Ž .Brain Research 859 2000 386–389www.elsevier.comrlocaterbres

Short communication

Effect of gastrin-releasing peptide on rat hippocampal extracellular GABAlevels and seizures in the audiogenic seizure-prone DBAr2 mouse

Nick Andrews ), Ben Davis, M. Isabel Gonzalez, Ryszard Oles, Lakhbir Singh,Alexander T. McKnight

Parke-DaÕis Neuroscience Research Centre, Cambridge UniÕersity ForÕie Site, Robinson Way, Cambridge CB2 2QB, UK

Accepted 28 December 1999

Abstract

Ž .Gastrin-releasing peptide GRP , a selective agonist for the BB subtype of bombesin receptor, is reported to depolarise GABAergic2

interneurons in the stratum oriens layer of the hippocampus. Such an action might lead to increased extracellular levels of GABA in thehippocampus, and result in an anti-convulsant effect with this peptide. We have tested this hypothesis by determining the effect of GRPon extracellular levels of GABA in the ventral hippocampus of the freely moving rat using in vivo microdialysis, and by intracerebroven-

Ž .tricular i.c.v. administration of GRP to audiogenic seizure-prone DBAr2 mice prior to exposure to the noise of an electric bell.Ž .Following local perfusion in the ventral hippocampus by reverse dialysis GRP 10 mM significantly raised levels of GABA in the

Ž .recovered dialysates by approximately 40%. In the seizure studies, GRP 30–300 ng increased the latency to tonic seizure, the number ofmice convulsing and reduced the incidence of lethality. In both dialysis and seizure studies, the effects of GRP were blocked by the

w 6 13x Ž .selective BB receptor antagonist, D-Phe , Leu-NHEt bombesin 6–13 . These experiments provide further functional evidence that2

activation of the BB receptor may modulate neurotransmission in the hippocampus, and that this action may confer anti-convulsant2

properties on agonists acting at the BB receptor in the brain. q 2000 Elsevier Science B.V. All rights reserved.2

Keywords: Bombesin; In vivo microdialysis; Rat; Audiogenic seizure; Mouse

1. Introduction

Bombesin BB and BB receptors have a heterogeneous1 2

distribution in the central nervous system, indicating thatthe endogenous ligands for these receptors may differen-

w xtially modulate neurotransmission 2 . Interestingly, it hasrecently been reported that BB receptors may be located2

on a class of GABAergic interneuron in the stratum oriensw xlayer of the hippocampus 3 . Thus bath application of

Ž .Gastrin-releasing peptide GRP induced a large inwardcurrent and produced a robust depolarisation of visuallyidentified interneurons within the stratum oriens layer ofhippocampal slices in vitro, under whole-cell clamp; acti-vation of the inward current was blocked by preincubation

w 6with the selective BB receptor antagonist D-Phe , Leu-213 x Ž .NHEt bombesin 6–13 . Significantly, the application of

GRP increased the frequency and amplitude of sponta-neous inhibitory post-synaptic currents in CA1 pyramidal

w xneurons, and this action was blocked by bicuculline 3 .

) Corresponding author. Fax: q44-1223-249106; e-mail:nick.andrews@wl.com

From these findings, we have postulated that this effectof BB receptor stimulation to increase GABA-mediated2

neurotransmission in the hippocampus, could be of func-tional significance in epilepsy since several forms of thedisease respond to treatments that modulate GABAergicneurotransmission such as GABA uptake inhibitors andallosteric modulators of the GABA receptor complex.A

We have evaluated this hypothesis by direct approaches inwhole-animal experiments: by investigating the effect onextracellular levels of GABA of local perfusion with GRPin the ventral hippocampus of the freely moving rat, andby testing for an action of GRP to protect against audio-genic seizures in the mouse.

2. Materials and methods

2.1. Animals

ŽHooded Lister rats 250–350 g males, Harlan Olac,.Bicester, UK and audiogenic seizure-prone DBAr2 mice

0006-8993r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0006-8993 00 02010-2

( )N. Andrews et al.rBrain Research 859 2000 386–389 387

Ž .males, 21–23 days of age, 6–10 g, Harlan, UK receivedfood and water ad libitum. They were housed in a holdingroom lit by low light levels.

2.2. In ÕiÕo microdialysis

2.2.1. SurgeryŽ .An intracerebral guide cannula BAS was implanted

Žusing coordinates from the atlas of Paxinos and Watsonw x.5 at a point 5.0 mm posterior to bregma, 4.8 mm lateralto the midline and 3.0 mm below the level of the dura,

Žusing stereotaxic techniques under general anaesthesia 3–. Ž4% isofluraner100% O . A microdialysis probe 4 mm2

.membrane length was then lowered down the guide can-nula and secured so that the active portion of the mem-

Žbrane extended into the ventral hippocampus probe tip.position 7.0 mm below dura . The probe was connected

via polyethylene to a Harvard precision pump for perfu-sion overnight at 1 mlrmin with artificial cerebrospinal

Ž .fluid aCSF consisting of 140 mM NaCl, 1.2 mM CaCl ,2

4 mM KCl and 11 mM glucose at pHs7.4.

2.2.2. HPLC analysisApproximately 18 h after surgery, dialysate samples

were collected every 30 min and assayed for their contentof GABA by high performance liquid chromatography

Ž .with electrochemical detection HPLC–ECD using a sys-tem comprising an ESA 580 solvent delivery pump, ESACoulochem II detector and ESA 5014b analytical cellŽ .E1sy200 mV, E2sq300 mV . The dialysate con-stituents were eluted using a Hypersil C18 5-mm columnŽ .150=4.6 mm in a mobile phase of 0.15 M sodium

Žacetater0.4 mM Na–EDTA in 40–45% acetonitrile pHs.5.5–6.4 at 1 mlrmin. Fresh standard solutions of GABA

dissolved in aCSF were prepared daily for peak verifica-tion and typically eluted around 18 min after injection ontothe column. Derivatisation and subsequent detection ofGABA was after pre-column derivatisation with o-phthal-

w xaldehyder2-methyl-2-propanethiol 1 .

2.2.3. ProcedureGRP in aCSF was perfused for 30 min after three

Ž .successive samples containing similar basal levels ofGABA had been attained. For antagonism studies aCSF

w 6containing the bombesin receptor antagonist D-Phe , Leu-13 x Ž .NHEt bombesin 6–13 was first perfused for 60 min.

This solution was then exchanged for one containing bothagonist and antagonist for 30 min, then finally, the perfu-sion solution was returned to antagonist alone for theremainder of the experiment.

2.3. Audiogenic seizures

Ž .Mice received intracerebroventricular i.c.v. injectionsof GRP dissolved in saline, or saline vehicle, under anaes-

Ž .thesia 3–4% isoflurane in 20% O , 80% N O . Injections2 2

Ž .5 ml were made at lambda with a 3.5-mm long 27-gaugeneedle attached to a 50-ml Hamilton syringe and thewound sealed with epoxy resin. For antagonist studies, the

Ž .antagonist was co-administered i.c.v. with a submaximaldose of GRP. Fifteen minutes after injection, tonic seizureswere induced by exposure to the noise of an electric bellŽ .125 dB, 1.4 kHz for 60 s. Latency to convulse wasmeasured in seconds, and the number of animals convuls-ing and the incidence of lethality were noted, and ex-pressed as a percentage. Animals failing to show a tonicseizure within 60 s were given a maximum latency scoreof 60 s.

2.3.1. Data analysisThe neurotransmitter peak was converted to a level of

GABA based on the calibration data from external GABAstandards; levels of GABA were expressed as a percentage

Ž .of basal mean of three samples prior to drug . Levels atindividual time points were compared with the last basalsample using the Student’s t-test.

For audiogenic seizure experiments, differences in la-tency to convulse were analysed by a Kruskall–Wallisfollowed by Dunn’s test. The percentage of animals con-vulsing and percentage lethality were analysed using a x 2

test followed by Fisher’s exact test.

3. Results

3.1. Effect of GRP on extracellular hippocampal GABAleÕels

Basal levels of GABA following 30 min of collectionŽ .were found to be 503"35.6 pg ns18 . GABA levels

were found to be significantly increased above basal levelsin both samples collected during the period in which GRP

Ž .Fig. 1. Effect of GRP 10 mM on extracellular ventral hippocampalGABA levels in rats. Results are expressed as the percentage change inGABA levels relative to levels of GABA prior to exposure to either GRPŽ . 6 13 x Ž . Ž . UUns6 or GRPqD-Phe , Leu-NHEt bombesin 6–13 ns6 . P -

0.01 Student’s t-test vs. 120-min time point.

( )N. Andrews et al.rBrain Research 859 2000 386–389388

Žwas present in the ventral hippocampus first sample:126.8"7.1% P-0.01, second sample: 142.9"11.5%

.P-0.01 relative to basal levels returning to basal levelsin the first sample where GRP was not present in the brainŽ .see Fig. 1 . This effect of GRP was completely blockedby co-administration of the selective BB receptor antago-2

w 6 13 x Ž . wnist, D-Phe , Leu-NHEt bombesin 6–13 ; ns6. D-6 13 x Ž .Phe , Leu-NHEt bombesin 6–13 by itself had no effect

Ž .on GABA levels data not shown indicating that there waslittle endogenous GRP-like tone acting at the BB recep-2

tors in the hippocampus.

( )3.2. Effect of GRP i.c.Õ. on audiogenic seizures

Ž .A single i.c.v. injection of GRP 30–300 ng , 15 minbefore starting the audiogenic stimulus, produced a dose-dependent increase in the latency to convulse, with a

Ž .minimum effective dose of 100 ng Fig. 2, panel a . At adose of 300 ng, GRP produced a near total blockade ofseizures with complete prevention of lethality. It is wellknown that administration of BB receptor agonists in-2

duces characteristic scratching typically using the hindlimbs; GRP induced this response at all doses tested.

Ž . ŽFig. 2. a The anti-convulsant effect of GRP on sound-induced seizures in DBAr2 mice. Values shown are the mean latency scores S.E.M. shown by. Ž . U UU Ž .vertical bars of ns6 per group top graph . P-0.05, P-0.01 vs. vehicle-treated animals Kruskall–Wallis followed by Dunn’s test . The middle

Ž .graph shows percentage of animals showing tonic seizure and the bottom graph percentage of animals dying within 60 s. b The antagonism of theŽ . w 6 13 x Ž . Žprotective effects of GRP against audiogenic seizures. GRP 100 ng was co-administered i.c.v. with D-Phe , Leu-NHEt bombesin 6–13 100–1000

. Ž . Ž .ng 15 min before exposure to sound. Values shown are the mean latency scores S.E.M. shown by vertical bars of ns8–19 per group top graph . Themiddle graph shows percentage of animals showing tonic seizure and the bottom graph shows percentage of animals dying within 60 s. U P-0.05,UU UUU ŽP-0.01, P-0.001 vs. vehicle-treated animals; sP-0.05, qqP-0.01, qqqP-0.001 vs. GRP-only-treated animals Kruskall–Wallis

2 .followed by Dunn’s test for latency, top graph; x followed by Fisher’s exact test for percentages, middle and bottom graphs .

( )N. Andrews et al.rBrain Research 859 2000 386–389 389

w 6Co-administration of the BB receptor antagonist D-Phe ,213 x Ž . ŽLeu-NHEt bombesin 6–13 which by itself did not

.induce scratching behaviour produced a dose-dependentŽ .100–1000 ng block of the anti-convulsant effect of GRPŽ . Ž Ž .100 ng compared to the group receiving GRP 100 ng

. Ž .and vehicle . The highest dose 1000 ng fully antagonisedŽ .the anti-convulsant effect of GRP Fig. 2, panel b .

4. Discussion

The results of this study show that GRP increasesextracellular levels of GABA in the hippocampus throughstimulation of BB receptors, thereby providing a func-2

tional correlate for observations showing that GRP depo-larises GABAergic interneurons within the hippocampusw x3 . Furthermore, GRP was found to be anti-convulsant inseizure-prone DBAr2 mice exposed to an audiogenicstimulus. It must be acknowledged however, that thesestudies do not allow us to say conclusively that the anti-convulsant activity of GRP was through an action withinthe hippocampus of the DBAr2 mouse, since the peptidewas delivered to the brain via the ventricles. It has beenreported that seizures in DBAr2 mice may actually becontrolled via the inferior colliculus, since this nucleus is

w xone of the main relays of the auditory system 4 , but weare not aware of any evidence for the presence of the BB2

receptor at this site. In order to characterise further theputative anti-epileptic potential of agonists acting at BB2

receptors, it will be necessary to assess their activity in

other tests of anti-convulsant activity, such as with the useof electroconvulsive shock.

In conclusion, the present results from in vivo experi-ments provide functional support for the findings reported

w xin the literature 3 and further, these data demonstrate thatGRP can reduce seizure activity following administrationof this peptide to the brain as a whole, via the ventricularsystem, suggesting that this action could be of wide-rang-ing use. Finally, the reduction of seizure severity may bedue to increased GABAergic function as suggested by theability of the BB agonist to increase extracellular GABA2

levels in the ventral hippocampus of the rat.

References

w x1 J. Kehr, U. Ungerstedt, Fast HPLC estimation of gamma-amino-butyric acid in microdialysis perfusates: effect of nipecotic and

Ž .3-mercaptopropionic acids, J. Neurochem. 151 1988 1308–1310.w x2 E.E. Ladenheim, R.T. Jensen, S.A. Mantey, P.R. McHugh, T.H.

Moran, Receptor heterogeneity for bombesin-like peptides in the ratŽ .central nervous system, Brain Res. 537 1990 233–240.

w x3 K. Lee, A.K. Dixon, I. Gonzalez, E.B. Stevens, S. McNulty, R.J.Oles, P.J. Richardson, R.D. Pinnock, L. Singh, Bombesin-like pep-tides depolarise hippocampal interneurones through interactions with

Ž . Ž .subtype 2 bombesin receptors, J. Physiol. London 518 1999791–802.

w x4 R. Marianowsky, H. Pollard, J. Moreau, G. Despres, Y. Ben-Ari, P.Tran-Ba-Huy, R. Romand, N-Methyl-D-aspartate receptor subunitsNR1 and NR2C are over-expressed in the inferior colliculus of

Ž .audiogenic mice, Neurosci. Lett. 189 1995 190–194.w x5 G. Paxinos, C.H. Watson, The Rat Brain in Stereotaxic Coordinates,

Academic Press, London, 1986.