The anti-epileptic actions of neuropeptide Y in the ...cnc.cj.uc.pt/BEB/private/pdfs/NeuroMod/ElBahh...

The anti-epileptic actions of neuropeptide Y in thehippocampus are mediated by Y2 and not Y5 receptors

Bouchaıb El Bahh,1* Silvia Balosso,2 Trevor Hamilton,1 Herbert Herzog,3 Annette G. Beck-Sickinger,4 Gunther Sperk,5

Donald R. Gehlert,6 Annamaria Vezzani2 and William F. Colmers11Department of Pharmacology, University of Alberta. Edmonton, Alberta, Canada T6G 2H72Department of Neuroscience, ‘Mario Negri’ Institute of Pharmacological Research, 20157 Milano, Italy3Neurobiology Program, Garvan Institute of Medical Research, NSW 2010 Sydney, Australia4Institute of Biochemistry, University of Leipzig, D04103 Leipzig, Germany5Institute for Pharmacology, Medical University Innsbruck, Peter-Mayr-Str. 1a, 6020 Innsbruck Austria6Neuroscience Division DC0510, Lilly Research Laboratories, Eli Lilly and Company, Lilly Corporate Center, Indianapolis,IN 46285-0510, USA

Keywords: BIIE0246, epilepsy, neuropeptide Y, presynaptic inhibition, Y2 knockout mouse, Y5 receptors

Abstract

Neuropeptide Y (NPY) potently inhibits glutamate release and seizure activity in rodent hippocampus in vitro and in vivo, but thenature of the receptor(s) mediating this action is controversial. In hippocampal slices from rats and several wild-type mice, a Y2-preferring agonist mimicked, and the Y2-specific antagonist BIIE0246 blocked, the NPY-mediated inhibition both of glutamatergictransmission and of epileptiform discharges in two different slice models of temporal lobe epilepsy, stimulus train-induced bursting(STIB) and 0-Mg2+ bursting. Whereas Y5 receptor-preferring agonists had small but significant effects in vitro, they were blocked byBIIE0246, and a Y5 receptor-specific antagonist did not affect responses to any agonist tested in any preparation. In slices from Y�=�

2mice, NPY was without effect on evoked potentials or in either of the two slice seizure models. In vivo, intrahippocampal injections ofY2- or Y5-preferring agonists inhibited seizures caused by intrahippocampal kainate, but again the Y5 agonist effects were insensitiveto a Y5 antagonist. Neither Y2- nor Y5-preferring agonists affected kainate seizures in Y�=�

2 mice. A Y5-specific antagonist did notdisplace the binding of two different NPY ligands in WT or Y�=�

1 mice, whereas all NPY binding was eliminated in the Y�=�2 mouse.

Thus, we show that Y2 receptors alone mediate all the anti-excitatory actions of NPY seen in the hippocampus, whereas our findingsdo not support a role for Y5 receptors either in vitro or in vivo. The results suggest that agonists targeting the Y2 receptor may beuseful anticonvulsants.

Introduction

The neuropeptide Y (NPY) system in the hippocampus and cortex hasbeen widely suggested to be involved in epileptogenesis and epilepsy.Although NPY is distributed broadly throughout the nervous system,seizure-related changes in NPY and NPY-receptor expression are seenmostly in brain areas, such as the hippocampus, involved in theinitiation and propagation of epileptic discharges (Sperk et al., 1992;Schwarzer et al., 1995). Furthermore, application of NPY potently andselectively inhibits excitatory synaptic transmission in the hippocam-pus, making the receptors mediating this of interest as anticonvulsantdrug targets.

NPY acts via at least five known receptor subtypes, all of whichbelong to the G protein-coupled receptor superfamily (Michel et al.,1998). Extensive studies in the hippocampus in vitro have demon-strated that NPY selectively inhibits glutamatergic synaptic transmis-sion in areas CA1 and CA3 (Colmers et al., 1987, 1991; Klapstein &Colmers, 1992; Greber et al., 1994; McQuiston & Colmers, 1996),

and suppresses glutamate release (Greber et al., 1994; Silva et al.,2003). This action results from the suppression of voltage-dependentCa2+ influx in presynaptic nerve terminals (Toth et al., 1993; Qianet al., 1997).The NPY system has been proposed to act as an endogenous

anticonvulsant (Vezzani & Sperk, 2004; Tu et al., 2005). Thus, in vivoseizures up-regulate NPY and Y2 receptor levels (Sperk et al., 1992;Gobbi et al., 1998; Schwarzer et al., 1998; Vezzani et al., 1999) butdown-regulate Y1 receptors (Kofler et al., 1997), and elevating NPYin the hippocampus elevates seizure thresholds, delays kindlingepileptogenesis, and results in fewer and briefer kindled seizures(Vezzani et al., 2002; Reibel et al., 2003; Richichi et al., 2004). NPY-deficient mice occasionally develop mild, spontaneous seizures andexhibit markedly enhanced susceptibility to motor seizures induced byconvulsant agents (Erickson et al., 1996; Baraban et al., 1997).Elevations in Y2 receptor levels have been observed in patients withtemporal lobe epilepsy (Furtinger et al., 2001), and NPY applicationsuppresses epileptiform activity in human epileptic dentate gyrus(Colmers et al., 1997; Patrylo et al., 1999).The nature of the NPY receptor mediating its inhibitory and

anticonvulsant actions has been controversial. In vitro studies inrodents and humans indicate a major role for the Y2 receptor(Colmers et al., 1991, 1997; El Bahh et al., 2002; Vezzani & Sperk,

Correspondence: Dr W. F. Colmers, as above.E-mail: [email protected]

*Present address: Department of Physiology, University of Montreal, CP 6128 Succ.Centre-Ville, Montreal QC H3C 3J7, Canada.

Received 14 March 2005, revised 22 July 2005, accepted 28 July 2005

European Journal of Neuroscience, Vol. 22, pp. 1417–1430, 2005 ª Federation of European Neuroscience Societies

doi:10.1111/j.1460-9568.2005.04338.x

2004), but some in vivo experiments, using agonist pharmacology(Woldbye et al., 1997) and knockout strategies (Marsh et al., 1999;Baraban, 2002), have implicated the Y5 receptor. Here, we combinein vitro and in vivo approaches in rats, and a spectrum of wild-type(WT) and Y-receptor knockout mice to study the site(s) and nature(s)of Y receptors that regulate glutamate release and, by extension,limbic seizures. Evidence here from all approaches taken, includingpharmacological agonists and antagonists, in vitro and in vivo, inrats, WT and receptor knockout mice, consistently supports a keyrole for the Y2 receptor, but provides no evidence for anyhippocampal NPY responses that were verifiably mediated by theY5 receptor.

Methods

Experimental procedures

All animal procedures were in accordance with Canadian Council ofAnimal Care guidelines (protocol approved by the University ofAlberta Health Sciences Laboratory Animal Committee) and relevantinternational laws and policies (EEC Council Directive 86 ⁄ 609, OJ L358, 1, 12 December 1987; Guide for the Care & Use of LaboratoryAnimals, US National Research Council, 1996).

Animals

The generation of Y�=�2 knockout mice on a mixed

C57BL ⁄ 6 · 129 ⁄ SvJ background was described previously (Sains-bury et al., 2002). Y�=�

5 mice on an inbred 129 ⁄ Sv background weregenerated using homologous recombination techniques as describedby Marsh et al. (1999) and maintained in a breeding colony at UCSF,and were a generous gift of Dr R. Palmiter. WT animals used in thein vitro studies were male Sprague–Dawley rats and C57BL ⁄ 6 (bothfrom the University of Alberta colonies). WT 129 ⁄ SvJ mice, of thesame genetic background as the Y�=�

5 animals, were purchased fromJackson Laboratories (Bar Harbor, ME, USA) andC57BL ⁄ 6 · 129 ⁄ SvJ came from the Garvan Institute (Sydney,Australia) and were identical to WT animals described previously(Sainsbury et al., 2002). The C57Bl ⁄ 6 mice used in the in vivoexperiments were from Charles River (Calco, Italy).

In vitro experiments

Slice preparation

Transverse hippocampal slices (400 lm) were prepared from young(3–5 weeks) animals, after decapitation, as described previously(Klapstein & Colmers, 1997; Ho et al., 2000; El Bahh et al., 2002)using a vibroslicer HR-2 (Sigmann-Elektronik, Huffenhardt, Ger-many), and submerged in a continuous flow (2–3 mL ⁄min) ofartificial cerebrospinal fluid (ACSF), composed of (mm): 124 NaCl,3 KCl, 1.4 NaH2PO4, 1.3 MgSO4, 26 NaHCO3, 1.5 CaCl2, 10 glucose,saturated with 5% CO2, 95%O2 (carbogen) at 32–34 �C. For stimulus-train-induced bursting (STIB) experiments, slicing procedures wereidentical, except that slices were cut at 600 lm thickness (Klapstein &Colmers, 1997; Ho et al., 2000; El Bahh et al., 2002).

Electrophysiology

Extracellular recordings of hippocampal areas CA1 and CA3 wereperformed as previously detailed (Klapstein & Colmers, 1992, 1997;Ho et al., 2000; El Bahh et al., 2002). Orthodromic populationexcitatory postsynaptic potentials (pEPSPs) were evoked via

electrodes placed in stratum radiatum of CA2 or the mossy fiberpathway. Field potentials were digitally averaged (n ¼ 3 for each datapoint) and analysed with pClamp 9 software (Axon Instruments,Union City, CA, USA). The initial linear slope of the pEPSP (50–80%of the maximum response in a slice) was measured (Klapstein &Colmers, 1992).The procedures both for 0-Mg2+ bursting and for STIB experiments

have been detailed previously (Stasheff et al., 1985; Klapstein &Colmers, 1997; Ho et al., 2000; El Bahh et al., 2002).

0-Mg2+-bursting

In brief, 400-lm-thick transverse hippocampal slices were prepared asabove, transferred to the perfusion chamber and an extracellularrecording electrode placed in stratum pyramidale of area CA3. Insome experiments, a bipolar stimulating electrode was placed on themossy fiber tract in the dentate hilus to deliver single stimuli (0.05 Hz)to optimize the placement of the recording electrode; stimuli were notapplied via this electrode during the remainder of the experiment. Theperfusion medium was then changed to one in which the MgSO4 wasomitted (0-Mg2+ ACSF), and the slice monitored for the occurrence ofspontaneous bursts (SBs). Experiments were initiated once the SBfrequency was stable (Klapstein & Colmers, 1997; El Bahh et al.,2002).

Stimulus train-induced bursting

Briefly, stratum radiatum of CA2 was stimulated and recordings madein area CA3 in 600-lm-thick transverse hippocampal slices. Sliceswere perfused with ACSF composed of (mm): 120 NaCl, 3.3 KCl, 0.9MgSO4, 1.6 CaCl2, 1.23 NaH2PO4, 25 NaHCO3, 10 glucose. After30 min perfusion, afterdischarges were elicited every 5 min by briefstimulus trains (four stimuli, 30 V, 0.1 ms, 100 Hz repeated at 5 Hz).The number of stimulus trains was set just above that needed to elicitan afterdischarge reliably (Klapstein & Colmers, 1997; Ho et al.,2000; El Bahh et al., 2002).

Drugs

Human NPY (hNPY) was purchased from Peptidec Technologies(Pierrefonds, Quebec, Canada). The receptor-preferring agonists:[ahx5)24]NPY, F7P34NPY and [hPP1)17,Ala31, Aib32]hNPY (Ala-AibNPY), were made by solid-state synthesis, as described previously(Rist et al., 1995; Cabrele et al., 2000; Soll et al., 2001). The ‘scrambled’NPYused as a control was similarly synthesized and contained the sameamino acids as NPY but in random sequence: SKPQRDANREP-TRYAIYDYSNPDIELHYLRPAYALG-NH2. BIIE0246 ((S)-N

2-[[1-[2-[4-[(R,S)-5,11-dihydro-6(6h)-oxodibenz[b,e]azepin-11-yl]-1-piperazi-nyl]-2-oxoethyl]cyclopentyl]acetyl]-N-[2-[1,2-dihydro-3,5(4H)-dioxo-1,2-diphenyl-3H-1,2,4-triazol-4-yl]ethyl]argininamid) (Doods et al.,1999) was a generous gift of Dr Henri Doods (Boehringer-Ingelheim).Novartis 1 ((trans-2-nitrobenzene-2-sulfonic acid (4-(2-naphthylmeth-ylamino) methyl) cyclohexyl methyl) amide) (Pronchuk et al., 2002)was a generous gift of Dr P. Hipskind (Lilly Research Laboratories), andGW438014A (N-[1-(2-phenylethyl)-1H-benzimidazol-2-yl]benzamidemethanesulfonate) (Daniels et al., 2002) was a generous gift of Dr AlexDaniels (GlaxoSmithKline). All other chemicals were obtained fromBDH Inc. (Toronto, Canada).All pharmacological agents were applied via bath perfusion. All

peptides and most drugs were stored as concentrated stock solutionsand diluted in ACSF prior to application, with the exception ofBIIE0246, which was dissolved in ethanol as a 1 mm stock solution(El Bahh et al., 2002) and diluted at least 10 000-fold in ACSF beforeapplication.

1418 B. El Bahh et al.

ª 2005 Federation of European Neuroscience Societies, European Journal of Neuroscience, 22, 1417–1430

The agonist and antagonist concentrations used throughout theseexperiments were chosen to maximize our opportunity to observespecifically responses that were not mediated by Y2 receptors. Ingeneral, peptide agonists were tested at a concentration of 1 lm,which, although relatively high in comparison with binding affinites,was never above the maximal effective concentration demonstrated foran NPY-related agonist in the hippocampus in vitro (Colmers et al.,1991; El Bahh et al., 2002). With agonists we had not tested in thehippocampus before (specifically AlaAibNPY and F7P34NPY), wemaintained a similar relationship between agonist concentration(1 lm) and the relative affinities of these agonists for their intendedreceptors (1 nm at Y5 and > 1 nm at Y1, respectively, Cabrele et al.,2000; Soll et al., 2001). Agonists were applied either alone or in thepresence of antagonists in 10 mL of saline. At the flow rates used,agonist application lasted between 3 and 5 min, while washout wascontinued until the reversal of agonist effects or a minimum of 20 min.

Concentrations of antagonists used were determined from previousexperiments (BIIE0246 – El Bahh et al., 2002; Novartis 1 – Pronchuket al., 2002), and were initially based on the known affinities of theantagonist for the target receptor. In the present experiments, we useda comparatively high (1 lm) concentration of the relatively potentNovartis 1 (2 nm affinity for Y5 receptors, see below) to ensureblockade of Y5 receptors, whereas we used substantially lowerconcentrations (30–100 nm) of BIIE0246, which has a similar (3 nm)affinity for the Y2 receptor. To assess antagonist effects, we firstestablished the effect of the agonist on a given preparation, then anantagonist was pre-applied for 15 min (in pEPSP experiments) or forat least 30 min (for STIB experiments), before agonists were againtested in the presence of the antagonist (El Bahh et al., 2002).

Statistical analyses

Each preparation served as its own control for comparison of agonistactions. Statistical comparisons were performed where appropriateusing a paired t-test (Graphpad Prism 4.0). For antagonist experi-ments, comparisons were made between responses taken after theequilibration period of the antagonist, immediately before applicationof the agonist. We never observed significant changes in synapticresponses with antagonist applications alone. Because of the very longduration of the in vitro experiments, usually only one slice per animalcould be analysed.

In vivo experiments

Animals

Male C57BL ⁄ 6, C57x129 and Y�=�2 mice (�60 days old) were used

for assessing seizure susceptibility to kainic acid (KA) and inexperiments involving intracerebral application of Y2 or Y5 receptorligands.

Intrahippocampal injections and EEG recordings

For intrahippocampal injections, mice were anesthetized withEquithesin (3.5 mL ⁄ kg: 1% phenobarbital ⁄ 4% chloral hydrate,Sigma, St Louis, MO, USA), and an injection guide cannula wasunilaterally positioned above the dura. For simultaneous electroen-cephalographic (EEG) recordings, nichrome-insulated bipolar depthelectrodes were implanted in the injected and contralateral hippo-campus [from bregma (mm): (nose bar at 0); AP )1.9; L ±1.5, 1.5below dura], and a ground lead was positioned over the nasal sinus.Cannula and electrodes were connected to a multipin socket andsecured to the skull with acrylic dental cement. Mice were allowed

3–5 days to recover from the surgical procedure before the start ofthe study (Vezzani et al., 2000).

Treatments

Drugs were dissolved in their respective vehicles and injected in avolume of 0.5 lL. The time course of drug injection was 1 min, afterwhich the needle was left in place for a further 1 min before removal(to avoid backflux). NPY agonists were infused intrahippocampally,5 min before a local injection of 33 pmol ⁄ 0.5 lL KA (Sigma) infreely moving mice, a dose which induces EEG seizures but notdamage (Vezzani et al., 2000). GW438014A (Daniels et al., 2002) wasinjected at 10 mg ⁄ kg intraperitoneally in a fine suspension of 0.5%methylcellulose containing 0.1% Tween 80, 30 min before adminis-tration of the Y5 receptor agonist and ⁄ or KA. Although considerablyless potent than Novartis 1, GW438014A has a high selectivity andmoderate potency for NPY-Y5 receptors (211 nm for Y5 vs.> 10 000 nm for Y1, Y2 and Y4), and has demonstrated bioavailabilitywhen administered intraperitoneally (Daniels et al., 2002), whichNovartis 1 does not. The dose and timing of the administration of thisantagonist has previously been shown to inhibit feeding mediated bycentrally injected NPY (Daniels et al., 2002). Control mice wereinjected intrahippocampally and ⁄ or systemically with the correspond-ing volumes of vehicle or scrambled NPY.

Seizure assessment

EEG recordings were carried out in the hippocampus of freely movingmice, as previously described (Vezzani et al., 2000). After baselinerecording of spontaneous EEG, drugs were injected through a needleextending 1.5 mm below the guide cannula into dorsal hippocampus.Continuous EEG recordings were made for at least 90 min after druginjection; EEG recordings were ended only if no seizures wereobserved for at least 30 min. Ictal episodes were characterized byhigh-frequency and ⁄ or multispike complexes and ⁄ or high-voltagesynchronized spikes simultaneously occurring in both hippocampi.Seizures were quantitated by determining the latency to the firstseizure (onset), the number of seizures, and the time spent in ictalactivity and spiking (interictal activity) (Vezzani et al., 2000).Three days after seizure assessments, mice were decapitated and

their brains postfixed in 4% buffered paraformaldehyde, cryoprotectedby immersion in 30% sucrose in phosphate-buffered saline, and Nisslstained to assess the correct positioning of the injection cannula andelectrodes. We observed nonspecific damage due to mechanicalinsertion of the electrodes and cannula only in KA-injected mice. Thispattern was similar to that observed in mice receiving NPY agonists orantagonists.

Statistical analysis

Data from experiments examining the actions of the Y2-selectiveagonist [ahx5)24]NPY on kainate-induced seizures were comparedusing Student’s t-test. Data from experiments examining the actions ofthe Y5 agonist and antagonist were compared using a one-way anova

followed by Tukey’s test, whereas data examining the effects of Y2

and Y5 agonists in the Y2 knockout mouse were compared using atwo-way anova followed by Tukey’s test.

Binding studies

Novartis 1 (Y5 antagonist) characterization

Binding assays were performed with crude membranes isolated fromcells transfected with cDNA constructs for the human Y1, Y2, Y4 or

NPY anti-epileptic effect is via Y2 not Y5 receptors 1419


Y5 receptors. Cells were scraped from the culture plates into PBS andpelleted in tubes. The homogenate binding studies were conducted aspreviously described (Gehlert et al., 1996, 2001). The cell pellets wereresuspended using a glass homogenizer in a 25 mm HEPES (pH 7.4)buffer containing 2.5 mm CaCl2, 1 mm MgCl2 and 2 g ⁄ L bacitracin.Cell membranes were incubated in a final volume of 200 lLcontaining 0.1 nm [125I]pPYY for Y1, Y2 and Y5 (SA 2200 Ci ⁄mmol,DuPont-NEN, Boston, MA, USA) or 0.1 nm [125I]PP (SA2200 Ci ⁄mmol, DuPont-NEN) for 2 h at room temperature. Nonspe-cific binding was defined as the amount of radioactivity remainingbound after incubating in the presence of 1 lm human NPY (Y1, Y2

and Y5 – Gehlert et al., 2001) or 1 lm bovine PP (Y4 – Gehlert et al.,1996). Incubations were terminated by rapid filtration through GF ⁄Cfilters (Wallac, Gaithersburg, MD, USA), which had been presoaked in0.3% polyethyleneimine (Sigma), using a TOMTEC (Orange, CT,USA) cell harvester. The filters were washed with 5 mL of 50 mm Tris(pH 7.4) at 4 �C and rapidly dried at 60 �C. The dried filters weretreated with MeltiLex A (Wallac), melt-on scintillator sheets, and theradioactivity retained on the filters counted using the Wallac 1205Betaplate counter. The results were analysed using the Excel(Microsoft Corp., Redmond, WA, USA) or Prism software packages(Graphpad, San Diego, CA, USA). Protein concentrations weremeasured with Commassie Protein Assay Reagent (Pierce, Rockford,IL, USA) using BSA for standards. This antagonist had no significantaffinity for any other Y receptor except Y5, where the Ki was about2 nm (Table 1).

Receptor autoradiography

This performed as described previously (Furtinger et al., 2001).Animals used in these studies (both Sprague–Dawley rats and mousestrains as indicated) were between 2 and 3 months of age.[125I]hPYY3)36 (Y2 + Y5) and [125I]Pro34hPYY (Y1 + Y5) (Neosys-tems, Strasbourg, France) were freshly iodinated and purified by HPLCas described previously (Furtinger et al., 2001). Frozen sections(20 lm) of the dorsal (coronal sections) and ventral (horizontalsections) hippocampus from WT, Y�=�

1 (Howell et al., 2003), Y�=�2

(Sainsbury et al., 2002) and Y�=�5 (Marsh et al., 1999) mice were pre-

incubated for 30 min at room temperature in 200 mLKrebs–Henseleit–Tris buffer (118 mm NaCl, 4.8 mm KCl, 1.3 mm MgSO4, 1.2 mm

CaCl2, 50 mm glucose, 15 mm NaHCO3, 1.2 mm KH2PO4, 10 mm

Tris, pH 7.3), then were incubated for 2 h in 20 mL buffer supple-mented with 0.1% bovine serum albumin, 0.05% bacitracin and 50 pmof the respective radioligand. Displacement studies were performedwith BIBO3304 (100 nm), Novartis 1 (30 and 100 nm), [d-Trp32]NPY(100 nm) and AlaAibNPY (30 and 100 nm). Nonspecific binding wasdetermined in the presence of 1 lm NPY. After incubation, sectionswere dipped twice and then washed for 30 s in ice-cold buffer, dippedin deionized water and rapidly dried under a stream of cold air. Theywere exposed to BioMax MR films (Amersham Pharmacia Biotech,Bucks., UK) for 2–3 days together with 125I-microscales.The autoradiograms were developed, digitized and analysed using a

computer-assisted image analysis system (Metamorph 3, VisitronSystem, Puchheim, Germany) equipped with a Zeiss CCD videocamera. Absorbance was measured in the molecular layer and the hilusof the dentate gyrus and in the stratum radiatum of area CA1. Grayvalues were converted to fmol ⁄mg wet weight using [125I]-micro-scales as standards. Specific binding was calculated by subtracting thenonspecific from total binding. Binding was performed on threesections from 4–6 animals for each animal and experimental conditionstudied. Comparisons were made using Student’s t-test.

Results

Evoked synaptic responses

NPY has potent inhibitory effects on evoked, excitatory synaptictransmission in both area CA1 and area CA3 of hippocampal slicesprepared from rats and a series of WT mice. Thus, the pan-agonistNPY (1 lm) suppressed pEPSPs in hippocampal area CA1 of rat,and WT mice having either a 129 ⁄ Sv, C57BL ⁄ 6 or mixedC57BL ⁄ 6 · 129 ⁄ SvJ background (C57 · 129) (Sainsbury et al.,2002) by 77–86%, with EC50 values ranging between 118 and 162 nm(Fig. 1a–c). In hippocampal area CA3, 1 lm NPY inhibited thepEPSPs by a smaller, but significant amount, 47 ± 4.7% (P < 0.02,n ¼ 7) in rat or WT mice, 32.7 ± 4.1% in 129 ⁄ SvJ WT (P < 0.01,n ¼ 4) and 31.5 ± 4.4% in C57 · 129 WT (P < 0.02, n ¼ 4).We also tested the actions of Y2 or Y5 receptor-preferring agonists in

both areas of the hippocampus in rats and threeWTstrains. In area CA1of rat and WT (129 ⁄ SvJ and C57 · 129) mice, the Y2-selective agonist[ahx5)24]NPY (1 lm – Rist et al., 1995; El Bahh et al., 2002) inhibitedthe pEPSPs by 51–67% (Fig. 1b), whereas the same concentration of thepotent, Y5-preferring agonist [hPP

1)17, Ala31Aib32]pNPY (AlaAibNPY– Cabrele et al., 2000) inhibited the pEPSPs only by 11–14% in rat andall three WT mouse strains (Fig. 1c).We determined the receptor selectivity of NPY and of the agonists

we used by comparing their effects in the absence and presence ofantagonists specific to the Y2 or Y5 receptors in both hippocampalareas. The effect of NPY in area CA1 of rat and WT (129 ⁄ SvJ andC57 · 129) mice was reduced by 60–80% by 30 nm of the Y2-specific

Table 1. Characterization of Novartis 1 affinity at human NPY ⁄ PP receptors

Receptor Ki (± SEM) (nm) n

hY1 > 1000 2hY2 > 1000 2hY4 > 1000 2hY5 1.72 ± 0.73 4

Fig. 1. Actions of NPY and agonists on excitatory synaptic transmission in vitro. Antagonists were uniformly without effects of their own when applied alone. (a)Concentration–response relationship for NPY actions on the pEPSP in area CA1 of hippocampal slice from rat and WT (C57Bl ⁄ 6 and C57 · 129) mice. (Inset)Representative traces showing the reversible inhibition of pEPSP in area CA1 by NPY in these species. Each point represents the mean ± SEM of 3–6 determinations.(b) Comparison of the effects of NPY, the Y2-preferring agonist, [ahx5-24]NPY, and the Y5-preferring agonist, AlaAibNPY, all applied at 1 lm, on the pEPSP inhippocampal area CA1 in rat and all three WT strains. (c–f) Pharmacological analysis of NPY responses in hippocampal slices. Data in (c–f) were from comparisonsin individual preparations of the effect of the agonist first in the absence, then in the equilibrated presence, of the antagonist. Effects are expressed as a percentage ofthe control response in the individual slice preparation. (c) Relative effect of NPY (1 lm) in the presence of antagonists to Y5 (Novartis 1, 1 lm) or Y2 (BIIE0246,30 nm) receptors in area CA1 of rat and 3 WT strains. (d) Relative effects of the Y5- (AlaAibNPY, 1 lm) and Y2-preferring ([ahx5-24]NPY, 1 lm) agonists in thepresence of the specific Y5 (Novartis, 1 lm) or Y2 (BIIE0246, 30 nm) antagonists in area CA1 of slices from rat and three WT strains of mice. Variability in relativeresponses to AlaAibNPY result from the small amplitude of the control responses. (e) Relative effects (compared with control responses) of the putative Y5 ⁄Y2-preferring agonist, NPY13-36, in the presence of either Novartis 1 or BIIE0246 in areas CA1 and CA3 of 129 ⁄ SvJ mice, and area CA3 of C57 · 129 mice. (f) Inhippocampal slices from Y �=�

2 mice, NPYand agonists with preferences for Y2, Y5 and Y1 receptors (all tested at 1 lm) are without significant effects in areas CA1or CA3, though baclofen (10 lm) was highly effective in both regions. Bars in each panel represent the means ± SEM for 4–10 slices per group. Statisticalcomparisons were between data points obtained from the same slice under different conditions, using the paired t-test. *P < 0.05, **P < 0.01.



antagonist BIIE0246 (Doods et al., 1999; El Bahh et al., 2002)(P < 0.01, n ¼ 4–8); the same concentration of BIIE0246 reduced theeffect of the Y2 preferring agonist [ahx5)24]NPY in rat and these twoWT mice by 70–86% (P < 0.01, n ¼ 4–11; Fig. 1d). By contrast, theY5-specific antagonist Novartis 1 (1 lm –Pronchuk et al., 2002) hadno significant influence on the actions of either NPYor AlaAibNPY inrat or in any of the WT mice tested (Fig. 1d and e). Interestingly, thesmall but significant inhibitory actions of the Y5 agonist observed inarea CA1 of rat and all three WT mice tested were not altered by the

Y5 antagonist, but were significantly reduced by 30 nm BIIE 0246(Fig. 1e, P < 0.05, n ¼ 4).Similar responses to agonists and antagonists were observed in area

CA3 of rat. Thus, whereas there were little or no significant changescaused by Novartis 1 in the responses to 1 lm applications of NPY,[ahx5)24]NPY, BIIE 0246 (30 nm) significantly reduced responses toboth agonists; the responses to 1 lm AlaAib NPY in rat area CA3were not significant (pEPSP inhibition of 6.7 ± 2.9%, n ¼ 6,P > 0.05) (Fig. 1f).



We also compared the actions in WT mice of the putatively mixedY2 ⁄Y5 -preferring agonist NPY13)36 (Woldbye et al., 1997; Baraban,2002). NPY13)36 (1 lm) inhibited evoked synaptic responses in areaCA1 of 129 ⁄ SvJ mice and in area CA3 of both 129 ⁄ SvJ andC57 · 129 mice. However, as observed with the other agonists, theeffect of NPY13)36 was not significantly inhibited by the Y5-specificantagonist Novartis 1 (1 lm), but was inhibited strongly by 30 nmBIIE0246 (Fig. 1g). Thus, this mixed agonist also appears to act onlythrough a Y2 receptor.We next tested NPY and related agonists in areas CA1 and CA3 of

the hippocampus from Y2 receptor-deficient mice. In contrast to allWT mice studied, 1 lm concentrations of NPY or agonists selectivefor the Y2 or Y5 receptor had no significant effect on pEPSPs evokedin both hippocampal areas of Y�=�

2 mice (Fig. 1h), nor did a Y1

receptor-selective agonist (F7P34NPY – Soll et al., 2001). However,the GABAB agonist baclofen (10 lm; Klapstein & Colmers, 1992;Qian et al., 1997) inhibited the pEPSPs in areas CA1 and CA3 of Y�=�

2

mice by 85% (n ¼ 10) and 50% (n ¼ 5), respectively (Fig. 1h),indicating that presynaptic neuromodulatory mechanisms at thesesynapses are not altered by the genetic inactivation of the Y2 receptor.These results indicate that NPY inhibits excitatory synaptic

transmission at Schaffer collateral–CA1 and mossy fiber–CA3synapses of all species tested by activation of Y2 receptors alone.

In vitro epileptiform discharges

To assess the contribution of specific receptor subtypes to theanticonvulsant actions of NPY we used two different models ofepileptiform activity in vitro, the 0-Mg2+ bursting model (Mody et al.,1987; Klapstein & Colmers, 1997; Marsh et al., 1999) and STIB, anin vitro model of limbic seizures (Stasheff et al., 1985; Klapstein &Colmers, 1997; Ho et al., 2000; El Bahh et al., 2002), which issensitive to therapeutic concentrations of anticonvulsant drugs effect-ive against partial complex seizures (Clark & Wilson, 1992). In the0-Mg2+ bursting model, which is insensitive to therapeutic levels ofanticonvulsants (Zhang et al., 1995), but which others have usedpreviously to test Y5 agonists (Marsh et al., 1999), we measured theactions of NPY and the Y5 agonist in slices from WT (C57 · 129)mice, in saline or in the presence of either Novartis 1 (1 lm) orBIIE0246 (100 nm). At 1 lm, NPY reduced bursting frequency byabout 40% in saline (P < 0.03, n ¼ 6) or in the presence of 1 lmNovartis 1 (P < 0.02, n ¼ 6) but its actions were blocked by the Y2

antagonist (P < 0.003). By contrast, the Y5 agonist showed nomeasurable effect under any of these conditions. (Fig. 2).We next assessed the response to NPY and selective agonists and

antagonists in STIB induced in area CA3 of hippocampal slices fromrat, WT (C57BL ⁄ 6 and C57 · 129) and Y�=�

2 mice. In slices from ratsand WT (C57BL ⁄ 6 and C57 · 129) mice, application of only 300 nmNPY induced a profound, prolonged but reversible suppression of theafterdischarges (Fig. 3a). Similar though less prolonged effects wereobserved with application of the Y2-preferring agonist [ahx5)24]NPY(1 lm). However, consistent with previous results (El Bahh et al.,2002), neither the Y1- nor the Y5-preferring agonists (applied at 1 lm)reduced the afterdischarges in either of the WT mouse strains used inthese experiments. Consistent with the results obtained on evokedsynaptic responses, pharmacological blockade of Y2 receptors with100 nm BIIE0246 also completely eliminated the actions of NPY andthe Y2-preferring agonist on STIB in slices from either WT strain(Fig. 3a). As expected, the Y2-specific antagonist did not affect thesuppression of STIB by baclofen or 2-CA (data not shown).Pharmacological blockade of Y5 receptors with Novartis 1 (1 lm)

Fig. 2. NPY inhibition of spontaneous (0-Mg2+) bursting in slices fromC57 · 129 WT mice is mediated by Y2 receptors. (a) Recordings ofspontaneous activity in a slice from a C57 · 129 WT mouse, each showinga 60-s segment of 0-Mg2+ bursting recorded in control, during drug applicationand after washout. Upper panel: effect of NPY alone. Middle panel: effect ofNPY (1 lm) in the continuing presence of BIIE 0246 (100 nm). Lower panel:effect of the Y5-preferring agonist AlaAibNPY. (b) Effects of NPY and AlaAibNPY on 0-Mg2+ bursting in slices from C57 · 129 hippocampus in control, inthe presence of the Y5 antagonist Novartis 1 (1 lm) or in the presence of BIIE0246 (100 nm). n ¼ 6 for all conditions. Paired-t-test was used to compareresponses as in Fig. 1: *P < 0.05, **P < 0.01.



Fig. 3. NPY inhibition of STIB epileptiform discharges in hippocampus requires the Y2 receptor. (a and b) Representative traces showing the effects of NPY onprimary afterdischarges (1�AD) induced in vitro by high-frequency stimulation (arrowheads) in area CA3 of the hippocampus. (a) NPY (300 nm) completelyblocked the 1�AD in slices from rat, C57BL ⁄ 6 and C57 · 129 WT mice; pretreatment of the same slice with the Y2 antagonist BIIE0246 (100 nm) completelyblocked the effect of NPY. Identical results were obtained in slices from 3–6 animals per group. (b) In hippocampal slices from Y�=�

2 mice, none of the NPYagonists tested affected the 1�AD, although the GABAB agonist baclofen was effective (lower right panel).



did not alter the effects of NPY or the Y2-preferring agonist in slicesfrom WT (129 ⁄ SvJ) mice. Neither the Y2 nor the Y5 antagonists bythemselves had significant effects on STIB.

In slices from Y�=�2 mice, application of NPY or any of the Y1, Y2,

and Y5 receptor-preferring agonists (all at 1 lm) had no effect on theoccurrence, latency or duration of the STIB afterdischarges (Fig. 3b),

Fig. 4. Representative EEG tracings from freelymoving mice injected with kainic acid unilaterallyinto the hippocampus, and responses to the Y2

agonist. RHP and LHP represent the injected(right) and contralateral (left) hippocampus,respectively. In each cluster of recordings in (a)–(c)the first two traces are baseline activity, the middletraces illustrate typical ictal episodes and the lasttwo traces illustrate interictal spiking activity. Thearrowheads below the middle traces indicate theduration of the ictal episodes. (a) WT (C57 · 129)mouse; (b) WT (C57BL ⁄ 6) mouse injected intra-hippocampally with 24 nmol [ahx5)24] NPYbefore kainic acid treatment; (c) Y�=�

2 mouse.Injection of the selective Y2 agonist into WTanimals significantly (�50%) reduced both thenumber of seizures and the total time spent in ictalactivity. Thus, the incidence of ictal episodes, butnot the duration of each ictal event, was signifi-cantly reduced by the treatment (b vs. a). Nosignificant effect on spiking activity was observedafter Y2 agonist injection. Y2 gene deletion did notsignificantly affect ictal activity but greatly en-hanced spiking activity in the hippocampi (c vs. a).The pattern of changes in EEG activity induced bykainic acid were similar in both strains of WT(C57 · 129 and C57BL ⁄ 6) mice, so only thetraces from the WT strain are shown here.



Fig. 5. Effects of Y2- and Y5-preferring agonists on kainic acid-induced EEG seizures in mouse hippocampus in vivo. (a) Intrahippocampal injection of the Y2-preferring agonist [ahx5)24]NPY 5 min before KA injection induced a significant inhibition of seizure number, increase in latency of seizure onset and reduction intime spent in seizures in C57BL ⁄ 6 mice compared with animals pretreated with ‘scrambled’ peptide. (b) Effects similar to those seen with the Y2 agonist wereobserved when the mice were pre-injected with AlaAibNPY. However, pretreatment of the animals with a Y5 antagonist (GW438014A) 30 min beforeintrahippocampal KA injection did not block the AlaAibNPY effects on seizures. (c) Effects of intrahippocampal injection of KA in WT (C57 · 129) and Y�=�

2mice. Pre-injection of either [ahx5)24]NPYor AlaAibNPY before KA injection had no effect on seizures in Y�=�

2 mice. Each bar represents the mean ± SEM for 8–25animals per group. Comparisons in (a) were made using Student’s t-test, comparisons in (b) were made using a one-way anova followed by Tukey’s test andcomparisons in (c) were made using a two-way anova followed by Tukey’s test. **P < 0.01 vs. control.



although they remained sensitive to 10 lm baclofen (Fig. 3b) and1 lm 2-CA (data not shown).

The results from these two in vitro models of epileptiform activityare consistent with the idea that the anti-epileptic effect of NPYin vitro requires the activation only of Y2 receptors.

In vivo seizures

Here we tested the anticonvulsant action of the intrahippocampalinjection of Y2 and Y5 receptor-preferring agonists on limbic motorseizures induced by local injection of KA in WT and Y2 knockoutmice (Sperk, 1994). Seizures were monitored behaviorally and withEEG recordings (Vezzani et al., 2000).

Although most Y�=�2 mice did not exhibit any behavioral pheno-

types, a small minority of mice demonstrated spontaneous seizuresafter handling, and in 2 ⁄ 10 mice spontaneous interictal spiking wasrecorded in the absence of any manipulation. Because of its rarity, wedid not study this effect in detail. WT (C57BL ⁄ 6 or C57 · 129) micewere never observed to display spontaneous or handling-inducedseizures.

As described previously (e.g. Vezzani et al., 2000), intrahippo-campal injection of 33 pmol KA results in epileptic-like discharges

(ictal activity) and interictal spiking in both ipsilateral and contra-lateral hippocampus (Fig. 4). No generalized behavioral convulsionswere observed; however, staring and immobility often preceded andaccompanied the occurrence of EEG ictal episodes. Prior injection of24 nmol of the Y2-preferring agonist [ahx5)24]NPY into the ipsilat-eral hippocampus of WT (C57BL ⁄ 6) mice (n ¼ 8–25) significantlyincreased the latency to onset of KA-induced seizures by134% (P < 0.01), reduced the number of seizure episodes by 50%(P < 0.01) and shortened the duration of epileptic activity by 64%(P < 0.01), compared with control animals injected with scrambledpeptide (Figs 4 and 5a). Injection of 6.8 nmol of the Y5-preferringagonist AlaAib NPY, a dose that induced food intake in rats (Cabreleet al., 2000), also reduced the number of seizure episodes by 44%(P < 0.01) and the time spent in seizure activity by 38% (P < 0.01)in KA-treated mice (n ¼ 7–12) (Figs 4 and 5b). However, pretreat-ment with intraperitoneal injection of 10 mg ⁄ kg of the Y5 receptor-specific antagonist GW438014A, a treatment that blocked feedinginduced by centrally administered NPY (Daniels et al., 2002), didnot alter the ability of the Y5-preferring agonist to reduce seizureactivity in WT (C57BL ⁄ 6) mice (Fig. 5b). Because of its localtoxicity, we could not test the effects of BIIE0246 in these in vivoexperiments.

Fig. 6. Binding of NPY ligands in mouse hippocampus is not displaced by a Y5 receptor antagonist or agonists. (a–c) Autoradiograms for the binding of the Y2 ⁄ 5receptor-preferring ligand [125I]hPYY3-36, and the Y1 ⁄ 5 receptor -preferring ligand [

125I]Pro34hPYY, and quantitative evaluation of binding. (a) Binding of the Y2 ⁄ 5receptor-preferring ligand [125I]hPYY3-36 (50 pm) in stratum radiatum of WT (C57 · 129) mice was not altered by AlaAibNPY (100 nm), [d-Trp32]NPY (30 nm)(c) or Novartis 1 (100 nm), but is not detectable in Y�=�

2 mice under the same conditions (a, c). In Y�=�5 mice (bottom panel of a), the binding of this ligand is

consistent with a Y2 profile. (b) Binding of the Y1 ⁄ 5 receptor-preferring ligand [125I]Pro34hPYY in the dentate gyrus of WT (C57 · 129) mice is almost entirelydisplaced by a saturating concentration of the Y1 antagonist BIBO3304 and in Y�=�

1 mice of the same background, but not by the Y5 antagonist Novartis 1 (100 nm)(b, c). In stratum radiatum of CA1, [125I]Pro34hPYY binding is reduced in Y�=�

1 mice by about 20% compared with the respective WT mice (c). In this area,saturating concentrations of the Y1 antagonist BIBO3304 did not influence Pro34hPYY binding significantly (c). The Y5 agonists AlaAib NPY (100 nm) and [d-Trp32]NPY (30 nm) and the Y5 receptor antagonist Novartis 1 (100 nm) only incompletely reduced [125I]hPro34hPYY binding in the presence of 1 lm BIBO3304 inWT and in Y�=�

1 mice (b and c). In the Y�=�5 mouse, binding of this ligand in dentate gyrus is consistent with a Y1 profile, but to some extent may also label Y2

receptors (lower panel in b). In all cases, NPY itself abolished all binding (a and c). (c) Quantitative binding data for different regions of mouse brain under theconditions indicated. Comparisons were made using Student’s t-test.



In response to intrahippocampal KA injection, both Y�=�2 and WT

(C57 · 129) mice (n ¼ 15–18) exhibited similar latency to onset,number of seizures and duration of ictal activity. The only differenceobserved was a greater duration of interictal activity seen in Y�=�

2 mice(P < 0.01, Fig. 5c). However, in contrast to the WT mice, pre-injection of either the Y2 or the Y5 agonist in Y�=�

2 animals did notsignificantly affect any of these indices of seizure activity (Fig. 5c).These results strongly suggest that the Y2 receptor is essential for

NPY suppression of limbic seizures in vivo.

Receptor autoradiography

The Y2 ⁄Y5-preferring agonist [125I]hPYY3-36 demonstrated pro-nounced labeling of the strata radiatum and oriens in areas CA1 toCA3 of WT and Y�=�

1 (Howell et al., 2003) but not Y�=�2 animals

(Fig. 6a). This binding was not significantly reduced by the Y5-preferring agonists AlaAibNPY (30 and 100 nm) and [d-Trp32]NPY(30 nm) or the Y5-specific antagonist Novartis 1 (30 and 100 nm),indicating that [125I]hPYY3-36 specifically labeled Y2 but not Y5

receptors. The Y1 ⁄Y5 receptor agonist [125I]Pro34hPYY (Dumont et al.,

1996; Marsh et al., 1999) bound intensely to receptors in the outer layerof the cerebral cortex, in the molecular layer of the dentate gyrus and inthe strata oriens and radiatumofCA1 toCA3ofWT,Y�=�

2 andY�=�5 mice

(Fig. 6b). Binding in the cortex and dentate gyrus molecular layer wasabolished when slices were incubated with 100 nm of the potent Y1

receptor-specific antagonist BIBO3304 (Wieland et al., 1998) and inY�=�

1 mice (Fig. 6b). Binding of [125I]Pro34hPYY in these regions wasunaffected byNovartis 1 (30 or 100 nm), AlaAibNPY (30 or 100 nm) or30 nm [d-Trp32]NPY, consistent with its labeling only Y1 receptorsthere. In strata oriens and radiatum of areas CA1 to CA3, binding of[125I]Pro34hPYY was not reduced either by 100 nm BIBO3304 or inY�=�

1 mice, consistent with the presence of Y5 binding sites. Here,however, we observed slight but measurable reductions in [125I]Pro34h-PYYbinding in the presence ofY5 ligands. Thus, AlaAibNPY (100 nm)reduced it by 21%, [d-Trp32]NPY (30 nm) reduced it by 9% andNovartis 1 (100 nm) reduced it by 12%. These results are at bestconsistent with very low levels of Y5 receptors; indeed, similar levels of[125I]Pro34hPYY binding were also seen in Y�=�

5 mice (Fig. 6b).Quantitative measurements are shown in Fig. 6c.

Discussion

Various neurotransmitter and neuromodulatory systems have beenfound to contribute to inhibition in the brain. Thus, enhancing theactivity of a given inhibitory system by either exogenous applicationof an agonist or increasing the endogenous release of the naturalagonist not surprisingly results in increased inhibition. However, itdoes not necessarily follow from this that the primary biological rolefor that system is to act as an ‘endogenous anticonvulsant’. Thus, NPYpotently suppresses seizure activity in vitro and in vivo. The presentstudy provides broad pharmacological and genetic evidence bothin vitro and in vivo that NPY and a wide variety of agonists modulateexcitatory synaptic transmission and epileptiform activity in thehippocampus by activation of exclusively the Y2 receptor subtype,while providing no evidence for a significant role of Y5 or other Yreceptors, despite conditions biased in favor of finding such evidence.However, consistent with the caveat above, our results do not indicatethat the ablation or pharmacological blockade of the Y2 receptorsignificantly alters excitability in vitro nor does ablation of thehippocampal Y2 receptor result in development of significant seizureactivity in mice in vivo.

We have previously shown in vitro that activation of presynapticNPY receptors reduces glutamate release from presynaptic terminalsby suppressing voltage-dependent Ca2+ currents (Toth et al., 1993;Qian et al., 1997) and that this action was sufficient to suppressepileptiform activity in vitro (Klapstein & Colmers, 1997). Here, wedemonstrate that the Y2 receptor is both necessary and sufficient tomediate this effect in hippocampus in vitro and in vivo. Geneticablation of Y2 receptors totally abolishes the anti-epileptic actions ofNPY and all agonists tested in vitro and in vivo, in contrast to previousstudies (Woldbye et al., 1997; Marsh et al., 1999; Baraban, 2002)which had suggested that Y5 receptors are essential for NPY’s actions.In a spectrum of WT animals, even micromolar concentrations of thepotent, Y5 receptor-specific antagonist Novartis 1 affected neither themodest effects of Y5-preferring agonists in vitro nor their moresignificant effects in WT mice in vivo. By contrast, the Y2 antagonistBIIE0246, at concentrations of only 30–100 nm, sharply reduced orfully blocked the in vitro effects of the Y5 receptor-preferring agoniststested in these same WT strains, including the 129 ⁄ SvJ strain used asthe WT for the Y5 receptor-knockout mouse (Marsh et al., 1999).Neither the knockout nor the pharmacological blockade of the Y2

receptor affected the ability of the neuromodulatory GABAB oradenosine A1 receptors, which also use the same intracellular signalingpathway as NPY to suppress glutamate release (Klapstein & Colmers,1992; Qian et al., 1997). Despite an experimental approach thatfavored the identification of Y5 receptor-mediated effects, we found noevidence to support a role of Y5 receptors in any of the presentexperiments. It is also important to note that even agonists which weredesigned to have both high affinity and specificity for the Y5 receptorsubtypes nevertheless still had measurable effects on Y2 receptors, asthese effects were blocked by the Y2 receptor antagonist (but not bythe Y5 antagonist) or were absent in Y�=�

2 mice. Thus, any conclusionsabout NPY receptor subtypes based on agonist responses alone mustbe interpreted with caution until properly verified with receptor-specific antagonists or knockout animals.Whereas knockout of Y2 receptors in these mice completely

prevented the effects of NPY on synaptic activity and in in vitroseizure models, the Y2 knockout mice showed remarkably littledifference from the WT in their responses to intrahippocampalapplication of KA. Indeed, the only difference in the response betweenY�=�

2 and WT mice to kainate was a slightly increased duration ofinterictal activity. Previous studies with the NPY knockout (Ericksonet al., 1996; Baraban et al., 1997) or the Y5 receptor knockout mouse(Marsh et al., 1999) reported the dramatic, highly lethal consequencesfor genetic alterations to the NPY system in the responses to kainate.The difference with the present results may arise from the in vivoexperimental approaches. Here we injected KA and receptor-selectiveagonists directly into the hippocampus, while in previous reports KAwas applied either peripherally or i.c.v., both of which would cause amuch greater volume of the brain to be affected by KA than in ourexperiments. In additon, the genetic background of the animalsappears to influence the seizure susceptibility in Y5 knockout mice(Marsh et al., 1999). In the present experiments, the in vitropharmacological results were remarkably consistent in rats and abroad spectrum of WT mice, including the 129 ⁄ SvJ (Marsh et al.,1999) and C57BL ⁄ 6 parent strains and the C57 · 129 animals whichhad the same backcrossed background as the Y�=�

2 (Sainsbury et al.,2002). The in vivo results were furthermore in complete agreementwith the results from the in vitro experiments. Our findings alsohighlight that some receptor-preferring agonists that appear veryselective at the low concentrations of radioligand employed in bindingexperiments can lose their selectivity when used at elevated concen-trations, which frequently occurs in in vivo experiments. One



important caveat in comparing the present results is that the in vitroelectrophysiological experiments were performed using youngeranimals (3–5 weeks) than were used for the other experiments (2–3 months). However, earlier results on the potent inhibition ofglutamate release by NPY and Y2 agonists (Greber et al., 1994),which were conducted in hippocampal slices from 2- to 3-month-oldrats, support and are entirely consistent with the present in vitro andin vivo findings.

Results from binding experiments were also fully consistent withthe functional observations. Thus, prominent populations of Y2 (strataoriens and radiatum of areas CA1 to CA3) and Y1 receptors (dentategyrus molecular layer) were observed. Ligands at these binding siteswere selectively displaced by Y2-and Y1-preferring ligands and wereabolished in Y�=�

2 and Y�=�1 mice, respectively, but were neither

displaced by Y5 ligands nor abolished in Y�=�5 mice, and thus

represent unambiguous Y2 and Y1 receptor sites. Thus, we found noevidence in these experiments for Y5 receptors, despite previousreports (e.g. Bregola et al., 2000). However, the nature of some of thebinding sites labeled by [125I]Pro34hPYY in strata radiatum and oriensis not clear. A significant population of these sites, seen in WT mice,which was not displaced by the Y5-preferring agonists AlaAibNPYand [d-Trp32]NPY, or the specific Y5 and Y1 receptor antagonistsNovartis 1 and BIBO3304, respectively, also remained in all threeknock-out lines, and were only reduced by only about 15–20% by theY4 ⁄Y5 ligand rPP, and by about 40% with the Y2-preferring agonist[ahx5)24]NPY (data not shown). Such [125I]Pro34hPYY-labeled siteswere not seen in the human hippocampus (Furtinger et al., 2001) andcould represent y6 sites in the mouse (Michel et al., 1998), orpotentially another, as yet uncharacterized, receptor (Lin et al., 2005).Regardless, the present binding data do not provide any evidence forthe presence of detectable Y5 receptors in mouse hippocampus.

Interestingly, the absence of significant Y5 receptor binding in therat is at odds with the high levels of Y5 mRNA observed in allhippocampal subfields (e.g. Kopp et al., 1999). Thus, Dumont et al.(1998) reported Y5-specific binding only in area CA3 of the ventralextension of rat hippocampus. Our present results also indicate thepresence of a small proportion of binding sites that are labeled bythe Y1 ⁄Y5-preferring ligand [125I]Pro34hPYY (50 pm) and that aredisplaced by the Y5-preferring agonist AlaAibNPY and theantagonist Novartis 1. Provided that the [125I]Pro34hPYY at aconcentration of 50 pm sufficiently labels Y5 sites in this tissue,these data argue for the presence of very low, but demonstrable,levels of Y5 binding sites in the hippocampus. By contrast, thebinding of [125I]Pro34hPYY that was not displaced by the Y1

antagonist BIBO3304 in the Y�=�5 mice may represent an additional

class of Y-receptors, as recently suggested (Dumont et al., 2005;Lin et al., 2005).

Although the present results make it certain that the Y2 receptor isessential for NPY regulation of limbic seizures originating in thehippocampus, different NPY receptor subtypes are known to regulateexcitability in other regions of the brain, including Y5 receptorsubtypes (e.g. Sun et al., 2001). Recent evidence in rats in vivo, usingthe same Y5 antagonist used here, indicates that the threshold forhippocampal rapid kindling can be reduced by Y5 agonists, and thatthis effect is blocked by the Y5 antagonist (Benmaamar et al., 2005).The site of this action is most likely extrahippocampal (Benmaamaret al., 2005). The effect of an acute functional inactivation of the Y2

receptor in adult animals remains unknown, but may be answeredwith the advent of Y2 antagonists with lower toxicity than that usedhere. This would also permit a test of the interesting question of thefunction of the Y2 receptor in the normal physiology of thehippocampus. Given the present results however, it is most unlikely

that NPY receptors other than Y2 significantly regulate excitabilitywithin the hippocampal formation, making this the most promisingY receptor as a target for potential new treatments for some focalepilepsies.

Acknowledgements

This research was supported by the Human Frontiers Science ProgramOrganization and the Canadian Institutes for Health Research. W.F.C. is aMedical Scientist of the Alberta Heritage Foundation for Medical Research. Wethank Dr Berrnhard Aray for assistance with the binding experiments. We aregrateful to Drs Henri Doods (Boehringer-Ingelheim), P. Hipskind (Eli Lilly)and Alex Daniels (GlaxoSmithKline) for their generous gifts of antagonists,and to Dr Richard Palmiter for his kind gift of Y�=�

5 mouse brains.

Abbreviations

EEG, electroencephalography; KA, kainic acid; NPY, neuropeptide Y; pEPSPs,population excitatory postsynaptic potentials; SBs, spontaneous bursts; STIB,stimulus train-induced bursting; WT, wild type.

References

Baraban, S.C. (2002) Antiepileptic actions of neuropeptide Y in the mousehippocampus require Y5 receptors. Epilepsia, 43 (Suppl. 5), 9–13.

Baraban, S.C., Hollopeter, G., Erickson, J.C., Schwartzkroin, P.A. & Palmiter,R.D. (1997) Knock-out mice reveal a critical antiepileptic role forneuropeptide Y. J. Neurosci., 17, 8927–8936.

Benmaamar, R., Richichi, C., Gobbi, M., Daniels, A.J., Beck-Sickinger, A.G.& Vezzani, A. (2005) Neuropeptide Y Y5 receptors inhibit kindlingacquisition in rats. Reg. Pept., 125, 79–83.

Bregola, G., Dumont, Y., Fournier, A., Zucchini, S., Quirion, R. & Simonato,M. (2000) Decreased levels of neuropeptide Y (5) receptor binding sites intwo experimental models of epilepsy. Neuroscience, 98, 697–703.

Cabrele, C., Langer, M., Bader, R., Wieland, H.A., Doods, H.N., Zerbe, O. &Beck-Sickinger, A.G. (2000) The first selective agonist for the neuropeptideY Y5 receptor increases food intake in rats. J. Biol. Chem., 275, 36043–36048.

Clark, S. & Wilson, W.A. (1992) Brain slice model of epilepsy: neuronalnetworks and actions of antiepileptic drugs. In: Faingold, C.L. & Fromm,G.H. (Eds), Drugs for Control of Epilepsy: Actions on Neuronal NetworksInvolved in Seizure Disorders. CRC Press, Boca Raton, Florida, pp. 89–123.

Colmers, W.F., Klapstein, G.J., Fournier, A., St-Pierre, S. & Treherne, K.A.(1991) Presynaptic inhibition by neuropeptide Y in rat hippocampal slice invitro is mediated by a Y2 receptor. Br. J. Pharmacol., 102, 41–44.

Colmers, W.F., Lukowiak, K. & Pittman, Q.J. (1987) Presynaptic action ofneuropeptide Y in area CA1 of the rat hippocampal slice. J. Physiol. (Lond.),383, 285–299.

Colmers, W.F., Pronchuk, N., Torok-Both, C., Ho, M., Aronyk, K., McKean, J.,Snyder, T., Sinclair, D.B., Javidan, M. & Beck-Sickinger, A.G. (1997)Neuropeptide Y2 and other receptors inhibit synaptic excitation in epileptichuman brain. Epilepsia, 38 (Suppl. 8), 12.

Daniels, A.J., Grizzle, M.K., Wiard, R.P., Matthews, J.E. & Heyer, D. (2002)Food intake inhibition and reduction in body weight gain in lean and obeserodents treated with GW438014A, a potent and selective NPY-Y5 receptorantagonist. Regul. Pept., 106, 47–54.

Doods, H., Gaida, W., Wieland, H.A., Dollinger, H., Schnorrenberg, G., Esser,F., Engel, W., Eberlein, W. & Rudolf, K. (1999) BIIE0246: a selective andhigh affinity neuropeptide Y Y(2) receptor antagonist. Eur. J. Pharmacol.,384, R3–R5.

Dumont, Y., Fournier, A. & Quirion, R. (1998) Expression and characterizationof the Neuropeptide Y Y5 receptor subtype in the rat brain. J. Neurosci., 18,5565–5574.

Dumont, Y., Fournier, A., St-Pierre, S. & Quirion, R. (1996) Autoradiographicdistribution of [125I]Leu31,Pro34]PYY and [125I]PYY3-36 binding sites in therat brain evaluated with two newly developed Y1 and Y2 receptorradioligands. Synapse, 22, 139–158.

Dumont, Y., Moyse, E., Fournier, A. & Quirion, R. (2005) Evidence for theexistence of an additional class of Neuropeptide Y receptor sites in the ratbrain. J. Pharmacol. Exp. Ther., in press. [doi:10.1124/jpet.105.089300]



El Bahh, B., Cao, J.Q., Beck-Sickinger, A.G. & Colmers, W.F. (2002)Blockade of neuropeptide Y(2) receptors and suppression of NPY’s anti-epileptic actions in the rat hippocampal slice by BIIE0246. Br. J.Pharmacol., 136, 502–509.

Erickson, J.C., Clegg, K.E. & Palmiter, R.D. (1996) Sensitivity to leptin andsusceptibility toseizuresofmice lackingneuropeptideY.Nature,381, 415–421.

Furtinger, S., Pirker, S., Czech, T., Baumgartner, C., Ransmayr, G. & Sperk, G.(2001) Plasticity of Y1 and Y2 receptors and neuropeptide Y fibers inpatients with temporal lobe epilepsy. J. Neurosci., 21, 5804–5812.

Gehlert, D.R., Schober, D.A., Beavers, L., Gadski, R.A., Hoffman, J.A.,Chance, R., Lundell, I. & Larhammar, D. (1996) Characterization of thepeptide binding requirements for the cloned human pancreatic polypeptidepreferring (PP1) receptor. Mol. Pharmacol., 50, 112–118.

Gehlert, D.R., Yang, P., George, C., Schober, D.A., Gackenheimer, S.L.,Johnson, D., Beavers, L.S., Gadski, R.A. & Baez, M. (2001) Cloning andcharacterization of rhesus neuropeptide Y Y1, Y2 and Y5 Receptors.Peptides, 22, 343–350.

Gobbi, M., Gariboldi, M., Piwko, C., Hoyer, D., Sperk, G. & Vezzani, A.(1998) Distinct changes in peptide YY binding to, and mRNA levels of, Y1and Y2 receptors in the rat hippocampus associated with kindlingepileptogenesis. J. Neurochem., 70, 1615–1622.

Greber, S., Schwarzer, C. & Sperk, G. (1994) Neuropeptide Y inhibitspotassium-stimulated glutamate release through Y2 receptors in rathippocampal slices in vitro. Br. J. Pharmacol., 113, 737–740.

Ho, M.W.Y., Beck-Sickinger, A.G. & Colmers, W.F. (2000) Neuropeptide Y5

receptors mediate inhibition of synaptic excitation in proximal subiculum,but do not suppress epileptiform activity in rat hippocampal slices.J. Neurophysiol., 83, 723–734.

Howell, O.W., Scharfman, H.E., Beck-Sickinger, A.G., Herzog, H. & Gray,W.P. (2003) Neuropeptide Y is neuroproliferative in primary hippocampalcultures. J. Neurochem., 86, 646–659.

Klapstein, G.J. & Colmers, W.F. (1992) 4-Aminopyridine and low Ca2+

differentiate presynaptic inhibition mediated by neuropeptide Y, baclofen and2-chloroadenosine in rat hippocampal CA1 in vitro. Br. J. Pharmacol., 105,470–474.

Klapstein, G.J. & Colmers, W.F. (1997) Neuropeptide Y suppresses epilepti-form activity in rat hippocampus in vitro. J. Neurophysiol., 78, 1651–1661.

Kofler, N., Kirchmair, E., Schwarzer, C. & Sperk, G. (1997) Altered expressionof NPY-Y1 receptors in kainic acid induced epilepsy in rats. Neurosci. Lett.,230, 129–132.

Kopp, J., Nanobashvili, A., Kokaia, Z., Lindvall, O. & Hokfelt, T. (1999)Differential regulation of mRNA’s for Neuropeptide Y and its receptorsubtypes in widespread areas of the rat limbic system during kindlingepileptogenesis. Brain Res. Mol. Brain Res., 72, 17–29.

Lin, S., Boey, D., Couzens, M., Lee, N., Sainsbury, A. & Herzog, H. (2005)Compensatory changes in [125I]-PYY binding in Y receptor knockout micesuggest the potential existence of further Y receptor(s). Neuropeptides., 39,21–28.

Marsh, D.J., Baraban, S.C., Hollopeter, G. & Palmiter, R.D. (1999) Role of theY5 neuropeptide Y receptor in limbic seizures. Proc. Natl Acad. Sci. USA,96, 13518–13523.

McQuiston, A.R. & Colmers, W.F. (1996) Neuropeptide Y2 receptors inhibitthe frequency of spontaneous but not miniature EPSCs in CA3 pyramidalcells of rat hippocampus. J. Neurophysiol., 76, 3159–3168.

Michel, M.C., Beck-Sickinger, A., Cox, H., Doods, H.N., Herzog, H.,Larhammar, D., Quirion, R., Schwartz, T. & Westfall, T. (1998) XVIInternational Union of Pharmacology recommendations for the nomenclatureof neuropeptide Y, peptide YY, and pancreatic polypeptide receptors.Pharmacol. Rev., 50, 143–150.

Mody, I., Lambert, J.D.C. & Heinemann, U. (1987) Low extracellularmagnesium induces epileptiform activity and spreading depression in rathippocampal slices. J. Neurophysiol., 57, 869–888.

Patrylo, P.R., van den Pol, A.N., Spencer, D.D. & Williamson, A. (1999) NPYinhibits glutamatergic excitation in the epileptic human dentate gyrus.J. Neurophysiol., 82, 478–483.

Pronchuk, N., Beck-Sickinger, A.G. & Colmers, W.F. (2002) Presynapticaction by multiple neuropeptide Y receptors modulates GABAergictransmission in paraventricular nucleus of rat hypothalamus. Endocrinol-ogy, 143, 535–543.

Qian, J., Colmers, W.F. & Saggau, P. (1997) Inhibition of synaptic transmissionby neuropeptide Y in rat hippocampal area CA1: modulation of presynapticCa2+ entry. J. Neurosci., 17, 8169–8177.

Reibel, S., Benmaamar, R., Le, B.T., Larmet, Y., Kalra, S.P., Marescaux, C. &Depaulis, A. (2003) Neuropeptide Y delays hippocampal kindling in the rat.Hippocampus, 13, 557–560.

Richichi, C., Lin, E.J., Stefanin, D., Colella, D., Ravizza, T., Grignaschi, G.,Veglianese, P., Sperk, G., During, M.J. & Vezzani, A. (2004) Anticonvulsantand antiepileptogenic effects mediated by adeno-associated virus vectorneuropeptide Y expression in the rat hippocampus. J. Neurosci., 24, 3051–3059.

Rist, B., Wieland, H.A., Willim, K.D. & Beck-Sickinger, A.G. (1995) A rationalapproach for the development of reduced-size analogues of neuropeptide Ywith high affinity to the Y1 receptor. J. Pept. Sci., 1, 341–348.

Sainsbury, A., Schwarzer, C., Couzens, M., Fetissov, S., Furtinger, S., Jenkins,A., Cox, H.M., Sperk, G., Hokfelt, T. & Herzog, H. (2002) Important role ofhypothalamic Y2 receptors in body weight regulation revealed in conditionalknockout mice. Proc. Natl Acad. Sci. USA, 99, 8938–8943.

Schwarzer, C., Kofler, N. & Sperk, G. (1998) Up-regulation of neuropeptide Y-Y2 receptors in an animal model of temporal lobe epilepsy.Mol. Pharmacol.,53, 6–13.

Schwarzer, C., Williamson, J.M., Lothman, E.W., Vezzani, A. & Sperk, G.(1995) Somatostatin, neuropeptide Y, neurokinin B and cholecystokininimmunoreactivity in two chronic models of temporal lobe epilepsy.Neuroscience, 69, 831–845.

Silva, A.P., Carvalho, A.P., Carvalho, C.M. & Malva, J.O. (2003) Functionalinteraction between neuropeptide Y receptors and modulation ofcalcium channels in the rat hippocampus. Neuropharmacology, 44, 282–292.

Soll, R.M., Dinger, M., Lundell, I., Larhammar, D. & Beck-Sickinger, A.G.(2001) Novel analogues of neuropeptide Y with preference for the Y1-receptor. Eur. J. Biochem., 268, 2828–2837.

Sperk, G. (1994) Kainic acid seizures in the rat. Prog. Neurobiol., 42, 1–32.Sperk, G., Marksteiner, J., Gruber, B., Bellmann, R., Mahata, M. & Ortler,

M. (1992) Functional changes in neuropeptide Y- and somatostatin-containing neurons induced by limbic seizures in the rat. Neuroscience,50, 831–846.

Stasheff, S.F., Bragdon, A.C. & Wilson, W.A. (1985) Induction of epileptiformactivity in hippocampal slices by trains of electrical stimuli. Brain Res., 344,296–302.

Sun, Q.Q., Akk, G., Huguenard, J.R. & Prince, D.A. (2001) Differentialregulation of GABA release and neuronal excitability mediated byneuropeptide Y1 and Y2 receptors in rat thalamic neurons. J. Physiol.(Lond.), 531, 81–94.

Toth, P.T., Bindokas, V.P., Bleakman, D., Colmers, W.F. & Miller, R.J. (1993)Mechanism of presynaptic inhibition by neuropeptide Yat sympathetic nerveterminals. Nature, 364, 635–639.

Tu, B., Timofeeva, O., Jiao, Y. & Nadler, J.V. (2005) Spontaneous release ofneuropeptide Y tonically inhibits recurrent mossy fiber synaptic transmissionin epileptic brain. J. Neurosci., 25, 1718–1729.

Vezzani, A., Michalkiewicz, M., Michalkiewicz, T., Moneta, D., Ravizza, T.,Richichi, C., Aliprandi, M., Mule, F., Pirona, L., Gobbi, M., Schwarzer, C. &Sperk, G. (2002) Seizure susceptibility and epileptogenesis are decreased intransgenic rats overexpressing neuropeptide Y. Neuroscience, 110, 237–243.

Vezzani, A., Moneta, D., Conti, M., Richichi, C., Ravizza, T., De Luigi, A., DeSimoni, M.G., Sperk, G., Andell-Jonsson, S., Lundkvist, J., Iverfeldt, K. &Bartfai, T. (2000) Powerful anticonvulsant action of IL-1 receptor antagoniston intracerebral injection and astrocytic overexpression in mice. Proc. NatlAcad. Sci. USA, 97, 11534–11539.

Vezzani, A. & Sperk, G. (2004) Overexpression of NPY and Y2 receptors inepileptic brain tissue: an endogenous neuroprotective mechanism in temporallobe epilepsy? Neuropeptides, 38, 245–252.

Vezzani, A., Sperk, G. & Colmers, W.F. (1999) Neuropeptide Y: emergingevidence for a functional role in seizure modulation. Trends Neurosci., 22,25–30.

Wieland, H.A., Engel, W., Eberlein, W., Rudolf, K. & Doods, H.N. (1998)Subtype selectivity of the novel nonpeptide neuropeptide Y Y1 receptorantagonist BIBO 3304 and its effect on feeding in rodents. Br. J. Pharmacol.,125, 549–555.

Woldbye, D.P., Larsen, P.J., Mikkelsen, J.D., Klemp, K., Madsen, T.M. &Bolwig, T.G. (1997) Powerful inhibition of kainic acid seizures byneuropeptide Y via Y5-like receptors. Nature Med., 3, 761–764.

Zhang, C.L., Dreier, J.P. & Heinemann, U. (1995) Paroxysmal epileptiformdischarges in temporal lobe slices after prolonged exposure to lowmagnesiumare resistant to clinically used anticonvulsants. Epilepsy Res., 20, 105–111.



The anti-epileptic actions of neuropeptide Y in the ...cnc.cj.uc.pt/BEB/private/pdfs/NeuroMod/ElBahh...

Documents

Transcript of The anti-epileptic actions of neuropeptide Y in the ...cnc.cj.uc.pt/BEB/private/pdfs/NeuroMod/ElBahh...