Calcium-permeable AMPA receptors trigger vesicular ...

Neuropharmacology 99 (2015) 396e407

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Calcium-permeable AMPA receptors trigger vesicular glutamate release from Bergmann gliosomes Chiara Cervetto a, b, Daniela Frattaroli a, Arianna Venturini a, Mario Passalacqua c, Mario Nobile d, Susanna Alloisio d, Carlo Tacchetti e, f, Guido Maura a, b, Luigi Francesco Agnati g, h, Manuela Marcoli a, b, * a

Department of Pharmacy, Section of Pharmacology and Toxicology, University of Genova, Viale Cembrano 4, 16148 Genova, Italy Centre of Excellence for Biomedical Research CEBR, University of Genova, Viale Benedetto XV, 5, 16132 Genova, Italy Department of Experimental Medicine, Section of Biochemistry, Italian Institute of Biostructures and Biosystems, University of Genova, Via L.B. Alberti 2, 16132 Genova, Italy d CNR, Biophysics Institute, Via de Marinis 6, 16146 Genova, Italy e Department of Experimental Medicine, University of Genova, Via L. B. Alberti 2, 16132 Genova, Italy f Experimental Imaging Center, Scientific Institute San Raffaele, Via Olgettina 60, 20132 Milano, Italy g Department of Biomedical, Metabolic Sciences and Neuroscience, University of Modena and Reggio Emilia, Via Campi 287, 41125 Modena, Italy h €g 8, Stockholm, Sweden Department of Neuroscience, Karolinska Institutet, Retzius va b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 June 2015 Received in revised form 4 August 2015 Accepted 5 August 2015 Available online 7 August 2015

The Bergmann glia is equipped with Ca2þ-permeable AMPA receptors for glutamate, indispensable for structural and functional relations between the Bergmann glia and parallel/climbing fibers-Purkinje cell synapses. To better understand roles for the Bergmann AMPA receptors, herein we investigate on gliotransmitter release and Ca2þ signals in isolated Bergmann glia processes obtained from adult rat cerebellum. We found that: 1) the rat cerebellar purified astrocyte processes (gliosomes) expressed astrocytic and Bergmann markers and exhibited negligible contamination by nerve terminals, microglia, or oligodendrocytes; 2) activation of Ca2þ-permeable AMPA receptors caused Ca2þ signals in the processes, and the release of glutamate from the processes; 3) effectiveness of rose bengal, trypan blue or bafilomycin A1, indicated that activation of the AMPA receptors evoked vesicular glutamate release. Cerebellar purified nerve terminals appeared devoid of glutamate-releasing Ca2þ-permeable AMPA receptors, indicating that neuronal contamination may not be the source of the signals detected. Ultrastructural analysis indicated the presence of vesicles in the cytoplasm of the processes; confocal imaging confirmed the presence of vesicular glutamate transporters in Bergmann glia processes. We conclude that: a vesicular mechanism for release of the gliotransmitter glutamate is present in mature Bergmann processes; entry of Ca2þ through the AMPA receptors located on Bergmann processes is coupled with vesicular glutamate release. The findings would add a new role for a well-known Bergmann target for glutamate (the Ca2þ-permeable AMPA receptors) and a new actor (the gliotransmitter glutamate) at the cerebellar excitatory synapses onto Purkinje cells. © 2015 Elsevier Ltd. All rights reserved.

Chemical compounds studied in this article: (S)-AMPA (PubChem CID: 158397) (S)-CPW 399 (PubChem CID: 657004) 1-Naphthylacetyl spermine trihydrochloride (PubChem CID: 16219727) GYKI 52466 (PubChem CID: 3538) Rose bengal (PubChem CID: 32343) Trypan blue (PubChem CID 9562061) Fura 2am (PubChem CID: 3086531) Keywords: Cerebellum Gliosomes Gliotransmitter release GluA2-lacking AMPA receptor Neuroglia

List of abbreviations: AMPA, a-amino-3-hydroxy-5-methyl-4-isoxazole propionate; BLBP, brain specific lipid binding protein; [Ca2þ]i, intragliosomal calcium concentration; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; (S)-CPW 399, (S)-a-amino-2,3,4,5,6,7-hexahydro-2,4-dioxo-1H-cyclopen tapyrimidine-1-propanoic acid; DL-TBOA, DL-threo-b-benzyloxyaspartic acid; EC50, half-maximum effective concentration; Emax, maximum effect; GFAP, glial fibrillary protein; GYKI 52466, 1-(4-aminophenyl)-4methyl-7,8-methylenedioxy-5H-(3-N-methylcarbamate)-2,3 benzodiazepine; IC50, half-maximum inhibitory concentration; MAP2, microtubule associated protein 2; NASPM, 1-Naphthylacetyl spermine trihydrochloride; ROI, regions of interest; VGLUT, vesicular glutamate transporter. * Corresponding author. Department of Pharmacy, Section of Pharmacology and Toxicology, University of Genova, Viale Cembrano 4, 16148 Genova, Italy. E-mail address: [email protected] (M. Marcoli). http://dx.doi.org/10.1016/j.neuropharm.2015.08.011 0028-3908/© 2015 Elsevier Ltd. All rights reserved.

C. Cervetto et al. / Neuropharmacology 99 (2015) 396e407

1. Introduction Bergmann cells, radial astrocytes with cell bodies in the Purkinje cell layer and projections extending radially in the molecular layer, represent the major glial cell type in the cerebellar cortex and are recognized to play active roles in the control of glutamatergic transmission onto Purkinje cells. About 87% of the climbing fiber and 65% of the parallel fiber glutamatergic synapses onto Purkinje cells are ensheathed by Bergmann processes (Xu-Friedman et al., 2001). Expression of Ca2þ-permeable AMPA receptors, lacking the GluA2 subunit, is a characteristic feature of Bergmann glia (Burnashev et al., 1992; Muller et al., 1992). The receptors are sited on Bergmann cell bodies and along processes that encapsulate the synapses between parallel or climbing fibers and Purkinje cells, showing preferential distribution in the processes surrounding synapses and in particular on the plasma membrane facing the presynaptic ending of parallel fibers (Matsui et al. 2005). These AMPA receptors can be activated by glutamate released from climbing and parallel fibers (Clark and Barbour, 1997; Bergles et al., 1997; see also Balakrishnan et al., 2014), and mediate detectable Ca2þ transients in Bergmann processes (Grosche et al., 1999; Beierlein and Regehr, 2006; Piet and Jahr, 2007). Activation of the Ca2þ-permeable AMPA receptors on Bergmann processes and the consequent Ca2þ signaling is essential to maintain structural and functional connections between the glial cell and the glutamatergic synapses made by parallel and climbing fibers onto Purkinje cells (Iino et al., 2001; see also Piet and Jahr, 2007; Hoogland and Kuhn, 2010; Saab et al., 2012). Nevertheless, the impact of Bergmann glia Ca2þ signaling on synaptic function is not fully understood. In particular, while it is recognized that Ca2þ signaling in astrocyte processes may trigger release of gliotransmitters, to our knowledge there is no information if localized Ca2þ responses trigger release of gliotransmitters from Bergmann glia. The sites where neuronal synapses are enwrapped by astrocytic processes are important for bidirectional communication between neurons and astrocytes. The astrocytic processes express the machinery for uptake and release of gliotransmitters, as well as receptors that can sense neurotransmitters and regulate gliotransmitter release (see Montana et al., 2006; Santello and Volterra, 2009; Bergersen and Gundersen, 2009; Parpura and Verkhratsky, 2011). Processes that contain gliotransmitter-loaded vesicles and are competent for gliotransmitter release can be isolated from central nervous system astrocytes; in the processes Ca2þ signals and gliotransmitter release can be directly measured (see Stigliani et al., 2006). Notably, the processes can be obtained from mature astrocytes from diverse adult animal brain regions: in situ maturation of astrocytes allow investigations on functioning of mature astrocytic processes and on the astrocyte contribution to synaptic transmission in mature brain. To better understand the mechanisms that might be involved in the Bergmann glia modulation of synaptic transmission (at synapses of parallel and climbing fibers onto Purkinje cells), herein we investigate on Ca2þ responses and gliotransmitter release from the purified astrocytic processes acutely prepared from adult rat cerebellum. We report evidence that processes of Bergmann glia are recovered in purified cerebellar gliosomes; AMPA receptors permeable to Ca2þ (GluA2-lacking) are located on Bergmann glia processes; activation of the AMPA receptors triggers Ca2þ entry and Ca2þ-dependent vesicular glutamate release from the processes; confocal imaging and ultrastructural analysis are consistent with the presence of glutamatergic vesicles in the processes. The findings provide the first direct evidence that Bergmann glia processes have the capacity to release glutamate. The AMPA-evoked vesicular release of glutamate from perisynaptic processes enwrapping excitatory synapses onto Purkinje cells might represent a

397

mechanism by which Bergmann glia microdomains can modulate glutamatergic transmission at individual Purkinje cells synapses in defined areas (see Grosche et al., 1999). 2. Materials and methods 2.1. Animals Adult male rats (Sprague Dawley 200e250 g), were bred at the animal care facility at DIFAR, University of Genova, Italy, housed at constant temperature (22 ± 1 C) and relative humidity (50%) under a regular lightedark schedule (lights on 7 AMe7 PM) and with free access to standard pellet diet and water. Experimental procedures and animal care complied with the European Communities Parliament and Council Directive of 22 September 2010 (2010/63/ EU) and with the Italian D.L. n. 26/2014, and were approved by the Italian Ministry of Health (protocol number 26768 of November 2012), in accordance with Decreto Ministeriale 116/1992. All efforts were made to minimize animal suffering and the number of animals used. 2.2. Preparation of purified astrocytic processes and nerve terminals After decapitation the cerebellum was rapidly removed and placed in ice-cold medium. Purified glial resealed particles were prepared according to Nakamura et al. (1993) with some modification; the particles (gliosomes) are astrocytic processes containing gliotransmitter-loaded vesicles and competent for gliotransmitter secretion; functional experiments support the morphological evidence that astrocyte processes are a preparation with negligible neuronal contamination (Stigliani et al., 2006). Briefly, the tissue was homogenized in 10 volumes of 10 mM Tris/HCl pH 7.4 containing 0.32 M sucrose, using a glass-Teflon tissue grinder (clearance 0.25 mm). The homogenate was centrifuged (5 min at 4 C and 1000 g) to remove nuclei and debris and the supernatant stratified on a discontinuous Percoll gradient (2, 6, 10 and 20% (v/v) in Trisbuffered sucrose) and centrifuged for 5 min at 4 C and 33 500 g. The layer between 2% and 6% (v/v) Percoll (gliosomal fraction; purified astrocyte processes) was collected and washed by centrifugation. In a subset of experiments, purified nerve terminals (purified synaptosomes) were prepared as previously reported (Cervetto et al., 2010) by collecting the layer between 10 and 20% Percoll. For release experiments, purified astrocyte processes (or nerve terminals) were suspended in standard HEPES medium (mM: NaCl 128, KCl 2.4, MgSO4 1.2, KH2PO4 1.2, CaCl2 1.0, and HEPES 10 with glucose 10, pH 7.4). For immunofluorescence confocal microscopy or calcium imaging, gliosomal or synaptosomal pellets were re-suspended in HEPES medium (1 mg proteinsml). Protein determinations were carried out according to Bradford (1976). 2.3. [3H]D-aspartate and endogenous glutamate release We studied release of the gliotransmitter glutamate by applying to astrocyte processes (gliosomes) the method for measuring the release of neurotransmitters from isolated superfused nerve terminals (synaptosomes); the method was repeatedly proven suitable to identify receptors located on the nerve terminals. Release monitoring from a superfused gliosomal (or synaptosomal) monolayer (see Raiteri et al., 1974), by removing any released compound, avoids formation of receptor biophase and enables ‘nude’ receptors to be exposed, preventing indirect effects. In these experimental conditions, only the targets located on the external membrane of astrocyte processes are selectively acted upon, allowing pharmacological characterization of release-regulating

398


receptors and direct investigation on modes for gliotransmitter release from the processes. After incubation with [3H]D-aspartate (0.03 mM, 15 min) gliosomes were transferred to parallel superfusion chambers at 37 C and superfused (0.5 ml/min) with standard medium as previously described (Marcoli et al., 2008; Cervetto et al., 2010). Briefly, after 33 min superfusion, superfusate fractions were collected in 3 min samples (from the first fraction, basal1, B1 to B7); after 38 min of superfusion, gliosomes were exposed (12 min) to ionotropic glutamate receptor agonists. The effect of cyclothiazide, concanavalin A, of ionotropic glutamate receptor antagonists, or of DL-threo-b-benzyloxyaspartic acid (DL-TBOA) was evaluated on the agonist response by adding the compounds 8 min before the agonist. Dependence of drug-evoked endogenous glutamate release or tritium release on external Ca2þ was assessed in gliosomes superfused with Ca2þ -free medium (supplemented with EGTA 0.5 mM) starting 18 min before addition of the agonist. In a subset of experiments, to block vesicular glutamate transporter (VGLUT) function, we preincubated gliosomes (30 min at 37 C) in the presence of rose bengal (0.5 mM), trypan blue (1 mM) or bafilomycin A1 (0.1 mM), before and during vesicle loading with [3H]Daspartate. Because bafilomycin A1 (as well as VGLUT inhibitors) prevents refilling of transmitter-empty vesicles, although it is without effect on transmitter-filled vesicles (Zhou et al., 2000), we challenged the processes with Kþ (KCl 25 mM substituting an equimolar concentration of NaCl; 15 min) to lead to discharge of the vesicles contents at the beginning of the incubation period with bafilomycin A1 or VGLUT inhibitors. In a subset of experiments, [3H] D-aspartate release was measured from purified synaptosomes, incubated with [3H]D-aspartate (0.03 mM, 15 min) transferred to parallel superfusion chambers at 37 C and superfused (0.5 ml/min) analogously to gliosomes (see Cervetto et al., 2010). In each experiment at least one chamber was used as a control for each condition and was superfused with standard medium or with medium appropriately modified. The amount of endogenous glutamate released in the fractions collected was measured by high-performance liquid chromatography, as previously described (Cervetto et al., 2012). The analytical method involved automatic precolumn derivatization (Waters 715 ultra wisp; Milford, MA) with o-phthalaldehyde, followed by separation on a C18 reversephase chromatography column (Chrompack International, 10 cm 4.6 mm, 3 mm) and fluorimetric detection. Homoserine was used as an internal standard. The detection limit was 100 fmol/ml. Protein determinations were carried out according to Bradford (1976). The radioactivity in gliosomes at the end of superfusion and in superfusate samples was determined by liquid scintillation counting. The amount of endogenous glutamate released in the fractions was expressed as pmol/mg protein. The efflux of radioactivity in each fraction was calculated as a percentage of the total radioactivity present at the onset of the fraction considered (fractional release). The mean endogenous glutamate or tritium fractional release in B1 and B2 fractions was taken as the 100% control value for each chamber; endogenous glutamate and tritium efflux in Bn fractions were evaluated as the percent variation with respect to the corresponding control value. The drug-evoked endogenous glutamate or tritium efflux was measured by subtracting the area under the curve of percent variations in endogenous glutamate release or in tritium fractional release in appropriate control chambers from the area under the curve of the percent variations in drug-treated chambers. 2.4. Calcium imaging The intragliosomal calcium concentration ([Ca2þ]i) was measured by the fura-2 acetoxymethylester (fura-2AM)

microfluorimetric technique essentially as described for single glued synaptosomes (Gualix et al., 2003; Marcoli et al., 2008; Cervetto et al., 2012). Gliosomes (15e20 mg) were deposited onto 20-mm glass coverslips coated with Cell-Tack (BD Biosciences, San Jose, CA, USA), centrifuged at 3000 rpm (3 min), loaded with 5 mM fura-2AM (45 min at 37 C) and then mounted in a microperfusion chamber on the stage of an inverted fluorescence microscope (Nikon TE200) equipped with a dual excitation fluorimetric Ca2þ imaging system (Hamamatsu, Milan, Italy) as described previously (Alloisio et al., 2008). Gliosomes were excited alternately at 340 and 380 nm with a sampling rate of 0.15 Hz, the resultant emission being collected above at 510 nm by a cooled digital CCD camera (C4880-91; Hamamatsu Photonics, Sunayama-Cho, Japan) and recorded by means of dedicated software (Aquacosmos; Hamamatsu Photonics). The changes in fluorescence ratio (F340/F380) were determined from selected regions of interest (ROI): isolated processes or microareas including 20e50 processes. The ratio F340/ F380 was used to indicate the changes in [Ca2þ]i. No calibration was performed in view of the many uncertainties related to the dissociation constant of fura-2 in the intracellular milieu (Karaki et al., 1997). However, in some experiments [Ca2þ]i was calculated according to Grynkiewicz et al. (1985), by using a Kd of 230 nM for the Ca2þ/fura-2 complex. Drugs were superfused at the rate of about 2.5 ml/min with HEPES medium and imaged through a Nikon 100 lens. Each experiment was carried out on at least three independent preparations. A pulse of high Kþ (35 mM) was applied at the end of each experiment to verify the viability of gliosomes. 2.5. Immunofluorescent confocal microscopy Gliosomes or synaptosomes (15e20 mg) were fixed with 2% paraformaldehyde, permeabilized with 0.05% Triton X-100 (5 min) and incubated 60 min with primary antibodies diluted in phosphate buffer saline (PBS) containing 3% albumin. The following antibodies were used: mouse anti-microtubule associated protein 2 (MAP2; 1:1000; SigmaeAldrich, St. Louis, MO, USA), mouse or rabbit anti-glial fibrillary protein (GFAP; 1:1000; SigmaeAldrich); rabbit anti-brain specific lipid binding protein (BLBP; 1:500; Millipore Corporation, Billerica, MA, USA) mouse anti-oligodendrocyte (RIP; 1:10,000; Millipore Corporation), and mouse anti-integrinaM (clone OX-42; 1:25; Millipore Corporation), guinea pig antiVGLUT1 (1:1000; Synaptic Systems, Goettingen, Germany) and mouse anti-VGLUT2 (1:500; Synaptic System). After washing with PBS, the preparations were incubated (60 min) with Alexa Fluor 488, 568 or 633 secondary antibodies conjugates (1:1000; Life Technologies Corporation, Carlsbad, CA, USA) in PBS containing 0.5% albumin. Images were collected by confocal microscopy using a three-channel TCS SP2 laser-scanning confocal microscope (Leica Wetzlar, Germany), equipped with 458, 476, 488, 514, 543 and 633 nm excitation lines. To assess the purity of each fraction, we estimated the percentage of GFAP, MAP-2, RIP and integrin-aMpositive particles in five non-overlapping fields from two different preparations of gliosomes and synaptosomes, respectively. We used ImageJ 1.44 software (Wayne Rasband, National Institutes of Health, Bethesda, MD, USA) to set the maximum and the minimum size of particles that were to be counted and to determine the fluorescence intensity above the background. Spatial colocalization was analyzed through two-dimensional correlation cytofluorograms obtained by means of macro routines integrated as plug-ins in ImageJ 1.34f software (Wayne Rasband, National Institutes of Health, Bethesda, MD, USA). Red and green labels were considered as co-localized in the same pixel if their respective intensities (0e255, eight bit) were strictly higher than the threshold of their channels, as determined by analyzing the color histograms. Data were collected from 10 to 12 fields from three different


preparations, and are expressed as mean ± SEM. 2.6. Western blot Proteins (1 mg/ml) were denatured in Laemli buffer and then subjected to a 10% SDS-polyacrylamide gel electrophoresis 200 V for 50 min (25 mg per lane; Mini Protean TGX Gel, Bio-Rad Laboratories, Hercules, CA, USA), followed by electroblotting (100 V for 50 min) on polyvinylidene difluoride membrane (Immobilon-P PVDF; Millipore Corporation). Immunodetection was performed using guinea pig anti-VGLUT1 (1:5000; Synaptic Systems) and rabbit anti-VGLUT2 (1:2000; Synaptic System). Primary antibodies were detected by using horseradish peroxidase-linked anti-guinea pig, or anti-rabbit (Cell Signaling Technology, Inc., Danvers, MA, USA) secondary antibodies, incubated for 1 h at room temperature. The specific bands were detected by means of an enhanced chemiluminescence system (Pierce). The membranes were stripped using Re-blot plus solution (Millipore Corporation) and re-probed with mouse anti-b-actin (1:1000; SigmaeAldrich). Developed films were analyzed using a specific software (GelDoc; Bio-Rad Laboratories).

399

1% osmium tetroxide in cacodylate buffer 0.1 M, pH 7.2, en bloc stained with a 1% aqueous solution of uranyl acetate and dehydrated through a graded ethanol series. Samples were then embedded in LX112 (Polysciences Inc., Warrington, PA, USA), polymerized for 12 h at 42 C, followed by 48 h at 60 C. Grey-silver ultrathin sections were obtained using a Leica Ultracut E microtome, stained with uranyl acetate and lead citrate, and analyzed with a FEI CM10 or Tecnai 12 G2 (FE1, Eindhoven, the Netherlands) electron microscope. 2.8. Calculations and statistical analysis Log concentration-response relationships and the agonist EC50 (half-maximum effective concentrations) or Emax (maximum effect) or the antagonist IC50 (half-maximum inhibitory concentrations) values were obtained by using a four-parameter logistic function-fitting routine (Sigma Plot software). Means ± SEM of the numbers of experiments (n) are indicated throughout. Significance of the difference was analyzed by the non-parametric ManneWhitney test, with statistical significance being taken at p < 0.05.

2.7. Electron microscopy

2.9. Materials

For ultrastructural analysis, purified gliosomes were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2, post-fixed in

[3H]D-aspartate was from Amersham Radiochemical Centre (Buckinghamshire, UK); a-amino-3-hydroxy-5-methyl-4-isoxazole

Fig. 1. Rat cerebellar astrocyte processes. Negligible contamination of astrocyte processes by nerve terminals, microglia, or oligodendrocytes. Immunofluorescence for GFAP (A, D, G) and for the neuronal marker MAP2 (B), the oligodendrocyte marker RIP (E), or the microglia marker integrin-aM (H), showing negligible contamination of purified astrocyte processes by non-astrocytic fractions. Bars indicate percentage of positive particles (% ± SEM of five non-overlapping fields from two different preparations): GFAP (C, F, I; solid bar) and MAP2, RIP or integrin-aM (C, F, I; empty bar), respectively.

400


Fig. 2. Rat cerebellar purified astrocyte processes contained Bergmann glia processes. Confocal microscopy analysis. Immunofluorescence for GFAP (A) and for the Bergmann glial marker BLBP (B); merge image (C). Co-localization of GFAP with BLBP in single processes (D); bars indicate percentage of co-localization (% ± SEM of 10 fields from three different preparations): GFAP-positive processes expressing also BLBP (solid bar) and BLBP-positive processes expressing also GFAP (empty bar).

propionate (AMPA), bafilomycin A1, (S)-a-amino-2,3,4,5,6,7hexahydro-2,4-dioxo-1H-cyclopen tapyrimidine-1-propanoic acid ((S)-CPW 399), 6-chloro-3,4-dihydro-3-(5-norbornen-2-yl)2H-1,2,4-benzothiazidiazine-7-sulfonamide-1,1-dioxide (cyclothiazide), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), 1-(4aminophenyl)-4-methyl-7,8-methylenedioxy-5H-(3-N-methylcarbamate)-2,3 benzodiazepine (GYKI 52466), 1-Naphthylacetyl spermine trihydrochloride (NASPM), DL-threo-b-benzyloxyaspartic acid (DL-TBOA) were from Tocris Bioscience (Bristol, UK); fura-2 acetoxymethylester (fura-2AM), concanavalin A, trypan blue and rose bengal were from SigmaeAldrich. When possible drugs were dissolved in distilled water or in physiological medium. Concanavalin A was dissolved in a buffer of 10 mM Hepes and 100 mM CaCl2 and then diluted 1:300. Stock solutions of cyclothiazide, bafilomycin A1, or CNQX were prepared in dimethyl sulfoxide and diluted at least 1:1000 in physiological medium; stock GYKI 52466 solutions were freshly prepared in HCl 0.1N and diluted 1:800; dimethyl sulfoxide diluted 1:1000 or HCl 0.1N diluted 1:800 had no effect on [3H]D-aspartate or endogenous glutamate release. All the salts were from SigmaeAldrich.

3. Results 3.1. Purified cerebellar astrocytic processes contain Bergmann glia processes Astrocytic processes were labeled with anti-GFAP antibody, a marker that identifies astrocytes, and proved negative for MAP2, integrin-aM, and RIP, which are markers for nerve terminals, microglia, and oligodendrocytes, respectively. This indicates negligible contamination of cerebellar astrocyte processes by nerve terminals, microglia, or oligodendrocytes (Fig. 1AeI). Expression of the Bergmann glia marker brain-specific lipid binding protein (BLBP) (Anthony et al., 2004) indicated that processes of the Bergmann glia are recovered in our preparation, representing more than 80% of the GFAP-positive processes (Fig. 2). Collectively the findings indicate that the processes of Bergmann glia represent the vast majority of astrocyte processes recovered from adult rat cerebellum.

Fig. 3. Activation of AMPA receptors evoked glutamate release from rat cerebellar astrocyte processes. (A) Log concentration-response curves for AMPA or (S)-CPW 399, in facilitating [3H]D-aspartate efflux. (B) Antagonism by CNQX and GYKI 52466 of the AMPA or (S)-CPW 399-evoked efflux of [3H]D-aspartate. Ineffectiveness of concanavalin A and cyclothiazide potentiation of AMPA effect; antagonism by GYKI 52466 of the AMPA-evoked [3H]D-aspartate efflux in the presence of cyclothiazide. Bars represent percentage of increase in [3H]D-aspartate in the presence of the drugs at the concentrations indicated. AMPA or (S)-CPW 399 was added for 12 min during superfusion; cyclothiazide, concanavalin A, or the antagonists were added 8 min before the agonist. Other experimental details in the Section 2.3. Data are means ± SEM (bars) of 3e12 experiments performed in triplicate. *p < 0.05 compared with the effect of the agonist alone; #p < 0.05 compared with the effect of AMPA plus cyclothiazide.


3.2. Glutamate releasing AMPA receptors are located on cerebellar astrocytic processes Glutamate release was studied by measuring tritium efflux from superfused astrocytic processes prelabeled with [3H]D-aspartate. To rule out the possibility that tritium release was an artifact resulting from the use of [3H]D-aspartate, in most experiments we monitored

401

in parallel the release of endogenous glutamate. In fact the behavior of endogenous glutamate was indistinguishable from that of tritium after labeling with [3H]D-aspartate, thus justifying the use of [3H]D-aspartate. The basal fractional [3H]D-aspartate outflow in the first two fractions collected from astrocytic processes amounted to 1.94 ± 0.09%/min (n ¼ 32); the basal outflow of endogenous glutamate in the first two fractions amounted to 1.26 ± 0.12 nmol/

Fig. 4. Ca2þ-permeable AMPA receptors accounted for AMPA-evoked glutamate efflux from rat cerebellar astrocyte processes. (A, B) Ca2þ dependence of AMPA-evoked glutamate release from rat cerebellar astrocyte processes; effect of external Ca2þ deprivation on the AMPA-evoked [3H]D-aspartate (A) or endogenous glutamate (B) efflux. (C, D) Effectiveness of NASPM against the [3H]D-aspartate (C) or endogenous glutamate (D) efflux evoked by AMPA, (S)-CPW 399 or AMPA plus cyclothiazide. Log concentration-response relationship of NASPM in inhibiting the [3H]D-aspartate (E) or endogenous glutamate (F) release evoked by AMPA (100 mM; ). Bars represent percentage increase in tritium or endogenous glutamate in the presence of the drugs at the concentrations indicated. Ca2þ-free EGTA (0.5 mM)-containing medium was added 18 min before the agonist. Other experimental details as in the legend to Fig. 3 or in the Section 2.3. Data are mean ± SEM of 3e14 independent experiments in triplicate. *p < 0.05 compared with the effect of the agonist alone in standard medium; #p < 0.05 compared with the effect of AMPA plus cyclothiazide.

▪

402


mg protein/min (n ¼ 19). AMPA increased [3H]D-aspartate efflux in a concentrationdependent manner (EC50 value 55.7 mM; Emax value 235% increase; Fig. 3A). (S)-CPW 399, a full AMPA receptor agonist that prevents receptor desensitization (Campiani et al., 2001), mimicked the effect of AMPA but exhibited greater efficacy (EC50 value 44.4 mM; Emax value 460% increase; Fig. 3A). Cyclothiazide (10 mM), which selectively inhibits AMPA receptor desensitization (Wong and Mayer, 1993), increased the releasing effect of AMPA, while concanavalin A (3 mM), which prevents kainate receptor desensitization while exerting minimal effects on AMPA receptors (Wong and Mayer, 1993), was ineffective (Fig. 3B). The AMPA/kainate re et al., 1988; 10 mM) or the selective ceptor antagonist CNQX (Honore AMPA receptor antagonist GYKI 52466 (Wilding and Huettner, 1995; 30 mM) inhibited the response to AMPA (in the presence or not of cyclothiazide) or to (S)-CPW 399 (Fig. 3B). Collectively the findings indicate the presence of functional AMPA receptors on astrocyte processes recovered from adult rat cerebellum, whose activation evokes release of the gliotransmitter glutamate. 3.3. Ca2þ-permeable AMPA receptors are located on cerebellar astrocytic processes The [3H]D-aspartate or endogenous glutamate releasing response to AMPA was almost abolished in the absence of extracellular Ca2þ (Fig. 4A,B). The effect of AMPA was not inhibited in the presence of DL-TBOA, a non-transportable blocker of glutamate plasmamembrane transporters (per cent increase of [3H]D-aspartate release: AMPA 50 mM, 94.8 ± 3.9; AMPA 50 mM þ DL-TBOA 10 mM, 91.2 ± 5.6; n ¼ 3), ruling out the possibility that Naþ influx through the AMPA receptors could induce reverse glutamate transport. The compound NASPM, a selective antagonist of the GluA2lacking, Ca2þ-permeable AMPA receptors (Tsubokawa et al., 1995;

Koike et al., 1997; Nilsen and England, 2007) abolished the AMPA-evoked [3H]D-aspartate or endogenous glutamate release (IC50 value 0.11 mM and 0.09 mM against [3H]D-aspartate and endogenous glutamate release, respectively; Fig. 4CeF); NASPM was also effective when the AMPA receptor desensitization was inhibited by cyclothiazide, or against the response to the nondesensitizing AMPA agonist (S)-CPW 399 (Fig. 4C,D). Addition of cyclothiazide, concanavalin A, or the ionotropic glutamate receptor antagonists at the concentrations used did not affect basal [3H]Daspartate or endogenous glutamate efflux (data not shown). Experiments on the purified synaptosomal fraction showed that the nerve terminals obtained from the cerebellum were almost devoid of contamination by astrocyte processes or other nonneuronal particles (Fig. 5AeI). The basal fractional [3H]D-aspartate outflow in the first two fractions collected from the nerve terminals amounted to 1.02 ± 0.23%/min (n ¼ 3); AMPA (50 mM) evoked a [3H] D-aspartate releasing response that was prevented by CNQX (10 mM) or GYKI 52466 (30 mM), while was unaffected by NASPM (Fig. 5J). The findings rule out the possibility that the NASPMsensitive AMPA response of the Bergmann processes could be due to neuronal contamination of the gliosome preparation. Isolated astrocyte processes and microareas containing 15e30 gliosomes were tested by the calcium imaging technique. When gliosomes were challenged with AMPA (50 mM), an increase in intragliosomal Ca2þ concentration, expressed as F340/F380 ratio, was observed. Fig. 6B,C shows typical AMPA-activated [Ca2þ]i responses obtained from a single astrocyte process and from a microarea during perfusion with 50 mM AMPA and subsequent 35 mM Kþ; the [Ca2þ]i response to depolarization (35 mM Kþ) at the end of each experiment confirmed the integrity of the gliosomal preparation. In fact, a calcium response to high Kþ was reported in in situ Bergmann glia cells (Shao and McCarthy, 1997). Interestingly, the finding confirms that processes acutely isolated from mature glial cells are capable of responding to Kþ depolarization (Paluzzi et al., 2007; see also Yaguchi and Nishizaki, 2010;

Fig. 5. Rat cerebellar nerve terminals. Negligible contamination by astrocyte processes, microglia, or oligodendrocytes, and response to AMPA. Immunofluorescence for the neuronal marker MAP2 (A, D, G) and for GFAP (B), the oligodendrocyte marker RIP (E), or the microglia marker integrin-aM (H), showing negligible contamination of purified nerve terminals by non-neuronal fractions. Bars indicate percentage of positive particles (% ± SEM of 5 non-overlapping fields from three different preparations): MAP2 (C, F, I; solid bar) and GFAP, RIP or integrin-aM (C, F, I; empty bar), respectively. (J) Ineffectiveness of NASPM and effectiveness of CNQX and GYKI 52466 against the [3H]D-aspartate efflux evoked by AMPA. Bars represent percentage increase in tritium in the presence of the drugs at the concentrations indicated. Other experimental details in the Section 2.4. Data are mean ± SEM of 3 independent experiments in triplicate. *p < 0.05 compared with the effect of AMPA.


403

Fig. 6. Intracellular calcium measurements in single gliosomes and in microareas of gliosomes. Fluorescence image showing fura-2AM loaded gliosomes; the regions of interest (ROI) are represented by numbered squares (A). Typical Ca2þ responses to 50 mM AMPA and 35 mM Kþ from a single gliosome (B) and in a microarea (C). Solid bars indicate the period during which substances were applied. (D) The AMPA-induced [Ca2þ]i signal was inhibited by 10 mM NASPM. Values of [Ca2þ]i increase are expressed as F340/F380 ratio. Data are mean ± SEM of 25 determinations in microareas from 3 independent experiments; *p < 0.05 vs. AMPA alone in standard medium.

Heinrich et al., 2012); this is shared by adult cultured astrocytes, at variance with cultured astrocytes from neonatal rodents that are indeed non-responsive to depolarization (see Paluzzi et al., 2007). The AMPA-induced [Ca2þ]i signal was inhibited by NASPM (10 mM; Fig. 6D). All together the results indicate that the AMPA receptors located on the external membrane of the astrocyte processes recovered from adult rat cerebellum are Ca2þ-permeable (GluA2-lacking).

3.4. Activation of the AMPA receptors evokes vesicular glutamate release from cerebellar astrocytic processes Inhibition of VGLUTs by the allosteric modulator rose bengal (Ogita et al., 2001) or by the competitive antagonist trypan blue (Roseth et al., 1998), or inhibition of the vacuolar type Hþ-ATPases (V-ATPases) by bafilomycin A1 (Bowman et al., 1988) prevented the endogenous glutamate as well as [3H]D-aspartate releasing response to AMPA (Fig. 7A,B). Preincubation with rose bengal, trypan blue or bafilomycin A1 at the concentrations used did not affect basal endogenous glutamate or [3H]D-aspartate efflux (data not shown). Expression of VGLUT1 and VGLUT2 in BLBP-positive processes was confirmed by immunofluorescence (Fig. 8). In particular, we found that about 80% of Bergmann glia (BLBP-positive) processes expressed VGLUT1 (Fig. 8AeD), and about 50% expressed VGLUT2 (Fig. 8EeH). Analysis of co-expression of VGLUT1 and VGLUT2 indicates that VGLUT2 was present in Bergmann glia (BLBP-positive) processes expressing VGLUT1, while about 25% of BLBP-positive processes expressed only VGLUT1 (Fig. 8IeM). Single Bergmann glia processes expressing both VGLUT1 and VGLUT2, or VGLUT1 only, are shown at higher magnification in Fig. 8NeU. Western Blot analysis confirmed expression of VGLUT1 and VGLUT2 in the processes (Fig. 8V). The presence of vesicles in the processes was confirmed by ultrastructural analysis: in Fig. 9 single processes equipped with smooth vesicles and with some clathrin-coated vesicles with a diameter of approximately 30 nm and with a scattered distribution within the cytoplasm, are shown. Collectively the findings indicate that Bergmann glia processes are endowed with glutamate loading vesicles, and Ca2þ-permeable AMPA receptors can activate vesicular gliotransmitter release.

4. Discussion The study provides the first direct evidence that the Bergmann glia processes are competent for vesicular exocytosis of the gliotransmitter glutamate, and that Ca2þ entry through Ca2þ-permeable AMPA receptors is coupled with vesicular glutamate release. The evidences were obtained by using astrocyte processes acutely prepared from adult rat cerebellum, for the most part originating from Bergmann cells and likely mirroring specific features of mature Bergmann glia processes. In fact, the processes were immunoreactive for GFAP indicating their astrocytic origin, and negligibly contaminated by nerve terminals, microglia, or oligodendrocytes. The Bergmann glia share with astrocytes the

Fig. 7. AMPA receptor activation evoked vesicular release of glutamate from rat cerebellar astrocyte processes. Effects of the VGLUT blockers rose bengal and trypan blue and of the V ATPase blocker bafilomycin A1 on the AMPA-evoked efflux of [3H]Daspartate or endogenous glutamate. Bars represent percentage increase in tritium (A) or endogenous glutamate (B); rose bengal, trypan blue or bafilomycin A1 was preincubated at the concentrations indicated for 30 min. Other experimental details as in the legend to Fig. 3 or in the Section 2.3. Data were mean ± SEM of 3e14 independent experiments in triplicate. *p < 0.05 when compared with control AMPA (preincubation in standard medium).

404


Fig. 8. Presence of VGLUTs in rat Bergmann glia processes. Immunofluorescence for BLBP (A, E), and for VGLUT1 (B) and VGLUT2 (F) and merge image showing co-expression of the markers (C, G). Co-localization of VGLUT1 and VGLUT2 with BLBP (D, H): bars indicate percentage (± SEM of 9e12 fields from three different preparations) of BLBP-positive processes endowed with VGLUT1 or VGLUT2 (D, H). Co-localization of VGLUT1 and VGLUT2 with BLBP in the processes (IeU): the images represent anti VGLUT1 (K, P, T), anti VGLUT2 (J, O, S), and anti-BLBP (I, N, R); merge image (L, Q, U). Schematic drawing of VGLUTs co-localization on BLBP-positive processes from rat cerebellum (M). Note a single process expressing both VGLUT1 and VGLUT2 (Q) and a process expressing VGLUT1 only (U). Western blot for VGLUT1 and VGLUT2 in cerebellar gliosome preparation (V).

specific marker GFAP (Roots, 1981); Bergmann glia also express BLBP both during development, and when the processes have ensheathed synapses (Yamada et al., 2000; Sottile et al., 2006;

Koirala and Corfas, 2010). Analysis of expression of BLBP in GFAPpositive processes indicates that Bergmann glia processes represent more than 80% of the astrocyte processes recovered from adult

Fig. 9. Electron micrographs of rat cerebellar gliosomes. Purified gliosomes were fixed, dehydrated and embedded in LX112. Ultrathin sections were stained with uranyl acetate and lead citrate, and analyzed with FEI CM10 or Tecnai 12 G2 electron microscopes. Single gliosomes are shown containing approximately 30-nm smooth and clathrin-coated (black arrow) vesicles scattered in the cytoplasm. Inset: The black boxed area is reproduced at higher magnification showing the smooth and clathrin-coated vesicles. Scale bar 0.205 mm.


rat cerebellum. We then used this preparation to investigate on gliotransmitter release and Ca2þ signals in Bergmann glia processes and on roles for the Bergmann AMPA receptors. The processes were equipped with release-stimulatory AMPA receptors; the glutamate releasing response to AMPA was lost in the absence of extracellular Ca2þ, indicating that it was dependent on AMPA-evoked Ca2þ entry. Direct evidence that Ca2þ enters the processes following AMPA receptor activation was obtained by measuring Ca2þ signals in processes. Sensitivity of the AMPAevoked Ca2þ signal or glutamate release to NSPAM, a selective inhibitor of GluA2-lacking Ca2þ-permeable AMPA receptors (Tsubokawa et al., 1995; Koike et al., 1997; Nilsen and England, 2007) is consistent with Ca2þ entering the astrocyte processes through the AMPA receptors. It is known that Ca2þ-permeable, GluA2-lacking AMPA receptors are almost exclusively expressed on Bergmann glia in cerebellum (Burnashev et al., 1992; Muller et al., 1992). The finding that functional Ca2þ-permeable AMPA receptors are located on astrocyte processes acutely prepared from adult rat cerebellum confirms that the processes are a viable preparation of Bergmann glia processes. The glutamate-releasing response to AMPA of the cerebellar nerve terminals, sensitive to CNQX or GYKI 52466 (see also Cervetto et al., 2010) but insensitive to NASPM, provides additional evidence that the NASPM-sensitive AMPA response of gliosomes is due to activation of the Ca2þpermeable AMPA receptors expressed on the Bergmann glia process membrane. The Ca2þ-permeable, NASPM-sensitive AMPA receptors on Bergmann glia processes are activated by ectopic release of glutamate from parallel fibers and from the climbing fibers (outside the synaptic cleft) (see Balakrishnan et al., 2014; Matsui and Jahr, 2003, 2004; Matsui et al., 2005), that can diffuse to the Bergmann AMPA receptors through volume transmission (see Agnati et al., 1992, 1995, 2014). Activation of such Ca2þ-permeable, NASPM-sensitive receptors appears important for modulating synapse transmission and for synapse plasticity (see Matsui and Jahr, 2003; Bellamy and Ogden, 2005; Piet and Jahr, 2007). Notably, the AMPA receptors seem to play a role as sensors of neural activity in synaptic signaling, to control coverage of synapses by the Bergmann glia processes, and therefore clearance of glutamate from the synapse through the transporters located on the processes themselves (see Saab et al., 2012; Buffo and Rossi, 2013 and references therein); by this way, quite interestingly, the AMPA receptors reached via volume transmission by ectopically released glutamate may function to modulate coverage of the synapses and then synapse privateness and volume transmission itself. We here propose that Bergmann glia AMPA receptors, besides signaling the presence of synapse and favoring “private” synaptic transmission by maintaining synaptic coverage by the glial processes, represent a new target for glutamate mechanism at the synapse. In fact, we here show that the AMPA receptors on Bergmann glia processes are coupled with localized Ca2þ signals and with the release of the gliotransmitter glutamate from the processes, which might provide an additional active mechanism for the Bergmann modulation of synapse transmission and plasticity. Interestingly, evidence was reported suggesting that Bergmann glia actively modulate neuronal activity and modify firing of Purkinje cells via the release of glutamate (Sasaki et al., 2012). Our findings indicate that activation of AMPA receptors on the Bergmann glia processes might be involved in such an active role, providing a mechanism by which the Bergmann glia sense synaptic activity and actively modulate neural activity at microdomains. Notably, functioning of the well-known main glutamate target on Bergmann glia process membranes (the glutamate transporter GLAST), and rapid clearance of glutamate is required to maintain the function of the fast kinetic AMPA receptors (see Attwell and Gibb, 2005). As far as the targets for the

405

gliotransmitter is concerned, glutamate released from Bergmann glia might trigger events at both postsynaptic and presynaptic levels through other glutamate receptors. Indeed, the release of glutamate depending on Bergmann glia photostimulation induced activation of AMPA receptors on Purkinje cells and long-term depression of parallel fibers to Purkinje cell synapses by activating metabotropic glutamate receptors (Sasaki et al., 2012). Also, the parallel fiber terminals were shown endowed with ionotropic glutamate kainate or AMPA (Ca2þ-impermeable AMPA, see Cervetto et al., 2010; present data) receptors that might be activated by the glutamate released from the Bergmann glia processes. Also, we must note that the climbing fiber signaling to interneurons, crucial for cerebellar circuit function, was through volume transmission (Szapiro and Barbour, 2007); activation of the Bergmann AMPA receptor by volume transmission from the climbing fibers, and consequent release of glutamate, might also be involved in the volume transmission to the interneurons (see Nishiyama and Linden, 2007). As a matter of fact, activation of the Bergmann AMPA receptors might be involved in triggering such diverse postsynaptic and presynaptic responses. Notably, our findings also indicate that Ca2þ entry through AMPA receptors triggered vesicular exocytotic glutamate release from Bergmann glia processes. Evidence has accumulated that astrocytes possess the machinery for exocytotic release of glutamate, glutamate being packaged into astrocytic small vesicles by VGLUTs that use the proton gradient generated by V-ATPase; blockage of VGLUTs by rose bengal, or of V-ATPase with bafilomycin A1 was shown to reduce glutamate release from astrocytes caused by various stimuli (Araque et al., 2000; Pasti et al., 2001; Montana et al., 2004; Crippa et al., 2006; see also Montana et al., 2006 and references therein). The vesicular glutamate transporters were expressed in astrocytic small vesicles, responsible for regulated vesicular glutamate release from astrocytes (Liu et al., 2011). While most studies used cultured astrocytes, evidence for the presence of VGLUTs in in situ astrocytes has been provided in the rodent hippocampus, frontal cortex, cerebellum or striatum (Bezzi et al., 2004; Zhang et al., 2004; Ormel et al., 2012a,b), suggesting that astrocytes in several brain regions are capable of vesicular release of glutamate. Ultrastructural analysis of the cerebellar glial processes showed several scattered small vesicles (uncoated or clathrin-coated) with diameters of ~30 nm within the cytoplasm. The findings are consistent with the morphological characteristics of vesicles competent for fusion in gliosomes from cerebral cortex of rats (Stigliani et al., 2006), and are at variance with the polarized clustered distribution of vesicles in the cerebrocortical nerve terminals obtained in parallel (Stigliani et al., 2006) or in ultrastructure of parallel fiber or climbing fiber terminals (Ango et al., 2008; Ormel et al., 2012b). We found that the (BLBP-positive) processes of mature Bergmann glia prepared from adult rat cerebellum expressed VGLUT1 and VGLUT2, indicating that they possess glutamatergic vesicles. We also obtained evidence, to the best of our knowledge for the first time, that the VGLUTs in Bergmann processes were functional, as the AMPA-evoked release was abolished by the VGLUT inhibitors rose bengal or trypan blue, or by the VATPases inhibitor bafilomycin A1. All together, the findings indicate that isolated processes of Bergmann glia are competent for vesicular exocytotic release of glutamate that can be triggered by activation of the Ca2þ-permeable AMPA receptors sited on the processes. Consistent with our findings, acutely isolated adult mouse Bergmann cells were reported to express components of regulated vesicular exocytosis, including synapsin I, synaptotagmins XI and XVI, syntaxins, synapsin, rim2, and Lgi3 (a syntaxin interactor and potential regulator of exocytosis) suggesting that Bergmann glia in vivo possess the machinery for regulated release of gliotransmitters (Koirala and Corfas, 2010).

406


5. Conclusions In conclusions, we here report the following novelties: - functional Bergmann glia processes can be recovered from adult rodent cerebellum, allowing functional investigation on mature Bergmann glia processes; - mature Bergmann glia processes possess vesicular mechanisms for release of glutamate; - activation of Ca2þ-permeable AMPA receptors triggers Ca2þ entry and vesicular glutamate release from the Bergmann glia processes. The capacity of Bergmann glia to release gliotransmitters as a consequence of parallel/ climbing fibers triggering Ca2þ signaling was hypothesized (see Grosche et al., 1999; Gallo and Ghiani, 2000; Bellamy, 2006), and release of the gliotransmitter glutamate from Bergmann glia was indeed proposed to contribute to active modulation of neuronal activity (Sasaki et al., 2012). We here provide direct evidence that AMPA-evoked Ca2þ signaling can trigger exocytotic release of glutamate from the Bergmann glia processes. We speculate that triggering glutamate release is a mechanism by which Ca2þpermeable GluA2-lacking AMPA receptors on Bergmann glia processes may contribute to compartmentalized signaling between parallel/climbing fibers and Purkinje cells in microdomains. This mechanism may add to the established roles of Bergmann glia in the control of synapse plasticity operated through glutamate transporters and AMPA-mediated signaling to transporters, and reveal a role for such receptors activated by dedicated glutamate release (see Matsui and Jahr, 2003, 2004; Matsui et al., 2005; see also Buffo and Rossi, 2013): sensing synaptic activity, maintaining the synapse enwrapping and controlling clearance of glutamate and therefore synaptic and extrasynaptic signals, but also providing an additional mechanism for regulation of the signals to the Purkinje cells in microdomains, meeting the requirements for fine motor control. Conflict of interest The authors have no conflict of interest to declare. Acknowledgments e The support of Italian Ministero dell’Istruzione, dell’Universita della Ricerca [Grant PRIN2008 to G.M]; and of University of Genoa, Progetto Ricerca Ateneo 2007, 2010 and 2011 [Grant 020301002054 to M.M., Grant D31J1100003005 and D31J1100161005 to C.C.] is gratefully acknowledged. References Agnati, L.F., Jelke, Band, Fuxe, K., 1992. Volume transmission in the brain. Do brain cells communicate solely through synapses? A new theory proposes that information also flows in the extracellular space. Am. Sci. 80, 362e374. Agnati, L.F., Zoh, M., Stromberg, I., Fuxe, K., 1995. Intercellular communication. In the brain wiring versus volume transmission. Neuroscience 69, 711e726. Agnati, L.F., Guidolin, D., Maura, G., Marcoli, M., Leo, G., Carone, C., De Caro, R., Genedani, S., Borroto-Escuela, D.O., Fuxe, K., 2014. Information handling by the brain: proposal of a new ‘‘paradigm’’ involving the roamer type of volume transmission and the tunneling nanotube type of wiring transmission. J. Neural. Transm. 121, 1431e1449. Alloisio, S., Cervetto, C., Passalacqua, M., Barbieri, R., Maura, G., Nobile, M., Marcoli, M., 2008. Functional evidence for presynaptic P2X(7) receptors in adult rat cerebrocortical nerve terminals. Febs Lett. 582, 3948e3953. Ango, F., Wu, C., Van der Want, J.J., Wu, P., Schachner, M., Huang, Z.J., 2008. Bergmann glia and the recognition molecule CHL1 organize GABAergic axons and direct innervation of Purkinje cell dendrites. PLoS Biol. 6 (4), e103. Anthony, T.E., Klein, C., Fishell, G., Heintz, N., 2004. Radial glia serve as neuronal progenitors in all regions of the central nervous system. Neuron 41, 881e890. Araque, A., Li, N., Doyle, R.T., Haydon, P.G., 2000. SNARE protein-dependent glutamate release from astrocytes. J. Neurosci. 20, 666e673. Attwell, D., Gibb, A., 2005. Neuroenergetics and the kinetic design of excitatory synapses. Nat. Rev. Neurosci. 6, 841e849. Balakrishnan, S., Dobson, K.L., Jackson, C., Bellamy, T.C., 2014. Ectopic release of

glutamate contributes to spillover at parallel fibre synapses in the cerebellum. J. Physiol. 592, 1493e1503. Bellamy, T.C., Ogden, D., 2005. Short-term plasticity of Bergmann glial cell extrasynaptic currents during parallel fiber stimulation in rat cerebellum. Glia 52, 325e335. Bellamy, T.C., 2006. Interactions between Purkinje neurones and Bergmann glia. Cerebellum 5, 116e126. Beierlein, M., Regehr, W.G., 2006. Brief bursts of parallel fiber activity trigger calcium signals in Bergmann glia. J. Neurosci. 26, 6958e6967. Bergersen, L.H., Gundersen, V., 2009. Morphological evidence for vesicular glutamate release from astrocytes. Neuroscience 158, 260e265. Bergles, D.E., Dzubay, J.A., Jahr, C.E., 1997. Glutamate transporter currents in Bergmann glial cells follow the time course of extrasynaptic glutamate. Proc. Natl. Acad. Sci. U. S. A. 94, 14821e14825. Bezzi, P., Gundersen, V., Galbete, J.L., Seifert, G., Steinhauser, C., Pilati, E., Volterra, A., 2004. Astrocytes contain a vesicular compartment that is competent for regulated exocytosis of glutamate. Nat. Neurosci. 7, 613e620. Buffo, A., Rossi, F., 2013. Origin, lineage and function of cerebellar glia. Prog. Neurobiol. 109, 42e63. Burnashev, N., Khodorova, A., Jonas, P., Helm, P.J., Wisden, W., Monyer, H., Seeburg, P.H., Sakmann, B., 1992. Calcium permeable AMPA receptors in fusiform cerebellar glial cells. Science 256, 1566e1570. Bowman, E.J., Siebers, A., Altendorf, K., 1988. Bafilomycins: a class of inhibitors of membrane ATPases from microorganisms, animal cells, and plant cells. Proc. Natl. Acad. Sci. U. S. A. 85, 7972e7976. Bradford, M.M., 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248e254. Campiani, G., Morelli, E., Nacci, V., Fattorusso, C., Ramunno, A., Novellino, E., Greenwood, J., Liljefors, T., Griffiths, R., Sinclair, C., Reavy, H., Kristensen, A.S., Pickering, D.S., Schousboe, A., Cagnotto, A., Fumagalli, E., Mennini, T., 2001. Characterization of the 1H-cyclopentapyrimidine-2,4(1H,3H)-dione derivative (S)-CPW 399, as a novel, potent and subtype-selective AMPA receptor full agonist with partial desensitisation-preventing properties. J. Med. Chem. 44, 4501e4504. Cervetto, C., Maura, G., Marcoli, M., 2010. Inhibition of presynaptic releasefacilitatory kainate autoreceptors by extracellular cyclic GMP. J. Pharmacol. Exp. Ther. 332, 210e219. Cervetto, C., Mazzotta, V., Frattaroli, D., Alloisio, S., Nobile, M., Maura, G., Marcoli, M., 2012. Calmidazolium selectively inhibits exocytotic glutamate release evoked by P2X7 receptor activation. Neurochem. Int. 60, 768e772. Clark, B.A., Barbour, B., 1997. Currents evoked in Bergmann glial cells by parallel fibre stimulation in rat cerebellar slices. J. Physiol. (Lond) 502, 335e350. Crippa, D., Schenk, U., Francolini, M., Rosa, P., Verderio, C., Zonta, M., Pozzan, T., Matteoli, M., Carmignoto, G., 2006. Synaptobrevin2-expressing vesicles in rat astrocytes: insights into molecular characterization, dynamics and exocytosis. J. Physiol. 570, 567e582. Gallo, V., Ghiani, C.A., 2000. Glutamate receptors in glia: new cells, new inputs and new functions. Trends Pharmacol. Sci. 21, 252e258. Gualix, J., Gomez-Villafuertes, R., Dıaz-Hernandez, M., Miras-Portugal, M.T., 2003. Presence of functional ATP and dinucleotide receptors in glutamatergic synaptic terminals from rat midbrain. J. Neurochem. 87, 160e171. €ller, T., Verkhratsky, A., Reichenbach, A., Kettenmann, H., Grosche, J., Matyash, V., Mo 1999. Microdomains for neuron-glia interaction: parallel fiber signaling to Bergmann glial cells. Nat. Neurosci. 2, 139e143. Grynkiewicz, G., Poenie, M., Tsien, R.Y., 1985. A new generation of Ca2þ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440e3450. Heinrich, A., Ando, R.D., Turi, G., Rozsa, B., Sperlagh, B., 2012. Kþ depolarization evokes ATP, adenosine and glutamate release from glia in rat hippocampus: a microelectrode biosensor study. Br. J. Pharmacol. 167, 1003e1020. , T., Davies, S.N., Drejer, J., Fletcher, E.J., Jacobsen, P., Lodge, D., Nielsen, F.E., Honore 1988. Quinoxalinediones: potent competitive non-NMDA glutamate receptor antagonists. Science 241, 701e703. Hoogland, T.M., Kuhn, B., 2010. Recent developments in the understanding of astrocyte function in the cerebellum in vivo. Cerebellum 9, 264e271. Iino, M., Goto, K., Kakegawa, W., Okado, H., Sudo, M., Ishiuchi, S., Miwa, A., Takayasu, Y., Saito, I., Tsuzuki, K., Ozawa, S., 2001. Glia-synapse interaction through Ca2þ-permeable AMPA receptors in Bergmann glia. Science 292, 926e929. Karaki, H., Ozaki, H., Hori, M., Mitsui-Saito, M., Amano, K., Harada, K., Miyamoto, S., Nakazawa, H., Won, K.J., Sato, K., 1997. Calcium movements, distribution, and functions in smooth muscle. Pharmacol. Rev. 49, 157e230. Koike, M., Iino, M., Ozawa, S., 1997. Blocking effect of 1-naphthyl acetyl spermine on Ca2þ-permeable AMPA receptors in cultured rat hippocampal neurons. Neurosci. Res. 29, 27e36. Koirala, S., Corfas, G., 2010. Identification of novel glial genes by single-cell transcriptional profiling of bergmann glial cells from mouse cerebellum. PLoS ONE 5, e9198. http://dx.doi.org/10.1371/journal.pone.0009198. Liu, T., Sun, L., Xiong, Y., Shang, S., Guo, N., Teng, S., Wang, Y., Liu, B., Wang, C., Wang, L., Zheng, L., Zhang, C.X., Han, W., Zhou, Z., 2011. Calcium triggers exocytosis from two types of organelles in a single astrocyte. J. Neurosci. 31, 10593e10601. Marcoli, M., Cervetto, C., Paluzzi, P., Guarnieri, S., Alloisio, S., Thellung, S., Nobile, M., Maura, G., 2008. P2X7 presynaptic receptors in adult rat cerebrocortical nerve terminals: a role in ATP-induced glutamate release. J. Neurochem. 105,

C. Cervetto et al. / Neuropharmacology 99 (2015) 396e407 2330e2342. Matsui, K., Jahr, C.E., 2003. Ectopic release of synaptic vesicles. Neuron 40, 1173e1183. Matsui, K., Jahr, C.E., 2004. Differential control of synaptic and ectopic vesicular release of glutamate. J. Neurosci. 24, 8932e8939. Matsui, K., Jahr, C.E., Rubio, M.E., 2005. High-concentration rapid transients of glutamate mediate neural-glial communication via ectopic release. J. Neurosci. 25, 7538e7547. Montana, V., Ni, Y., Sunjara, V., Hua, X., Parpura, V., 2004. Vesicular glutamate transporter-dependent glutamate release from astrocytes. J. Neurosci. 24, 2633e2642. Montana, V., Malarkey, E.B., Verderio, C., Matteoli, M., Parpura, V., 2006. Vesicular transmitter release from astrocytes. Glia 54, 700e715. Muller, T., Moller, T., Berger, T., Schniter, J., Kettenmann, H., 1992. Calcium entry through kainate receptors and resulting potassium channel blockade in Bergmann glial cells. Science 256, 1563e1566. Nakamura, Y., Iga, K., Shibata, T., Shudo, M., Kataoka, K., 1993. Glial plasmalemmal vesicles: a subcellular fraction from rat hippocampal homogenate distinct from synaptosomes. Glia 9, 48e56. Nilsen, A., England, P.M., 2007. A subtype-selective, use-dependent inhibitor of native AMPA receptors. J. Am. Chem. Soc. 129, 4902e4903. Nishiyama, H., Linden, D.J., 2007. Pure spillover transmission between neurons. Nat. Neurosci. 10, 675e677. Ogita, K., Hirata, K., Bole, D.G., Yoshida, S., Tamura, Y., Leckenby, A.M., Ueda, T., 2001. Inhibition of vesicular glutamate storage and exocytotic release by Rose Bengal. J. Neurochem. 77, 34e42. Ormel, L., Stensrud, M.J., Bergersen, L.H., Gundersen, V., 2012a. VGLUT1 is localized in astrocytic processes in several brain regions. Glia 60, 229e238. Ormel, L., Stensrud, M.J., Chaudhry, F.A., Gundersen, V., 2012b. A distinct set of synaptic-like microvesicles in astroglial cells contain VGLUT3. Glia 60, 1289e1300. Paluzzi, S., Alloisio, S., Zappettini, S., Milanese, M., Raiteri, L., Nobile, M., Bonanno, G., 2007. Adult astroglia is competent for Naþ/Ca2þ exchanger-operated exocytotic glutamate release triggered by mild depolarization. J. Neurochem. 103, 1196e1207. Parpura, V., Verkhratsky, A., 2011. The astrocyte excitability brief: from receptors to gliotransmission. Neurochem. Int. 61, 610e621. Pasti, L., Zonta, M., Pozzan, T., Vicini, S., Carmignoto, G., 2001. Cytosolic calcium oscillations in astrocytes may regulate exocytotic release of glutamate. J. Neurosci. 21, 477e484. Piet, R., Jahr, C.E., 2007. Glutamatergic and purinergic receptor-mediated calcium transients in Bergmann glial cells. J. Neurosci. 27, 4027e4035. Raiteri, M., Angelini, F., Levi, G., 1974. A simple apparatus for studying the release of neurotransmitters from synaptosomes. Eur. J. Pharmacol. 2, 411e414. Roots, B.I., 1981. Comparative studies on glial markers. J. Exp. Biol. 95, 167e180. Roseth, S., Fykse, E.M., Fonnum, F., 1998. Uptake of L-glutamate into synaptic vesicles: competitive inhibition by dyes with biphenyl and amino- and sulphonic

407

acid-substituted naphthyl groups. Biochem. Pharmacol. 56, 1243e1249. Saab, A.S., Neumeyer, A., Jahn, H.M., Cupido, A., Simek, A.A., Boele, H.J., Scheller, A., Le Meur, K., Gotz, M., Monyer, H., Sprengel, R., Rubio, M.E., Deitmer, J.W., De Zeeuw, C.I., Kirchhoff, F., 2012. Bergmann glial AMPA receptors are required for fine motor coordination. Science 337, 749e753. Santello, M., Volterra, A., 2009. Synaptic modulation by astrocytes Ca2þedependent glutamate release. Neuroscience 158, 253e259. Sasaki, T., Beppu, K., Tanaka, K.F., Fukazawa, Y., Shigemoto, R., Matsui, K., 2012. Application of an optogenetic byway for perturbing neuronal activity via glial photostimulation. Proc. Natl. Acad. Sci. U. S. A. 109, 20720e20725. Shao, Y., McCarthy, K.D., 1997. Responses of Bergmann glia and granule neurons in situ to N-methyl-D-aspartate, norepinephrine, and high potassium. J. Neurochem. 68, 2405e2411. Sottile, V., Li, M., Scotting, P.J., 2006. Stem cell marker expression in the Bergmann glia population of the adult mouse brain. Brain Res. 1099, 8e17. Stigliani, S., Zappettini, S., Raiteri, L., Passalacqua, M., Melloni, E., Venturi, C., Tacchetti, C., Diaspro, A., Usai, C., Bonanno, G., 2006. Glia re-sealed particles freshly prepared from adult rat brain are competent for exocytotic release of glutamate. J. Neurochem. 96, 656e668. Szapiro, G., Barbour, B., 2007. Multiple climbing fibers signal to molecular layer interneurons exclusively via glutamate spillover. Nat. Neurosci. 10, 735e742. Tsubokawa, H., Oguro, K., Masuzawa, T., Nakaima, T., Kawai, N., 1995. Effects of spider toxin and its analogue on glutamate-activated currents in the hippocampal CA1 neuron after ischemia. J. Neurophysiol. 74, 218e225. Wilding, T.J., Huettner, J.E., 1995. Differential antagonism of alpha-amino-3hydroxy-5-methyl-4-isoxazolepropionic acid-preferring and kainatepreferring receptors by 2,3-benzodiazepines. Mol. Pharmacol. 47, 582e587. Wong, L.A., Mayer, M.L., 1993. Differential modulation by cyclothiazide and concanavalin A of desensitization at native alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid- and kainate-preferring glutamate receptors. Mol. Pharmacol. 44, 504e510. Xu-Friedman, M.A., Harris, K.M., Regehr, W.G., 2001. Three-dimensional comparison of ultrastructural characteristics at depressing and facilitating synapses onto cerebellar Purkinje cells. J. Neurosci. 21, 6666e6672. Yaguchi, T., Nishizaki, T., 2010. Extracellular high Kþ stimulates vesicular glutamate release from astrocytes by activating voltage-dependent calcium channels. J. Cell. Physiol. 225, 512e518. Yamada, K., Fukaya, M., Shibata, T., Kurihara, H., Tanaka, K., Inoue, Y., Watanabe, M., 2000. Dynamic transformation of Bergmann glial fibers proceeds in correlation with dendritic outgrowth and synapse formation of cerebellar Purkinje cells. J. Comp. Neurol. 418, 106e120. Zhang, Q., Pangrsic, T., Kreft, M., Krzan, M., Li, N., Sul, J.Y., Halassa, M., Van Bockstaele, E., Zorec, R., Haydon, P.G., 2004. Fusion-related release of glutamate from astrocytes. J. Biol. Chem. 279, 12724e12733. Zhou, Q., Petersen, C.C.H., Nicoll, R.A., 2000. Effects of reduced vesicular filling on synaptic transmission in rat hippocampal neurones. J. Physiol. 525, 195e206.