α-Synuclein expression is induced by depolarization and cyclic AMP in enteric neurons

Authors


Address correspondence and reprint requests to Pascal Derkinderen, Inserm U913, 1, place Alexis Ricordeau, 44093 Nantes Cedex 1, France. E-mail: derkinderenp@yahoo.fr; pascal.derkinderen@chu-nantes.fr

Abstract

J. Neurochem. (2010) 115, 694–706.

Abstract

Accumulated evidence emphasizes the importance of α-synuclein expression levels in Parkinson’s disease (PD) pathogenesis. PD is a multicentric disorder that affects the enteric nervous system (ENS), whose involvement may herald the degenerative process in the CNS. We therefore undertook the present study to investigate the mechanisms involved in the regulation of expression of α-synuclein in the ENS. The regulation of α-synuclein expression was assessed by qPCR and western blot analysis in rat primary culture of ENS treated with KCl and forskolin. A pharmacological approach was used to decipher the signaling pathways involved. Intraperitoneal injections of Bay K-8644 and forskolin were performed in mice, whose proximal colons were further analyzed for α-synuclein expression. Depolarization and forskolin increased α-synuclein mRNA and protein expression in primary cultures of ENS, although L-type calcium channel and protein kinase A, respectively. Both stimuli increased α-synuclein expression through a Ras/extracellular signal-regulated kinases pathway. An increase in α-synuclein expression was also observed in vivo in the ENS of mice injected with Bay K-8644 or forskolin. In conclusion, we have identified stimuli leading to α-synuclein over-expression in the ENS, which could be critical in the initiation of the pathological process in PD.

Abbreviations used:
DMEM

Dulbecco’s modified Eagle’s medium

ENS

enteric nervous system

ERK

extracellular signal-regulated kinases

PBS

phosphate-buffered saline

PD

Parkinson’s disease

PKA

protein kinase A

ROS

reactive oxygen species

TBS

Tris-buffered saline

α-Synuclein is a neuronal protein that has been linked both to normal synaptic function and to neurodegeneration. Missense mutations of α-synuclein are responsible for rare autosomal dominant forms of Parkinson’s disease (PD) (see Waxman and Giasson 2009 for review) and aggregated α-synuclein has been shown to be the main component of the pathological hallmark of sporadic PD, namely Lewy bodies (Spillantini et al. 1997). There is a large body of evidence implicating the expression level of α-synuclein in the pathogenesis of PD. Duplications (Chartier-Harlin et al. 2004) and triplications (Singleton et al. 2003) of the α-synuclein gene have been identified in familial forms of PD. In animal models, over-expression of α-synuclein reproduces some of the cardinal pathological, neurochemical, and behavioral features of the human disease (Chesselet 2008). These studies indicate that over-expression of α-synuclein is sufficient to cause PD, and that its transcriptional regulation may be critically involved in the development of sporadic cases of the disease (Scherzer et al. 2008).

The traditional assumption that PD is a primary disorder of the substantia nigra has been challenged over the last years. It has indeed become increasingly evident that the pathological process of PD affects several neuronal structures outside the substantia nigra (Braak et al. 2003), among which is the enteric nervous system (ENS) (Braak et al. 2006; Lebouvier et al. 2008). Remarkably, from analyses of the temporal and spatial patterns of the spread of Lewy aggregates throughout the central and peripheral nervous systems, Braak et al. (2006) have determined that the appearance of α-synuclein aggregates occurs in the ENS during the earliest stage of PD, even before the substantia nigra. This led Braak to put forth the general proposal that PD pathology may begin in the gastrointestinal tract and that the pathological process further spreads to the CNS via the vagal innervation of the gut (Braak et al. 2006). A recent and thorough survey of the expression of α-synuclein in the ENS and in its vagal connections in rats has shown that α-synuclein is expressed in a subset of enteric neurons that are synaptically linked with α-synuclein-positive vagal neurons (Phillips et al. 2008). These results, along with the anatomical observations from Braak, offer a mechanism for the development and spread of the Lewy pathology in PD, in which both the ENS and α-synuclein play a crucial role.

Given the importance of the expression levels of α-synuclein for developing PD on one hand and the putative key role of the ENS in the pathophysiology of the disease on the other hand, we undertook the present study to investigate the mechanisms involved in the regulation of expression of α-synuclein in a model of primary culture of ENS (Chevalier et al. 2008) and in vivo. To this end, we used two distinct stimuli, membrane depolarization and forskolin, because many important physiological and pathological events in the ENS are regulated by neuronal activity and cyclic AMP (Howe et al. 2006; Neylon et al. 2006; Chevalier et al. 2008).

Material and methods

Reagents and antibodies

KCl, forskolin, nifedipine, Bay K-8644 (−) were purchased from Sigma (Saint Quentin Fallavier, France). Omega-conotoxin and omega-agatoxin were purchased from Alomone (Jerusalem, Israel). PD98059, U0126 and FTI-277 were purchased from Calbiochem (Meudon, France). CM-H2DCFDA was purchased from Invitrogen (Cergy-Pontoise, France). The following commercially available antibodies were used for western blotting: mouse monoclonal anti-α-synuclein (1 : 500; BD Bioscience; Le Pont-De-Claix, France) and rabbit polyclonal anti-α-synuclein (1 : 500; Santa Cruz Biotechnology, Heidelberg, Germany), rabbit polyclonal anti-phospho-extracellular signal-regulated kinases (ERK) (1 : 2000; Cell Signaling; Ozyme, Saint Quentin en Yvelines, France), reacting with active ERK1/2 (doubly phosphorylated on the tyrosine and threonine residues of the activation loop), total anti-ERK1/2 (1 : 1000; Cell Signaling), mouse monoclonal anti-HSP 70 (1 : 1000; Cell Signaling) and mouse monoclonal anti-PGP 9.5 (1 : 1000; Ultraclone limited, Isle of Wight, UK). For immunocytochemistry, mouse monoclonal anti-α-synuclein (1 : 500; BD Bioscience) and rabbit polyclonal anti-α-synuclein (1 : 500; Santa Cruz Biotechnology), mouse monoclonal anti-β III tubulin (1 : 500; Sigma), rabbit polyclonal anti-NF200 (1 : 500; Millipore; Molsheim, France), rabbit polyclonal anti-glial fibrilary acidic protein and rabbit polyclonal anti-S100β (1 : 500; Dako, Trappes, France) were used.

Primary cultures of ENS

Small intestine of rat embryos E15 (35–45 per isolation from three pregnant Sprague–Dawley rats (CERJ, Le Genest St Isle, France) were removed and finely diced in Hank's Buffered Salt Solution (Sigma). Tissue fragments were collected in 5 mL of medium [Dulbecco’s modified Eagle’s medium (DMEM)-F12 (1 : 1) medium] and digested at 37°C for 15 min in 0.1% trypsin (Sigma). The trypsin reaction was stopped by adding 10 mL of medium containing 10% fetal calf serum and then treated by Dnase I 0.01% (Sigma) for 10 min at 37°C. After triturating with a 10 mL pipette, cells were centrifuged at 500 g for 10 min. Cells were counted and then seeded at a density of 2.4 × 105 cells/cm2 on 24-well plates previously coated for 6 h with a solution of gelatin (0.5%;Sigma) in sterile phosphate-buffered saline (PBS). After 24 h, the medium was replaced with a serum-free medium [DMEM-F12 (1 : 1) containing 1% of N-2 supplement (Life Technologies, Cergy Pontoise, France)]. Cells were maintained in culture for 15 days. Half of the medium was replaced every 2 days (Chevalier et al. 2008).

Immunohistochemistry

After the fixation procedure (1 h in 0.1 M PBS containing 4% paraformaldehyde at 25°C), cells seeded on glass slide or tissues were washed in PBS and then permeabilized for 30 min in PBS/NaN3 containing 1% Triton X-100 and 4% horse serum before being incubated with the primary antibodies diluted in PBS/NaN3, 4% horse serum, and 1% Triton X-100 for 90 min at 25°C for cells and overnight at 4°C for tissues. When biotinylated α-synuclein antibody was used (mouse monoclonal antibody, BD Biosciences, biotinylated with EZ-Link Sulfo-NHS-LC-Biotinylation Kit from Thermo scientific), endogenous peroxidase activity was blocked by incubating preparations with 3% hydrogen peroxide for 20 min. Endogenous biotin was blocked with a commercial streptavidin/biotin blocking kit (Vector laboratories, Burlingame, CA, USA) according to the manufacturer’s instruction. Following incubation with primary antisera, cells or tissue were washed three times with PBS and incubated respectively for 30 and 90 min with secondary antibodies coupled to fluorophores: donkey anti-rabbit, anti-goat or anti-mouse IgG conjugated to carboxymethylindocyanine 3 or 5 (CY3 or CY5) (1 : 500; Jackson Laboratories, Immunotech, Marseille, France), donkey anti-rabbit or anti-mouse IgG conjugated to FluoProbes®488 (1 : 500; Interchim, Montluçon, France) or streptavidin coupled to CY3 (Invitrogen). Nuclei were stained with a 4′,6′-diamidino-2-phenylindole for 15 min (1 : 500; Sigma). After a final wash, samples were laid flat on a microscope slide and mounted in an aqueous fluorescence mounting medium (Dako). Specimens were viewed under a Zeiss Axiovert 200 mol/L microscope fluorescence microscope associated with the APOTOME mode (confocal like) (Carl Zeiss S.A.S., Le Pecq, France) and images were analyzed with axiovision 4.8 software (Carl Zeiss) and further treated with the Image J software (National Institute of Health, Bethesda, MD, USA).

Western blot

Primary culture of ENS were harvested in NETF buffer (100 mM NaCl, 2 mM EGTA, 50 mM Tris–Cl, pH 7.4, and 50 mM NaF) containing 1% (v/v) Nonidet P-40, 2 mM orthovanadate, phosphatase inhibitor cocktail II (Roche, Neuilly sur Seine, France) and a protease inhibitors cocktail (Roche). Tissues were lysed in NETF buffer with ‘Precellys 24’ tissue homogenizer (Bertin technologies, St Quentin-en-Yvelines, France). Equal amounts of lysate were separated using the Invitrogen NuPage Novex Bis Tris MiniGels™ before electrophoretic transfer with the iBlot™ Dry Blotting System also from Invitrogen. Membranes were blocked for 1 h at 25°C in Tris-buffered saline (TBS) (100 mM NaCl, 10 mM Tris, pH 7.5) with 5% non-fat dry milk. Membranes were incubated overnight at 4°C with the primary antibodies. Bound antibodies were detected with horseradish peroxidase-conjugated anti-rabbit or anti-mouse antibodies (Amersham, Les Ulis, France; diluted 1 : 5000) and visualized by enhanced chemiluminescent detection (ECL plus, Amersham). When necessary, membranes were stripped for 10 min in Reblot buffer (Millipore, Molsheim, France) followed by extensive washing in TBS before reblocking for 30 min in TBS with 5% non-fat dry milk and reprobing. The relevant immunoreactive bands were quantified with laser-scanning densitometry and analyzed with NIH Image J software. To allow comparison between different autoradiographic films, the density of the bands was expressed as a percentage of the average of controls (untreated). The value of α-synuclein immunoreactivity was normalized to the amount of PGP 9.5 immunoreactivity in the same sample and expressed as a percentage of controls.

Quantitative PCR analysis

RNA extraction from enteric primary culture was performed with RNAeasy Minikit (Qiagen S.A., Courtaboeuf, France) according to the manufacturer’s instructions. For reverse transcription, 1 μg of purified total RNA was denatured and subsequently processed for reverse transcription using SuperScript III Reverse Transcriptase (Invitrogen) according to the manufacturer’s recommendations. PCR amplifications were performed using the Absolute Blue SYBR green fluorescein kit (ABGENE, Courtaboeuf, France) according to the manufacturer’s protocol and run on MyiQ thermocycler (Bio-Rad, Marnes la coquette, France). The mRNA level of expression was determined using the formula of the comparative cycle theshold: (Ct): ΔCt, where ΔCt = (Ct,α-synuclein − Ct,PGP 9.5) sample − (Ct, α-synuclein − Ct,PGP 9.5) calibrated as previously described (Livak and Schmittgen 2001).

Primers were generated by the oligo 4.0 S software (National Biosciences, Plymouth, MN, USA) based on their Tm (melting temperature) as calculated by the nearest neighbor method (as close as possible to 60°C) with less than 2°C difference between them and all the primer duplexes kept to a minimum (less than four nucleotides) and no G nor C nor GC stretches longer than four nucleotides. Primers were also chosen on separate exons to amplify cDNA but not genomic DNA. Then, the primers were submitted to BLASTn analysis (NCBI) to confirm their specificity. The following primers were used: for α-synuclein, forward: 5′-CACAAGAGGGAATCCTGGAA-3′; reverse: 5′-TCATGCTGGCCGTGAGG-3′; PGP 9.5, forward: 5′-CCCCGAGATGCTGAACAAGTG-3′; reverse: 5′-CGATCACTGCTGATGGAAGA-3′.

Reactive oxygen species and neuron-specific enolase assays

After pharmacological treatments, primary cultures were loaded with pre-warmed Hank's Buffered Salt Solution containing 5 μM of CM-H2DCFDA (Invitrogen) for 15 min at 37°C then followed by a 10 min incubation at 37°C in DMEM medium without phenol red prior to microscopy analysis. For quantification of fluorescence, cells were lyzed with 100 μL of NETF/NP40 lysis buffer and fluorescence was read at 517 nm. Neuron-specific enolase release into culture medium was assessed as described previously (Abdo et al. 2010)

In vivo experiments

Male C57BL6N mice (Janvier, France) weighing 21–23 g were housed in cage in temperature-controlled room (21 ± 1°C), one week before the experiments. The mice were given access to food and water ad libitum and were maintained on 12 h light/dark cycle. Animal care was conducted in accordance with standard ethical guidelines and approved by the local ethic committee. Animals received a daily intraperitoneal (i.p.) injection of Bay K-8644 (2 mg/kg) or forskolin (2 mg/kg) or vehicle (10% ethanol) for 3 days. Animals were killed 24 h after the last i.p. injection and the proximal colons were taken and analyzed by immunoblot and immunohistochemistry.

Statistical analysis

All data are given as the mean ± standard error of the mean (SEM). Comparisons of mean values between groups were performed by Student’s t-test for unpaired data or by analysis of variance followed by Dunnett’s test. When data were not normally distributed, a Mann–Whitney test was performed. Differences were considered statistically significant if p < 0.05.

Results

α-Synuclein is expressed by neurons in primary culture of rat ENS

Following 14 days of culture, enteric neurons were organized in ganglia connected to each other by interganglionic fiber strands as evidenced by immunostaining using β III tubulin antibody (Fig. 1a) (Chevalier et al. 2008). Enteric glial cells, identified by glial fibrilary acidic protein immunostaining, were also present in enteric ganglia and along interganglionic fiber strands (Fig. 1b). α-Synuclein immunostaining revealed that α-synuclein was present in the cytoplasm of the somata as well as in the fibers of enteric neurons (Fig. 1a). In contrast, enteric glial cells did not express α-synuclein (Fig. 1b).

Figure 1.

 Expression of α-synuclein in primary culture of ENS. (a) The presence of enteric neurons in primary culture of ENS was assessed by β III tubulin Immunostaining (β tubulin III); nuclei were stained with 4′,6′-diamidino-2-phenylindole (DAPI). (b) The presence of enteric glial cells within primary culture of ENS was assessed by glial fibrilary acidic protein immunostaining (GFAP); nuclei were stained with 4′,6′-diamidino-2-phenylindole (DAPI). (a, b) Immunostaining with antibodies against α-synuclein demonstrate that α-synuclein colocalized primarily with enteric neurons and not with enteric glial cells (merge). Scale bar represents 20 μm.

Expression of α-synuclein is increased in enteric neurons following KCl-induced depolarization and forskolin challenge

Membrane depolarization elicited by 40 mM KCl induced a significant increase in the protein level of α-synuclein in primary culture of rat ENS (Fig. 2a). A twofold increase in α-synuclein expression was observed after 24 h of treatment, reaching 3.35-fold after 72 h (Fig. 2a and b). To rule out a non-specific osmotic effect of KCl, primary culture of rat ENS were treated with an equimolar concentration of mannitol. Such a treatment did not induce synuclein expression as compared to control either at 24 or 72 h (Fig. 2a and b). Treatment of primary culture of rat ENS with 20 μM forskolin, which increases intracellular cyclic AMP, induced an almost sixfold significant increase in the expression of α-synuclein after 72 h (Fig. 2a and b). The increase in α-synuclein expression induced by KCl was associated with a significant 1.9-fold increase in the corresponding transcript at 24 h (Fig. 2c). A significant 2.1- and 1.8-fold increase in α-synuclein transcripts was observed following treatment with 20 μM forskolin at 12 and 24 h respectively (Fig. 2c).

Figure 2.

 Effects of KCl-induced membrane depolarization and forskolin on α-synuclein expression in primary culture of ENS. (a) After 14 days in culture, primary culture of ENS were treated with vehicle (control), 40 mM mannitol, 40 mM KCl, and 20 μM forskolin for 24 or 72 h. Cells were harvested and homogenized in NETF/NP40 (1%) buffer and 35 μg of protein per sample were subjected to immunoblot analysis using antibodies specific for α-synuclein. After stripping, membranes were reprobed with anti-PGP 9.5 antibodies to ensure equal loading of neuronal proteins. (b) For quantification, the optical densities of α-synuclein immunoreactive bands were measured, normalized to the optical densities of PGP 9.5 immunoreactive bands in the same samples, and expressed as percentages of controls. Data correspond to mean ± SEM of 6–15 samples per condition. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control, *p < 0.01). (c) Quantitative PCR analysis of α-synuclein mRNA in primary culture of ENS treated with vehicle (control), 40 mM KCl (black bars) or forskolin (gray bars) for 3, 6, 12, 24, 48 and 72 h. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control, *p < 0.01). Data correspond to mean ± SEM of 6 samples per condition. (d) After 14 days in culture, primary culture of ENS were treated with vehicle (control), 40 mM KCl, or 20 μM forskolin for 72 h. Cells were loaded with CM-H2DCFDA (non-fluorescent) which is oxidized in DCF (fluorescent) by ROS. (e) For quantification, the relative fluorescence of DCF was measured using a microplate reader at 517 nM. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test. Data correspond to mean ± SEM of 6 samples per condition. (f) After 14 days in culture, primary culture of ENS were treated with vehicle (control), 40 mM KCl, and 20 μM forskolin for 72 h or with 100 μM H2O2 for 6 h. Cells were harvested and homogenized in NETF/NP40 (1%) buffer and 35 μg of protein per sample were subjected to immunoblot analysis using antibodies specific for Hsp70. After stripping, membranes were reprobed with anti-β actin antibodies to ensure equal loading of proteins. The autoradiograms are representative of three independent experiments.

In some instances, treatments with depolarizing agents have been associated with neuronal injury (Ramnath et al. 1992). As α-synuclein expression can be increased by cell stress (Gomez-Santos et al. 2003), we have performed a set of experiments to determine whether a treatment with KCl or forskolin provoke neuronal oxidative stress and/or neuronal cell death. First, by using the reactive oxygen species (ROS)-specific fluorescent dye, CM-H2DCFDA (Sung et al. 2001), we found that the ROS fluorescence was mainly detectable within neurons in treated and untreated primary culture of ENS (Fig. 2d) and that the amount of intracellular ROS was comparable between control and KCl or forskolin-treated cells (Fig. 2e). Second, the expression level of the 70 kDa heat-shock protein (Hsp70), a chaperone protein whose expression is up-regulated in neuronal cells following an oxidative injury (Shyu et al. 2004), was assessed in both treated and untreated cells. Treatment with either 40 mM KCl or 20 μM forskolin did not change the expression level of Hsp70 protein (Fig. 2e) whereas hydrogen peroxide which was used as positive control, induced a reproducible increase in the expression of Hsp70 as compared with control (Fig. 2f). Third, neuron-specific enolase release into the culture medium was used to estimate neuronal injury after treatments with high-KCl and forskolin. We have recently shown that this technique enables a reliable and specific assessment of neuronal cell death in primary culture of ENS (Abdo et al. 2010). The amount of neuron-specific enolase released in the culture medium was comparable between cells treated with 40 mM KCl, 20 μM forskolin and controls (2.3 ± 0.7 ng/mL for controls, 1.7 ± 0.6 ng/mL for KCl-treated cells and 3.2 ± 0.7 ng/mL for forskolin-treated cells, n = 8, p > 0.05 vs. control).

Taken together, our results demonstrate that depolarization and forskolin challenge of enteric neurons result in induction of α-synuclein expression at both the transcript and protein levels. The effects of depolarization and forskolin were specific and not a consequence of cell injury.

Induction of expression of α-synuclein by depolarization is mediated through L-type calcium channels

Voltage-operated calcium channels are critical in the regulation of gene expression by depolarization in the CNS (Flavell and Greenberg 2008). These channels have been classified by electrophysiological and pharmacological means into L-, N-, P-, Q-, R- and T-type channels (Catterall 2000). Within the enteric nervous system, L-, N-, P- and Q-type Ca2+ channels have been identified (Smith et al. 2003). Treatment of primary culture of ENS with nifedipine (1 μM), a specific antagonist of L-type calcium channels (Catterall 2000), completely prevented the effects of depolarization on α-synuclein induction (Fig. 3a and b). In contrast, pre-treatment with 0.1 μM of omega-conotoxin and omega-agatoxin, which inhibit specifically N- and P/Q-type calcium channels respectively (Catterall 2000), had no effects on the expression of α-synuclein elicited by depolarization (Fig. 3a and b).

Figure 3.

 Pharmacological characterization of the voltage operated calcium channels involved in the effects of depolarization on α-synuclein expression. (a) After 14 days in culture primary culture of ENS were treated with 40 mM KCl for 72 h in the absence or presence of 1 μM nifedipine, 0.1 μM omega-conotoxin and 0.1 μM omega-agatoxin applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. (b) Quantification of the results for α-synuclein protein level was assessed as described in the legend to Fig. 2b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01; treated in the presence of 1 μM nifedipine vs. in its absence: °p < 0.01). Data correspond to mean ± SEM of 3–9 samples per condition. (c) After 14 days in culture, primary culture of ENS were treated with 1 μM Bay K-8644 for 72 h. Immunoblots were performed as described in the legend to Fig. 2a. (d) Quantification of the results for α-synuclein protein level was assessed as described in the legend to Fig. 2b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01). Data correspond to mean ± SEM of 3–15 samples per condition.

To establish whether a selective L-type calcium channel agonist alone is able to induce α-synuclein expression, we used Bay K-8644, a selective agonist of these channels. Incubation of primary cultures of ENS with 1 μM of Bay K-8644 for 72 h significantly increased the expression of α-synuclein (Fig. 3c and d).

Collectively, these results demonstrate a critical role for L-type calcium channels in depolarization-induced α-synuclein expression.

Induction of expression of α-synuclein by forskolin is mediated through PKA activation and L-type calcium channels

By increasing the intracellular level of cyclic AMP, forskolin is able to activate protein kinase A (PKA). PKA has been shown to be involved in some of the effects depolarization in neurons (Grewal et al. 2000). This logically led us to use a PKA inhibitor, H89 (Chijiwa et al. 1990), to study the role of PKA on the effects of both forskolin and depolarization. Pre-treatment of primary culture of ENS with 2 μM H89 completely prevented the forskolin-induced increase in α-synuclein expression (Fig. 4a) but did not alter the effects of KCl-induced depolarization (Fig. 4b).

Figure 4.

 Role of PKA in the regulation of α-synuclein expression by KCl-induced membrane depolarization and forskolin. (a) After 14 days in culture, primary culture of ENS were treated 40 mM KCl for 72 h in the absence or presence of 2 μM H89 applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. (b) After 14 days in culture, primary culture of ENS were treated with 20 μM forskolin for 72 h in the absence or presence of 2 μM H89 applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. (c) After 14 days in culture, primary culture of ENS were treated with 20 μM forskolin for 72 h in the absence or presence of 1 μM nifedipine applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. The autoradiograms are representative of at least 4 independent experiments.

Regulation of gene expression by forskolin in neurons is either L-type calcium channels-dependent (Konradi et al. 2003) or -independent (Cigola et al. 1998). We thus studied the effects of nifedipine on forskolin-induced expression of α-synuclein and showed that this inhibitor of L-type calcium channels completely prevented the effects of forskolin (Fig. 4c).

These results suggest that both PKA activity and L-type calcium channels are required for the induction of expression of α-synuclein by forskolin. In contrast, PKA is not involved in the depolarization-induced α-synuclein expression.

Activation of ERK is required for depolarization and forskolin-induced expression of α-synuclein

Extracellular signal-regulated kinases play a pivotal in the regulation of gene expression in neurons (Grewal et al. 1999). ERK phosphorylation and activation can be achieved though several signaling pathways including depolarization and increase in intracellular cyclic AMP (Derkinderen et al. 1999). We therefore assessed the role of the ERK pathway in the induction of α-synuclein expression in enteric neurons. We first studied whether KCl-induced depolarization and forskolin were able to activate and thus to phosphorylate ERK in enteric neurons. Treatment of primary culture of ENS with either KCl or forskolin induced a rapid and monophasic activation of ERK as assessed by immunoblotting with antibodies specifically reacting with the active phosphorylated form of the kinase (Fig. 5a–d). The activation of ERK by depolarization appeared to be more sustained than the one induced by forskolin (Fig. 5a–d).

Figure 5.

 Role of ERK signaling pathway in the regulation of α-synuclein expression by KCl-induced membrane depolarization and forskolin. (a, b) After 14 days in culture, primary culture of ENS were treated either with 40 mM KCl or with 20 μM forskolin for 5, 15, 30, 60 or 120 min. Cell lysates were subjected to immunoblot analysis using antibodies specific for the dually phosphorylated (active) forms of ERK1 and ERK2 (P-ERK1/2). After stripping, the membranes were reprobed with anti-ERK (ERK1/2) antibodies. Immunoblots were performed as described in the legend to Fig. 2. (c, d) The values of active phospho-ERK1 and 2 were normalized to the amount of total ERK1 and 2 in the same sample and expressed as a percentage of controls. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01) (e, f) After 14 days in culture, primary culture of ENS were treated with either 40 mM KCl or 20 μM forskolin for 72 h respectively in the absence or presence of 10 μM U0126 applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. (g, h) Quantification of the results for α-synuclein protein level was assessed as described in the legend to Fig. 2b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01; treated in the presence of U0126 vs. in its absence: °p < 0.01). Data correspond to mean ± SEM of 6–12 samples per condition.

As activation of ERK results from the phosphorylation by the dual-specificity mitogen-activated protein kinases/ERK kinases, we have used PD98059 (Alessi et al. 1995) and U0126 (Favata et al. 1998), two mitogen-activated protein kinases/ERK kinases inhibitors. Pre-treatment of enteric neurons with 10 μM U0126 completely prevented the effects of both depolarization and forskolin on α-synuclein expression (Fig. 5e–h). Similar results were obtained using 50 μM PD98059 as a pre-treament (data not shown).

Ras is required for depolarization and forskolin-induced expression of α-synuclein

Membrane depolarization and cyclic AMP are capable of activating ERK through diverse signaling pathways (Derkinderen et al. 1999). The small G protein Ras is a point of convergence for the two stimuli to activate ERK (Obara et al. 2007). We thus studied the effects of a Ras inhibitor, the farnesyl transferase inhibitor FTI-277 (Lerner et al. 1995), on the expression of α-synuclein in enteric neurons. Pre-treatment with 10 μM FTI-277 significantly decreased the induction of α-synuclein expression elicited by both forskolin and depolarization (Fig. 6a–d). This suggests that depolarization and forskolin act through a common signaling pathway to activate ERK and in turn to induce α-synuclein expression. In line with this hypothesis, a simultaneous treatment with forskolin and KCl was no more efficient than forskolin and KCl alone to induce α-synuclein (Fig. 6e and f).

Figure 6.

 Role of Ras signaling pathway in the regulation of α-synuclein expression by KCl-induced membrane depolarization and forskolin. (a, c) After 14 days in culture, primary culture of ENS were treated either with 40 mM KCl or 20 μM forskolin for 72 h in the absence or presence of 10 μM FTI-277 applied 30 min before. Immunoblots were performed as described in the legend to Fig. 2a. b and d, quantification of the results for synuclein protein level was assessed as described in the legend to Fig. 2b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01). (e) After 14 days in culture, primary culture of ENS were treated with either 40 mM KCl or 20 μM forskolin alone or with both treatment concomitantly for 72 h. Immunoblots were performed as described in the legend to Fig. 2a. (f) Quantification of the results for synuclein protein level was assessed as described in the legend to Fig. 2b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01). Data correspond to mean ± SEM of six samples per condition.

Taken as a whole, our results demonstrate a critical role for the Ras/ERK pathway in the regulation of α-synuclein expression by forskolin or depolarization.

Bay K-8644 and forskolin increase α synuclein expression in enteric neurons in vivo

We eventually sought to determine whether the effects of depolarization and forskolin on α-synuclein expression could be also observed in vivo. We first studied the expression of α-synuclein in the colonic ENS of mice. Whole mount immunofluorescence experiments showed that the expression of α-synuclein in the colonic myenteric plexus of mice was restricted to neurons (Fig. 7a). No α-synuclein staining was observed in enteric glial cells (Fig. 7a). To study the effects of depolarization and forskolin in living mice, we used Bay K-8644 and forskolin, two compounds that have already been shown to be efficient when intraperitoneally administered (Jinnah et al. 1999; Melis et al. 2002). Using western blot analysis, we have shown that treatment of mice with 2 mg/kg of Bay K-8644 or with 2 mg/kg of forskolin every day for 3 days induced a significant increase in the expression of α-synuclein in the proximal colon as compared with controls (Fig. 7b). Immunofluorescence experiments revealed that the increase in α-synuclein expression induced by Bay K-8644 and forskolin occurred in neurons and mainly in their somata (Fig. 7c).

Figure 7.

 Bay K-8644 and forskolin increase α synuclein expression in enteric neurons in mice. (a) Double immunohistochemical labeling of myenteric neurons and enteric glial cells from proximal colon of mice. Myenteric neurons were identified with anti-NF200 antibodies and stained with anti-synuclein antibodies. Enteric glial cells were labeled using S100β antibodies and stained with anti-synuclein antibodies. Nuclei were stained with DAPI. Scale bar represents 20 μm. (b) Mice were i.p. injected daily with Bay K-8644 (2 mg/kg) during 3 days. Their proximal colon was taken and the amount of α-synuclein was assessed by immunoblots that were performed and quantified as described in the legend to Fig. 2a and b. Statistical analysis was performed with anova followed by Dunnett’s multiple comparison test (treated vs. control: *p < 0.01). Data correspond to mean ± SEM of 5 animals per condition. (c) Mice were i.p. injected daily with forskolin (2 mg/kg) during 3 days. Their proximal colon was taken, dissected and α-synuclein immunostaining was performed on myenteric plexus. Nuclei were stained with DAPI. Data are representative of five different animals per condition. Scale bar represents 20 μm.

Discussion

The three main outcomes of the present survey are (i) the induction of α-synuclein expression, a key protein in the pathophysiology of PD, by cyclic AMP and depolarization; (ii) the critical role of ERK in the regulation of α-synuclein expression; (iii) the identification of the enteric neurons as a subset of neurons in which α-synuclein expression can be regulated.

Although the expression of α-synuclein has been suggested to be involved in the pathogenesis of PD, only a few studies to date have addressed the specific issue of the extracellular stimuli capable of modulating α-synuclein expression in neurons. An increase in α-synuclein expression occurs in response to growth factors such as nerve growth factor and basic fibroblast growth factor (Stefanis et al. 2001; Clough and Stefanis 2007). This study is the first to show that stimuli linked to neuronal activity, namely intracellular cyclic AMP and membrane depolarization can induce the expression of α-synuclein. Remarkably, a previous report performed in PC12 cells failed to demonstrate any effect of a non-hydrolyzable analogue of cyclic AMP, whereas nerve growth factor induced a robust increase in α-synuclein expression (Stefanis et al. 2001). Two reasons can be put forward to explain such a discrepancy in the effects of cyclic AMP on α-synuclein expression. First, we have used a primary culture model in this study instead of a cell line. Second, the non-hydrolyzable analogue of cyclic AMP was applied for 10 days in PC12 cells instead of the 72 h treatment with forskolin used in the present study. Given the time course of α-synuclein mRNA induction observed in the present study, it is likely that a 10-day treatment with forskolin would have also failed to elicit an increase in α-synuclein expression.

Deciphering the signaling pathways, we have underscored two key elements involved in the effects of membrane depolarization and forskolin on α-synuclein expression. First, we have shown L-type calcium channels were critically involved in the effects of depolarization and forskolin (Fig. 8). This is in accordance with the large body of studies performed in the CNS which have demonstrated that L-type calcium channels are the cornerstone in signaling mechanisms linking neuronal activity to gene expression (Flavell and Greenberg 2008). Recent evidence has emerged that a similar role for these channels also exists in the ENS (Chevalier et al. 2008). Remarkably, the involvement of L-type calcium channels in PD has been recently addressed both in experimental Parkinsonism and in an epidemiological survey. A dysregulation of L-type calcium channels occurs in rodents treated with either 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine or rotenone, and treatment with a channel blocker prevents the development of neurodegeneration in these animals (Chan et al. 2007). In line with these results obtained in animals, current long-term use of calcium channel blockers for hypertension in humans is associated with a significantly reduced risk of PD (Becker et al. 2008). Second, we have demonstrated that the Ras/ERK signaling pathway is necessary for α-synuclein expression following both depolarization and forskolin challenge (Fig. 8). This concurs with the results of Clough and Stefanis (2007), who also showed that Ras and ERK were two critical steps in the induction of α-synuclein by growth factors. From a pathological point of view, patients with PD exhibit cytoplasmic aggregates of activated forms of ERK within their nigral neurons (Zhu et al. 2002) and 6-hydroxydopamine elicits a sustained ERK activation that contributes to neuronal cell death in vitro (Kulich and Chu 2001), raising the possibility that abnormal patterns of ERK activation may contribute to dopaminergic neuronal cell death. In addition, a recent report has demonstrated that the product of the leucine-rich repeat kinase 2 gene, whose mutations account for frequent autosomal-dominant PD, induces α-synuclein expression via ERK (Carballo-Carbajal et al. 2010). Taken as whole, these results, along with the one obtained in the present study, strongly suggest that a dysregulation of both L-type calcium channels and of the Ras/MAP kinase pathway are present in neurons from PD patients and that such phenomenon are likely to occur not only in central neurons but also in enteric neurons.

Figure 8.

 Signaling pathways involved in the cAMP- and membrane depolarization-induced expression of α-synuclein in enteric neurons. Depolarization induced by KCl increased α-synuclein expression through activation of L-type calcium channels. The effects of forskolin, an activator of adenylyl cyclase (AC) on synuclein expression are mediated through protein kinase A (PKA) and L-type calcium channels. Both stimuli converge on a Ras/ERK pathway. Black arrows indicate activation. Red lines indicate blockade by inhibitors.

Our findings may be relevant to the pathogenesis of PD. We have used in this study our recently developed model of primary culture of ENS, an in vitro model that recapitulates the main features of the ENS (Chevalier et al. 2008). The results obtained in this model were reinforced by the fact that Bay K-8644 and forskolin also induced a significant increase in synuclein expression in vivo. The ENS has received great deal of interest over the last years for its role in the pathophysiology of PD (Lebouvier et al. 2009). It has been suggested that the lesions in the ENS occur at a very early stage of the disease, even before the involvement of the CNS (Braak et al. 2006). This led to the postulate that the enteric nervous system is likely to be critical in the pathophysiology of PD as it could represent a route of entry for a putative environmental factor to initiate the pathological process (Braak’s hypothesis) then spreading to the CNS via vagal connections (Braak et al. 2006). In this context, several recent reports strongly support that α-synuclein is pivotal in the spread of the pathological process from the ENS to the CNS. -α-Synuclein has been shown to be secreted by neuronal cells in vitro and this secreted α-synuclein is prone to aggregate (Lee et al. 2005). Aggregates of α-synuclein can be taken up from the extracellular space by neighboring neurons thereby triggering neuronal cell death and the formation of Lewy body-like intracellular inclusions (Desplats et al. 2009), supporting the hypothesis that α-synuclein behaves like prion protein. Remarkably, the amount of α-synuclein secreted in the extracellular space is likely to be correlated with the quantity of α-synuclein present in the intracellular space (Lee et al. 2005). Altogether, these data suggest that the increase in the intracellular protein level of α-synuclein within enteric neurons may be the first critical step in the development of PD.

Eventually, our results have implications that go beyond PD. An emerging concept in gastroenterology is that a wide range of diseases, such as motility disorders, can be considered in part as enteric neuropathies. In particular, aging is associated with a variety of motility disorders or the gut including delays in gastric emptying and longer intestinal transit time (Camilleri et al. 2008). Aged rats display neuronal loss as well as changes in the neurochemical phenotype in the ENS, which are likely to result in motility disorders (Phillips et al. 2007). Remarkably, along with neuronal loss, these rats exhibit dystrophic enteric neurons that contains α-synuclein aggregates reminiscent of Lewy pathology (Phillips et al. 2009). This suggests that the presence of pathogens or xenobiotics in the gastrointestinal tract can convert normal aging into pathological aging associated not only with PD but also to a larger concept of enteric neuropathies/synucleinopathies.

Acknowledgements

This work was supported by a grant from France Parkinson. Work in Michel Neunlist’s lab is supported by Fondation de France, France Parkinson, CECAP (Comité d’Entente et de Coordination des Associations de Parkinsoniens), ADPLA (Association des Parkinsoniens de Loire Atlantique), FFPG (Fédération française des groupements parkinsoniens), Parkinsoniens de Vendée, GFNG (groupe français de neurogastroentérologie). TL is a recipient of poste d’accueil Inserm. MN and PDe are both recipients of contrats d’Interface Inserm. The authors are grateful to the Cellular imaging platform PiCell, IFR26, Nantes, France for Apotome pictures. The authors declare no conflicts of interest.

Ancillary