Specific neuromodulatory actions of amyloid-β on dopamine release in rat nucleus accumbens and caudate putamen

Journal of Alzheimer’s Disease 19 (2010) 1041–1053 1041DOI 10.3233/JAD-2010-1299IOS Press

Specific Neuromodulatory Actions ofAmyloid-β on Dopamine Release in RatNucleus Accumbens and Caudate Putamen

Elisa Muraa,1, Stefania Predaa,1, Stefano Govonia,∗, Cristina Lannia, Luigia Trabaceb, Massimo Grillic,Federica Lagomarsinoc, Anna Pittalugac,d and Mario Marchic,daDepartment of Experimental and Applied Pharmacology, Centre of Excellence in Applied Biology, University ofPavia, Pavia, ItalybDepartment of Biomedical Sciences, Faculty of Medicine c/o OO.RR., University of Foggia, Foggia, ItalycSection of Pharmacology and Toxicology, Department of Experimental Medicine, University of Genoa, Genoa,ItalydCenter of Excellence for Biomedical Research, University of Genoa, Italy

Accepted 5 October 2009

Abstract. We previously demonstrated that amyloid-β (Aβ) has a neuromodulatory action in the nucleus accumbens (NAc).In this area of the brain, the peptide disrupts the cholinergic control of dopamine (DA) release both in vivo and in vitro. Theaim of the present work was to extend the research on the neuromodulatory effect of Aβ1−40 on DA transmission to differentrelease stimuli and to another dopaminergic brain area, the caudate putamen (CPu), in order to clarify whether the effect ofthe peptide is stimulus- or brain area-selective. We performed both in vivo (microdialysis associated to HPLC) and in vitrostudies (synaptosomes in superfusion). Both in NAc and in CPu and both in vivo and in vitro, Aβ did not affect either basal orpotassium-stimulated DA release. In CPu, the Aβ ability to impair the DA release evoked by the cholinergic agonist carbachol,observed in NAc, was confirmed only in vitro. Moreover, in vitro Aβ affected a specific component of the DA overflow evokedby the non-selective metabotropic glutamate receptors agonist t-ACPD. Altogether, these results show that Aβ may have differentneuromodulatory actions depending upon the secretory stimulus and, in vivo, the brain area investigated.

Keywords: Amyloid-β, dopamine, microdialysis, nucleus accumbens, striatum, synaptosomes

INTRODUCTION

According to the “amyloid cascade hypothesis”, ex-cessive production and deposition of amyloid-β (Aβ)in extracellular sites are responsible for a concatenateseries of events resulting in neurotoxicity and conse-quent neuronal death [1–3].

However, literature data demonstrate that amyloid-βprotein precursor metabolism is a normal and ubiqui-

1Both authors contributed equally.∗Correspondence to: Prof. Stefano Govoni, Viale Taramelli 14,

27100 Pavia, Italy. Tel.: +39 0382 987394; Fax: +39 0382 987405;E-mail: [email protected].

tous process [4] and that Aβ is a normal product of suchmetabolism [5,6], and is present in plasma and cere-brospinal fluid (CSF) of healthy subjects [7]. Altogeth-er these observations suggest that the peptide may haveso far unforeseen physiological roles, some of whichdirectly link to synaptic activities. Just to quote someexamples, the presence of the peptide is important forneuronal cell viability [8], low levels of intracellularAβ could have a physiological role in the modulationof the cyclic AMP response element (CRE)-dependentgene expression and, as a consequence, a positive in-fluence on synaptic plasticity [9]. This view is furthersupported by the results published by Puzzo and col-leagues [10], who showed that picomolar Aβ positive-

ISSN 1387-2877/10/$27.50 2010 – IOS Press and the authors. All rights reserved

1042 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

ly modulates synaptic plasticity in hippocampus. Alsovery intriguing are the observations that Aβ may exertsynaptic activities in absence of neurotoxicity [11,12].A putative Aβ-mediated synaptic modulation is partic-ularly interesting as it might suggest the existence ofAβ-driven functional alterations of neurotransmission,in turn producing subtle cognitive and behavioral dis-turbances in addition and before the well known neu-rodegenerative events.

A disease that in part fits the pattern described aboveis Alzheimer’s disease (AD), whose clinical presenta-tion commonly includes cognitive deficits, as well asemotional and behavioral disturbances, and, less fre-quently, depressive episodes, psychosis with halluci-nations [13], and sleep disorders [14]. Thus, complexsymptoms of AD can only be explained by the singleneurotransmitter deficit model, although undoubtedlycholinergic transmission impairment plays an impor-tant role in the etiology of this disease. More specif-ically, some of the cognitive and executive functionsprofoundly affected by AD are predominantly regulat-ed by dopamine (DA) in limbic areas that play a criticalrole in elaborating thoughts, behavior, and emotionsusing representational knowledge. These operationsare often referred to as working memory.

In accord with all these observations, we previouslyinvestigated the effect of an acute Aβ administrationon DA release in the limbic area of the nucleus accum-bens (NAc) [15], and we demonstrated that in this brainarea, Aβ disrupts the cholinergic control of DA re-lease. Within this context, Trabace and colleagues [16]showed that Aβ also alters DA transmission in pre-frontal cortex.

The purpose of the present work was to extend ourresearch on the neuromodulatory action of Aβ on DAtransmission, in order to clarify whether the effect ofAβ differs according to the brain area or the secretorystimulus investigated. We compared the effect of Aβon DA release in NAc and in the dorsal striatum (cau-date putamen, CPu). The choice of CPu was justifiedby the consideration that, contrary to NAc, in CPu thedopaminergic system regulates motor functions whoseimpairment is considered either a late and minor com-ponent of the AD symptomatology or a consequence ofcomorbidity.

The second important choice was to decide whichpeptide to use for the experiments. The two mainisoforms of Aβ found in AD brains are Aβ1−40 andAβ1−42 that have different profiles of aggregation [17].In fact, Aβ1−42 has been reported to aggregate fasterthan Aβ1−40 and thus it is considered the most neuro-

toxic species [18]. On the other hand, physiologicallythe 40-amino-acid long peptide is the most abundantform [12,19,20], since the concentration of secretedAβ1−42 is about 10% that of Aβ1−40 [21]. Hence,Aβ1−40 and Aβ1−42 may have different biological ac-tions [22], and the ratio of their production may bealtered in pathological conditions, including familialAD [23]. Accordingly, we decided to first examine theeffect of the physiologically more abundant Aβ1−40.Moreover, in our previous in vivo experience in NAc,at variance with Aβ1−40, Aβ1−42 was ineffective sinceit was retained inside the dialysis probe and did notreach the brain tissue, as shown by immunohistochem-ical analysis [15]. For all these reasons, in the presentresearch we limited our study to the neuromodulatoryeffects of Aβ1−40 with only few comparative in vitroexperiments using Aβ1−42.

Finally, our dual approach consisted of in vivo studiesusing the microdialysis technique associated to HPLCand in vitro studies by superfusion of isolated synapto-somes.

MATERIALS AND METHODS

Materials

Atropine (Atr), Aβ1−40, Aβ40−1, Aβ1−42, carbamy-lcholine chloride (carbachol, CCh), 4-[N-(3-Chloroph-enyl)carbamoyloxy]-2-butynyltrimethylammoniumchloride (McN A-343), and mecamylamine (Mec) wereobtained from Sigma Aldrich (Milan, Italy); trans-1-amino-cyclopentane-1,3-dicarboxylic acid (t-ACPD),(2S)-2-Amino-2-[(1S,2S)-2-carboxycycloprop-1-yl]-3-(xan th-9-yl) propanoic acid (LY341495), and (1R,4R,5S,6R)-4-Amino-2-oxabicyclo[3.1.0]hexane-4,6-dicarboxylic acid (LY379268) were provided by TocrisCookson (Bristol, UK); E)-Ethyl1,1a,7,7a-tetrahydro-7-(hydroxyimino)cyclopropa[b]chromene-1a-carboxy-late (CPCOOEt) and 2-Methyl-6-(phenylethynyl)pyri-dine hydrochloride (MPEP) were obtained from AscentScientific (Weston Super-Mare, UK).

In vivo experiments

Animals. Young, male Wistar rats (275–300 g;Harlan, Udine Italy), housed in standard conditions(temperature 23 ± 1◦C; humidity 50%) with 12:12light/dark cycles, water and food ad libitum, were usedeither for microdialysis experiments or as brain tissuesource for the in vitro experiments. All experimental

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1043

procedures were performed according to internation-al regulations for the care and treatment of laboratoryanimals and in accordance with the Italian Act (DL n116, GU, suppl 40, 18 February, 1992) and with EECCouncil Directive (86/609, OJ L 358, 1, 12 December,1987).

Microdialysis probe implantation. Rats were anes-thetized with Equithesin 3 ml/kg (pentobarbital 9.7 g,chloral hydrate 42.5 g, MgSO4 21.3 g for 1L, 10%ethanol, 40% propylene glycole v/v) administered in-traperitoneally and placed in a stereotaxic apparatus(David Kopf Instruments, Tujunga, CA, USA). Theskin was shaved, disinfected, and cut with a sterilescalpel to expose the skull. A hole was drilled to allowthe implantation of the probe into the brain parenchy-ma. The probe was implanted in the shell of NAc (AP +1.6 mm, ML ± 0.8 mm from bregma and DV −8.0 mmfrom dura) or in CPu (AP + 0.7 mm, ML ± 3.0 mmfrom bregma and DV −7.0 mm from dura), accordingto the Paxinos and Watson atlas [24], and secured tothe skull with one stainless steel screw and dental ce-ment. All in vivo experiments were performed usingmicrodialysis probes, made in our laboratory accord-ing to the original method described by Di Chiara etal. [25] (Emophan Bellco Artificial OR-internal diam-eter 200 µm, cutoff 40 KDa; Bellco, Mirandola, Italy),with a nominal active length depending upon the se-lected brain area (2 mm for NAc and 3 mm for CPu).Finally, the skin was sutured, and the rats were allowedto recover from anesthesia for at least 24 h before theneurotransmitter release study.

Microdialysis samples collection. Microdialysis ex-periments were performed on conscious freely movingrats. On the day of the experiments (24 h after the sur-gical procedure), the probe was perfused with artificialCSF containing 145 mM NaCl, 3.0 mM KCl, 1.26 mMCaCl2, 1.0 mM MgCl2, 1.4 mM Na2HPO4, bufferedat pH 7.2–7.4 and filtered through a Millipore 0.2 µmpore membrane. In all experiments, the microdialysismembrane was allowed to stabilize for 1 h at the flowrate of 2 µl/min, without collecting samples. At the endof the stabilization period, three samples were collectedto evaluate baseline release of DA and then the specifictreatment started. All treatments were administered bymanually switching syringes and tubing connections toallow drugs diluted in artificial CSF to flow through theprobes. Tubing switches were performed taking careto maintain constant flow rates and collection volumes.Both basal and treatment samples were collected ev-ery 20 min in 100 µl Eppendorf tubes at a flow rateof 2 µl/min, using a 1000 µl syringe (Hamilton) and a

microinjection pump (CMA/100, CMA/MicrodialysisAB). In vitro recovery of the probe for DA was about30%.

Each rat was used for only one microdialysis ses-sion. At the end of each experiment the position of themicrodialysis probe was verified by histological pro-cedures. Only data from rats in which probe trackswere exactly located in the target area were used forstatistical analysis.

Treatments. The drugs used (CCh, McN A-43, Mec,Atr, t-ACPD) and Aβ1−40 were dissolved to the appro-priate concentration in artificial CSF and delivered tobrain tissue by reverse dialysis. Potassium (K+) wasadministered at two different concentrations, 60 mMand 100 mM, depending upon the experimental con-ditions. Although these concentrations may seem toohigh in comparison with treatments of slices or isolatedsynaptosomes, they are currently used in microdialysisexperiments in order to evaluate neurotransmitter re-lease [16,26,27]. Under these conditions, our exper-imental experience suggests that the concentration ofthe ion reaching the brain tissue is about one third ofthat administered.

Chromatography. In vivo dialysate samples were an-alyzed for DA concentration using HPLC with electro-chemical detection (ESA Coulochem II Bedford-MA)using a Supelcosil LC-18-DB Column 15 cm× 4.6 mmpacked with a 5-µm C18 stationary phase (Supelco)and a sample injection Rheodyne 7125 with 20 µl loop.The pH 5.5 mobile phase consisted of 0.71 g Na2HPO4,6 g NaH2PO4, 15% methanol (v/v), 33.6 mg EDTA,and 116 mg octyl sodium sulfate dissolved in a finalvolume of 1L solution. The phase was degassed un-der vacuum and simultaneously filtered through a Mil-lipore 0.2 µm pore membrane and was delivered witha Waters (Millipore, Mildford, MA, USA) Model 510pump at flow rate of 1.3 ml/min. DA was detected at2 nA with a Model 5011 dual-electrode analytical cell.The detector was operated at a nominal potential of+130 mV applied to the first electrode and a potentialof −125 mV applied to the second electrode. Only thesignal generated at the second electrode was routinelyrecorded. The analytical cell signal was recorded ona Houston Instruments (Gistel, Belgium) Omniscriberecorder. Dialysate samples (20 µl) were manually in-jected into the Rheodyne injector using a 50 µl Hamil-ton syringe. The sensitivity of the assay for DA was2 pg per sample.

Statistics. Values were expressed as percentage ofbasal values. In this case, the average concentration ofthree consecutive samples immediately preceding the

1044 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

drug dose (less than 10% variation) was considered asthe basal value and was defined as 100%. For statis-tical analysis we used the GraphPad Instat statisticalpackage (version 3.05 GraphPad software, San Diego,CA, USA). The data were analyzed by analysis of vari-ance (ANOVA) followed, when significant, by an ap-propriate post hoc comparison test. When the Bartlett’stest indicated nonhomogeneity of variance, a nonpara-metric test, the Kruskal-Wallis ANOVA followed byDunn’s test, was applied. Data were considered signif-icant for p < 0.05. The reported data are expressed asmeans ± S.E.M. The number of animals used for eachexperiment is reported in the legend to Figs 2–4.

Immunohistochemical analysis

Immunohistochemical analysis has been performedto verify the presence of Aβ in the perfused tissueand to confirm (according to HOECHST 33342 stain-ing) the absence of neurotoxic-induced apoptotic phe-nomenon. Brain tissue samples were frozen and storedat−80◦C. For immunodetectionof infused Aβ peptide,20 µm coronal sections (obtained on a cryostat LeicaCM 1510) were incubated with a primary monoclonalantibody recognizing Aβ protein (clone 4G8; Chemi-con International). Sections were then incubated with amouse anti-IgG antibody RPE conjugated (Dako). Af-ter the fluorescent labeling procedures, sections werefinally counterstained for DNA with HOECHST 33342and mounted in a drop of Mowiol (Calbiochem, InalcoSpA, Milan, Italy).

In vitro experiments

Superfusion of isolated nerve endings. Crude synap-tosomes from NAc or CPu were prepared as previous-ly described and labeled with [3H]dopamine ([3H]DA;final concentration 0.05 µM in the presence of 0.1 µMof the serotonin uptake blocker 6-nitroquipazine and0.1 µM of the noradrenaline blocker nisoxetine respec-tively to avoid false labeling of serotonergic or nora-drenergic terminals [28]. Incubation was performed at37◦C, for 15 min, in a rotary water bath. After thelabeling period, identical portions of the synaptoso-mal suspensions were layered on microporous filters atthe bottom of parallel chambers in a Superfusion Sys-tem [29] (Ugo Basile, Comerio, Varese, Italy) main-tained at 37◦C, and synaptosomes were then super-fused at 0.5 ml/min with standard physiological solu-tion as above. The system was first equilibrated during37 min of superfusion; subsequently, six consecutive

1-min fractions were subsequently collected. Synapto-somes were exposed to agonists at the end of the firstfraction collected (t = 38 min) until the end of the su-perfusion, while antagonists and Aβ were added 8 minbefore agonists and maintained throughout the super-fusion. When studying the effect of Aβ on the releaseof neurotransmitter evoked by high K+, synaptosomeswere transiently (90 s) exposed, at the end of the firstfraction collected, to 15 mM KCl-containing medium(NaCl substituting for an equimolar concentration ofKCl).

Fractions collected and superfused synaptosomeswere counted for radioactivity. The amount of radioac-tivity released into each superfusate fraction was ex-pressed as a percentage of the total synaptosomal tri-tium content at the start of the fraction collected (frac-tional efflux). When expressed as overflow (%), drugeffects were estimated by subtracting the neurotrans-mitter content into the fractions corresponding to thebasal release from those corresponding to the evokedrelease. Values in the text are reported as mean % ±S.E.M.

Statistics. Analysis of variance was performed byANOVA followed by Bonferroni multiple-comparisonstest, as appropriate (see legend to the figures). Da-ta were considered significant for p < 0.05 (softwareGraph Pad Prism version 4.03). Appropriate controlswere always run in parallel.

RESULTS

Immunohistochemistry

To test the effect of Aβ1−40 infusion on DA re-lease, we first studied whether Aβ (1 µM) adminis-tration through the dialysis probe allowed the deliveryof the peptide to the tissue in both NAc and CPu. InNAc, immediately after the perfusion of the peptide,Aβ immunoreactivity was detected within the brainarea where the dialysis probe was located. At this timethe peptide showed a range of diffusion comparable tothe end of the tip (200 µm) (Fig. 1a). Also in CPuAβ1−40 infused through the dialysis probe reached thetissue (Fig. 1b). Moreover, in both NAc and CPu, nogross signs of neurodegeneration were observed with-in the area of amyloid diffusion as shown by Hoechst33342 staining indicating the absence of apoptotic phe-nomena.

In vivo results

We first analyzed the effects of K+ depolarization onDA release and the sensitivity to Aβ treatment in NAc.

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1045

Fig. 1. Aβ delivered through the dialysis probe was found within the brain. Coronal sections indicating the brain areas in which the dialysisprobe was implanted and fluorescence micrographs showing the presence of human Aβ. A) Aβ immunoreactivity (red-PE staining, white arrow)in nucleus accumbens (NAc) immediately after perfusion of 1 µM Aβ1−40. B) Presence of Aβ also in caudate putamen (CPu) immediately after1 µM Aβ1−40 infusion. Nuclear DNA was counterstained with Hoechst 33342 (blue staining). Magnification: x 20.

60 mM K+ evoked an increase in DA release that wastoo small to be detected (not shown). Hence, we tried100 mM K+ that elicited about a 17 fold increase ofDA release (Fig. 2a). The effect leveled off 40 min fol-lowing interruption of K+ delivery (not shown). Theaction of K+ was insensitive to 1 µM Aβ1−40 treatment(Fig. 2a). The choice of using a 1 µM Aβ concentrationwas derived from our previous experience [16] show-ing that 1 µM Aβ is the minimal active concentrationthat produced the maximal effect on DA release. Theconcentration of the injected peptide may give origin tobrain levels of Aβ that are well above those observed inthe control human brain [30] suggesting caution in theinterpretation of the results. Preliminary concentration-response curves of Aβ against CCh elicited-DA releasewere superimposable to the previously published onesand the use of higher Aβ concentrations (up to 10 µM)did not affect in any case either basal or K+ stimulated-DA release (not shown). In the same brain area, thecholinergic agonist CCh (100 µM) elicited a sharp in-crease of DA release (up to 11 times the basal level).The effect of CCh was at least partially mediated by

muscarinic cholinergic receptors since it was inhibited(75%) by Atr (1 µM) administration but only marginal-ly (14%) by the administration of Mec (50 µM). 1 µMAβ1−40 did not affect DA basal release. Nevertheless,the peptide greatly inhibited the DA release evoked byCCh (up to 72%, Fig. 2b), as previously shown [15].

In CPu, 100 mM K+ evoked an increase in DA re-lease that was sustained but very variable among differ-ent animals (3353 ± 1941); hence, we used 60mM K+

that reproducibly induced an over 6 fold increase of DArelease. Both the basal and the K+-stimulated releasewere insensitive to 1 µM Aβ1−40 treatment (Fig. 3).Since in NAc Aβ affected DA outflow elicited by CCh,we decided to test the effect of the peptide on thischolinergic agonist also in CPu. Unfortunately, in CPu100 µM CCh was ineffective in enhancing DA releasefrom basal values (Fig. 3). The lack of effect of CChmay be due to the fact that in the striatum this agonistacts on both M1 and M2 muscarinic receptors that haveopposite effects on DA release. In fact, in this brainarea M1 receptors enhance while M2 receptors inhib-it DA release [31,32]. Hence, we tried the selective

1046 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

Fig. 2. In vivo in nucleus accumbens Aβ did not affect K+-stimulated dopamine (DA) release but completely abolished that evoked by carbachol(CCh). A) 1 µM Aβ1−40 effect on 100 mM K+-stimulated DA release. * p < 0.05 vs. Basal; + p > 0.05 vs. K+ 100 mM (Kruskal–WallisANOVA followed by Dunn’s test). B) Inhibition of 100 µM CCh-elicited DA release by atropine (Atr, 1 µM) or mecamylamine (Mec, 50 µM)and Aβ1−40 (1 µM) effects on basal and 100 µM CCh-evoked DA release. ∆ p > 0.05 vs. Basal; *** p < 0.001 vs. Basal; ◦ p > 0.05 vs.CCh; p < 0.05 vs. CCh (Kruskal – Wallis ANOVA followed by Dunn’s test). Data are expressed as mean ± SEM of 5–8 individual rats for eachexperimental group in each panel.

Fig. 3. Lack of effect of in vivo Aβ on both basal and K+-stimulated dopamine (DA) release in caudate putamen. Absence of effect of 100 µMcarbachol (CCh), enhancement of DA release by 60 mM K+ and effect Aβ1−40 (1 µM) on both basal and K+-stimulated DA release. ∆p >0.05 vs. Basal; *** p < 0.001 vs. Basal; +p > 0.05 vs. K+ 60 mM. (One-way ANOVA followed by Bonferroni Multiple Comparisons Test).Data are expressed as mean ± SEM of 5–8 individual rats for each experimental group.

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1047

Fig. 4. In caudate putamen in vivo Aβ did not affect t-ACPD-evoked dopamine (DA) release. Enhancement of DA release by 3 mM t-ACPDand effect of 1 µM Aβ1−40 on t-ACPD evoked DA release. ** p < 0.01 vs. Basal. (One-way ANOVA followed by Tukey-Kramer MultipleComparisons Test). Data are expressed as mean ± SEM of 5–8 individual rats for each experimental group.

M1 agonist McN A-343; even if literature data show amodest (roughly 20%) effect of this compound on DArelease [33], in our experimental conditions the effectof McN A-343 was similar as extent but it was not sig-nificant (130 ± 25). Searching for a non-depolarizingstimulus able to enhance DA release in CPu, we infusedan agonist of group I and II metabotropic glutamatereceptors (mGluRs), i.e., t-ACPD. Published data showthat in striatum 3 mM t-ACPD enhances DA release upto 367% [34]. However, in our study 3 mM t-ACPDenhanced DA release only by about 75% over basalvalues. The increase of DA release elicited by 3 mMt-ACPD was not inhibited by 1 µM Aβ1−40 (Fig. 4).

In vitro results

Synaptosomes prepared from rat NAc and CPu andprelabeled with [3H]-DA were exposed in superfusionto K+ (15 mM) and to Aβ (100 nM). Figure 5 (a andc) shows that the K+-evoked [3H]-DA overflow in ac-cumbal and CPu nerve endings was very similar (2.85± 0.12 and 3.2 ± 0.31, respectively) and were unaf-fected by the presence of 100nM Aβ in the superfusionmedium. The time courses of [3H]-DA release evokedby K+ in the NAc and CPu nerve endings are illustratedin Fig. 5 (b and d). The peak effect was observed atmin 40 of superfusion. Figure 5b and 5d also show that100 nM Aβ did not modify the basal [3H]-DA releasein both areas studied.

CCh was able to significantly enhance the [3H]-DArelease both from accumbal and striatal nerve endings(Fig. 6). The [3H]-DA overflow evoked by 30 µM

CCh in accumbal isolated nerve endings (2.02 ± 0.1)was significantly inhibited by 100nM Aβ (-32%) while10 nM Aβ was ineffective. Atr (100 nM) inhibit-ed the [3H]-DA overflow to a similar extent (−30%).The CCh evoked [3H]-DA overflow was completelyblocked in presence of Atr (100 nM) and Mec (20 µM;Fig. 6a). The time course of [3H]-DA release evokedby CCh in presence or in absence of 100 nM Aβ isreported in Fig. 6b. The peak effect was observedat min 40 of superfusion. The effect of CCh on the[3H]-DA overflow in CPu is illustrated in Fig. 6c. Thestimulatory effect of 30 µM CCh on [3H]-DA over-flow was quantitatively very similar to that reported inNAc (Fig. 6a; 2.25± 0.19). The CCh-evoked [3H]-DAoverflow was partially inhibited by 10 and 100 nM Aβ(−24%; −35%, respectively) while 1 nM Aβ was in-effective. The inhibitory effects of Atr and Atr + Mecwere very similar to those obtained in NAc (Fig. 6a).The time course of [3H]-DA release evoked by CCh inpresence or in absence of Aβ is reported in Fig. 6d.

In addition to Aβ1−40, both in NAc and CPu, wetested in vitro Aβ1−42 and the reverse peptide Aβ40−1.In NAc Aβ1−42 (100 nM) had a similar effect thanAβ1−40 in inhibiting the CCh induced-DA overflow,while in CPu it was ineffective (Fig. 6a, c). Moreover,the reverse peptide Aβ40−1 (100 nM) did not modifythe CCh induced-DA overflow both in the NAc and inthe CPu (Fig. 6a, c), suggesting that the effect of 1-40 and 1-42 peptides is specific not simply due to theaddition of a peptide solution.

Figure 7a shows that t-ACPD (100 µM) was unableto elicit a significant [3H]-DA release from CPu synap-

1048 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

Fig. 5. In vitro both in nucleus accumbens (NAc) and in caudate putamen (CPu) Aβ did not affect neither basal nor K+-stimulated dopamine(DA) release. A) Effect of 100 nM Aβ on 15 mM K+ evoked [3H]-DA overflow from rat accumbal synaptosomes. B) Time course of 15 mMK+ induced [3H]-DA release from rat accumbal synaptosomes in presence and in absence of 100 nM Aβ. C) Effect of 100 nM Aβ on 15 mMK+ evoked [3H]-DA overflow from rat CPu synaptosomes. D) Time course of 15 mM K+ induced [3H]-DA release from rat CPu synaptosomesin presence and in absence of 100 nM Aβ. Data are means ± SEM of four experiments run in triplicate. *** p < 0.001 vs. Aβ alone. One wayANOVA followed by Bonferroni post hoc test. Synaptosomes prepared from one animal were needed to perform each single experiment.

tosomes and Aβ1−40 (100 nM) did not influence itstime-course.

To shed light on the possible effect of t-ACPD on the[3H]-DA release and on the downstream mechanismsin the CPu synaptosomes, we performed experiments tostudy the effect of this drug and of some mGluRs antag-onists (CPCOOEt, MPEP, and LY341495, respectivelynon competitive antagonists of mGlu1 and mGlu5 re-ceptor subtypes and competitive antagonist of mGlu2/3receptors) on the K+-evoked release of [3H]-DA fromCPu nerve endings. Figure 7b shows that K+-evoked

[3H]-DA overflow was significantly inhibited by t-ACPD (100 µM; −21%) and MPEP (1 µM; −19%).Aβ (100 nM) left unaffected the inhibitory effect oft-ACPD. The effect of t-ACPD was further increasedin presence of CPCOOET (100 nM; −43%), a selec-tive mGluR1 blocker. This inhibitory effect was notmodified by 100 nM Aβ1−40. The inhibitory effect oft-ACPD was significantly counteracted by LY341495(30 nM). The [3H]DA overflow elicited by K+ in pres-ence of t-ACPD + LY341495 was significantly reducedby 100 nM Aβ1−40 (−36%). The selective mGluR 2/3

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1049

Fig. 6. Both in nucleus accumbens (NAc) and caudate putamen (CPu) Aβ1−40 inhibited in vitro the carbachol (CCh) evoked [3H]-dopamine(DA) overflow. A) Effect of Aβ1−40 (10–100 nM), Aβ1−42 and Aβ40−1 (100 nM) and cholinergic antagonists (100 nM atropine, Atr; 20 µMmecamylamine, Mec) on 30 µM CCh evoked [3H]-DA overflow from rat accumbal synaptosomes. Data are means ± SEM of four experimentsrun in triplicate. * p < 0.05, *** p < 0.001 vs. CCh alone. One way ANOVA followed by Bonferroni post hoc test. B) Time course of 30 µMCCh induced [3H]-DA release from rat accumbal synaptosomes in presence and in absence of 100 nM Aβ. * p < 0.05 vs. CCh alone. Two wayANOVA followed by Bonferroni post hoc test. C) Effect of Aβ1−40 (1–100 nM), Aβ1−42 and Aβ40−1 (100 nM) and cholinergic antagonists(100 nM Atr and 20 µM Mec) on 30 µM CCh evoked [3H]-DA overflow from rat CPu synaptosomes. Data are means ± SEM of four experimentsrun in triplicate. * p < 0.05, *** p < 0.001 vs. CCh alone. One way ANOVA followed by Bonferroni post hoc test. D) Time course of 30 µMCCh induced [3H]-DA release from rat CPu synaptosomes in presence and in absence of 100 nM Aβ. * p < 0.05 vs. CCh alone. Two wayANOVA followed by Bonferroni post hoc test. Synaptosomes prepared from one animal were needed to perform each single experiment.

agonist LY379268 tested at 100nM concentration, sig-nificantly inhibited the [3H]-DA overflow elicited byK+; this effect was not modified by 100 nM Aβ1−40.

DISCUSSION

Our results show the Aβ ability to exert neuromodu-latory effect in absence of gross neurotoxicity and sup-

port the observation that the neuromodulatory effect ofAβ on DA release is consistent in the various brain ar-eas examined, although important differences emergedwhen considering the secretory stimulus and the in vivoand in vitro approach.

Following Aβ administration, immunohistochemi-cal inspection indicated that under our experimentalconditions, Aβ diffuses through the dialysis membrane

1050 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

Fig. 7. In caudate putamen synaptosomes the [3H]-dopamine (DA) overflow elicited by K+ in presence of t-ACPD + LY341495 was significantlyreduced by Aβ. A) Time course of 100 µM t-ACPD induced [3H]-DA release from rat caudate putamen synaptosomes in presence and in absenceof 100 nM Aβ. B) Effect of 100 nM Aβ1−40 and metabotropic glutamate receptors antagonists and agonist (30 nM LY341495; 1 µM MPEP;100 nM CPCOOEt; 100 nM LY379268) on 15 mM K+ evoked [3H]-DA overflow from rat caudate putamen synaptosomes. Data are means ±SEM of four experiments run in triplicate. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. K+ alone; $ p < 0.05, $$ p < 0.01 vs. K+ plus t-ACPD;## p < 0.01 vs. K+ plus t-ACPD plus LY341495. One way ANOVA followed by Bonferroni post hoc test. Synaptosomes prepared from oneanimal were needed to perform each single experiment.

to the tissue where it is found in proximity of the dial-ysis probe (Fig. 1 a and b). No gross signs of neurode-generation were observed within the area of amyloiddiffusion as shown by Hoechst immunocytochemistryand as expected within the time frame (from 40 min toa few hours) using a low micromolar freshly preparedAβ1−40 solution. Moreover, we used a dialysis mem-brane with a cutoff size of 40 000 DA to allow the pas-sage through the dialysis fiber of soluble Aβ monomersor small molecular weight oligomers and avoid highmolecular weight oligomers (the neurotoxic species asshown by [35]). We cannot exclude, however, the pres-ence, also at this early time, after Aβ treatment, of moresubtle signs of toxicity such as synaptic degenerationand neurite retraction.

The in vivo administration of Aβ produced a dra-matic decrease of the CCh-induced DA release in NAc.In CPu, where no effect of CCh was observed, an in-teraction between Aβ and CCh co-treatments was al-so absent. In both areas Aβ was ineffective on K+

stimulated-DA release. The in vitro data on synap-tosomes confirmed the lack of effect of Aβ on K+

stimulated-DA release in both areas and the inhibito-ry activity of Aβ on CCh stimulated-DA release inNAc. Moreover, at variance with the in vivo obser-vations, in vitro CCh also elicited DA release in CPusynaptosomes and this effect was sensitive to Aβ in-hibition. Altogether, the data suggest that, at least in

these two brain areas, the effect of Aβ on the releaseof DA from dopaminergic terminals when assayed invitro is dependent upon the secretory stimulus select-ed rather than from the brain area used as a source ofdopaminergic nerve endings. When comparing in vitroAβ1−40 and Aβ1−42, both peptides had a similar effectin NAc, but in CPu Aβ1−42 was ineffective. Presently,we are unable to explain the observed in vitro differ-ential sensitivity of NAc and CPu to Aβ1−42. We mayonly speculate on the existence of separate binding sitesfor the two peptides and on their differential distribu-tion/abundance on accumbal and CPu isolated nerveendings. The reverse peptide Aβ40−1 was ineffectivein all the condition tested.

Interestingly, the effect of the cholinergic stimula-tion upon DA release depends upon both muscarinicand nicotinic cholinergic receptors. However, the com-ponent sensitive to Aβ inhibition appears to be the onemediated by muscarinic receptors (see Fig. 6) both inCPu and in NAc (for this area, see also [15]), as sug-gested by the observation that, in in vitro studies, Aβquantitatively mimics atropine in preventing the CCh-induced DA release in both areas.

The reason for the sensitivity to Aβ of the CChstimulated-DA release may reside in the inhibitory ac-tion that Aβ exerts on CCh stimulated-protein kinaseC (PKC) activation downstream the cholinergic mus-carinic receptors as previously shown [15]. A number

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1051

of experimental observations do indicate that Aβ mayimpair PKC activation [11,36,37]. This view seemsfurther supported by in vitro data showing that the re-lease of DA elicited by K+/t-ACPD in the presence ofthe LY341495 compound is sensitive to Aβ inhibition(see below).

The release of neurotransmitter evoked by a depolar-izing stimulus (such as [K+]out or CCh itself) is a com-plex process that depends on Ca2+ ions entering thenerve terminals through voltage sensitive Ca2+ chan-nels (VSCCs). However, it is now clear that a stimulus-induced, Ca2+-dependent release of neurotransmitterinvolves more than one protein kinase family, whoseactivation results in different signaling pathways, mostof which positively convey on neurotransmitter exo-cytosis. For instance, once entered the nerve termi-nals, Ca2+ ions can trigger the functioning of cytoso-lic Ca2+-sensitive adenylyl cyclase (AC) isoforms, lo-cated in the cytosol of nerve terminals [38] or activateCa2+-dependent PKC, also present at the presynapticlevel [39]. AC may in turn catalyze the conversion ofcytosolic ATP into cyclic AMP, which then activatescAMP-dependent protein kinase A (PKA) [40]. BothPKA and PKC are known to positively reverberate onvesicular exocytosis of neurotransmitters, then facili-tating neurotransmitter release [39,41,42] and most ofthe presynaptic G-protein coupled receptors (such asthe group I and II metabotropic glutamate receptors)eventually exert a presynaptic control on neurotrans-mitter release by hampering these intraterminal phos-phorylative pathways [43–46].

The role of the above-mentioned Ca2+-dependentintraterminal pathways in the release of DA is indirect-ly unveiled by the effects of the broad spectrum agonistt-ACPD. On one hand, the inhibitory effect exerted bythe broad spectrum agonist t-ACPD on DA exocytosisas well as its reversal by LY341495 is consistent withthe existence of mGlu2/3 presynaptic heteroreceptorson striatal dopaminergic terminals [47] negatively cou-pled to AC and indirectly claims for a AC/cAMP/PKA-dependent component in the K+-evoked release of DAat this level. On the other hand, the observation thatMPEP mimics t-ACPD while CPCCOEt reinforces itsinhibitory effect suggests the presence of presynapticgroup I mGlu receptors on striatal dopaminergic termi-nals [34] while favoring a PKC-mediated componentin DA exocytosis. We do not know whether mGlu 1/5and mGlu2/3 receptor subtypes colocalize on the samedopaminergic terminals, or if a mixed population ofdopaminergic terminals bearing the different receptorsubtypes exists. What is evident is that both recep-

tor subfamilies participate in DA exocytosis by mod-ulating distinct phosphorylative processes and when t-ACPD-induced PKC-mediated events are favored, be-cause of the blockade of presynaptic group II mGlu re-ceptors, the K+-evoked release of DA became sensitiveto Aβ. Indeed, t-ACPD in the presence of LY341495compound has been shown to be associated with PKCactivation. This view is confirmed by the fact that thepeptide did not modify the inhibition on DA releaseexerted by the group II mGluR agonist LY379268.

However, it should be mentioned that the same inhi-bition of DA release following the administration of thebroad-spectrum agonist t-ACPD was not observed inthe in vivo experiments. The differences observed in thein vivo experiments compared to the in vitro data, con-cerning t-ACPD and CCh responses in CPu, may havevarious explanations. First, all the in vitro data obtainedon perfused synaptosomes are due to the direct effectsof the added drugs, which have to act upon receptorsor modulatory sites located on the same synaptosomefrom which occurs the release of the transmitter [29].In contrast, in vivo data are the resultant of the actionof the administered substance also at more distant sitesand, possibly, to the interplay of different transmitters.Moreover, in vivo a differential-regional specific regu-lation of neurotransmission may take place, as in thecase of the cholinergic regulation of DA release, oc-curring in NAc and absent in CPu [48]. In addition,in vivo metabolic and clearance phenomena of the ad-ministered substance may take place affecting the finalresult.

Thus, the acute neuromodulatory activity of Aβ mayhave important regulatory roles, yet to be defined, ver-sus various neurotransmitters and in various brain ar-eas in different physiopathological conditions. Thisview is at least partially supported by the observationby Brody and colleagues [30] showing rapid changesof Aβ concentrations following acute brain injury inpatients. Under these conditions the observed increaseof Aβ in extracellular fluid is thought to contribute tothe clinical stabilization of the patients’ brain activi-ty. These observations as well as ours do fit in theKamenetz model [12] suggesting the participation ofAβ in a short loop feed-back mechanism damping downon synaptic activity when trespassing a given threshold.

According to the present results and to the previouslycited published data, the effect of Aβ on synaptic trans-mission may be perceived as a continuum ranging frommodulation of synaptic activity to neurodegenerationas the concentration and the aggregation state of thepeptide increase; an event perhaps favored by an altered

1052 E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release

ratio Aβ1−40/Aβ1−42 occurring as a consequence ofthe disease. These effects may differ from area to areaand also as a function of the prevalent peptide synthe-sized (1-40, 1-42, as well as other fragments). Severalaspects of the general reference frame above depictedhave yet to be studied. However, the available datashowing the effect of amyloid in areas scarcely affectedby the disease raise questions on the physiological roleof the peptide in normal brain activity rather than or inaddition to its effects in the diseased brain.

ACKNOWLEDGMENTS

This work was supported by grants from the Ital-ian Ministero Universita Ricerca to Prof. Stefano Gov-oni, Mario Marchi and Anna Pittaluga (prot. N◦

2007HJCCSF, prot. N◦ 20072BTSR2 002 and prot. N◦

200728AA57 002), from Compagnia di San Paolo andfrom University of Genoa ‘Progetto Ricerca Ateneo’ toProf. Mario Marchi. The authors wish to thank Prof.Pietro Frattini for his help and advice about the HPLCtechnique, Ms. Maura Agate for careful editorial as-sistance and Silvia E. Smith (University of Utah) forreviewing the manuscript.

Authors’ disclosures available online (http://www.j-alz.com/disclosures/view.php?id=175).

REFERENCES

[1] Hardy JA, Higgins GA (1992) Alzheimer’s disease: the amy-loid cascade hypothesis. Science 256, 184-185.

[2] Pimplikar SW (2008) Reassessing the amyloid cascade hy-pothesis of Alzheimer’s disease. Int J Biochem Cell Biol 41,1261-1268.

[3] Hardy J (2009) The amyloid hypothesis for Alzheimer’s dis-ease: a critical reappraisal. J Neurochem 110, 1129-1134.

[4] Busciglio J, Gabuzda DH, Matsudaira P, Yankner BA (1993)Generation of beta-amyloid in the secretory pathway in neu-ronal and nonneuronal cells. Proc Natl Acad Sci U S A 90,2092-2096.

[5] Shoji M, Golde TE, Ghiso J, Cheung TT, Estus S, ShafferLM, Cai XD, McKay DM, Tintner R, Frangione B, Younkin S(1992) Production of the Alzheimer amyloid beta protein bynormal proteolytic processing. Science 258, 126-129.

[6] Haass C, Schlossmacher MG, Hung AY, Vigo-Pelfrey C, Mel-lon A, Ostaszewski BL, Lieberburg I, Koo EH, Schenk D,Teplow DB, Selkoe D (1992) Amyloid beta-peptide is pro-duced by cultured cells during normal metabolism. Nature359, 322-325.

[7] Pearson HA, Peers C (2006) Physiological roles for amyloidbeta peptides. J Physiol 575, 5-10.

[8] Plant LD, Boyle JP, Smith IF, Peers C, Pearson HA (2003) Theproduction of amyloid beta peptide is a critical requirementfor the viability of central neurons. J Neurosci 23, 5531-5535.

[9] Cuello AC, Canneva F (2008) Impact of intracellular β-amyloid in transgenic animals and cell models. NeurodegenerDis 5, 146–148.

[10] Puzzo D, Privitera L, Leznik E, Fa M, Staniszewski A, PalmeriA, Arancio O (2008) Picomolar amyloid-beta positively mod-ulates synaptic plasticity and memory in hippocampus. J Neu-rosci 28, 14537-14545.

[11] Zhong P, Gu Z, Wang X, Jiang H, Feng J, Yan Z (2003) Im-paired modulation of GABAergic transmission by muscarinicreceptors in a mouse transgenic model of Alzheimer’s disease.J Biol Chem 278, 26888-26896.

[12] Kamenetz F, Tomita T, Hsieh H, Seabrook G, Borchelt D,Iwatsubo T, Sisodia S, Malinow R (2003) AβPP processingand synaptic function. Neuron 37, 925-937.

[13] Assal F, Cummings JL (2002) Neuropsychiatric symptoms inthe dementias. Curr Opin Neurol 15, 445-450.

[14] Yesavage JA, Friedman L, Ancoli-Israel S, Bliwise D, SingerC, Vitiello MV, Monjan AA, Lebowitz B (2003) Develop-ment of diagnostic criteria for defining sleep disturbance inAlzheimer’s disease. J Geriatr Psychiatry Neurol 16, 131-139.

[15] Preda S, Govoni S, Lanni C, Racchi M, Mura E, Grilli M,Marchi M (2008) Acute beta amyloid administration disruptsthe cholinergic control of dopamine release in the nucleusaccumbens. Neuropsychopharmacology 33, 1062-1070.

[16] Trabace L, Kendrick KM, Castrignano S, Colaianna M, DeGiorgi A, Schiavone S, Lanni C, Cuomo V, Govoni S (2007)Soluble amyloid beta1-42 reduces dopamine levels in rat pre-frontal cortex: relationship to nitric oxide. Neuroscience 147,652-663.

[17] Kirkitadze MD and Kowalska A (2005) Molecular mecha-nisms initiating amyloid β-fibril formation in Alzheimer’s dis-ease. Acta Biochim Pol 52, 417-423.

[18] Verdile G, Fuller S, Atwood CS, Laws SM, Gandy SE, MartinsRN (2004) The role of beta amyloid in Alzheimer’s disease:still a cause of everything or the only one who got caught?Pharmacol Res 50, 397-409.

[19] Terai K, Iwai A, Kawabata S, Tasaki Y, Watanabe T, MiyataK, Yamaguchi T (2001) Amyloid-deposits in transgenic miceexpressing human β-amyloid precursor protein have the samecharacteristics as those in Alzheimer’s disease. Neuroscience104, 299-310.

[20] Walsh DM, Selkoe DJ (2007) Aβ Oligomers-a decade of dis-covery. J Neurochem 101, 1172-1184.

[21] Bitan G, Kirkitadze MD, Lomakin A, Vollers SS, Benedek GB,Teplow DB (2003) Amyloid beta-protein (Abeta) assembly:abeta 40 and Abeta 42 oligomerize through distinct pathways.Proc Natl Acad Sci U S A 100, 330-335.

[22] Zou K, Kim D, Kakio A, Byun K, Gong JS, Kim J, Kim M,Sawamura N, Nishimoto S, Matsuzaki K, Lee B, YanagisawaK, Michikawa M (2003) Amyloid beta-protein (Abeta)1-40protects neurons from damage induced by Abeta1-42 in cultureand in rat brain. J Neurochem 87, 609-619.

[23] Mayeux R, Tang MX, Jacobs DM, Manly J, Bell K, MerchantC, Small SA, Stern Y, Wisniewski HM, Mehta PD (1999)Plasma amyloid beta-peptide 1-42 and incipient Alzheimer’sdisease. Ann Neurol 46, 412-416.

[24] Paxinos G, Watson C (1998) The Rat Brain in StereotaxicCoordinates. Academic Press: New York, London.

[25] Di Chiara G (1990) In-vivo brain dialysis of neurotransmitters.Trends Pharmacol Sci 11, 116-121.

[26] Shi XR, Chang J, Ding JH, Fan Y, Sun XL, Hu G (2008) Kir6.2knockout alters neurotransmitter release in mouse striatum:an in vivo microdialysis study. Neurosci Lett 439, 230-234.

E. Mura et al. / Amyloid-β Impairs Rat Brain Dopamine Release 1053

[27] Yoshimoto K, Nishimura A, Hattori H, Joyce JN, Yoshida T,Hioki C, Kogure A, Ueda S (2006) In vivo brain microdialysisstudies on the striatal dopamine and serotonin release in zittermutant rats. Mech Ageing Dev 127, 628-632.

[28] Marchi M, Paudice P, Bonanno G, Raiteri M (1985) Presynap-tic muscarinic receptor activation enhances striatal dopaminerelease evoked by depolarization but not that induced by non-depolarizing stimuli. Neurochem Int 7, 137–141.

[29] Raiteri L, Raiteri M (2000) Synaptosomes still viable after 25years of superfusion. Neurochem Res 25, 1265-1274.

[30] Brody DL, Magnoni S, Schwetye KE, Spinner ML, EsparzaTJ, Stocchetti N, Zipfel GJ, Holtzman DM (2008) Amyloid-beta dynamics correlate with neurological status in the injuredhuman brain. Science 321, 1221-1224.

[31] Xu M, Mizobe F, Yamamoto T, Kato T (1989) Differentialeffects of M1- and M2-muscarinic drugs on striatal dopaminerelease and metabolism in freely moving rats. Brain Res 495,232-242.

[32] De Klippel N, Sarre S, Ebinger G, Michotte Y (1993) Effectof M1- and M2-muscarinic drugs on striatal dopamine releaseand metabolism: an in vivo microdialysis study comparingnormal and 6-hydroxydopamine-lesioned rats. Brain Res 630,57-64.

[33] Xu M, Yamamoto T, Kato T (1990) In vivo striatal dopaminerelease by M1 muscarinic receptors is induced by activationof protein kinase C. J Neurochem 54, 1917-1919.

[34] Bruton RK, Ge J, Barnes NM (1999) Group I mGlu receptormodulation of dopamine release in the rat striatum in vivo. EurJ Pharmacol 369, 175-181.

[35] Sabella S, Quaglia M, Lanni C, Racchi M, Govoni S, Cac-cialanza G, Calligaro A, Bellotti V, De Lorenzi E (2004) Cap-illary electrophoresis studies on the aggregation process of β-amyloid 1-42 and 1-40 peptides. Electrophoresis 25, 3186–3194.

[36] Lee W, Boo JH, Jung MW, Park SD, Kim YH, Kim SU, Mook-Jung I (2004) Amyloid beta peptide directly inhibits PKCactivation. Mol Cell Neurosci 26, 222-231.

[37] Olariu A, Yamada K, Mamiya T, Hefco V, Nabeshima T (2002)Memory impairment induced by chronic intracerebroventric-ular infusion of beta-amyloid (1-40) involves downregulationof protein kinase C. Brain Res 957, 278-286.

[38] Wang HB, Storm DR (2003) Calmodulin-regulated adenylylcyclases: Cross-talk and plasticity in the central nervous sys-

tem. Mol Pharmacol 63, 463-468.[39] Ryves WJ, Dekker LV, Brammer MJ, Campbell IC (1996)

PKC in rat cortical synaptosomes: effects of depolarization.Neuroreport 8, 323–327.

[40] Greengard P (2001) Neuroscience – The neurobiology of slowsynaptic transmission. Science 294, 1024-1030.

[41] Chavez-Noriega LE, Stevens CF (1994) Increased transmitterrelease at excitatory synapses produced by direct activationof adenylate cyclase in rat hippocampal slices. J Neurosci 14,310-317.

[42] Trudeau LE, Emery DG, Haydon PG (1996) Direct modulationof the secretory machinery underlies PKA-dependent synapticfacilitation in hippocampal neurons. Neuron 17, 789-797.

[43] Grilli M, Raiteri L, Pittaluga A (2004) Somatostatin inhibitsglutamate release from ouse cerebrocortical nerve endingstrough presynaptic sst2 receptor linked to the adenylyl cyclase-protein linase A pathway. Neuropharmacology 46, 388-396.

[44] Longordo F, Feligioni M, Chiaramonte G, Sbaffi PF, RaiteriM, Pittaluga A (2006) The human immunodeficiency virus-1protein transactivator of transcription up-regulates N-methyl-D-aspartate receptor function by acting at metabotropic gluta-mate receptor 1 receptors coexisting on human and rat brainnoradrenergic neurones. J Pharmacol Exp Ther 317, 1097-1105.

[45] Wang SJ, Sihra TS (2004) Noncompetitive metabotropic glu-tamate5 receptor antagonist (E)-2-methyl-6-styryl-pyridine(SIB1893) depresses glutamate release through inhibition ofvoltage-dependent Ca2+ entry in rat cerebrocortical nerve ter-minals (synaptosomes). J Pharmacol Exp Ther 309, 951-958.

[46] Corti C, Battaglia G, Molinaro G, Riozzi B, Pittaluga A, Cor-si M, Mugnaini M, Nicoletti F, Bruno V (2007) The use ofknock-out mice unravels distinct roles for mGlu2 and mGlu3metabotropic glutamate receptors in mechanisms of neurode-generation/neuroprotection. J Neurosci 27, 8297-8308.

[47] Karasawa J, Yoshimizu T, Chaki S (2006) A metabotropic glu-tamate 2/3 receptor antagonist, MGS0039, increases extracel-lular dopamine levels in the nucleo accumbens shell. NeurosciLett 393, 127-130.

[48] O’Connor WT, Drew KL, Ungerstedt U (1995) Differentialcholinergic regulation of dopamine release in the dorsal andventral neostriatum of the rat: an in vivo microdialysis study.J Neurosci 15, 8353-8361.