional activation of the AChR subunit gene, GABP recruits the coactivator p300, whereas it recruits HDAC1, a well-known repressor, in the absence of AChR activation (Figure 8).

The decreased interaction between GABP and HDAC1, as well as the reduced recruitment of HDAC1 on the AChR subunit promoter in the presence of neuregulin could result from the concomitant reduction of the amount of HDAC1 protein in the cells. Such a regulation of HDAC1 is not unprecedented, as it has been shown that quinidin or TNF treatment induce a diminution of the HDAC1 protein levels in breast cells (Zhou et al, 2000; Vashisht Gopal et al, 2006).

Trancriptional regulation by GABP involves chromatin acetylation

Both factors recruited by GABP are enzymes that regulate the level of histone acetylation. This suggests that the transcriptional regulation of the AChR subunit by GABP involves chromatin acetylation, which is consistent with the observation that histones are hyperacetylated in the nuclei, which express AChR genes.

In innervated muscle, the level of chromatin acetylation in myonuclei thus correlates with the expression of the AChR. Consistently, in a previous work, we have shown that in response to electrical activity, chromatin acetylation was low in extrasynaptic nuclei (Mejat et al, 2005). Upon denervation, a situation in which AChR expression is activated throughout the muscle fibres, histone acetylation was strongly increased. Therefore, a relationship exists between the status of histone acetylation and the expression of AChR genes in myonuclei.

Moreover, our results demonstrate that in innervated muscle, the HAT activity of p300 is required for AChR expression at the NMJ, whereas in the extrasynaptic regions, we previously showed that HDACs repressed AChR expression. Histone hyperacetylation is thus not only a marker of the myonuclei, which express AChR genes, but also plays a functional role as it is required for AChR expression, both at the NMJ in innervated muscles and in extrasynaptic nuclei in denervated muscle.

Histone phosphoacetylation is not restricted to immediate-early genes

ChIP experiments are not yet applicable to subsynaptic nuclei because of the amounts of purified subsynaptic nuclei they require. Alternatively, we have tried to perform ChIP experiments on muscles injected with neural agrin, but the amount of nuclei activated by agrin in injected muscles did not allow a sufficient enrichment to see the acetylation of the AChR promoter raise above background. To visualize histone acetylation by ChIP on the AChR promoter, we therefore had to use cultured myotubes. The absence of the effect of agrin on endogenous AChR expression in C2C12 myotubes prevented the possibility to directly demonstrate that agrin induced histone acetylation on the AChR promoter in cultured myotubes. However, the results obtained with neuregulin can most likely be extrapolated to agrin, as in myotubes, agrin and MuSK activate AChR expression via neuregulin receptors (Meier et al, 1998).

In cultured myotubes treated with neuregulin, immediate-early genes are transiently activated before synaptic genes. This transient activation seems to be important for the later activation of synaptic genes, as blocking the action of cJun with a dominant-negative mutant resulted in the inhibition of AChR subunit activation (Si et al, 1999). These observations were made using neuregulin, but could be relevant for agrin and MuSK as they also activate JNK (Lacazette et al, 2003) and therefore regulate cJun. Our results show that activation of immediate-early genes by neuregulin correlates with histone H3 hyperphosphoacetylation on the cJun promoter. EGF and neuregulin bind to receptors of the erbB family, respectively erbB1 and erbB2, 3 and/or 4, which trigger similar intracellular signalling pathways. This is consistent with the results of the groups of Mahadevan, Allis and Sassone Corsi, who demonstrated that activation of cJun expression by EGF was associated with histone H3 phosphoacetylation. These works demonstrated that the phosphorylation of histone H3S10 favoured the acetylation of the neighbouring lysine residues (Cheung et al, 2000; Clayton et al, 2000; Lo et al, 2000). It is therefore predictable that H3S10 phosphorylation upon neuregulin treatment and at the NMJ synergizes with H3 acetylation.

Our observation that H3 phosphoacetylation also affects the AChR subunit gene 4 h after neuregulin treatment demonstrates that H3 phosphoacetylation is not restricted to the activation of immediate-early genes and can affect other genes. This is corroborated by the immunofluorescence observations at the NMJ, which show that H3 is stably phosphoacetylated at many loci. In future experiments, it will be interesting to determine if the phosphorylation of histone H3 on AChR genes is triggered by the same kinases that phosphorylate histone H3 on immediate-early genes (Sassone-Corsi et al, 1999; Thomson et al, 1999), or if specific kinases are recruited at different genes or in different situations of transcriptional activation.

ErbB receptors are known proto-oncogenes in breast cancer (Linggi and Carpenter, 2006). The finding that erbB signalling induces histone phosphorylation and acetylation on its target genes could also be relevant in this context.

Neuregulin in vitro, agrin in vivo?

Recent publications suggest that although neuregulin and its receptors are critical for the activation of AChR expression in cultured myotubes, they only play a marginal role in the activation of synaptic genes in vivo that would exclusively rely on a neuregulin-independent pathway triggered by MuSK (Escher et al, 2005; Jaworski and Burden, 2006). This conclusion was drawn from the observation that in erbB2/erbB4-/- muscle (Escher et al, 2005), as well as in mice lacking neuregulin in motoneurons and muscle (Jaworski and Burden, 2006), NMJ can develop normally. However, these approaches still leave open the possibility that in the normal situation, neuregulin signalling could participate in AChR activation. Therefore, the hypothesis initially raised by Brenner and co-workers (Lacazette et al, 2003), who proposed that agrin would be crucial for NMJ formation and neuregulins would rather participate in its maintenance, could still be valid if one considers the possibility that in the absence of neuregulin signalling, the muscle could adapt to rely solely on the neuregulin-independent pathway triggered by MuSK.

Consistent with a role of neuregulin after agrin action, neuregulin and its receptor accumulate at agrin-induced ectopic synapses (Jones et al, 1999). In addition, neuregulins can activate synaptic genes expression in vivo, as shown on utrophin in 1999 (Gramolini et al, 1999).

Subynaptic chromatin is profoundly remodelled

Neuregulin-induced histone hyperacetylation was detectable in Western blot, suggesting that it affected many loci. The vast number of genes that have to be turned on to generate the postsynaptic scaffold could explain the large distribution of chromatin acetylation in subsynaptic nuclei. Consistently, the ChIP experiments performed with anti-acetylated histone H3 antibodies demonstrated that acetylation occurs on genes activated by neural factors.

The finely punctuated immunolabelling of acetylated and phosphoacetylated histones was distributed all over the nucleoplasm of subsynaptic nuclei, and was reminiscent of the electron micrographs that revealed a global decompaction of chromatin. The hyperacetylation of the muscle genome and synaptic genes activation in subsynaptic nuclei is therefore associated with a broad remodelling of the genome. In addition to chromatin decompaction, such a remodelling likely involves the specification of particular distributions of specific genes to the nucleoplasm.

The ultrastructure of muscle subsynaptic nuclei is quite similar to that of central neurons. In central neurons, chromatin acetylation has also been shown to play a crucial role in translating environmental signals into a physiological response. This is, for example, the case for the regulation of clock genes expression by the circadian rhythm, in which the transcriptional regulator CLOCK has recently been shown to be a HAT (Doi et al, 2006). Histone acetylation has also been shown to be involved in stable modifications of neuronal properties such as long-term potentiation and long-term depression, which involve histone hyper- or hypoacetylation on specific genes involved in these processes via the p300 closely related HAT CBP and HDAC5 (Guan et al, 2002).

Materials and methods

Electron microscopy

Tissue samples were fixed in 2% (w/v) glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 2 h at room temperature, post-fixed in buffered osmium tetroxide and embedded in epoxy resin after ethanol dehydration.

Ultrathin sections were collected on formvar-coated copper grids, contrasted in uranyl acetate and lead citrate and viewed by transmission electron microscopy (Philips CM 120) at the Centre Technologique des Microstructures, Claude Bernard University, Lyon, France.

Constructs and antibodies

The constructs used were pcDNA3-HD1 for human HDAC1 with the Flag epitope (Yang et al, 1997), pCI-Flag-p300 (Ogryzko et al, 1996), EB GABP (Schaeffer et al, 1998) and the epsilon-luciferase reporter vector (Duclert et al, 1993).

The antibodies used were GABP (Schaeffer et al, 1998), Flag M2 and -tubulin (Sigma), AcH3(K9K14), PAcH3 (K14S10), PH3 (S10), HDAC1 (Upstate Biotechnology), and p300 (sc-584) (Santa Cruz Biotechnology, Santa Cruz, CA, USA).

Cell lines, culture conditions and transfections

C2C12 cells were maintained as myoblasts in growth medium (1/2 DMEM and 1/2 HAM F12, Gibco BRL, supplemented with 13% foetal calf serum, HyClone Perbio). Cells were differentiated in differentiation medium (DMEM medium supplemented with 2% horse serum, Bio Media Canada).

The C2C12 myotubes were treated 48 h after being switched to differenciating medium, with neuregulin (recombinant neuregulin-1, EGF domain, Upstate) to induce AChR gene expression.

Cell transfections were performed with Lipofectamine Plus (Life Technologies, Gibco BRL) according to the manufacturer's instructions.

In vivo electroporation

Operative procedure was performed using aseptic techniques and according to the local ethical committee recommendations (Comité Rhône Alpes d'Ethique pour l'Expérimentation Animale). Five-week-old OF1 male mice were anaesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg) to obtain a deep state of general anesthesia. A 30 l volume of 0.9% NaCl containing 5 g of DNA, Lys and CoA or LysCoA was injected into Tibialis anterior muscles. Injected muscles were then electroporated with 1 cm² plaque electrodes placed on each side of the leg and eight 200 V cm^-1 pulses of 20 ms applied at 2 Hz (BTX ECM 830).

Ectopic synapse induction

To induce ectopic synapses, 2 g of purified agrin N25C95-A4B8-His (Scotton et al, 2006) was injected into mouse soleus muscles. Two weeks after injection, muscle fibres were isolated under a microscope (SZX 12, Olympus) using ultrafine forceps (Moria, No. 5).

Immunofluorescence

Immunofluorescence analyses were performed as follows. C2C12 cells or isolated muscle fibres were fixed for 5 and 20 min respectively in 3.7% formalin TBS (Tris-buffered saline: 150 mM NaCl in 50 mM Tris, pH 7.4) before permeabilization for 5 or 30 min respectively with 1% Triton in TBS. Cells were incubated with the primary antibodies diluted with 1% BSA in TBS overnight at 4°C, then washed in TBS, and incubated for 1 h with biotinylated anti-rabbit antibodies (Amersham). Biotinylated antibodies were revealed after a 1 h incubation with streptavidin–Texas red (Amersham). The NMJ was stained with -Bungarotoxin–Alexa488 (Molecular Probes). Nuclei were stained with Hoechst reagent 33258 (1 g/ml bisbenzimidine, Sigma). Fluorescent images were visualized by microscopy on a Zeiss Axioplan 2 microscope. Images were captured with a Photometrics CoolSNAP fx camera and processed with Photoshop 7 (Adobe Systems).

Co-immunoprecipitation and Western blotting

Whole-cell extracts were prepared from C2C12 cells in NP-40 buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1% NP-40 and Complete protease inhibitors (Roche)), and precleared with protein-A sepharose (Pharmacia Biotech). 10% of extracts were kept as input sample. Then extracts were incubated with specific antibodies overnight at 4°C, followed by incubation with protein-A sepharose at 4°C for 1 h and were washed three times. The immunoprecipitated material was boiled in 2 loading buffer, electrophoresed by SDS–PAGE and immunodetected.

Chromatin immunoprecipitation

Neuregulin-treated and untreated C2C12 myotubes were fixed in 3.7% formaldehyde in culture medium for 10 min at 37°C. Chromatin was prepared using a kit from Upstate Biotechnology according to the recommendations of the manufacturer, with ten 10-s sonication pulses, which yielded chromatin fragments of an apparent size of 800 bp (as monitored on agarose gels; data not shown). Equivalent amounts of chromatin were immunoprecipitated using specific antibodies. Formaldehyde-induced crosslinking was reversed (4 h at 65°C), and a sequence of the AChR promoter, cJun promoter or H4 promoter (as an internal control) was detected by quantitative PCR. The results of the cJun and AChR promoters quantification were normalized to the levels of histone H4 promoter. The primers used were as follows: for AChR (amplified region -161 to +68, 5'-GATGACAGGCCTTGTGGATT-3' forward and 5'-GACAAGCTTGAGGGAACAGG-3' reverse); for c-jun (5'-TACTCTCAAGCCCGCTCAAC-3' forward and 5'-CCGAGAAAGGGCTGAATGAT-3' reverse); for H4 (amplified region -161 to +69, 5'-GACACCGCATGCAAAGAATAGCTG-3' forward and 5'-CTTTCCCAAGGCCTTTACCACC-3' reverse).

RNA preparation, reverse transcription and real-time quantitative PCR

Total mRNA was extracted from homogenized (FastPrep, Bio 101, in RLT buffer, Qiagen) whole Tibialis anterior muscles using the RNeasy mini RNA extraction kit (Qiagen) with the additional proteinase K and DNase treatments. First strand cDNA was synthesized from 250 ng of total RNA using the Superscript II (Invitrogen). Gene expression was evaluated by real-time quantitative PCR (qPCR) (LightCycler, Roche) using the LightCycler FastStart DNA Master SYBR Green PCR kit (Roche) according to the manufacturer's instructions. The sequences of the primers were as follows: -actin forward 5'-CCCTGTATGCCTCTGGTCGT-3', reverse 5'-ATGGCGTGAGGGAGAGCAT-3'; AChR forward 5'-CTTGGTGCTGCTCGCTTACTT-3', reverse 5'-CGTTGATAGAGACCGTGCATT-3'; ErbB3 forward 5'-CTTACGGGACACAATGCTGA-3', reverse 5'-GCATGGCTGGAGTTGGTATT-3'. The measures were normalized to -actin RNA levels or ErbB3.

Luciferase reporter assay

2 g of epsilon-luciferase reporter vector and 3 g of pcDNA3 carrier DNA were co-electroporated in Tibialis anterior muscle with increasing dose of LysCoA (produced as described by Lau et al, 2000). 100 l of muscle homogenates (FastPrep, Bio 101, in passive lysis buffer, Promega) was mixed in an equal volume of luciferase substrate solution (0.8 M tricine pH 7.8, 1.5 M MgCO₃, 273 mM MgSO₄, 27 mM Coenzyme A, 0.5 M ATP, 10 mM luciferin and 0.5 M EDTA) and placed in a Microplate Luminometer (Veritas Turner BioSystem) to measure light production for 10 s.

Acknowledgements

We are grateful to Guy Nicaise and Dominique Leguellec for providing the electron micrographs, and to Denis Villemagne for precious technical assistance with the ChIP experiments. We are grateful to MA Ruegg, who kindly provided purified agrin and advised on ectopic synapses production. This work was supported by grants from the French Muscle Dystrophy association (AFM), the CNRS, the ANR (Neuroscience program) and the MENRT (ACI jeune chercheur).

References

Bannert N, Avots A, Baier M, Serfling E, Kurth R (1999) GA-binding protein factors, in concert with the coactivator CREB binding protein/p300, control the induction of the interleukin 16 promoter in T lymphocytes. Proc Natl Acad Sci USA 96: 1541–1546 | Article | PubMed | ChemPort |

Briguet A, Ruegg MA (2000) The Ets transcription factor GABP is required for postsynaptic differentiation in vivo. J Neurosci 20: 5989–5996 | PubMed | ISI | ChemPort |

Bush TS, St CM, Resendes KK, Rosmarin AG (2003) GA-binding protein (GABP) and Sp1 are required, along with retinoid receptors, to mediate retinoic acid responsiveness of CD18 (beta 2 leukocyte integrin): a novel mechanism of transcriptional regulation in myeloid cells. Blood 101: 311–317 | Article | PubMed | ChemPort |

Chan HM, La Thangue NB (2001) p300/CBP proteins: HATs for transcriptional bridges and scaffolds. J Cell Sci 114: 2363–2373 | PubMed | ISI | ChemPort |

Cheung P, Tanner KG, Cheung WL, Sassone-Corsi P, Denu JM, Allis CD (2000) Synergistic coupling of histone H3 phosphorylation and acetylation in response to epidermal growth factor stimulation. Mol Cell 5: 905–915 | Article | PubMed | ISI | ChemPort |

Clayton AL, Rose S, Barratt MJ, Mahadevan LC (2000) Phosphoacetylation of histone H3 on c-fos- and c-jun-associated nucleosomes upon gene activation. EMBO J 19: 3714–3726 | Article | PubMed | ISI | ChemPort |

Crosio C, Fimia GM, Loury R, Kimura M, Okano Y, Zhou H, Sen S, Allis CD, Sassone-Corsi P (2002) Mitotic phosphorylation of histone H3: spatio-temporal regulation by mammalian Aurora kinases. Mol Cell Biol 22: 874–885 | PubMed | ISI | ChemPort |

de Kerchove A, Cartaud J, Ravel-Chapuis A, Seroz T, Pasteau F, Angus LM, Jasmin BJ, Changeux JP, Schaeffer L (2002) Expression of mutant Ets protein at the neuromuscular synapse causes alterations in morphology and gene expression. EMBO Rep 3: 1075–1081 | Article | PubMed |

Doi M, Hirayama J, Sassone-Corsi P (2006) Circadian regulator CLOCK is a histone acetyltransferase. Cell 125: 497–508 | Article | PubMed | ChemPort |

Duclert A, Savatier N, Changeux JP (1993) An 83-nucleotide promoter of the acetylcholine receptor epsilon-subunit gene confers preferential synaptic expression in mouse muscle. Proc Natl Acad Sci USA 90: 3043–3047 | Article | PubMed | ChemPort |

Duclert A, Savatier N, Schaeffer L, Changeux JP (1996) Identification of an element crucial for the sub-synaptic expression of the acetylcholine receptor epsilon-subunit gene. J Biol Chem 271: 17433–17438 | Article | PubMed | ISI | ChemPort |

Escher P, Lacazette E, Courtet M, Blindenbacher A, Landmann L, Bezakova G, Lloyd KC, Mueller U, Brenner HR (2005) Synapses form in skeletal muscles lacking neuregulin receptors. Science 308: 1920–1923 | Article | PubMed | ChemPort |

Fischle W, Wang Y, Allis CD (2003) Histone and chromatin cross-talk. Curr Opin Cell Biol 15: 172–183 | Article | PubMed | ISI | ChemPort |

Fromm L, Burden SJ (1998) Synapse-specific and neuregulin-induced transcription require an ets site that binds GABPalpha/GABPbeta. Genes Dev 12: 3074–3083 | PubMed | ISI | ChemPort |

Fromm L, Burden SJ (2001) Neuregulin-1-stimulated phosphorylation of GABP in skeletal muscle cells. Biochemistry 40: 5306–5312 | PubMed | ChemPort |

Galasinski SC, Resing KA, Goodrich JA, Ahn NG (2002) Phosphatase inhibition leads to histone deacetylases 1 and 2 phosphorylation and disruption of corepressor interactions. J Biol Chem 277: 19618–19626 | Article | PubMed | ISI | ChemPort |

Gramolini AO, Angus LM, Schaeffer L, Burton EA, Tinsley JM, Davies KE, Changeux JP, Jasmin BJ (1999) Induction of utrophin gene expression by heregulin in skeletal muscle cells: role of the N-box motif and GA binding protein. Proc Natl Acad Sci USA 96: 3223–3227 | Article | PubMed | ChemPort |

Guan Z, Giustetto M, Lomvardas S, Kim JH, Miniaci MC, Schwartz JH, Thanos D, Kandel ER (2002) Integration of long-term-memory-related synaptic plasticity involves bidirectional regulation of gene expression and chromatin structure. Cell 111: 483–493 | Article | PubMed | ISI | ChemPort |

Gurley LR, D'Anna JA, Barham SS, Deaven LL, Tobey RA (1978) Histone phosphorylation and chromatin structure during mitosis in Chinese hamster cells. Eur J Biochem 84: 1–15 | Article | PubMed | ISI | ChemPort |

Hsu JY, Sun ZW, Li X, Reuben M, Tatchell K, Bishop DK, Grushcow JM, Brame CJ, Caldwell JA, Hunt DF, Lin R, Smith MM, Allis CD (2000) Mitotic phosphorylation of histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell 102: 279–291 | Article | PubMed | ISI | ChemPort |

Jaworski A, Burden SJ (2006) Neuromuscular synapse formation in mice lacking motor neuron- and skeletal muscle-derived Neuregulin-1. J Neurosci 26: 655–661 | Article | PubMed | ChemPort |

Jones G, Moore C, Hashemolhosseini S, Brenner HR (1999) Constitutively active MuSK is clustered in the absence of agrin and induces ectopic postsynaptic-like membranes in skeletal muscle fibers. J Neurosci 19: 3376–3383 | PubMed | ISI | ChemPort |

Koike S, Schaeffer L, Changeux JP (1995) Identification of a DNA element determining synaptic expression of the mouse acetylcholine receptor delta-subunit gene. Proc Natl Acad Sci USA 92: 10624–10628 | Article | PubMed | ChemPort |

Lacazette E, Le CS, Gajendran N, Brenner HR (2003) A novel pathway for MuSK to induce key genes in neuromuscular synapse formation. J Cell Biol 161: 727–736 | Article | PubMed | ChemPort |

Lau OD, Kundu TK, Soccio RE, it-Si-Ali S, Khalil EM, Vassilev A, Wolffe AP, Nakatani Y, Roeder RG, Cole PA (2000) HATs off: selective synthetic inhibitors of the histone acetyltransferases p300 and PCAF. Mol Cell 5: 589–595 | Article | PubMed | ISI | ChemPort |

Linggi B, Carpenter G (2006) ErbB receptors: new insights on mechanisms and biology. Trends Cell Biol 16: 649–656 | Article | PubMed | ChemPort |

Lo WS, Trievel RC, Rojas JR, Duggan L, Hsu JY, Allis CD, Marmorstein R, Berger SL (2000) Phosphorylation of serine 10 in histone H3 is functionally linked in vitro and in vivo to Gcn5-mediated acetylation at lysine 14. Mol Cell 5: 917–926 | Article | PubMed | ISI | ChemPort |

Margueron R, Trojer P, Reinberg D (2005) The key to development: interpreting the histone code? Curr Opin Genet Dev 15: 163–176 | Article | PubMed | ISI | ChemPort |

Meier T, Masciulli F, Moore C, Schoumacher F, Eppenberger U, Denzer AJ, Jones G, Brenner HR (1998) Agrin can mediate acetylcholine receptor gene expression in muscle by aggregation of muscle-derived neuregulins. J Cell Biol 141: 715–726 | Article | PubMed | ISI | ChemPort |

Mejat A, Ramond F, Bassel-Duby R, Khochbin S, Olson EN, Schaeffer L (2005) Histone deacetylase 9 couples neuronal activity to muscle chromatin acetylation and gene expression. Nat Neurosci 8: 313–321 | Article | PubMed | ChemPort |

Mejat A, Ravel-Chapuis A, Vandromme M, Schaeffer L (2003) Synapse-specific gene expression at the neuromuscular junction. Ann NY Acad Sci 998: 53–65 | Article | PubMed | ChemPort |

Murnion ME, Adams RR, Callister DM, Allis CD, Earnshaw WC, Swedlow JR (2001) Chromatin-associated protein phosphatase 1 regulates aurora-B and histone H3 phosphorylation. J Biol Chem 276: 26656–26665 | Article | PubMed | ISI | ChemPort |

Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y (1996) The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87: 953–959 | Article | PubMed | ISI | ChemPort |

Pflum MK, Tong JK, Lane WS, Schreiber SL (2001) Histone deacetylase 1 phosphorylation promotes enzymatic activity and complex formation. J Biol Chem 276: 47733–47741 | Article | PubMed | ISI | ChemPort |

Polesskaya A, Naguibneva I, Fritsch L, Duquet A, it-Si-Ali S, Robin P, Vervisch A, Pritchard LL, Cole P, Harel-Bellan A (2001) CBP/p300 and muscle differentiation: no HAT, no muscle. EMBO J 20: 6816–6825 | Article | PubMed | ISI | ChemPort |

Qiu Y, Zhao Y, Becker M, John S, Parekh BS, Huang S, Hendarwanto A, Martinez ED, Chen Y, Lu H, Adkins NL, Stavreva DA, Wiench M, Georgel PT, Schiltz RL, Hager GL (2006) HDAC1 acetylation is linked to progressive modulation of steroid receptor-induced gene transcription. Mol Cell 22: 669–679 | Article | PubMed | ChemPort |

Resendes KK, Rosmarin AG (2006) GA-binding protein and p300 are essential components of a retinoic acid-induced enhanceosome in myeloid cells. Mol Cell Biol 26: 3060–3070 | Article | PubMed | ChemPort |

Sanes JR, Lichtman JW (1999) Development of the vertebrate neuromuscular junction. Annu Rev Neurosci 22: 389–442 | Article | PubMed | ISI | ChemPort |

Sanes JR, Lichtman JW (2001) Induction, assembly, maturation and maintenance of a postsynaptic apparatus. Nat Rev Neurosci 2: 791–805 | Article | PubMed | ISI | ChemPort |

Sassone-Corsi P, Mizzen CA, Cheung P, Crosio C, Monaco L, Jacquot S, Hanauer A, Allis CD (1999) Requirement of Rsk-2 for epidermal growth factor-activated phosphorylation of histone H3. Science 285: 886–891 | Article | PubMed | ISI | ChemPort |

Schaeffer L, de Kerchove A, Changeux JP (2001) Targeting transcription to the neuromuscular synapse. Neuron 31: 15–22 | Article | PubMed | ISI | ChemPort |

Schaeffer L, Duclert N, Huchet-Dymanus M, Changeux JP (1998) Implication of a multisubunit Ets-related transcription factor in synaptic expression of the nicotinic acetylcholine receptor. EMBO J 17: 3078–3090 | Article | PubMed | ChemPort |

Scotton P, Bleckmann D, Stebler M, Sciandra F, Brancaccio A, Meier T, Stetefeld J, Ruegg MA (2006) Activation of MuSK and binding to dystroglycan is regulated by alternative mRNA splicing of agrin. J Biol Chem 281: 36835–36845 | Article | PubMed | ChemPort |

Si J, Wang Q, Mei L (1999) Essential roles of c-JUN and c-JUN N-terminal kinase (JNK) in neuregulin-increased expression of the acetylcholine receptor epsilon-subunit. J Neurosci 19: 8498–8508 | PubMed | ISI | ChemPort |

Thomson S, Clayton AL, Hazzalin CA, Rose S, Barratt MJ, Mahadevan LC (1999) The nucleosomal response associated with immediate-early gene induction is mediated via alternative MAP kinase cascades: MSK1 as a potential histone H3/HMG-14 kinase. EMBO J 18: 4779–4793 | Article | PubMed | ISI | ChemPort |

Vashisht Gopal YN, Arora TS, Van Dyke MW (2006) Tumour necrosis factor-alpha depletes histone deacetylase 1 protein through IKK2. EMBO Rep 7: 291–296 | Article | PubMed | ChemPort |

Wolffe AP, Kurumizaka H (1998) The nucleosome: a powerful regulator of transcription. Prog Nucleic Acid Res Mol Biol 61: 379–422 | PubMed | ISI | ChemPort |

Yang WM, Yao YL, Sun JM, Davie JR, Seto E (1997) Isolation and characterization of cDNAs corresponding to an additional member of the human histone deacetylase gene family. J Biol Chem 272: 28001–28007 | Article | PubMed | ISI | ChemPort |

Zhou Q, Melkoumian ZK, Lucktong A, Moniwa M, Davie JR, Strobl JS (2000) Rapid induction of histone hyperacetylation and cellular differentiation in human breast tumor cell lines following degradation of histone deacetylase-1. J Biol Chem 275: 35256–35263 | Article | PubMed | ISI | ChemPort |