neuromedin c decreases potassium conductance and increases a non-specific conductance in rat...

Ž .Brain Research 750 1997 67–80

Research report

Neuromedin C decreases potassium conductance and increases a non-specificconductance in rat suprachiasmatic neurones in brain slices in vitro

Tracy Reynolds 1, Robert D. Pinnock )

Parke DaÕis Neuroscience Research Center, UniÕersity of Cambridge ForÕie Site, Robinson Way, Cambridge CB2 2QB, UK

Accepted 29 October 1996

Abstract

Whole-cell recordings were made, in both current and voltage clamp, from suprachiasmatic neurones maintained in coronal rat brainŽ .slices. In current clamp doses of 10 and 100 nM neuromedin C NMC were shown to increase basal firing rate in 9 out of 32 neurones.

ŽThe excitatory responses to 100 nM NMC were accompanied by small increases in neuronal input resistance 25.0"9.9% in 4 out of 7. Ž .neurones tested and depolarisations of membrane potential 9.8"3.4 mV in 4 out of 7 neurones tested . However, 10 nM NMC caused

no changes in either neuronal input resistance or membrane potential despite the clear increases in neuronal firing rate. Whenvoltage-clamped at y60 mV, 100 nM NMC induced an inward current of 14.8"1.2 pA in 46 of 210 neurones. The NMC-induced

Ž .inward current was shown to be unaffected by perfusion with 1 mM tetrodotoxin TTX . The inward current recorded at y60 mV wastypically associated with a decrease in membrane conductance. Construction of current–voltage relationships in the absence and presenceof 100 nM NMC showed that with the majority of the NMC-sensitive neurones the inward current either reversed polarity close to thepotassium reversal potential or decreased at hyperpolarised potentials. This reversal potential was shifted to more depolarised potentialswhen the extracellular concentration of potassium was increased. The NMC-induced inward current was unaffected by reduction of theextracellular concentration of sodium or by addition of 0.2 mM cadmium. In potassium-free conditions, in both the dialysing pipettesolution and perfusing saline, NMC was still able to induce an inward current. The additional reduction of the extracellular concentrationof sodium, whilst recording in potassium-free conditions, was also unable to abolish the inward current. Recordings made with anelectrode containing the non-hydrolysable guanosine triphosphate analogue, guanosine 5X-thio-triphosphate, resulted in NMC-inducedinward currents which failed to recover to baseline. It is concluded that NMC excites a subpopulation of suprachiasmatic neurones bydecreasing a resting potassium conductance and increasing a non-specific conductance, via a G-protein link.

Keywords: Bombesin; Gastrin-releasing peptide; Neuromedin C; Suprachiasmatic; Periventricular; Whole-cell; Potassium conductance; Non-specificconductance

1. Introduction

Ž .Neuromedin C NMC is one of three mammaliancounterparts to the bombesin-like peptide family, whichincludes peptides derived from amphibian and mammaliantissues. All three mammalian peptides, which includes

Ž . Ž .gastrin-releasing peptide GRP and neuromedin B NMB ,have been isolated from porcine spinal cord and non-antral

w xgastric tissue 18,19,24–26 . The peptide NMC is thecarboxy terminal decapeptide fragment of the mammalian

Ž .counterpart gastrin-releasing peptide GRP . Bombesin anda number of related amphibian peptides were themselves

) Corresponding author. E-mail: [email protected] Present address: Merck Sharpe and Dohme Neuroscience Research

Center, Terlings Park, Harlow, Essex, UK

isolated from the skin of a number of species of Europeanw xand American frogs 2,10,11 . Both amphibian and mam-

malian members of the bombesin-like peptide family havebeen shown to evoke a range of biological responses invivo which includes effects on feeding, thermoregulation,blood pressure, locomotor activity and circadian rhythmsw x1,3,8,10,12,21 . To date, two mammalian bombesin recep-tor subtypes have been cloned, termed the GRP and NMB

w xreceptor subtypes 6,41 . In situ hybridisation studies haveshown that each receptor subtype displays a distinct distri-bution throughout the rat CNS, with overlap in some

w xregions 5 .Evidence supports the view that the biological clock or

circadian pacemaker is located within the suprachiasmaticŽ .nucleus SCN , a structure located in the rostral hypothala-

w x Ž .mus 20,24,27,38 . Albers et al. 1991 have recently

0006-8993r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0006-8993 96 01332-7

( )T. Reynolds, R.D. PinnockrBrain Research 750 1997 67–8068

demonstrated that GRP is able to induce phase shifts incircadian rhythms, noted as delays in the onset of wheel-

w xrunning activity in hamsters 1 . This evidence supports arole for the mammalian form of bombesin in the re-en-trainment of circadian rhythms. In a previous study, wereported the use of the three mammalian counterparts and

wthe se lec tive G R P -recep to r an tagon ist D -6 x Ž .Phe bombesin 6–13 ethylamide to examine the phar-

macology of the bombesin receptors expressed within thew xSCN 32 . The data obtained were supportive of the pre-

dominant expression of the GRP-preferring bombesin re-ceptor subtype by a subpopulation of suprachiasmatic neu-rones. In this present study using NMC we have examinedthe ionic mechanism which underlies the acute excitatoryresponses to the bombesin-like peptides. The peptide GRP,and the decapeptide fragment NMC, have been shown tobe expressed by neurones which appear to be intrinsic to

w xthe nucleus 24,40 . This aim has been achieved by appli-cation of the whole-cell recording technique to acutelyprepared rat brain slices.

2. Materials and methods

2.1. Preparation of brain slices

Coronal brain slices containing the suprachiasmatic nu-w xclei were prepared from rats as previously described 30 .

Prior to sacrifice, male Wistar rats were housed in a 12-hlight–dark cycle. During the subjective day only, ratsŽ .50–150 g lightly anaesthetized with ether were killed bycervical dislocation and the brain was rapidly removed.

Ž .Coronal slices 350–400 mm were cut at the level of theŽ .rostral hypothalamus in a vibratome Oxford instruments

Ž .filled with artificial cerebrospinal fluid ACSF at a tem-perature of 18–258C. A single slice was submerged in a

Ž .tissue chamber volume approx. 0.5 ml which was contin-uously perfused at a rate of 2 mlrmin with ACSF pre-heated to a temperature of 378C. The ACSF was composed

Ž .of the following concentrations in mM : NaCl, 126; KCl,2; MgCl , 1.3; CaCl , 2; KH PO P3H O, 1.2; glucose,2 2 2 4 2

11; NaHCO , 25; this solution was saturated with 95% O3 2

and 5% O . The pH of the solution at 378C was 7.2.2

2.2. Whole-cell recordings

A period of at least 60 min was allowed for equilibra-tion of the brain slices, prior to making recordings. Allwhole-cell recordings were made in the dorsal region ofthe nucleus, on the border of the periventricular andsuprachiasmatic nuclei, between 1 to 11 h after preparationof the brain slices. In the initial control experiments, patchelectrodes were filled with intracellular solution of the

Ž .following composition mM : potassium gluconate, 125;NaCl, 15; MgCl , 2; EGTA, 11; HEPES, 10; K ATP, 1.5;2

Na GTP, 0.2; the pH of the solution was adjusted to 7.2

with KOH. The resistance of the electrodes determined inACSF, prior to cell attachment, were between 4 to 10MV . Compensation for the series resistance presented bythe electrode was achieved by adjusting a bridge balancecircuit, prior to advancing the electrode into the brain slice.After advancing the electrode into the slice, cell attach-ment was then achieved using the blind technique. Sealresistances were found to be between 5 to 10 GV whileneurones were cell-attached. Application of further nega-tive pressure to the electrode served to rupture the mem-brane to obtain the whole-cell configuration. Cells wereabandoned if no action potentials could be detected incurrent clamp, despite the injection of positive depolarisingcurrent, suggesting that the cell was not a neurone.

Membrane potentials and currents were recorded on-lineŽ .on a personal computer IBM AT using the software

Ž .VCAN Dr. J. Dempster, University of Strathclyde, UK .The raw current and voltage signals were also stored indigitised form on video tape using a modified Sony PCMŽ .Fentronics and Sony b-max video recorder. In the volt-

Ž .age-clamp configuration, current–voltage IrV relation-ships were constructed from voltage step commands of700 to 1000 ms duration from a holding potential of y60mV to voltages ranging from y140 to y40 mV, using theVCAN software. Steady-state currents were measured dur-ing the last 500 ms of each step, unless otherwise stated.Neurones were voltage-clamped using both continuous anddiscontinuous single electrode voltage clamps, using sam-pling rates of between 1 to 10 kHz in the latter configura-tion. No significant differences in either the size of theinward currents induced by NMC or the appearance of theIrV relationships were observed when using discontinuousas opposed to continuous voltage clamp. At the end of therecording, corrections were made for voltage errors arisingdue to junction potentials by noting the electrode potentialfollowing retraction into the perfusing ACSF and modified

Ž .extracellular solutions. Resting membrane potentials RMPwere calculated from control IrV relationships determinedin voltage clamp, where the RMP was the voltage at whichno holding current was required. Neuronal input resis-tances were also calculated in voltage clamp from theinverse of the slope of the control IrV relationshipsbetween y60 and y70 mV. Neuronal input resistancesŽ .R , membrane potentials and currents are expressed asIN

Ž .means"standard error of the mean S.E.M. . Compar-isons of sets of data were made using the Student’s t-test.The level of statistical significance of differences betweendata was assessed using the RS1 version 12.10E softwareŽ .BBN Software, Cambridge, Massachusetts, USA , withP-0.05 being considered as statistically significant.

2.3. Composition of modified intracellular and extracellu-lar solutions

A number of modifications to the ionic compositionwere made to the perfusing ACSF. The low sodium solu-

( )T. Reynolds, R.D. PinnockrBrain Research 750 1997 67–80 69

tion was made by replacing the sodium chloride withequimolar choline chloride or N-methyl-D-glucamine, leav-ing the remaining 25 mM sodium bicarbonate. In the caseof the choline substitution, atropine was also added to theACSF to a final concentration of 1 mM in order to blockany actions of choline on muscarinic receptors. Whencadmium and nickel chloride were added to the ACSF at aconcentration of 0.2 mM, the potassium phosphate bufferwas replaced with equimolar potassium chloride.

During experiments where recordings in potassium-freeŽconditions in both the intracellular and extracellular solu-

.tions were made, potassium chloride in the ACSF wasreplaced with equimolar cesium chloride. Membrane po-tentials were recorded using modified intracellular solutionwhere potassium gluconate was replaced with equimolarcesium gluconate, the pH of the final pipette solution wasadjusted to 7.2 with cesium hydroxide. When the ACSFwas changed to a modified solution during a whole-cellrecording, a period of at least 5 min was allowed forequilibration. The experiments were only continued whenthe baseline holding current had stabilised.

In another series of experiments the control potassiumgluconate-based intracellular solution was modified to in-clude 0.2 mM of the tetralithium salt of the non-hydrolysa-

Ž . Xble guanosine triphosphate GTP analogue, guanosine-5 -Ž .thiotriphosphate GTPg S , in place of hydrolysable GTP.

2.4. Drugs

Drugs were stored at y208C in mM stock solutionswhich were diluted in ACSF and applied to the slice bysuperfusion. The peptide NMC was applied for periods of60 s, with between 20 to 30 min being allowed forrecovery of responses before reapplication of the peptide.All drugs used in this study were obtained from Sigma,with the exception of the peptide NMC which was ob-tained from Genosys.

3. Results

All whole-cell recordings were made in the dorsalregion of the nucleus, on the border of the SCN andperiventricular nucleus of the hypothalamus. In our previ-

w xous report 30 this region was referred to as the PVNrSCNregion.

3.1. Neuronal membrane properties

A total of 210 neurones were sampled under voltage-clamp conditions, of which 46 were shown to be sensitiveto NMC. The RMP and R values determined for bothIN

NMC-sensitive and NMC-insensitive neurones did not dis-play any obvious differences. The RMP and R valuesIN

for the 46 NMC-sensitive neurones were calculated asy52.9"1.2 mV and 639"32 MV respectively. Hyper-polarising step commands from y60 mV to )y80 mV

were shown to activate a slow inward current whichreached steady-state within 200 ms in 16 out of 46 neu-

Ž .rones Fig. 10A . This ionic current resembled, in terms ofbo th k inetics and vo ltage sensitiv ity , the

w xhyperpolarisation-activated cation current or I 22 . Ath

y130 mV the magnitude of inward rectifying current wasŽ .determined to be 41.6"5.4 pA ns16 . This value was

calculated by subtracting the current measured immedi-ately after the settling of the capacitance transient fromthat measured during the last 500 ms of the voltage pulseŽ .denoted by I and I respectively in Fig. 10B .ins ss

An additional transient ionic current was noted at y60mV when hyperpolarising prepulses were applied to atleast y90 mV, peaking in 20 ms and lasting for 50 ms in

Ž .duration Fig. 10B . This inward current was shown to beŽ .blocked by the addition of 0.2 mM nickel ns2 . There-

fore in terms of pharmacology, kinetics and voltage sensi-tivity, this transient current resembled the T-type calcium

w xcurrent 17 . Furthermore, this T-type calcium current wasobserved in virtually the same subpopulation of neuroneswhich also expressed I , i.e. 11 out of 46 neurones.h

Whole-cell recordings were made largely during the sub-jective day. However, during this time there was no evi-dence of any circadian variation in neuronal parameters orexpression of either I or the T-type calcium current.h

3.2. Responses to NMC in current clamp

In a previous study, we reported that the three mam-malian counterparts of the bombesin-like peptide familyŽ .i.e. NMC, GRP and NMB were able to evoke excitatoryresponses from a subpopulation of suprachiasmatic neu-

w xrones 30 . The mean EC value reported for NMC in the50

SCN approximated to 10 nM. Application of 10 nM NMCfor 60 s to suprachiasmatic neurones, during whole-cellrecordings, induced excitatory responses from 9 out of 32neurones. The magnitude and duration of the responses,which lasted for 5 to 15 min, were comparable to those

w xreported in the previous extracellular study 30 . Despitethe notable increases in neuronal firing rate induced by 10nM NMC, no changes in either membrane potential or

Ž .neuronal input resistance were observed Fig. 1 . The latterparameter was assessed from the size of the hyperpolaris-ing electrotonic potentials arising from the periodic injec-tion of negative current via the recording electrode. In 7 ofthese 9 recordings, NMC was reapplied at a concentrationof 100 nM following a recovery period of at least 20 min.In only 4 of these 7 NMC-sensitive neurones, the excita-tory responses to 100 nM NMC were associated with anincrease in the basal neuronal input resistance of 25.0"

Ž .9.9% ns4 and a small depolarisation of membraneŽ .potential 9.8"3.4 mV, ns4 .

3.3. Responses to NMC in Õoltage clamp

Application of 100–1000 nM NMC to a total of 210neurones voltage-clamped at y60 mV, induced an inward


Fig. 1. A whole-cell recording obtained from an NMC-sensitive suprachiasmatic neurone, in current clamp. Application of 10 nM NMC to the brain sliceby superfusion resulted in an increase in neuronal firing rate which recovered within 10 min. The action potentials have been truncated by the cut-offfrequency imposed by the pen recorder and were in fact 55 mV in height. The increase in firing rate was not associated with any changes in membranepotential or neuronal input resistance. The latter neuronal parameter was assessed from the size of the electrotonic hyperpolarising potentials resulting fromthe injection of negative current into the recording electrode, for at least 500 ms to ensure complete charging of membrane capacitance. The lower part ofthe figure shows a ratemeter histogram recording which shows the number of action potential spikes detected in successive 10-s periods.

Fig. 2. Voltage-clamp data obtained from a single NMC-sensitive suprachiasmatic neurone. A: when voltage-clamped at y60 mV in the presence of 1 mMTTX, application of 1 mM NMC induced a 50-pA inward current which was associated with a decrease in membrane conductance. The decrease inmembrane conductance was denoted by the decrease in the size of the current pulses required to apply periodic voltage step excursions from y60 to y70

Ž . Ž . Ž .mV. B: construction of current–voltage IrV relationships in the absence open circles and at the peak of the inward current closed circles determinedŽ .where the concentration of potassium in the perfusing ACSF was 3.2 mM. C: new IrV relationships determined in the absence open circles and presence

Ž .closed circles of the peptide NMC following an increase in the concentration of potassium to 13.2 mM in the perfusing ACSF. D: the net leak currentinduced by NMC in both conditions. In control potassium conditions the NMC-induced inward current reversed at y99 mV. Following the increase in theextracellular concentration of potassium by 10 mM, the reversal potential was observed to shift to y64 mV. Thus the reversal of the NMC-induced currentwas observed to shift in the direction predicted by the Nernst equation.


Ž .Fig. 3. The peptide NMC 100 nM induced an inward current of 14.8"1.2 pA in 46 out of a total of 210 neurones which were voltage-clamped at y60mV, in the presence of TTX. Construction of IrV relationships in the absence and presence of NMC suggested that the NMC responses could besubdivided into 3 subgroups which are each illustrated by net leak IrV relationships determined from expermental data. A: in a minority of the responses

Ž . Ž .studied in voltage clamp i.e 6 out of 46 the NMC-induced inward current displayed a single reversal potential at y98.5"1.4 mV, ns6 which wasŽ . Žclose to E y98 mV . B: in the majority of responses, the NMC-induced inward current either reversed at potentials close to E y104.7"1.9 mV,K K

. Ž .ns12 or decreased with hyperpolarisation ns25 . There was also an additional reversal potential evident at more depolarised potentials at around y30Ž .mV. C: in even fewer responses ns3 the NMC-induced inward current increased with hyperpolarisation.

current of 14.8"1.2 pA in only 46 suprachiasmatic neu-rones. The NMC-induced inward currents were associatedwith at the most a 50% decrease basal membrane conduc-tance, manifested by a decrease in the current required to

Žapply periodic voltage steps from y60 to y70 mV Fig..2, Fig. 4 . The duration of the inward currents was be-

tween 5 to 15 min, which was comparable to the durationof the responses observed in current clamp. The size of theinward current was unaffected by the addition of 1 mM

Ž . Ž .TTX ns46 , 10 mM ouabain ns2 or 0.2 mM cad-Ž .mium chloride ns3 to the perfusing ACSF. However,

despite the clear excitatory responses evoked by 10 nM

Fig. 4. Voltage-clamp data obtained from a single NMC-sensitive neurone which was clamped at y60 mV in the presence of TTX. A: current traceshowing the 29-pA inward current induced by 100 nM NMC which was associated with a 50% decrease in basal membrane conductance. B: increasing theconcentration of potassium in the perfusing ACSF by 10 mM resulted in a decrease in the NMC-induced current to 10 pA. C: construction of IrV

Ž . Ž .relationships from the data displayed in part A, in the absence open circles and presence of NMC closed circles . D: construction of IrV relationshipsŽ .from the data in part B. E: the net leak current induced by NMC, where the concentration of potassium in the ACSF is 3.2 mM closed circles and 13.2

Ž .open circles .


NMC in current clamp, no inward currents could be de-tected when neurones were voltage-clamped.

Construction of IrV relationships between the voltagesy140 to y40 mV suggested the existence of 3 groups ofNMC responses. Fig. 3 displays examples of the net leakcurrent induced by NMC versus voltage from experimentaldata which is typical of each group. The net NMC-inducedleak current was calculated by the subtraction of thesteady-state currents measured in the absence and presenceof NMC at each command voltage. The leak IrV relation-ships determined from the minority of NMC responsesŽ .ns6 displayed a single reversal potential at y98.5"1.4

Ž .mV ns6 , which is close to the theoretical reversalpotential for potassium calculated from the Nernst equa-

Ž . Ž .tion E sy98 mV Fig. 3A . Furthermore, the NMC-K

induced inward current did not display any rectification orvoltage dependence. The slope conductance calculated forthe NMC-induced current in 5 of these 6 experiments,

Ž .using 100 nM NMC, was found to be 0.4"0.1 nS ns5 .These values were calculated from the slope of the IrVrelationships. Whilst in only 3 experiments, the NMC-in-duced current displayed a linear increase with membrane

Ž .hyperpolarisation Fig. 3C . The slope conductance of thisŽ .current was calculated as 0.4"0.2 nS ns3 . In the

Ž .majority of NMC responses ns37 the leak IrV rela-tionships displayed either a decrease in current with mem-

Ž . Ž .brane hyperpolarisation ns25 Fig. 4 or an actual

reversal of current at hyperpolarised potentials, i.e. atŽ . Ž .y104.7"1.9 mV ns12 Fig. 3B . A second reversal

potential was also observed at approximately y30 mV, inthe majority of the 37 responses to NMC.

Application of 10 mM baclofen by superfusion for 60 swas shown to inhibit the spontaneous firing of two neu-rones tested in current clamp. When 10 mM baclofen wasapplied to 6 neurones voltage-clamped at y60 mV, an

Ž .outward current of 25.3"10.0 pA ns6 was observed,which was associated with an increase in membrane con-

Ž .ductance Fig. 5A,C,E . The outward current was unaf-fected by the addition of 1 mM TTX. Construction of IrVrelationships in the absence and presence of baclofenrevealed a single reversal potential at y105.0"3.9 mVŽ . Ž .ns6 which is close to E y98 mV .K

3.4. EÕidence for NMC decreasing a potassium conduc-tance

Leak IrV relationships constructed from 18 of the 46NMC responses studied in voltage clamp displayed rever-sal potentials which were close to E . In a further 25 ofK

the 46 responses there was evidence of a decrease in thesize of the NMC-induced inward current with membranehyperpolarisation. The reversal of the NMC-induced cur-rent to the outward direction at membrane potentials hy-perpolarised to E , suggests that NMC decreases a potas-K

Ž .Fig. 5. Voltage-clamp data obtained from the same NMC-sensitive neurone displayed in Fig. 4. where baclofen 10 mM induced a suppression of neuronalfiring rate. A: the current trace shows that 60 s superfusion with baclofen induced an outward current of 22 pA at y60 mV. B: increasing the concentrationof potassium in the perfusing ACSF by 10 mM resulted in baclofen inducing an inward current of 14 pA. C: construction of the IrV relationships

Ž . Ž .determined in part A, in the absence open circles and presence closed circles of baclofen. D: construction of the IrV relationships determined in part B,Ž . Ž .in the absence open circles and presence closed circles of baclofen. E: the net leak current induced by baclofen, where the concentration of potassium in

Ž . Ž .the ACSF is 3.2 mM closed circles and 13.2 mM open circles .


sium conductance. To further investigate this possibility,the extracellular concentration of potassium was increasedby 10 mM from 3.2 mM to 13.2 mM in 3 NMC-sensitiveneurones. The leak IrV relationships determined in con-trol potassium conditions, where the extracellular concen-tration of potassium was 3.2 mM, displayed a reversal

Ž .potential at y88.3"12.7 mV ns3 . Increasing theextracellular concentration of potassium to 13.2 mM pro-duced a depolarising shift in the reversal potential ob-served for the new leak IrV relationships to y65.3"9.8

Ž .mV ns3 . Furthermore, the reversal potential was shiftedin the direction predicted by the Nernst equation, whereE would be expected to shift from y98 to y60 mV.K

Nevertheless, the magnitude of shift in the mean reversalpotential for all 3 neurones was clearly smaller than thatpredicted by the Nernst equation. Fig. 2 illustrates the IrVrelationships constructed from one of these experimentswhere the reversal potential for the NMC-induced inwardcurrent was shifted in the depolarising direction by 35 mV,from y99 mV to y64 mV.

In an another neurone which was representative of themajority of NMC-sensitive neurones, NMC-induced an

Fig. 6. Intracellular dialysis of neurones with a cesium gluconate-basedpipette solution whilst perfusing with potassium-free ACSF was unable tocompletely abolish the NMC-induced current in voltage clamp. A: an

Žexample of the Ir V relationships constructed in the absence open. Ž .circles and presence of NMC closed circles from 1 of 8 experiments.

B: the net leak current induced by NMC in potassium-free conditionstypically displayed an increase with membrane hyperpolarisation.

Fig. 7. Whilst recording from an NMC-sensitive neurone in potassium-freeconditions, the ACSF was further modified such that the concentration ofsodium was reduced from 151 to 25 mM, replacing with equimolarcholine chloride plus the addition of 1 mM atropine. However, partialreplacement of sodium was unable to block the remaining NMC-inducedcurrent. A: construction of Ir V relationships from 1 of 2 neurones, in

Ž . Ž .the absence open circles and presence closed circles of NMC. B: thenet leak current induced by NMC increases with membrane hyperpolari-sation in an apparently voltage-independent manner. Thus the inwardcurrent remaining after replacement of potassium appears to result fromthe activation of a non-specific conductance.

inward current which decreased with hyperpolarisation butno actual reversal of current was observed where the

Žextracellular concentration of potassium was 3.2 mM Fig..4A,C,E . Increasing the extracellular concentration of

potassium by 10 mM resulted in the decrease in the size ofthe inward current measured at y60 mV, from 29 to 10pA, and the appearance of a reversal potential at y104

Ž .mV Fig. 4B,D,E . Application of baclofen to the sameneurone induced an outward current of 22 pA at y60 mVwhich reversed to the inward direction at y107 mV.When the extracellular concentration of potassium wasincreased by 10 mM, baclofen induced an inward currentof 14 pA at y60 mV which reversed direction at y44 mVŽ .Fig. 5 .

3.5. Reduction of the extracellular concentration of sodium

In a total of 5 experiments, the concentration of sodiumin the perfusing ACSF was reduced from 151 to 25 mM,


Ž .being replaced with equimolar choline ns3 or N-Ž .methyl-D-glucamine ns2 . In control sodium conditions,

application of 100 nM NMC induced an inward current aty60 mV of 12.0"2.3 pA in 3 NMC-sensitive neurones

Ž .and a mean inward current of 8.5 pA 8 and 9 pA in twoother neurones. Reapplication of NMC following the re-placement of 126 mM sodium chloride in the perfusing

Ž .ACSF with choline chloride plus 1 mM atropine waswithout any notable affect on the size of the inward current

Ž .at y60 mV i.e. 11.1"2.1 pA, ns3 . Similarly withN-methyl-D-glucamine substitution, there was no notableaffect on the size of the inward current measured at y60

Ž Ž ..mV i.e. mean inward current of 6.5 pA 6 and 7 pA .Using the paired Student’s t-test to compare the two setsof data showed that no statistically significant differencescould be detected between the NMC-induced inward cur-rents measured in either control or sodium substituted

Ž .conditions P)0.05, 95% confidence limits, ns5 .

3.6. Replacement of potassium in the intracellular andextracellular solutions

Whole-cell recordings were made from 8 NMC-sensi-tive neurones using the cesium gluconate-based intra-cellular solution whilst perfusing the slice with cesium-substituted ACSF. Following the initial stabilisation of the

Ž .voltage clamp at y60 mV, NMC 100 nM was appliedwhilst perfusing with control potassium-containing ACSFand shown to induce an inward current of 12.2"1.4 pAŽ .ns8 . After allowing for the recovery of the initial

response to the peptide, the perfusing ACSF was changedto a potassium-free solution. The change in composition ofthe perfusing solution was generally made after periods ofbetween at least 30 to 45 min following the time of initialpatch rupture, to obtain the whole-cell configuration. Neu-ronal input resistances measured in potassium-substitutedconditions were higher than those measured using a potas-

Žsium gluconate-based pipette solution i.e. 1154.6"140.6.MV , ns8 versus 639.4"32.9 MV , ns46 . These

differences were shown to be statistically significant usingŽ .the Student’s t-test unpaired , where P-0.0001 using

95% confidence limits. Reapplication of NMC, duringperfusion with potassium-free ACSF, induced an inward

Ž .current of 8.6"1.3 pA ns8 which was not associatedwith any obvious decreases in membrane conductance. Acomparison of the magnitude of the NMC-induced inward

Žcurrent in potassium-free to control conditions i.e. 14.8"

. Ž .1.2 pA, ns46 , using the Student’s t-test unpaired ,showed that removal of potassium significantly reducedthe size of the inward current induced by NMC at y60

Ž .mV P-0.05, 95% confidence limits . Furthermore, con-struction of IrV relationships revealed that in potassium-free conditions there was no evidence for a reversal of theNMC-induced current at hyperpolarised potentials. In factthe leak IrV relationships typically displayed an increasein the size of the inward current with membrane hyperpo-

Ž .larisation Fig. 6 . In 4 of the 8 whole-cell recordingsmade in potassium-free conditions, the extracellular con-centration of sodium was further reduced to 25 mM by

Ž .replacing with choline chloride plus 1 mM atropine . In 2

Fig. 8. In two other experiments performed in potassium-free conditions, the further partial replacement of sodium in the ACSF with choline resulted in areversible block of the NMC-induced current at all voltages from y140 to y40 mV. A: current trace from one of these experiments showing that inpotassium-free conditions NMC induced an inward current of 16 pA which was not associated with any notable changes in membrane conductance. B:

Ž .partial replacement of sodium in the potassium-free ACSF with choline plus atropine completely blocked the inward current. C: the tissue slice wassubsequently washed with potassium-free ACSF containing 151 mM sodium for 30 min. Reapplication of NMC showed a recovery of the response.


Ž .of 4 of these experiments, NMC 100 nM induced a meanŽ .inward current of 10.8 pA 13.3 and 8.3 pA whilst

perfusing with potassium-free ACSF, with a sodium con-centration of 151 mM. The additional partial replacementof extracellular sodium chloride reduced the size of the

Ž .mean inward current to 5.7 pA 5.0 and 6.3 pA . The leakIrV relationships constructed in potassium-free and lowsodium conditions showed that the NMC-induced inwardcurrent increased with hyperpolarisation, apparently in a

Ž .voltage-independent fashion Fig. 7 . Addition of 0.2 mMŽ .cadmium chloride ns2 was without effect on either the

size of the inward current measured at y60 mV or theappearance of the leak IrV relationship. The mean rever-sal potential for the resulting conductance was found to be

Ž .y10.7 mV y2.9 and y18.5 mV . These values weredetermined either directly from the IrV relationship or bylinear regression. However, in the other two experimentsperformed in potassium-free conditions, the additional par-tial replacement of extracellular sodium with choline re-sulted in the reversible block of the inward current at all

Ž .Ž .voltages examined i.e. y140 to y40 mV Fig. 8 .

3.7. Lack of eÕidence to support NMC modulating achloride conductance

Application of the GABA agonist muscimol to a totalA

of 14 suprachiasmatic neurones evoked small outwardcurrents of 8.6"3.6 pA, in every neurone tested, whichwere associated with an increase in membrane conduc-

Ž .tance Fig. 9 . The muscimol-induced outward currentswere unaffected by the addition of 1 mM TTX. Construc-tion of IrV relationships in the absence and presence ofmuscimol revealed a single reversal potential at y57.9"

Ž .2.2 mV ns14 which is close to the theoretical reversalŽ .potential for chloride E sy52 mV . Both the theoreti-Cl

cal and experimentally derived values for E clearly differCl

from the reversal potential observed for the ionic conduc-tance remaining in potassium-free, low sodium conditions,whose mean reversal potential was y10.7 mV.

3.8. Effects of NMC on T-type calcium currents and Ih

Only 16 out of 46 of the NMC-sensitive neuronesexpressed a slow inward current activated by hyperpo-

Fig. 9. Voltage-clamp data obtained from a suprachiasmatic neurone which was clamped at y60 mV. A: current trace showed that application of muscimolŽ . Ž .10 mM for 60 s induced an increase in basal membrane conductance. B: construction of IrV relationships in the absence open circles and presenceŽ .closed circles of muscimol. C: the net leak current induced by muscimol displayed a single reversal potential at y54 mV, which is close to E .Cl


Fig. 10. A comparison of the hyperpolarisation-activated cation current or I in the absence and presence of NMC. The magnitude, kinetics and voltages ath

which this current was first activated, were unaffected by NMC. Voltage-clamp data obtained from a single suprachiasmatic neurone. A series of voltagestep excursions were applied from a holding potential of y60 mV to y130 mV. The corresponding current steps demonstrate the development of a slow

Ž .inward current termed I , at potentials negative to y80 mV, which reaches steady-state within 200 ms. The magnitude of inward rectification washŽ .calculated by substraction of current measured after the settling of the capacitance transient I from that measured during the last 500 ms of the stepins

Ž .I . A transient inward current which peaked within 20 ms of onset, is also observed following the return to y60 mV from hyperpolarising voltage stepsss

to at least y90 mV. This transient current is referred to as the T-type calcium current.

larised potentials, also referred to as I . Of the 16h

suprachiasmatic neurones expressing I , no differences inh

either the magnitude, kinetics or voltages at which inwardrectification was first apparent, were detected in the pres-

Ž .ence of NMC Fig. 10 . As stated earlier, 11 of the 16neurones which expressed I also exhibited T-type cal-h

cium currents. No change in either the kinetics or voltagesensitivity were observed for this calcium current in thepresence of NMC. Furthermore, NMC did not increase theproportion of neurones exhibiting the T-type calcium cur-rent.

3.9. Intracellular dialysis with GTPgS

A total of 14 suprachiasmatic neurones were dialysedwith an intracellular pipette solution containing the non-hydrolysable GTP analogue, GTPg S. No obvious changesin either the baseline holding current or basal membraneconductance were observed following the initial stabilisa-

Ž .tion of the voltage clamp ns14 . Perfusion of NMCŽ .100 nM induced an inward current of 14.0"4.1 pA aty60 mV, in 4 of the 14 neurones. The inward current wasassociated with a decrease in membrane conductance. In

Fig. 11. Current trace from a suprachiasmatic neurone which was voltage-clamped at y60 mV and was dialysed with intracellular pipette solutionŽ .containing the non-hydrolysable GTP analogue, GTPg S. Perfusion of NMC 100 nM for 60 s induced an inward current of 26 pA which was associated

with a decrease in membrane conductance. Furthermore, both the inward current and decrease in membrane conductance were shown to persist for theŽ .duration of the recording i.e. 60 min . Reapplication of NMC was without further effect.


all 4 neurones both the peptide-induced inward current anddecrease in membrane conductance were shown to persistfor the entire length of the recordings, i.e. between 60 to

Ž . Ž .90 min Fig. 11 . Reapplication of NMC 100 nM follow-ing stabilisation of the inward current was without further

Ž .effect ns2 . As a control, application of the ligand-gatedŽ .ion channel agonist muscimol 10 mM was shown to

induce a reversible increase in membrane conductanceŽ .ns4 .

4. Discussion

Using the whole-cell recording technique we havedemonstrated that nanomolar concentrations of NMC canevoke excitatory responses from a subpopulation ofsuprachiasmatic neurones. A comparison of the presentcurrent clamp recordings of the potency, magnitude andduration of the responses to the bombesin related peptides,

w xwith our previously reported study 30 , suggests that theresponses to NMC were little affected by the invasivenature of the whole-cell recording technique. The mostnotable feature of these current clamp recordings made inwhole cell configuration was the lack of consistent changesin neuronal input resistance associated with the large in-creases in neuronal firing rate evoked by NMC. Such anobservation could result either from modulation of an ionicpump or a combination of at least two ionic conductances.However, the excitatory responses evoked by 100 nM

ŽNMC, equivalent to the submaximal concentration inw xprevious studies 30 , were associated with a relatively

Ž .small membrane depolarisation 9.8"3.4 mV, ns4 andŽincrease in basal neuronal input resistance 25.0 "

.9.9%,M V ns4 . This finding is more consistent withmodulation of at least two ionic conductances than either asingle conductance or a pump.

When the NMC-sensitive neurones were voltage-clamped at y60 mV, 100 nM NMC was found to cause aninward current which was of similar duration to the depo-larization and increase in firing observed in current clamp.However, no inward currents could be detected upon appli-cation of 10 nM NMC, despite the excitatory responsesobserved in current clamp. This finding is probably aconsequence of the very small currents required to producea 1 to 3 mV depolarization and consequent large increasein firing rate. Using Ohm’s law, the mean inward current

Ž .induced by 100 nM NMC i.e. 14.8 pA in an averagesuprachiasmatic neurone, with an input resistance of 600MV , would be expected to cause a 9 mV depolarisation ofmembrane potential. This value compares well with the

Žactual depolarisation observed in current clamp i.e. 9.8"

.3.4 mV, ns4 . Hence the slow inward currents measuredin voltage clamp when 100 nM NMC was applied weresufficient to evoke the excitatory responses observed incurrent clamp. The peptide-induced current required toproduce a 3 mV depolarization would be barely detectablefrom the baseline holding current.

Addition of TTX to the perfusing ACSF was withouteffect on size of the inward current evoked by 100 nMNMC. This evidence and the persistence of inward currentduring exposure to cadmium supports the view that NMCdepolarises suprachiasmatic neurones via a direct post-synaptic action, rather than an indirect modulation ofneurotransmission. Modulation of neurotransmission couldbe achieved by either a presynaptic action or coagonistaction with other neuropeptides. This is in contrast to anearlier extracellular study in which co-application of thepeptides vasoactive intestinal peptide and peptide histidineisoleucine along with GRP was necessary to evoke excita-

w xtory responses 1 .Construction of IrV relationships before and during a

response to NMC showed that in the majority of neuronesthe NMC-induced current decreased at hyperpolarised po-tentials and in some cases reversed to the outward direc-tion at potentials close to E . In some of these experi-K

ments, increasing the extracellular concentration of potas-sium produced a shift of the reversal potentials in thedepolarising direction. Furthermore, whilst perfusing theslice with nominally potassium-free ACSF, the NMC-in-duced current increased as opposed to decreased withmembrane hyperpolarisation. This evidence suggests that adecrease in a potassium conductance underlies most of theresponses to NMC. This shows similarities to a previousintracellular recording study of dorsal raphe neurones, thatdemonstrated NMB receptor coupling to a decrease in a

w xpotassium conductance 31 .The IrV relationships determined from the majority of

NMC responses studied in voltage clamp also displayedanother reversal potential closer to 0 mV. This observationsuggested the involvement of an additional ionic conduc-tance. Although artefacts due to poor voltage clamp maycause inaccurate determination of reversal potentials, inad-equate voltage clamp at distal dendrites would be expectedto shift the reversal of the NMC-induced current to morenegative potentials rather than the positive potentials thatwe observed in our recordings. Furthermore, intracellulardialysis of neurones with a solution in which cesiumreplaced potassium would effectively increase the lengthconstant, hence improving the voltage clamp at distalsynapses. Thus the fact that NMC was still able to inducean inward current in potassium-free conditions, suggeststhat this additional ionic conductance did not arise frompoor voltage clamp of potassium events.

In support of this proposal the GABA agonist ba-B

clofen demonstrated a direct postsynaptic inhibitory actionon neurones within the SCN, the baclofen-induced outwardcurrent was shown to reverse at potentials close to E .K

Within the dorsal raphe baclofen has similar effects byw xactivating a potassium conductance 42 . Increasing the

extracellular concentration of potassium resulted in a shiftin the reversal potential in the depolarising direction. Thusthe data in this present study is consistent with GABA B

receptors coupling to the increase in a potassium conduc-


tance. However, the demonstration of the existence ofGABA receptors on postsynaptic membranes is in con-B

trast to a recent report in which only a presynaptic site ofaction of baclofen to reduce excitatory neurotransmissionbetween the retinohypothalamic tract and neurones within

w xthe ventral SCN could be demonstrated 15 . In the presentstudy, recordings were made in the more dorsal aspects ofthe nucleus where GABA receptors appear to be locatedB

on postsynaptic neuronal membranes, mediating suppres-sion of neuronal firing. The close agreement between thebaclofen reversal potentials and E also supports the caseK

that the neurones were adequately voltage-clamped.An alternative explanation for the leak IrV relation-

ships obtained for NMC is the modulation of an ionicw xpump 9 . This suggestion is also consistent with the

failure to observe actual reversals of the NMC-inducedinward current in the majority of the neurones. However,addition of ouabain, an inhibitor of the sodium–potassiumŽ .Na–K ATPase pump, was without effect on the NMC-in-duced current at y60 mV. Removal of potassium from theperfusing ACSF was unable to completely abolish theNMC-induced inward current which suggests that theNaq–Kq ATPase pump does not make a major contribu-tion to the excitatory action of NMC. Furthermore, in-creases in the extracellular concentration of potassiumgenerally reduced the inward current recorded at y60 mVwhich is also inconsistent with the involvement of theelectrogenic pump. This evidence is consistent with NMCmodulating two ionic conductances.

In some experiments partial replacement of sodium inthe perfusing ACSF with choline or N-methyl-D-glucaminewas without consistent effects on the additional ionicevent, in either normal or potassium-free conditions.Therefore the additional ionic conductance modulated byNMC does not appear to predominantly utilise sodium as acharge carrier in these cases. The mean reversal potentialmeasured for the additional conductance in potassium-freeconditions was found to be approximately y10.7 mV.This value clearly differs from the chloride reversal poten-tial determined by applying the GABA agonist musci-A

mol. Hence, NMC appears to activate a non-specific con-ductance as opposed to a chloride conductance. Neverthe-less the permeability of this non-specific conductance hasyet to be established. The addition of 0.2 mM cadmiumwas without effect on this non-specific conductance, whichsuggests that entry of calcium via voltage-activated cal-cium channels was not required for activation. The non-specific conductance was equally unaffected by cholineversus N-methyl-D-glucamine substitution for sodium. Thissuggests that the non-specific conductance activated by

Ž .NMC is not a calcium-dependent non-specific CANcation conductance. Studies of CAN currents in a range oftissues have suggested that these ion channels are perme-able to many large organic cations, with the exception of

w xN-methyl-D-glucamine 4,29 .In contrast, partial replacement of extracellular sodium

in the remaining potassium-free experiments reversiblyabolished the inward current, indicating that the conduc-tance used sodium. However, in control potassium condi-tions, partial replacement of extracellular sodium waswithout effect on the inward current or the appearance ofthe IrV relationship. Therefore NMC appears to increasean ionic conductance which can be primarily carried bysodium ions in only a minority of cases. This finding issimilar to the action of substance P in the locus coeruleus,where partial replacement of extracellular sodium was

w xshown to unmask the potassium event 35 . However, inthe present study when potassium on both sides of themembrane was at physiological concentrations, reductionof extracellular sodium did not unmask the potassiumevent modulated by NMC. Other workers have reported agreater sodium dependence for cation conductances acti-vated by G-protein coupled receptors in other CNS neu-

w xrones than we have here 13,16,23,28,33,35,37 .A single reversal potential was observed close to E inK

a minority of the responses to NMC studied in voltageclamp. The NMC-induced current in these experimentsdisplayed little voltage dependence y140 and y40 mV.Furthermore, NMC did not appear to affect the hyperpolar-isation-activated potassium conductance or I . Addition ofh

0.2 mM cadmium was without effect on either the size ofthe inward current or appearance of the IrV relationship,which suggests that the potassium conductance is notcalcium-dependent. Therefore the peptide NMC may besuppressing a voltage-independent leak potassium conduc-tance. A number of other G-protein-linked receptors suchas metabotropic glutamate and a -adrenergic receptors,1

have also been reported to modify a voltage-independentleak potassium conductance which could not be adequatelycharacterised using potassium channel toxins and blockersw x14,28 .

The evidence from voltage-clamp data suggests thatNMC decreases a potassium conductance and increases anon-specific conductance. The varying contribution of eachconductance to the NMC responses in different neuronesgave rise to different IrV relationships that were encoun-tered under voltage-clamp conditions. There was no evi-dence of any circadian variation in these curves that wouldexplain the range of either leak IrV relationships or NMCcurrents observed.

The non-hydrolysable GTP analogue, GTPg S, is knownw xto irreversibly activate G-proteins 7 . Intracellular dialysis

with intracellular pipette solution containing GTPg S, re-sulted in NMC-inducing irreversible inward currents, thusconfirming the existence of a G-protein link betweenreceptor and ionic events. This finding is consistent withpredictions made from the sequence analysis of both of the

w xrecently cloned bombesin receptor subtypes 6,41 . Elec-trophysiological studies of other G-protein-linked peptidereceptors such as substance P in the locus coeruleus andneurotensin in the ventral tegmental area, have demon-strated similar ionic mechanisms to those in the present


w xstudy for NMC in the SCN 16,36 . The general patternappears to consist of coupling to a decrease in a potassiumconductance and the increase in an additional conductance,either non-specific or sodium-selective.

Biochemical studies in peripheral tissues such as pan-creatic acini, cell lines such as murine Swiss 3T3 fibroblastcells and brain slices suggest that activation of bombesinreceptors results in hydrolysis of membrane phosphatidyli-nositol with the subsequent generation of inositol trisphos-

w xphate and diacylglycerol 34 . Activation of protein kinaseC by the lipophilic second messenger diacylglycerol, hasbeen proposed to mediate the suppression of an inwardrectifying potassium conductance induced by application

w xof substance P to neurones from the nucleus basalis 39 .Modulation of the two potassium conductances by a -1

adrenoreceptors in neurones of the dorsal raphe does notappear to involve protein kinase C. In contrast to the actionof substance P in the nucleus basalis, a -adrenoreceptor-1

mediated activation of protein kinase C appears to lead tow xreceptor desensitisation 28 . However, the existence and

relevance of such intracellular pathways in mediating theeffects of NMC in the SCN remains to be established.

In conclusion, NMC has been shown to excite a sub-population of suprachiasmatic neurones by exerting a di-rect depolarising influence. The neuronal excitation ap-pears to be the result of a decrease in a resting potassiumconductance and increase in a non-specific conductance,with varying contributions of each conductance. Further-more, there is evidence of a G-protein link between recep-tor activation and the ionic events.

References

w x1 Albers, H.E., Liou, S.-Y, Stopa, E.G. and Zoeller, R.T., Interactionof colocalized neuropeptides: functional significance in the circadian

Ž .timing system, J. Neurosci., 11 1991 846–851.w x2 Anastasi, A., Erspamer, V. and Bucci, M., Isolation and structure of

bombesin and alytesin, two analogous active peptides from the skinof the European amphibians Bombina and Alytes, Experientia, 27Ž .1970 166–167.

w x3 Babcock, A.M., Baker, D.E. and Moody, T.W., Bombesin-inducedhypothermia: a dose–response and receptor antagonist study, Phar-

Ž .macol. Biochem. BehaÕ., 43 1992 957–959.w x4 Bal, T. and McCormick, D.A., Mechanisms of oscillatory activity in

guinea-pig nucleus reticularis thalami in vitro: a mammalian pace-Ž .maker, J. Physiol., 468 1993 669–691.

w x5 Battey, J.F. and Wada, E., Two distinct receptor subtypes forŽ .mammalian bombesin-like peptides, Trends Neurosci., 14 1991

524–528.w x6 Battey, J.F., Way, J.M., Corjay, M.H., Shapira, H., Kusano, K.,

Harkins, R., Wu, J.M., Slattery, T., Mann, E. and Feldman, R.I.,Molecular cloning of the bombesinrgastrin-releasing peptide recep-

Ž .tor from Swiss 3T3 cells, Proc. Natl. Acad. Sci. USA, 88 1991395–399.

w x7 Brown, D.A., G-proteins and potassium currents in neurones, Annu.Ž .ReÕ. Physiol., 52 1990 215–242.

w x8 Brown, M.R., Allen, R., Villareal, J., Rivier, J. and Vale, W.,Bombesin: central nervous system actions to affect the autonomic

Ž .nervous system, Ann. NY Acad. Sci., 547 1988 174–182.

w x9 De Weer, P., Voltage dependence of the Na–K pump, Annu. ReÕ.Ž .Physiol., 50 1988 225–241.

w x10 Erspamer, V., Discovery, isolation and characterization of theŽ .bombesin-like peptides, Ann. NY Acad. Sci. USA, 547 1988 3–9.

w x11 Erspamer, V., Erspamer, G.F., Inselvini, M. and Negri, L., Occur-rence of bombesin and alytesin in extracts of the skin of threeEuropean discoglossid frogs and pharmacological actions ofbombesin on extravascular smooth muscle, Br. J. Pharmacol., 45Ž .1972 333–348.

w x12 Erspamer, V., Melchiorri, P. and Sopranzi, N., The action ofbombesin on the systemic arterial blood pressure of some experi-

Ž .mental animals, Br. J. Pharmacol., 45 1972 442–450.w x13 Farkas, R.H., Nakajima, S., Nakajima, Y., Neurotensin excites basal

forebrain cholinergic neurons: ionic and signal–transduction mecha-Ž .nisms. Proc. Natl. Acad. Sci. USA, 91 1994 2853–7.

w x14 Guerineau, N.C., Bossu, J.-L., Gahwiler, B.H. and Gerber, U.,´Activation of a non-selective cationic conductance by metabotropicglutamatergic and muscarinic agonists in CA3 pyramidal neurones

Ž .of the rat hippocampus, J. Neurosci., 15 1995 4395–4407.w x15 Jiang, Z.-G., Allen, C.N. and North, R.A., Presynaptic inhibition by

baclofen of retinohypothalamic excitatory synaptic transmission inŽ .rat suprachiasmatic nucleus, Neuroscience, 64 1995 813–819.

w x16 Jiang, Z.-G., Pessia, M. and North, R.A., Neurotensin excitation ofŽ .rat ventral tegmental neurones, J. Physiol., 474 1994 119–129.

w x17 Llinas, R. and Yarom, Y., Electrophysiology of mammalian inferior´olivary neurones in vitro. Different types of voltage-dependent ionic

Ž .conductances, J. Physiol., 315 1981 549–567.w x18 Matozaki, T., Zhu, W.-Y., Tsunoda, Y., Goke, B. and Williams,

J.A., Intracellular mediators of bombesin action on rat pancreaticŽ .acinar cells, Am. J. Physiol., 260 1991 G858–G864.

w x19 McDonald, T.J., Jornvall, H., Nisson, G., Vagne, M., Ghatei, M.,Bloom, S.R. and Mutt, V., Characterization of a gastrin-releasingpeptide from porcine non-antral gastric tissue, Biochem. Biophys.

Ž .Res. Commun., 90 1979 227–233.w x20 Meijer, J.H. and Rietveld, W.J., Neurophysiology of the suprachias-

Ž .matic circadian pacemaker in rodents, Physiol. ReÕ., 69 1989671–707.

w x21 Merali, Z. and Piggins, H., Effects of dopamine D1 and D2 receptoragonists and antagonists on bombesin-induced behaviours, Eur. J.

Ž .Pharmacol., 191 1990 281–293.w x22 Mercuri, N.B., Bonci, A., Calabresi, P., Stefani, A. and Bernardi, G.,

Properties of the hyperpolarisation-activated cation current I in rathŽ .midbrain dopaminergic neurones, Eur. J. Neurosci., 7 1995 462–

469.w x23 Mercuri, N.B. Stratta, F. Calabresi, P. Bonci, A. Bernardi, G.,

Activation of metabotropic glutamate receptors induces an inwardcurrent in rat dopamine mesencephalic neurons, Neuroscience, 56Ž .1993 399–407.

w x24 Mikkelsen, J.D., Larsen, P.J., O’Hare, M.T. and Wiegand, S.J.,Gastrin-releasing peptide in the rat suprachiasmatic nucleus: animmunohistochemical, chromatographic and radioimmunological

Ž .study, Neuroscience, 40 1991 55–66.w x25 Minamino, N., Kangawa, K. and Matsuo, H., Neuromedin B: a

novel bombesin-like peptide identified in porcine spinal cord,Ž .Biochem. Biophys. Res. Commun., 114 1983 541–548.

w x26 Minamino, N., Kangawa, K. and Matsuo, H., Neuromedin C: abombesin-like peptide identified in porcine spinal cord, Biochem.

Ž .Biophys. Res. Commun., 119 1984 14–20.w x27 Moore, R.Y. and Eichler, V.B., Loss of circadian adrenal cortico-

sterone rhythm following suprachiasmatic nucleus lesions in the rat,Ž .Brain Res., 42 1972 201–206.

w x28 Pan, Z.Z., Grudt, T.J. and Williams, J.T., a -Adrenoceptors in rat1

dorsal raphe neurones: regulation of two potassium conductances, J.Ž .Physiol., 478 1994 437–447.

w x29 Partridge, L.D. and Swandulla, D., Calcium-activated non-specificŽ .cation channels, Trends Neurosci., 11 1988 69–72.

w x30 Pinnock, R.D., Reynolds, T. and Woodruff, G.N., Different types of


bombesin receptors on neurons in the dorsal raphe nucleus and therostral hypothalamus in rat brain slices in vitro, Brain Res., 653Ž .1994 199–124.

w x31 Pinnock, R.D. and Woodruff, G.N., Bombesin excites a subpopula-tion of 5-hydroxytryptamine-sensitive neurones in the rat dorsal

Ž .raphe nucleus in vitro, J. Physiol., 440 1991 55–65.w x32 Prasad, V.N.H. and Moody, T.W., Bombesin-like peptides stimulate

Ž .phosphatidylinositol turnover in rat brain slices, Peptides, 9 19891345–1349.

w x33 Raggenbass, M. Goumaz, M. Sermasi, E. Tribollet, E. Dreifuss, J.J.,Vasopressin generates a persistent voltage-dependent sodium current

Ž .in a mammalian motoneuron. J. Neurosci., 11 1991 1609–16.w x34 Rozengurt, E., Bombesin-induction of cell proliferation in 3T3 cells,

Ž .Ann. NY Acad. Sci., 547 1988 277–292.w x35 Shen, K.Z., North, R.A., Muscarine increases cation conductance

and decreases potassium conductance in rat locus coeruleus neu-Ž .rones. J. Physiol., 455 1992 471–85.

w x36 Shen, K.-Z. and North, R.A., Substance P opens cation channels andcloses potassium channels in rat locus coeruleus neurones, Neuro-

Ž . Ž .science, 50 2 1992 345–353.w x37 Spigelman, I., Puil, E., Ionic mechanism of substance P actions on

Ž .neurons in trigeminal root ganglia, J. Neurophysiol., 64 1990273–81.

w x38 Stephan, F.K. and Zucker, I., Circadian rhythms in drinking be-haviour and locomotor activity of rats are eliminated by hypothala-

Ž .mic lesions, Proc. Natl. Acad. Sci. USA, 69 1972 1583–1586.w x39 Takano, K., Stanfield, P.R., Nakajima, S. and Nakajima, Y., Protein

kinase C-mediated inhibition of an inward rectifier potassium chan-Ž .nel by substance P in nucleus basalis neurones, Neuron, 14 1995

999–1008.w x40 Wada, E., Way, J., Lebacq-Verheyden, A.M. and Battey, J., Neu-

romedin B and gastrin-releasing peptide mRNAs are differentiallyŽ .distributed in the rat nervous system, J. Neurosci., 10 1990

2917–2930.w x41 Wada, E., Way, J., Shapira, H., Kusano, K., Lebacq-verheyden,

A.M., Coy, D., Jensen, R. and Battey, J., cDNA cloning, characteri-zation and brain region-specific expression of a neuromedin-B-pre-

Ž .ferring bombesin receptor, Neuron, 6 1991 421–430.w x42 Williams, J.T., Colmers, W.F. and Pan, Z.Z., Voltage and ligand-

activated inwardly rectifying currents in dorsal raphe neurones inŽ .vitro, J. Neurosci., 8 1988 3499–3506.

neuromedin c decreases potassium conductance and increases a non-specific conductance in rat...

Documents