Address correspondence and reprint requests to Pr Mohamed Amri, Laboratory of Functional Neurophysiology and Pathology, UR/11ES09, Department of Biological Sciences, Faculty of Science of Tunis, University Tunis El Manar, 2092 Tunis, Tunisia. E-mail: firstname.lastname@example.org and Dr Hubert Vaudry, Inserm U982, Laboratory of Neuronal and Neuroendocrine Communication and Differentiation, International Associated Laboratory Samuel de Champlain, Regional Platform for Cell Imaging of Haute-Normandie (PRIMACEN), Institute for Medical Research and Innovation (IRIB), University of Rouen, 76821 Mont-Saint-Aignan, France. E-mail: email@example.com
Oxidative stress, induced by various neurodegenerative diseases, initiates a cascade of events leading to apoptosis, and thus plays a critical role in neuronal injury. In this study, we have investigated the potential neuroprotective effect of the octadecaneuropeptide (ODN) on 6-hydroxydopamine (6-OHDA)-induced oxidative stress and apoptosis in cerebellar granule neurons (CGN). ODN, which is produced by astrocytes, is an endogenous ligand for both central-type benzodiazepine receptors (CBR) and a metabotropic receptor. Incubation of neurons with subnanomolar concentrations of ODN (10−18 to 10−12 M) inhibited 6-OHDA-evoked cell death in a concentration-dependent manner. The effect of ODN on neuronal survival was abrogated by the metabotropic receptor antagonist, cyclo1–8[DLeu5]OP, but not by a CBR antagonist. ODN stimulated polyphosphoinositide turnover and ERK phosphorylation in CGN. The protective effect of ODN against 6-OHDA toxicity involved the phospholipase C/ERK MAPK transduction cascade. 6-OHDA treatment induced an accumulation of reactive oxygen species, an increase of the expression of the pro-apoptotic gene Bax, a drop of the mitochondrial membrane potential and a stimulation of caspase-3 activity. Exposure of 6-OHDA-treated cells to ODN blocked all the deleterious effects of the toxin. Taken together, these data demonstrate for the first time that ODN is a neuroprotective agent that prevents 6-OHDA-induced oxidative stress and apoptotic cell death.
Reactive oxygen species (ROS) and free radicals, such as hydrogen peroxide (H2O2), hydroxyl radicals, nitric oxide, and superoxide anions, are known to induce a strong oxidative stress responsible for the oxidation of membrane lipids, cellular proteins and DNA (Valko et al. 2007), which are toxic for the cells and can trigger apoptosis. ROS are continuously generated by cellular metabolism, and antioxidant molecules or enzymes (reduced glutathione, catalase, superoxide dismutase, and glutathione peroxidase) prevent tissue damages under normal conditions (Fatokun et al. 2007). Excessive production of ROS is observed in many neurodegenerative disorders, including Alzheimer's disease, Parkinson's disease, and ischemia (Valko et al. 2007), and it has been established that oxidative stress plays a major role in neuronal cell death associated with these pathologies.
The octadecaneuropeptide ODN, originally isolated from the rat central nervous system (CNS), is generated by the proteolytic cleavage of an 86-amino acid precursor called diazepam-binding inhibitor (DBI). DBI and its derived peptides, including ODN, are collectively termed endozepines (Tonon et al. 2006). In the brain of vertebrates, DBI mRNA is predominantly expressed by glial cells (Tonon et al. 2006) and the occurrence of DBI-like immunoreactivity has been visualized in astroglial cells in various regions of the CNS, notably in ependymal cells bordering the cerebral ventricles, in tanycytes of the median eminence, in pituicytes of the pituitary, and in Bergmann cells of the cerebellum (Vidnyanszky et al. 1994; Tonon et al. 2006; Matsuda et al. 2007; Alfonso et al. 2012). It was initially reported that the endozepines ODN and DBI act as inverse agonists of central-type benzodiazepine receptors (CBR) (Ferrero et al. 1986). More recently, it has been shown that DBI-derived peptides can activate a metabotropic receptor positively coupled to either phospholipase C (PLC) (Leprince et al. 2001) or adenylyl cyclase (AC) (Lesouhaitier et al. 2000; Matsuda et al. 2007; Hamdi et al. 2012a). The primary structure of ODN has been strongly preserved during evolution (Tonon et al. 2006), suggesting that this peptide plays important biological functions. Indeed, behavioral studies have shown that ODN, acting through CBR, induces anxiety and pro-conflict behavior (Ferrero et al. 1986; Tonon et al. 2006), attenuates pentylenetetrazol-evoked convulsions (Tonon et al. 2006), shortens pentobarbital-induced sleeping time (Dong et al. 1999), and reduces fluid intake in rodents (Manabe et al. 2001). Besides, ODN acting through a metabotropic receptor exerts a potent anorexigenic effect (Do-Rego et al. 2007). At the cellular level, ODN regulates glial cell activity as well as neuronal function, that is, cell proliferation and intracellular calcium concentration in astrocytes (Gandolfo et al. 1997, 1999), neurosteroid biosynthesis in intact nerve cells (Do-Rego et al. 2001), and neuropeptide expression in neurons (Compère et al. 2005). In addition, recent data indicate that ODN exerts a strong protective activity and attenuates ROS accumulation in glial cells (Hamdi et al. 2011), but its potential neuroprotective activity has never been tested.
Cultured rat cerebellar granule neurons (CGN) are a useful model for the identification of neuroprotective compounds and the elucidation of their mechanisms of action (Botia et al. 2007; Polazzi et al. 2009). It has notably been shown that transforming growth factor beta-2 (Polazzi et al. 2009) prevents 6-hydroxydopamine (6-OHDA) toxicity in CGN, and that pituitary adenylate cyclase-activating polypeptide (PACAP) protects CGN against H2O2-induced oxidative stress (Masmoudi-Kouki et al. 2011). It is well established that 6-OHDA provokes neuronal apoptosis through the production of ROS leading to the expression of the pro-apoptotic protein Bax (Heaton et al. 2006) and the uncoupling mitochondrial oxidative phosphorylation resulting in energy deprivation, mitochondrial perforation and activation of caspase-3 (Heaton et al. 2006). Extracellular auto-oxidation of 6-OHDA may account for ROS production in cultured rat CGN, since the neurotoxin generates p-quinones and an array of other free radical species, notably H2O2, superoxide anions, and hydroxyl radicals in culture media (Hanrott et al. 2006).
There is evidence indicating that peptides, such as PACAP, protect both neurons and glial cells against oxidative stress (Botia et al. 2007; Masmoudi-Kouki et al. 2011). We have thus hypothesized that the gliopeptide ODN, which exerts a potent protective effect on astroglial cells (Hamdi et al. 2011), may also protect neurons against oxidative stress. Therefore, the aim of this study was to investigate the possible neuroprotective activity of ODN against 6-OHDA-induced cell death in rat CGN, and to characterize the transduction pathways mediating the anti-apoptotic effect of the gliopeptide.
Materials and methods
Wistar rats (Pasteur Institute, Tunis, Tunisia, and Charles River Laboratories, St Germain sur l'Arbresle, France) were kept in a temperature-controlled room (21 ± 1°C) under an established photoperiod (lights on from 7:00 am to 7:00 pm) with free access to food and water. Experiments were performed in accordance with American Veterinary Medical Association. Approval for these experiments was obtained from the Medical Ethical Committee for the Care and Use of Laboratory Animals of Pasteur Institute of Tunis. Approval Nu FST/LNFP/Pro152012.
Dulbecco's modified Eagle's medium (DMEM), F-12, l-glutamine, fetal bovine serum (FBS), insulin-transferrin-selenium supplement, antibiotic-antimycotic solution, and trypsin-EDTA buffer were purchased from Gibco (Invitrogen, Grand Island NY, USA). 6-OHDA, chelerythrine, H89, isobutylmethylxanthine (IBMX), U73122, trichloroacetic acid (TCA), cytosine-β-d-arabinofuranoside (β-ARAC), fluorescein diacetate-acetoxymethyl ester (FDA-AM), N-acetyl cysteine (NAC), bovine serum albumin (BSA), and Triton X-100 were obtained from Sigma Aldrich (St. Louis, MO, USA). The lactate dehydrogenase (LDH) kit assay was commercialized by Bio-Maghreb (Tunis, Tunisia). 5-6-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA), calcein-AM, and 5,5″,6,6″-tetrachloro-1,1″,3,3″-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) were from Molecular Probes (Eugene, OR, USA). U0126 and the Apo-ONE homogeneous caspase-3/7 assay kit were purchased from Promega (Charbonnières, France). Ro 15788 was a generous gift from Hoffmann-La Roche (Basle, Switzerland). Myo-[3H]inositol (100 Ci/mmol) was from PerkinElmer (Courtaboeuf, France). Rat ODN (QATVGDVNTDRPGLLDLK) and the ODN antagonist (cyclo1-8[DLeu5]OP) were synthesized by using the standard fluorenylmethyloxycarbonyl (Fmoc) procedure, as previously described (Leprince et al. 2001).
CGN primary culture
Granule cell suspensions were prepared from cerebella of 8-day-old Wistar rats of both sexes as previously described (Solecki et al. 2001). Pups were decapitated and cerebellums were collected in a Petri dish containing DMEM. The meninges were removed and the cerebellums were finely chopped with scissors. Tissue fragments were washed once in 10 mL of Krebs solution supplemented with 0.3% BSA and 0.02% MgSO4 0.02% (KBM). After centrifugation at 100 g for 10 s, the supernatant was discarded and replaced by 10 mL of KBM supplemented with 2.5 mg trypsin. After 3 min of enzymatic digestion, 10 mL of Krebs solution supplemented with 3.2 mg trypsin inhibitor and 0.5 mg DNase was added to stop the reaction. The suspension was centrifuged at 100 g for 10 s, the supernatant was discarded, and replaced by 10 mL of KBM supplemented with 5.2 mg trypsin inhibitor and 0.8 mg DNase. The suspension was then pipeted up and down with a tapered glass Pasteur pipette 10 times to ensure a mechanical dissociation. The suspension was then transferred to a tube containing 5 mL of KBM containing 0.001% CaCl2 and centrifuged at 200 g for 5 min. Cells were re-suspended in culture medium and seeded in multi-well plates (Costar, Tewksbury, USA), at a density of 1 × 106 cells/mL. Cells were cultured in a chemically defined medium consisting of 75% DMEM and 25% Ham's F-12 supplemented with 2 mM glutamine, 25 mM KCl, 1% insulin-transferrin-selenium supplement, 1% antibiotic-antimycotic solution, and 10% FBS. After 24 h, 10 μM β-ARAC was added to the medium. Cells were grown at 37°C in a humidified incubator with an atmosphere of 5% CO2/ 95% O2. In all experiments, CGN were used after 7 days of culture.
Cell cytotoxicity measurement
The cytotoxicity of 6-OHDA on CGN was determined by measuring LDH activity (LHD assay kit, Bio-Maghreb, Ariana, Tunisia) in the culture medium at the end of the experiments. The results were expressed as a percentage of total LDH release after cell lysis with 1% Triton X-100 in saline phosphate buffer (PBS, 0.1 M, pH 7.4).
Cell survival measurement
The surviving of CGN exposed to 6-OHDA was quantified by measuring FDA in cultured CGN. At the end of the experiments, cells were incubated in the dark with FDA (15 μg/mL, 8 min), rinsed once with PBS, and lysed with a 10 mM Tris-HCl solution containing 1% sodium dodecyl sulfate (SDS). Fluorescence intensity (λ excitation = 485 nm and λ emission = 538 nm) was measured with a FL800TBI fluorescence microplate reader (Bio-Tek Instruments, Winooski, VT, USA).
Polyphosphoinositide turnover measurement
The effect of ODN of polyphosphoinositide turnover was quantified by measuring inositol incorporation into CGN phospholipids. Cells were incubated for 3 h in fresh serum-free medium containing myo-[3H]inositol (100 Ci/mmol, PerkinElmer, Courtaboeuf, France). Inositol incorporation was stopped by removing the medium and adding 10% (w/v) ice-cold TCA. The cells were homogenized and centrifuged (13 000 g, 10 min, 4°C). [3H]polyphosphoinositides ([3H]PIPs) were extracted from the pellets with a chloroform/methanol solution (2 : 1, v/v) and counted in a beta counter (LKB 1217 Rack Beta, EG and G Wallac, Evry, France).
Cells were pre-incubated for 30 min with fresh serum-free medium containing 100 μM IBMX. Then, the cells were incubated for 3 h with graded concentrations of ODN. cAMP formation was stopped by removing the medium and adding ice-cold ethanol (70%) on CGN. Cells were homogenized, centrifuged (14 000 g, 10 min, 4°C) and the supernatants were evaporated by vacuum centrifugation overnight. The dried extracts were then reconstituted in RIA buffer (0.05 M sodium acetate, pH 5.8), and cAMP content was measured with a commercial cAMP RIA kit (Institute of Isotopes, Budapest, Hungary).
Western blot analysis
Total cellular proteins from CGN were extracted by using a 50 mM Tris/HCl lysis buffer containing 1% Triton X-100 and 10 mM EDTA. After centrifugation (12 000 g, 15 min, 4°C), the proteins contained in the supernatant were precipitated by addition of ice-cold 10% TCA, and washed three times with alcohol/ether (70/30) solution. Proteins were denatured (100°C, 5 min) in 50 mM Tris/HCl (pH 7.5) containing 20% (v/v) glycerol, 0.7 M 2-mercaptoethanol, 0.004% (w/v) bromophenol blue and 3% (w/v) SDS, electrophorezed on a 10% SDS–polyacrylamide gel and electrically transferred onto a nitrocellulose membrane (Amersham, Les Ulis, France). The membrane was incubated with the blocking solution (4% milk in 50 mM Tris saline buffer containing 0.05% Tween 20) at 25°C for 1 h, and revealed with antibodies against α-tubulin (Promega) or antibodies against the phosphorylated and total forms of ERK (Santa Cruz Biotechnology, Santa Cruz, CA, USA), using a chemiluminescence detection kit (ECL System; Amersham). Signals on autoradiographic films were quantified using an image analysis system (Samba Technologies, Châtillon, France).
Quantitative PCR analysis
Previous data have shown that, in CGN, factors acting through the MAPK pathway exert similar effects on Bcl-2 and Bax gene expression on the one hand and the corresponding protein levels on the other hand (Falluel-Morel et al. 2004; Aubert et al. 2008). Thus, the effect of ODN on Bax and Bcl-2 mRNA levels was performed by quantitative RT-PCR. Total RNA was isolated from CGN using the NucleoSpin kit (Macherey-Nagel, Hoerd, France) and 3–4 μg were used for cDNA synthesis using ImProm II Promega kit (Promega). PCR amplifications were performed with an ABI PRISM 7500 Sequence Detection System (Applied Biosystems, Courtaboeuf, France) using 5 ng cDNA, 1X Fast SYBR Green universal PCR Mastermix (Applied Biosystems) and 300 nM forward (TGCAGAGGATGATTGCTGATGT) and reverse (CAGCTGCCACACGGAAGAA) Bax primers or forward (GGCTGGGATGCCTTTGTG) and reverse (CAGCCAGGAGAAATCAAACAGA) Bcl-2 primers, under standard running conditions as suggested by the manufacturer. The amount of cDNA in each sample was calculated by the comparative quantification cycle (Cq) method and expressed as 2exp(−ΔΔCq) using glyceraldehyde-3-phosphate dehydrogenase as an internal control.
The effect of ODN on 6-OHDA-induced ROS formation was performed by measuring the fluorescence of DCF. At the end of the experiments, CGN were incubated for 30 min in fresh medium containing 10 μM CM-H2DCFDA and then washed twice with PBS. DCF fluorescence (λ excitation = 485 nm and λ emission = 538 nm) was measured with a FL800TBI fluorescence microplate reader.
The effect of ODN on 6-OHDA-induced decrease of reduced glutathione, which represents the majority of intracellular free thiols in cells, was measured by using the VitaBright-48 dye (Chemometec A/S, Allerød, Denmark). At the end of the experiments, CGN were detached by trypsination and the cell suspension was mixed with the vitaBright-48 probe. Stained cells were immediately loaded into NC-Slide A8™ (Chemometec A/S, Allerød, Denmark). Fluorescence intensity (λ excitation = 360 nm and λ emission = 485 nm) was measured with a NucleoCounter NC-3000™ (Chemometec A/S).
The effect of ODN on 6-OHDA-induced decrease of mitochondrial membrane potential was measured by using the JC-1 probe. At the end of the experiments, CGN were incubated for 15 min with the JC-1 probe (7.5 μg/mL) and then washed with PBS. Fluorescence intensity was measured with a FL800TBI fluorescence microplate reader and expressed as a ratio of the fluorescence emission at 590 nm (orange, intact mitochondrial membrane potential) versus 530 nm (green, collapsed mitochondrial membrane potential). Images were acquired with an eclipse E-600 microscope (Nikon, Champigny-sur-Marne, France) equipped with a 3 CCD Sony DXC950 camera interfaced with the Visiolab computerized program (Biocom, Les Ulis, France).
Caspase-3 activity measurement
Previous data have demonstrated that, on CGN, caspase-3 immunobloting and caspase-3 fluorogenic substrate assays yield similar results (Vaudry et al. 2003). Thus, the effect of ODN on 6-OHDA-induced increase of caspase-3 activity was measured by using the Apo-ONE Homogeneous Caspase-3/7 kit (Promega). At the end of the experiments, CGN were washed twice with PBS and re-suspended in DMEM (100 μL) mixed with 1X caspase assay buffer containing 25 μM caspase-3 substrate. Caspase-3 activity was calculated from the slope of the fluorescence intensity (λ excitation = 485 nm and λ emission = 530 nm) measured every 15 min for 3 h.
Data are expressed as the mean ± SEM from three independent experiments. Statistical analysis of the data was performed by using anova, followed by Bonferroni's test. p-value of 0.05 or less was considered statistically significant.
Protective effect of ODN against 6-OHDA-induced CGN death
Incubation of cultured CGN with graded concentrations of 6-OHDA (10 to 120 μM) provoked a dose- and time-dependent decrease of the proportion of surviving cells (Fig. 1a). The concentration of 30 μM 6-OHDA, which killed 50% of the cells after 72 h of treatment (Fig. 1a inset), was used in all subsequent experiments. Co-incubation of CGN with 6-OHDA and graded concentrations of ODN (10−18 to 10−8 M) for 72 h provoked a biphasic mirror effect on neuronal survival and LDH release (Fig. 1b). At very low concentrations (10−18 to 10−12 M), ODN prevented the deleterious effect of 6-OHDA on CGN survival and suppressed the stimulatory effect of the toxin on LDH release. At higher concentrations (10−11 to 10−8 M), the effects of ODN gradually vanished. The protective effect of ODN on CGN was also visualized by staining cells with calcein-AM. After 72 h of incubation in control conditions or with 10−14 M ODN, cells were alive and calcein-labeled cells displayed the typical shape of healthy differentiated granule neurons with bipolar fusiform cell bodies and long neurites organized in a network (Fig. 1ci and ii). In contrast, incubation with 30 μM 6-OHDA induced profound morphological changes with a shrinkage of cell bodies, a disruption of the neurite network and a decrease in the number of calcein-labeled cells (Fig. 1ciii). Treatment of cells with 10−14 M ODN totally prevented the morphological alterations induced by 6-OHDA (Fig. 1civ).
Receptor and signal transduction pathways involved in the neuroprotective effect of ODN on 6-OHDA-induced CGN death
In contrast to ODN, [Ala15]ODN (10−14 M), an analog devoid of agonistic activity on the ODN metabotropic receptor (Leprince et al. 2001), had no protective action on 6-OHDA-evoked cerebellar granule cell death (Fig. 2a). Pre-incubation of CGN for 30 min with cyclo1-8[DLeu5]OP (cDOP; 10−6 M), a selective ODN metabotropic receptor antagonist, which had no effect by itself on cell survival, totally abolished the neuroprotective action of ODN (Fig. 2a). In contrast, the central-type benzodiazepine receptor antagonist Ro 15788 (10−6 M) and the selective GABAA receptor antagonist bicuculline (10−9 M) did not affect the neuroprotective action of ODN on 6-OHDA-induced toxicity.
We next investigated the signaling cascade involved in the neuroprotective action of ODN on CGN. The PLC inhibitor U73122 (10−7 M) and the PKC inhibitor chelerythrine (10−7 M) totally abrogated the neuroprotective action of ODN against 6-OHDA-evoked cell death (Fig. 2b). In contrast, the selective PKA inhibitor H89 (2 × 10−5 M) did not modify the effect of ODN on 6-OHDA-induced cell death. In addition, blockage of ERK phosphorylation with the mitogen-activated protein kinase (MEK) inhibitor U0126 (10−6 M) suppressed the neuroprotective action of ODN (Fig. 2b).
Involvement of phospholipase C activity and ERK phosphorylation in the neuroprotective effect of ODN on 6-OHDA-induced CGN death
Incubation of granule cells with graded concentrations of ODN (10−18 M to 10−8 M) for 3 h evoked a dose-dependent increase in [3H]inositol incorporation into polyphosphoinositides (PIPs; Fig. 3a) and the maximum effect (140% over control; p <0.001) was observed at a concentration of 10−12 M. Pre-incubation of CGN with the ODN metabotropic receptor antagonist cyclo1-8[DLeu5]OP (10−6 M) or the PLC inhibitor U73122 (10−7 M) had no consequence on basal incorporation of [3H]inositol into PIPs, but totally suppressed the stimulatory effect of ODN (10−12 M; Fig. 3b). Consistent with the absence of action of H89 on the neuroprotective activity of the gliopeptide, ODN (10−18 M to 10−8 M) had no significant effect on cAMP production, whereas the AC activator forskolin (3 × 10−5 M) induced a fivefold increase in cAMP content (Fig. 3c).
Exposure of CGN to 10−14 M ODN stimulated ERK phosphorylation within 10 min (335% over control; p <0.001; Fig. 4a) and a significant effect was still observed after 6 h of treatment. Incubation of CGN with graded concentrations of ODN (10−18 to 10−10 M) for 3 h induced a dose-related stimulation of ERK phosphorylation (Fig. 4b). Pre-treatment of CGN with U73122 (10−7 M) or U0126 (10−6 M), which had no effect by themselves on ERK phosphorylation, abolished the stimulatory action of ODN (Fig. 4c and d). In contrast, H89 (2 × 10−5 M) did not modify ODN-induced ERK phosphorylation (Fig. 4e).
Effect of ODN on ROS accumulation and glutathione depletion
To examine whether ODN could block 6-OHDA-induced intracellular ROS accumulation, CGN were labeled with CM-H2DCFDA, a probe which generates the fluorescent compound dichlorofluorescein (DCF) upon oxidation with ROS. Incubation of cells with 6-OHDA for 72 h induced a 60% increase in DCF fluorescence intensity (Fig. 5a). Co-incubation of cells with ODN (10−14 or 10−12 M) did not modify DCF fluorescence intensity, but totally abolished the effect of 6-OHDA on DCF formation (Fig. 5a). In contrast, at concentrations of 10−18 M and 10−10 M, ODN was unable to attenuate ROS formation induced by 6-OHDA (Fig. 5a).
Glutathione (GSH), the most abundant intracellular scavenger, plays an important role in the protection of cells against ROS toxicity (Han et al. 2004). The effect of ODN on the intracellular content of GSH was assessed using the VitaBright-48 dye, a probe which forms a fluorescent compound when conjugated with free thiols such as GSH. Incubation of CGN with 30 μM 6-OHDA reduced by 43% VitaBright-48 fluorescence intensity (Fig. 5b). In the presence of low concentrations of ODN (10−14 and 10−12 M), the deleterious effect of 6-OHDA was abolished.
The transduction pathways involved in the inhibitory action of ODN on 6-OHDA-induced intracellular ROS formation and GSH depletion were then examined. Incubation of CGN with U73122, chelerythrine or U0126 totally abrogated the effects of ODN on 6-OHDA-induced DCF formation and VitaBright-48 fluorescence decrease (Fig. 5c and d). In contrast, H89 was unable to block the effects of the peptide.
Effect of ODN on mitochondrial potential, caspase-3 activation and Bcl-2 gene expression
Considering the major effect of ROS in the permeabilization of the mitochondrial outer membrane, the ability of ODN to prevent damages induced by 6-OHDA on mitochondrial integrity was examined. Control and ODN (10−14 M)-treated CGN exhibited many active mitochondria (red fluorescence of the JC-1 probe reflecting a high mitochondrial membrane potential) in the cell bodies and along neurites (Fig. 6ai and 6aii). Treatment of CGN with 6-OHDA resulted in a marked decrease of the red signal in the mitochondria and the labeling of most cell bodies in green, indicating that mitochondrial integrity was severely impaired by 6-OHDA (Fig. 6aiii). Incubation of cells with 10−14 M ODN suppressed the deleterious effect of 6-OHDA on mitochondrial membrane potential and only few cells remained labeled in green (Fig. 6aiv). Quantitative analysis indicated that 6-OHDA induced a significant reduction (−56%) of the 590/530 nm fluorescent ratio, and that ODN restored the ratio to control values (Fig. 6b).
To further explore the mechanisms involved in the protective action of ODN, the effect of the peptide on the expression of the anti-apoptotic gene Bcl-2 and of the pro-apoptotic gene Bax was investigated. Exposure of CGN to 6-OHDA produced a decrease (−28%) of Bcl-2 mRNA level and a concomitant increase (+102%) of Bax mRNA levels (Fig. 6c and d). Addition of neuroprotective concentrations of ODN (10−14 and 10−12 M) to the culture medium restored the expression of Bcl-2 and Bax genes to control levels. We next investigated the effect of ODN on caspase-3 activity. Incubation of CGN with 6-OHDA provoked a 60% increase of caspase-3 activity. Co-incubation of the cells with neuroprotective concentrations of ODN (10−14 and 10−12 M) totally blocked 6-OHDA-induced caspase-3 activation (Fig. 7a). Treatment of CGN with U73122, chelerythrine or U0216, which had no effect by themselves on caspase-3 activity, abrogated the ability of ODN to block caspase-3 activation induced by 6-OHDA, whereas H89 had no effect (Fig. 7b).
Finally, to confirm that ODN exerted a neuroprotective activity on ROS-dependent apoptotic cell death, the effect of NAC, a potent free radical scavenger, was examined on 6-OHDA-evoked antioxidant defense impairment and apoptosis. As illustrated in Fig. 8, NAC (5 × 10−3 M) totally abrogated the action of 6-OHDA in CGN on ROS formation (Fig. 8a), GSH depletion (Fig. 8b), caspase-3 activation (Fig. 8c), and cell death (Fig. 8d).
The main finding of this study is that the gliopeptide ODN protects CGN against apoptosis induced by 6-OHDA exposure. Although ODN was found to be effective at very low concentrations (in the femtomolar range), at higher doses the neuroprotective action of the gliopeptide gradually declined. Such a bell-shaped concentration–response curve has already been observed with ODN on [3H]thymidine incorporation (Gandolfo et al. 1999) and intracellular calcium mobilization (Leprince et al. 2001) in astrocytes. Consistent with these data, several studies have shown that subfemtomolar concentrations of two astroglial-derived proteins, that is, activity-dependent neurotrophic factor and activity-dependent neuroprotective protein, provide neuroprotection against various neurotoxins including 6-OHDA (Dejda et al. 2005) and brain injuries induced by ischemia or Parkinson's disease (Dejda et al. 2005). As in the case of ODN, the potency and efficacy of activity-dependent neurotrophic factor and activity-dependent neuroprotective protein in preventing neuronal cell death vanished at high doses with no remaining protective effect at subnanomolar concentrations (Dejda et al. 2005). It is noteworthy that ODN was still effective to protect CGN from the cytotoxic effect of 6-OHDA after 72 h of treatment, indicating that the gliopeptide exhibits a sustained neuroprotective activity. ODN also exerts a glioprotective effect against oxidative stress (Hamdi et al. 2011) but at concentrations which are 1000 times higher than those found here to have a neuroprotective effect. As a matter of fact, the concentration of ODN needed to prevent the deleterious effects of 6-OHDA on CGN is in the same range as that present in the culture medium of astrocytes (Masmoudi et al. 2005). Taken together, these data strongly suggest that ODN, specifically produced by glial cells in the brain (Tonon et al. 2006; Compère et al. 2010), exhibits a potent neuroprotective activity that surpasses by far its glioprotective effect. In agreement with these data, there is now mounting evidence indicating that astrocytes are not only detrimental in pathologies of the CNS, but may exert beneficial effects on neurons (Sofroniew and Vinters 2010) via the release of neuroprotective compounds.
The effects of ODN are mediated through either CBR associated with the GABAA receptor (Ferrero et al. 1986) or a metabotropic G protein-coupled receptor (Tonon et al. 2006). In CGN, the effect of ODN on 6-OHDA-induced cell death was blocked by a selective metabotropic endozepine receptor antagonist (Leprince et al. 2001) but was not affected by a CBR antagonist nor by a GABAA receptor antagonist, indicating that the neuroprotective effect of ODN is mediated via a G protein-coupled receptor. Previous studies have shown that the endozepine metabotropic receptor can activate either the PLC/PKC (Leprince et al. 2001) or the AC/PKA signaling pathways (Lesouhaitier et al. 2000; Hamdi et al. 2012a). Here, we show that, in CGN, ODN stimulates polyphosphoinositide turnover but does not affect cAMP formation. In addition, treatment of CGN with PLC or PKC inhibitors totally abrogated the anti-apoptotic activity of ODN, while a PKA inhibitor had no effect. Altogether, these observations indicate that the protective action of ODN against 6-OHDA-induced cell death can be specifically ascribed to the activation of the PLC/PKC signaling pathway (Fig. 9). In agreement with this notion, recent studies have shown that activation of PKC can rescue hippocampal neurons from global cerebral ischemia-induced apoptosis (Sun and Alkon 2010). In addition, siRNA knockdown of PKC blocks the neuroprotective effect of mechano-growth factor against 6-OHDA-induced apoptosis in the SH-SY5Y cell line (Quesada et al. 2011). It has already been reported that signaling events downstream of PKC activation lead to phosphorylation of ERK1/2 in various cell types (Noh et al. 2012) and that sustained activation of MAPK, similar to the one observed with ODN, promotes neuronal survival (Roovers and Assoian 2000). Thus, activation of the MAPK/ERK signaling cascade contributes to the neuroprotective action of compounds such as brain-derived neurotrophic factor (Han and Holtzman 2000) or PACAP (Vaudry et al. 2003) in CGN. Moreover, the MEK blocker U0126 abrogated the effect of ODN on 6-OHDA-induced cell death and suppressed the stimulatory action of ODN on ERK phosphorylation. The fact that ODN-induced ERK activation was blocked by the metabotropic receptor antagonist cyclo1–8[DLeu5]OP or the PLC inhibitor chelerythrine, together with the observation that MEK inhibition by U0126 abolished the protective effect of ODN, provide evidence for the implication of the ODN metabotropic receptor coupled to the PKC/ERK pathway in the anti-apoptotic action of ODN (Fig. 9).
Previous studies have demonstrated that 6-OHDA generates an array of free radical species by extracellular auto-oxidation (Hanrott et al. 2006). These ROS, which can easily cross the plasma membrane (Pedroso et al. 2009; Kaczara et al. 2010), impair cellular antioxidant defenses leading to cell death (Hamdi et al. 2011, 2012b). Consistent with this notion, this study has shown that 6-OHDA induced a strong increase of ROS production which was significantly reduced by ODN. The inhibitory effect of ODN on ROS formation is probably a key mechanism in its neuroprotective action. Indeed, we have observed that, in 6-OHDA-treated cells, ODN increases the cellular content of GSH, a free radical scavenger which plays a major role in the brain antioxidative defense (Han et al. 2004). The involvement of GSH in the neuroprotective effect of ODN is supported by data showing that inhibition of GSH synthesis in neuroblastoma cells induces an increase in the production of ROS and in the vulnerability of the cells to oxidative injury (Miyama et al. 2011). NAC, a GSH precursor (Shimizu et al. 2002) and a scavenger known to slow down 6-OHDA auto-oxidation (Shimizu et al. 2002), also provided complete protection against 6-OHDA-induced CGN death. The fact that ODN attenuates 6-OHDA-induced reduction of GSH in a PKC/MAPK-dependent manner strongly suggests that ODN, through activation of its metabotropic receptor, reduces intracellular ROS accumulation. Indeed, we have recently found that ODN is able to increase, via the metabotropic receptor, the level of GSH and the activity of the antioxidant enzymes SOD and catalase in cultured astrocytes (Hamdi et al. 2012a,b). Furthermore, there is evidence that activation of the PKC/ERK transduction pathway contributes to restore the GSH cellular content and to protect cells against oxidative stress-induced apoptosis in a neuronal hippocampal cell line (Luo and DeFranco 2006) and in cultured astrocytes (Masmoudi-Kouki et al. 2011). Altogether, these data indicate that ODN inhibits apoptosis of CGN by increasing GSH production, through a PKC- and ERK/MAP kinase-dependent transduction cascade, which in turn attenuates ROS formation. It is established that cellular ROS accumulation and GSH depletion exert opposite effects on the expression of pro- and anti-apoptotic factors that belong to the Bax/Bcl-2 family leading to activation of mitochondria-dependent apoptotic pathways (Lu et al. 2011) and several studies have determined that ERK1/2 phosphorylation can control the expression and/or activation of Bcl-2 family members to promote cell survival (Balmanno and Cook 2009). This study reveals that ODN increased the expression of the anti-apoptotic gene Bcl-2 and simultaneously blocked the 6-OHDA-induced expression of the pro-apoptotic gene Bax. As a consequence, ODN prevented the deleterious effect of 6-OHDA on mitochondrial potential and caspase-3 activity in a PKC- and ERK-dependent manner (Fig. 9).
Several lines of evidence suggest that the protective action of the gliopeptide ODN may have a physiopathological significance in neurodegenerative disorders. In particular, the concentration of endozepines is elevated in the cerebrospinal fluid of Alzheimer's and Parkinson's patients (Ferrarese et al. 1990), and it has been reported that β-amyloid peptide, the major constituent of senile plaques, stimulates endozepine biosynthesis and release from cultured astrocytes (Tokay et al. 2005). Moreover, ODN exerts a protective effect upon the deleterious action of oxidative stress by activating the endogenous antioxidant systems of astrocytes (Hamdi et al. 2011). It is now established that reactive astrocytes contribute to the defense of surrounding neurons under moderate oxidative stress by releasing neuroprotective factors, suggesting that overproduction of endozepines by astroglial cells observed in neurodegenerative disorders may be involved in the protection of neurons. Besides its neuroprotective effect, a recent study reported that ODN stimulates neurogenesis in adult mouse brain, and that DBI inhibition with siRNA leads to growth arrest and death of neural progenitors (Alfonso et al. 2012). Altogether, these data suggest that ODN might have a therapeutic potential for treatment of cerebral injuries involving oxidative neurodegeneration. Furthermore, the fact that ODN can inhibit neuronal cell death and stimulate neurogenesis suggests that it could also play a key role during brain development. In support of this hypothesis, it has been shown that, in the cerebellar cortex, endozepines are exclusively expressed by Bergmann glia (Vidnyanszky et al. 1994; Tonon et al. 2006) which control CGN migration. The possible effect of ODN on survival, differentiation and/or migration of CGN thus deserves further investigations.
In conclusion, this study has demonstrated that ODN, acting through its metabotropic receptor, exerts a potent protective action against 6-OHDA-induced CGN death. The anti-apoptotic activity of ODN on granule neurons is mediated through the PKC and MAP-kinase transduction pathways, and can be accounted for by inhibition of ROS-induced mitochondrial dysfunctions and caspase-3 activation.
The authors thank Mr Samir Elbahi for skillful technical assistance. H.K. and Y.H were recipients of fellowships from the University of Tunis El Manar and a France-Tunisia exchange programs CMCU-Utique and Inserm-DGRS. This study was supported by the Research Unit UR/11ES09, an Inserm-DGRS program (to M.A. and M.C.T.; grant number M 10/M), Inserm (U982), the Institute for Medical Research and Innovation (IRIB), the LARC-Neuroscience network and the Region Haute-Normandie. No conflict of interest.