190, 55–77 (1994) 55 printed in great britain © the ...(lp) neuron. these effects were blocked by...

55J. exp. Biol. 190, 55–77 (1994)Printed in Great Britain © The Company of Biologists Limited 1994

MULTIPLE RECEPTORS MEDIATE THE MODULATORYEFFECTS OF SEROTONERGIC NEURONS IN A SMALL

NEURAL NETWORK

BING ZHANG AND RONALD M. HARRIS-WARRICK*

Section of Neurobiology and Behavior, Seeley G. Mudd Hall, Cornell University, Ithaca,NY 14853, USA

Accepted 7 February 1994

SummaryThe gastropyloric receptor (GPR) cells are a set of cholinergic/serotonergic

mechanosensory neurons that modulate the activity of neural networks in the crabstomatogastric ganglion (STG). Stimulation of these cells evokes a variety of slowmodulatory responses in different STG neurons that are mimicked by exogenouslyapplied serotonin (5-HT); these responses include tonic inhibition, tonic excitation andinduction of rhythmic bursting. We used pharmacological agonists and antagonists toshow that these three classes of modulatory response in the STG neurons are mediated bydistinct 5-HT receptor subtypes. GPR stimulation or application of 5-HT or 2-me-5HT(a vertebrate 5-HT3 agonist) inhibited the pyloric constrictor (PY) neurons; these actionswere selectively antagonized by gramine. GPR stimulation or application of 5-HTinduced rhythmic bursting in the electrically coupled anterior burster (AB) and pyloricdilator (PD) neurons; these effects were antagonized by the 5-HT1c/2 antagonistcinanserin and by atropine at concentrations that do not block muscarinic cholinergicreceptors in the crab STG. The 5-HT agonists 5-CT (5-HT1) and a-me-5HT (5-HT2) alsoinduced AB/PD bursting, which was blocked by cinanserin, but not by atropine. GPRstimulation or application of 5-HT and 5-CT evoked tonic excitation of the lateral pyloric(LP) neuron. These effects were blocked by cinanserin. Several other 5-HT agonists andnearly all the vertebrate 5-HT antagonists we tested had little or no effect on the crabpyloric 5-HT receptors. These results provide further evidence that the modulatorysensory GPR neuron uses serotonin to evoke multiple modulatory responses via multiple5-HT receptors. However, the 5-HT receptors in the crab STG neurons are notpharmacologically similar to vertebrate 5-HT receptors.

Introduction

Serotonin, or 5-hydroxytryptamine (5-HT), is known to be a neurotransmitter in bothvertebrate and invertebrate nervous systems (Amin et al. 1954; Brodie and Shore, 1957;

*To whom reprint requests should be addressed.

Key words: serotonin, 5-HT receptor, agonist, antagonist, stomatogastric ganglion, pyloric, neuralnetwork, crab, Cancer borealis.

Welsh and Moorehead, 1960; Maynert et al. 1964; Gerschenfeld and Paupardin-Tritsch,1974a,b; Beltz et al. 1984; Katz et al. 1989; Katz and Harris-Warrick, 1989, 1990a). Ourknowledge of the mechanisms controlling the action of 5-HT has been enhanced throughstudies of its receptor types using selective ligands, especially in the vertebrates. SinceGaddum and Picarelli (1957) first proposed two kinds of 5-HT receptor, studies invertebrate tissues have demonstrated at least four major 5-HT receptor types and anumber of subtypes (Bradley et al. 1986; Schmidt and Peroutka, 1989; Andrade andChaput, 1991a). It is now generally believed in vertebrates that different excitatory orinhibitory responses are mediated by distinct 5-HT receptors (Aghajanian, 1981; Bobkerand Williams, 1990; Andrade and Chaput, 1991b). However, little is known about 5-HTreceptor types in the nervous systems of invertebrates.

We have been studying the physiological roles of serotonin in the crab stomatogastricganglion (STG). The STG contains about 30 neurons comprising two central patterngenerator (CPG) networks controlling the rhythmic movements of crab foregut muscles(Selverston and Moulins, 1987). A set of four peripheral mechanosensory neurons, calledthe gastropyloric receptor (GPR) cells, is activated by tension at the gastropyloric borderof the foregut. The GPR cells contain both 5-HT and choline acetyltransferase, thesynthetic enzyme for acetylcholine (ACh). It has been proposed that the GPR cells useboth 5-HT and ACh as co-transmitters to modulate the motor patterns generated by theSTG (Katz et al. 1989; Katz and Harris-Warrick, 1989, 1990a, 1991). The traditional roleof mechanosensory cells is to provide rapid phasic feedback to the motor network. TheGPRs accomplish this by releasing ACh, which evokes fast nicotinic cholinergicexcitatory postsynaptic potentials in a variety of postsynaptic neurons. A newneuromodulatory role for mechanosensory cells was observed when the GPRs inducedprolonged modulatory changes in the motor patterns generated by the crab CPGs (Katzand Harris-Warrick, 1989, 1990a). These modulatory effects are mimicked by exogenous5-HT and vary among the different STG neurons. Some neurons respond to GPR or 5-HTwith tonic inhibition; others exhibit tonic excitation, rhythmic bursting or plateaupotentials.

These different cellular responses to 5-HT could be explained by two general classes ofmechanisms. First, the different target neurons could possess different 5-HT receptorsthat evoke different electrophysiological responses. Second, the neurons could share acommon 5-HT receptor, with the signal being interpreted differently by the differenttarget neurons because of differences in intrinsic cellular response properties, G proteinsubtypes or coupled second-messenger mechanisms. These two mechanisms are notmutually exclusive, and further permutations are possible. In this paper, we presentevidence from electrophysiological and pharmacological studies that crab STG neuronspossess multiple receptors recognized by 5-HT. These 5-HT receptors can be classifiedinto several distinguishable subtypes in different pyloric neurons; each is characterizedby selective agonists and antagonists. Neurons that have different physiologicalresponses to GPR stimulation and to application of 5-HT or its agonists are sensitive todifferent 5-HT antagonists. These results also provide further support for our hypothesisthat 5-HT is a transmitter used by the GPR cells.

Part of this work has appeared in abstract form (Zhang and Harris-Warrick, 1991).

56 B. ZHANG AND R. M. HARRIS-WARRICK

Materials and methods

Animals

Crabs (Cancer borealis) were purchased from suppliers in Boston, Massachusetts. Thecrabs were maintained in Instant Ocean aquaria at 12–14 ˚C.

Saline and chemicals

The crab saline contained (mmol l21): 440 NaCl, 11 KCl, 13 CaCl2, 26 MgCl2,8 glucose, 11 Trizma base buffer, and 5 maleic acid. The pH was 7.4. The followingdrugs were tested: atropine sulfate, cyproheptadine hydrochloride and gramine[3-(dimethylaminomethyl) indole] (purchased from Sigma); MDL 72222 (3-tropanyl-3,5-dichlorobenzoate), pirenperone, mCPP [1-(3-chlorophenyl)piperazine dihydro-chloride], 5-CT (5-carboxamidotryptamine maleate), CGS-12066B dimaleate, a-me-5HT [(±)-a-methyl-5HT maleate], 2-me-5HT [(±)-2-methyl-5-HT maleate] and1-phenylbiguanide (purchased from Research Biochemicals Inc.) The following drugswere generous gifts from the pharmaceutical companies indicated in parentheses:cinanserin hydrochloride (The Squibb Institute for Medical Research); metergoline(Farmitalia Carloerba); metoclopramide hydrochloride (A. H. Robins Co.); methysergidemaleate (Sandoz Pharmaceuticals); mianserin hydrochloride (Organon Inc.); andspiperone (Janssen Pharmaceutica Inc.). All drug solutions were prepared immediatelybefore use. Atropine, cinanserin and all agonists were dissolved directly into the saline.Gramine, MDL 72222, mianserin and metoclopramide were dissolved in ethanol at10 mmol l21. The other drugs were dissolved in either methanol or dimethyl sulfoxide(DMSO) at 10 mmol l21. The final concentration was obtained by diluting each stock intosaline. Ethanol, DMSO and methanol had no effects of their own at the final dilutionsused (maximum 0.2 %). 5-HT (serotonin creatinine sulfate, Sigma) was applied bypressure ejection from an adjacent micropipette at 1 mmol l21 to 1 mmol l21 (see below).

PhysiologyThe dissection of the stomatogastric nervous system followed the methods described

by Selverston et al. (1976). The dissection was usually performed on the day of theexperiment. However, some dissections were performed up to the desheathing of the STGthe previous evening; this did not affect the results in any way.

Physiological recordings were performed as described by Katz et al. (1989; Katz andHarris-Warrick, 1989, 1990a). Extracellular recordings were made from identified motornerves with bipolar stainless-steel pin electrodes. Intracellular recordings were made withglass microelectrodes (15–25 MV) filled with 4 mol l21 potassium acetate plus0.3 mol l21 potassium chloride, or 3 mol l21 potassium chloride. STG neurons wereidentified by (1) matching action potentials recorded from the appropriate motor nerveroot and intracellularly from the soma, (2) the characteristic shape of membrane potentialoscillations and spike amplitudes, and (3) the timing of spike activity within the pyloricrhythm. Following neural identification, input from the anterior ganglia was blocked byfilling a Vaseline ring on the desheathed stomatogastic nerve (stn) with isotonic sucrosesolution in 50 mmol l21 Tris–maleate buffer. Sometimes, a Vaseline chamber

57Multiple 5-HT receptors in the crab STG neurons

surrounding the anterior ganglia was filled with the isotonic sucrose solution or salinecontaining 1 mmol l21 tetrodotoxin (TTX), and the stn was then desheathed.

The STG (bath volume 0.2–0.5 ml) was superfused with saline at 2 ml min21. The 5-HT agonists were bath-applied for various times (2–5 min) before the results wererecorded and washed for 30–50 min afterwards. Serotonin antagonists were allowed tosuperfuse the ganglion for at least 15 min to ensure equilibration before any recordingswere made and washed for at least 30 min between drug applications. All the drug effectsreversed upon removal of the drug. Unless noted otherwise, all experiments wereperformed in saline containing picrotoxin (PTX, Sigma, 5 mmol l21) to eliminate theglutamate synapses between neurons (Bidaut, 1980; Eisen and Marder, 1982; Marder andEisen, 1984; Marder, 1987). This reduced the synaptic interactions between the STGneurons, facilitating the measurement of direct cellular responses to GPR stimulation or5-HT (Fig. 1; Katz and Harris-Warrick, 1989, 1990a).

Pressure ejection was used to ‘puff’ small volumes of 5-HT solution onto the neuropilof the STG. The puffer micropipettes had tip diameters of about 2 mm and were locatedapproximately 80 mm downstream from the STG. Calibrated pressure pulses


a-me-5HT (+)Cinanserin (−)

Atropine (−)

5-HT, 5-CT

5-HT2-me-5HT (+)

Gramine (−)

5-HT, 5-CT (+)Cinanserin (−)

Excitatory

Inhibitory

Inhibitory and excitatory

(+) Agonist(s)

5-HT (+)

(−) Antagonist(s)

4GPR

2PD

2LPG

LP

AB

IC VD

5PY

Fig. 1. The pyloric circuit, adapted from Katz and Harris-Warrick (1990a). Inhibitorychemical synapses are denoted by circles. The filled circles are picrotoxin-sensitive glutamatesynapses, while the open circles are cholinergic synapses. Electrotonic junctions are presentedas resistors. Serotonergic inputs from the gastropyloric receptor (GPR) cells to stomatogastricganglion neurons are denoted by triangles and bars. The triangles represent excitatorysynapses, while the bars are inhibitory. Serotonergic agonists and antagonists that act atdifferent GPR synapses are indicated.

(300–3000 ms; 35–138 kPa) ejected the 5-HT solution over the whole ganglion. FastGreen FCF (1 mg ml21, Sigma) was used as an indicator for the 5-HT ejection; the dyealone had no effect on the STG.

The identification and stimulation of the GPR cells followed the methods described byKatz et al. (1989).

Data analysis

Data presented here were collected from experiments conducted with sixty animals.Each drug was tested at least five times. Values are expressed as mean ± S.E.M. A pairedStudent’s t-test was used for data analysis. The difference was considered significantwhen the P<0.05.

Results

Katz and Harris-Warrick (1989, 1990a) identified four major classes of response toGPR stimulation among neurons of the pyloric network (Fig. 1). These include transientinhibition of PY, the induction or enhancement of rhythmic bursting in the AB/PDneurons, tonic excitation of LP and biphasic responses (inhibition followed by excitation)in the inferior cardiac neuron (IC). Bath application of 5-HT mimicked the GPR responsein several of these cases, although the time course of the response was very differentbecause of the prolonged application of the amine. We have studied 5-HT receptorpharmacology in the first three classes of responses (PY, AB/PD and LP). Here we showthat a brief (300–3000 ms) application of 5-HT by pressure ejection, or ‘puffing’, canmore closely approximate the time course of the modulatory component of GPRstimulation. Bath application of different 5-HT agonists mimics the modulatory effects ofGPR stimulation and 5-HT application in different neurons. Further, the effects of GPRstimulation and application of 5-HT or its agonists are selectively blocked by selective5-HT antagonists in different STG neurons.

5-HT and 2-me-5HT mimic the effects of GPR in the PY neuron

GPR stimulation and a brief ‘puff’ of 5-HT (0.1 mmol l21, 1–3 s) both induced atransient inhibition of bursting in the PY neuron (Katz and Harris-Warrick, 1989, 1990a;Fig. 2). Note that during this inhibition the PY oscillations were selectively reduced intheir peak amplitude and slowed in their rate of post-inhibitory rebound; the trough of theburst (due to synaptic inhibition from the AB/PD neurons, see Fig. 1) was unaffected.Since PY receives glutamatergic inhibition from LP, the GPR-and 5-HT-evoked transientinhibition could result from an indirect effect from the excited LP. To eliminate thispossibility, we perfused the ganglion with PTX-containing saline, which blocks thesynaptic inhibition from LP (Fig. 1). Under this condition, we still observed an inhibitionof PY by GPR stimulation and application of 5-HT (Fig. 2B), confirming a direct effect ofGPR and 5-HT on PY (Katz and Harris-Warrick, 1989, 1990a).

This inhibition of the PY cell was selectively mimicked by bath application of the 5-HT3 agonist 2-me-5HT (Fig. 3). The number of spikes per PY burst was reduced (similarto the effect of bath-applied 5-HT; Katz and Harris-Warrick, 1989, 1990a), and the peak


amplitude of the PY oscillation was reduced, with little change in the synaptically evokedtrough of the burst. Table 1 shows that 2-me-5HT significantly reduced the number ofaction potentials per PY burst in a dose-dependent manner, essentially eliminating PYfiring at 10 mmol l21. A number of other 5-HT agonists were tested, but only 2-me-5HTaffected the PY neurons. 2-me-5HT had no effect on other pyloric neurons (see Table 4).

Gramine blocks GPR, 5-HT and 2-me-5HT inhibition of the PY neuron

Gramine (1–100 mmol l21) blocked the GPR-evoked inhibition of the PY neuron


B

A

Without picrotoxin

With picrotoxin

20 Hz

10 mV

1 s

PY

PY

PY

PY

20 Hz

20 Hz

20 Hz

GPR 5-HT

Control

Control

Gramine

Gramine

Fig. 2. Gramine blocks inhibition of PY neurons by GPR stimulation or 5-HT pressureejection. GPR stimulation (15–20 Hz) and a brief 5-HT (0.1 mmol l21) puff cause a transientinhibition of the PY neuron. Gramine (10 mmol l21) abolishes both inhibitory responses. Thepresence of picrotoxin (5 mmol l21) does not change the effects of 5-HT and gramine (A andB). The most hyperpolarized potentials for the PY neurons in A and B were 255 mV and258 mV, respectively.

(Fig. 2). Similarly, gramine blocked the inhibition evoked by a brief puff of 5-HT(0.1 mmol l21; Fig. 2) and reduced the inhibition evoked by bath application of theagonist 2-me-5HT (Fig. 3). The effects of gramine completely reversed upon washout.Gramine alone had little or no effect on existing PY activity. Various other antagonistswere also tested: they had no antagonistic effects on either 5-HT- or GPR-inducedinhibition of PY (see Table 5). Gramine had no effect on 5-HT or GPR-evoked responsesin other neurons.

5-HT, 5-CT and a-me-5HT mimic the effects of GPR in the AB/PD neurons

GPR stimulation and 5-HT application also induce or enhance rhythmic bursting in theelectrically coupled AB/PD neurons (Katz and Harris-Warrick, 1989, 1990a). After asucrose block has been placed on the desheathed stn to block modulatory inputs to theSTG from the anterior ganglia, the electrically coupled conditional burster PD and AB


B

C D

A Control 2-me-5HT

2-me-5HT + gramine Wash

15 mV

5 s

Fig. 3. Selective inhibition of PY by 2-me-5HT. The 5-HT3 agonist 2-me-5HT (1 mmol l21,bath-applied) selectively inhibits PY (B). This inhibition of PY is reduced by gramine(10 mmol l21, C). The experiment was carried out in the presence of picrotoxin. The mosthyperpolarized potential of PY is 250 mV.

Table 1. Dose-dependent effect of 2-me-5HT on PY

[2-me-5HT] Reduction in the number of(mmol l−1) action potentials per burst (%)

1 44±40 (N=7; P<0.05)5 67±33 (N=3; P<0.05)

10 98±4 (N=4; P<0.01)

Values are mean ± S.E.M.

neurons usually slow down or cease their rhythmic bursting, as do all the STG neurons(Russell, 1979). As shown in Fig. 4, a brief puff of 5-HT (1 mmol l21, 3 s) or a short trainof GPR stimuli (20 Hz, 3 s) induced rhythmic bursting in a quiescent AB cell. 5-HTstimulation of AB/PD bursting could also be observed during recordings from theelectrically coupled PD neurons (Figs 5 and 6A). The time course of the responses toGPR stimulation or 5-HT puff was somewhat different: GPR stimulation inducedbursting with only a short delay, while the 5-HT puff (1 mmol l21) induced an initialhyperpolarization in AB/PD neurons before bursting began (Figs 4 and 5). In asynaptically intact preparation, part of this inhibition is due to synaptic inhibition by the


4 mV

5 s

GPR

AB

GPR

5-HT

Control

Atropine

5-HT

B

A

Fig. 4. Atropine blockade of AB bursting induced by GPR stimulation or puffed 5-HT. Undercontrol conditions (A), a brief puff of 5-HT (1 mmol l21) or GPR stimulation (20 Hz) inducesrhythmic bursting in the AB neuron. The induction of bursting in AB by both 5-HT and GPRdisappears when atropine (0.1 mmol l21) is added to the control saline (B). The restingpotential for the AB neuron is 263 mV, which remains unchanged in the presence of atropine.Note that an initial hyperpolarization of the AB typically follows the puff of 5-HT. Thishyperpolarization is caused in part by IPSPs (arrows) from simultaneous excitation of the LPneuron (not shown). Atropine selectively blocks the induction of bursting in AB, but not theexcitation in the LP neuron.

Fig. 5. Atropine blockade of 5-HT-induced bursting in the PD neuron. (A) A brief puff of 5-HT (0.1 mmol l21) enhances rhythmic bursting in the PD neuron (control). In the intactpreparation, PD responds to 5-HT with a short delay due to LP excitation and consequentsynaptic inhibition. (B) This delay becomes shorter when LP is hyperpolarized (arrow) toremove the glutamatergic inhibition from LP to PD. (C) Atropine (50 mmol l21) reduces orblocks the PD response without affecting LP excitation. (D) The same result is seen when theLP neuron is hyperpolarized by current injection (arrow). The most hyperpolarized potentialsare 253 mV for PD and 257 mV for LP.


B

C

D

AControl

Atropine

LP

PD

LP

PD

LP

PD

LP

PD

5-HT

5-HT 10 mV

5 s

5-HT

5-HT

Fig. 5

LP neuron (Fig. 1). This can be seen in Figs 4 and 5, whose recordings were madewithout PTX in the saline, so that glutamatergic inhibition of the AB/PD neurons by LPcould occur (see arrows in Fig. 4 indicating LP-evoked IPSPs; this is observed directly in


A

Control

Cinanserin

Control

Control

GPR(20 Hz)

GPR(20 Hz)

PD

PD

PD

PD

5-HT

5-HT

15 mV

1 s

15 mV

5 s

B

C

5-CT 5-CT+cinanserin

a-me-5HTa-me-5HT+cinanserin

Fig. 6

Fig. 5). When the LP neuron was hyperpolarized to remove its inhibition of the AB/PDneurons, AB/PD responded to 5-HT with a shorter delay (Fig. 5; compare the onset of PDbursting after 5-HT puff in traces A and B).

Enhancement of AB/PD bursting by GPR stimulation and 5-HT application wasmimicked by bath application of two vertebrate 5-HT agonists. They are the 5-HT1

agonist 5-CT and the 5-HT2 agonist a-me-5HT (Fig. 6B,C). a-me-5HT is a selectiveagonist acting only on AB/PD, while 5-CT also excited the LP neuron (see below). Theeffects of 5-CT and a-me-5HT were dose-dependent (Tables 2 and 3). 5-CT at5 mmol l21 and 10 mmol l21 significantly increased the bursting frequency of the AB/PDneurons by 173 % and 89 % respectively. a-me-5HT increased AB/PD bursting by 37 %,167 % and 91 % at 1, 5 and 10 mmol l21, respectively. It appears that the effects of 5-CTand a-me-5HT were maximal at 5 mmol l21 and declined at supramaximalconcentrations. No other agonists tested affected the AB/PD neurons (Table 4).

Atropine blocks 5-HT/GPR-induced rhythmic bursting in the AB/PD neurons

The induction and enhancement of AB/PD bursting by GPR and 5-HT was blocked by


Fig. 6. GPR, 5-HT, 5-CT and a-me-5HT enhancement of AB/PD bursting and blockade bycinanserin. (A) The 5-HT1c/2 antagonist cinanserin (20 mmol l21) blocks rhythmic bursting ofPD by GPR stimulation (20 Hz) and 5-HT (10 mmol l21). The resting potential of PD is257 mV. Cinanserin itself has little effect on the resting potentials. (B) The 5-HT1 agonist 5-CT (10 mmol l21, bath-applied) enhances PD bursting. This effect is blocked by cinanserin(10 mmol l21). The most hyperpolarized potentials of PD are 258 mV (control and 5-CT+cinanserin) and 262 mV (5-CT). (C) The 5-HT2 agonist a-me-5HT (10 mmol l21, bath-applied) enhances PD bursting. This enhancement is blocked by cinanserin (10 mmol l21). Themost hyperpolarized potentials of PD are 257 mV (control), 259 mV (a-me-5HT) and258 mV (a-me-5HT+cinanserin). Picrotoxin (5 mmol l21) was present throughout theexperiments. Cells in A, B and C are different PD cells.

Table 2. Dose-dependent effects of 5-CT on PD and LP

[5-CT] Increase in PD cycling Increase in the number of action (mmol l−1) frequency (%) potentials per LP burst (%)

1 35±28 (N=5; P>0.05) 156±124 (N=4; P<0.05)5 173±130 (N=3; P<0.05) 195±180 (N=3; P<0.05)

10 89±56 (N=5; P<0.05) 289±209 (N=5; P<0.01)


Table 3. Dose-dependent effect of a-me-5HT on PD

[a-me-5HT] Increase in PD cycling (mmol l−1) frequency (%)

1 37±58 (N=3; P>0.05)5 167±153 (N=3; P<0.05)

10 91±110 (N=8; P<0.05)


atropine (10 mmol l21 to 0.1 mmol l21; Figs 4 and 5). For example, as shown in Fig. 4,atropine eliminated the GPR/5-HT-evoked bursting response in the AB neuron, butsimultaneous excitation of the LP neuron was unaffected by atropine, as seen by theIPSPs it evoked in the AB neuron (arrows, Fig. 4). 5-HT and GPR stimulation continuedto evoke bursting in the AB/PD neurons when the LP neuron was hyperpolarized toremove its synaptic inhibition of the AB/PD neurons, and this bursting was abolished byatropine (Fig. 5). Fig. 7 shows a quantitative analysis of atropine reduction of GPR- and


Table 4. Effects of serotonergic agonists on the crab stomatogastric ganglion neurons

Maximal concentration LP AB/PD

Receptor applied Selected PY tonic rhythmic AntagonizedAgonists subtypesa (mmol l−1) referencesb inhibition excitation bursting by

mCPP 1 50 2, 3 −c − − −5-CT 1, 1a>1b 10 1, 4, 5, 6 − + + CinanserinCGS-12066B 1b 100 7 − − − −a-me-5HT 2 10 3, 8, 9 − − + Cinanserin2-me-5HT 3 10 3, 8, 9, 10 + − − Gramine1-Phenylbiguanide 3 100 11 − − − −

a5-HT receptor subtypes in vertebrates.b(1) Peroutka (1988); (2) Lucki et al. (1989); (3) Holohean et al. (1990); (4) Schmidt and Peroutka (1989);

(5) Roberts et al. (1988); (6) Rasmussen and Aghajanian (1990); (7) Neale et al. (1987); (8) Mawe et al.(1986); (9) Richardson et al. (1985); (10) Yankel and Jackson (1988); (11) Ireland and Tyers (1987).

c+ indicates that the agonist mimics the effects of GPR and 5-HT; − indicates that the agonist has no effect.

1.0

0.8

0.6

0.4

0.2

0.0

*

Cyc

le f

requ

ency

(H

z)

Control Atropine

Saline

5-HT

Fig. 7. A plot of pyloric cycle frequency enhancement by 5-HT pressure ejection and itsblockade by atropine. Values were collected from four experiments. A paired Student’s t-testshows that 5-HT significantly (P<0.05) enhances the PD bursting frequency (marked with anasterisk). The response to 5-HT is blocked by atropine (0.1 mmol l21).

5-HT-induced AB/PD bursting. A paired Student’s t-test indicates a significant blockadeby atropine. Thus, atropine selectively blocks GPR- and 5-HT-induced rhythmic burstingexcitation of the AB neuron. Atropine did not block either the responses of other neuronsto 5-HT (Table 5, Fig. 5) or the responses of AB/PD to the 5-HT agonists 5-CT and a-me-5HT (not shown).

The muscarinic agonist pilocarpine and the monoamine dopamine also induce


A

Control

Pilocarpine

Pilocarpine+atropine

10 mV

1 s

B

C

Wash

D

Fig. 8. The effects of atropine on pilocarpine-induced AB/PD bursting. The muscarinicagonist pilocarpine (0.1 mmol l21) induces fast and regular bursting recorded in the PD neuron(B). Addition of atropine (0.1 mmol l21) to the pilocarpine-containing saline only slightlyslows PD bursting (C). The effects are reversible (D). The most hyperpolarized membranepotential is 254 mV.

rhythmic bursting in the AB/PD neurons in the lobster Panulirus interruptus and in thecrab C. borealis (Marder, 1987; Harris-Warrick and Flamm, 1987). In contrast to itscomplete blockade of GPR- and 5-HT-induced AB/PD bursting, atropine (0.1 mmol l21)only slightly reduced pilocarpine-induced bursting (Fig. 8) and did not affect dopamine-induced AB/PD bursting (data not shown). Thus, atropine at these concentrations is arelatively selective 5-HT receptor antagonist and does not effectively block muscariniccholinergic receptors or cause a non-selective inhibition of the bursting mechanism.

Cinanserin blocks AB/PD bursting evoked by GPR, 5-HT, 5-CT and a-me-5HT

The vertebrate 5-HT1c/2 antagonist cinanserin (10–20 mmol l21) blocked theenhancement of AB/PD bursting by GPR stimulation and 5-HT (Fig. 6A) and by bathapplication of the 5-HT agonists 5-CT and a-me-5HT (Fig. 6B,C). This blockade wascompletely reversed upon washout. Cinanserin by itself had no effect on the AB/PDbaseline activity. Other vertebrate 5-HT antagonists (except atropine) had no effects on 5-HT- or GPR-induced rhythmic bursting of the AB/PD neurons (Table 5).

To show that cinanserin is a selective antagonist for 5-HT, we tested its effects onAB/PD bursting induced by the muscarinic agonist pilocarpine. Fig. 9 shows thatcinanserin had no effect on pilocarpine-induced AB/PD bursting.


Table 5. Effects of serotonergic antagonists on the crab STG neurons

Maximalconcentration LP PD/AB

Receptor applied Selected PY tonic rhythmicAntagonists subtypesa (mmol l−1) referencesb inhibition excitation bursting

Metergoline 1c/d, 2 100 1, 2, 3, 4 −c − −Spiperone 1a, 2 10 3, 4, 5, 6 − − −Methysergide 1c, 2 10 7, 8, 9 − − −Cinanserin 2, 1c 20 1, 7 − + +Mianserin 2, 1c 10 3, 4, 5 − − −Pirenperone 2 20 15 − − −Cyproheptadine 2 100 1, 11 − − 0MDL72222 3 100 3, 4, 5 − − −Zacopride 3 20 3, 4 − − 0Metoclopramide 3 100 11, 12 0 − 0Atropine u 100 13, 14 − − +Gramine u 100 13 + − −

a5-HT receptor subtypes in vertebrates. Atropine and gramine are classified as undeterminedsubtypes.

b(1) McCall and Aghajanian (1980); (2) White and Neuman (1983); (3) Peroutka (1988); (4) Schmidtand Peroukta (1989); (5) Holohean et al. (1990); (6) Felder et al. (1990); (7) Proudfit and Anderson(1973); (8) Roberts et al. (1988); (9) McCall and Aghajanian (1979); (10) Barbeau and Rossignol(1990); (11) Yakel et al. (1988); (12) Yakel and Jackson (1988); (13) Gaddum and Picarelli (1957);(14) Gerschenfeld and Stefani (1966); (15) Colpaert et al. (1982).

c− and + indicate non-blockade and blockade, respectively; 0 means the drug was not tested.

5-HT and 5-CT mimic the effects of GPR on the LP neuron

GPR stimulation and 5-HT application evoked tonic depolarization and high-frequencyspiking in the LP neuron (Figs 5, 10A). During the 5-HT puff or GPR stimulation, LPfired a tonic burst of action potentials at high frequency that effectively inhibitedoscillations in the pacemaker AB/PD neurons (see Figs 4 and 5; Katz and Harris-Warrick, 1990a). Following this, the LP fired at high frequency but was rhythmicallyinhibited by the AB/PD group.


Fig. 9. The effects of cinanserin on pilocarpine-induced AB/PD bursting. The muscarinicagonist pilocarpine (0.1 mmol l21) induces rhythmic bursting in AB/PD, as recorded in the PDneuron (B). Cinanserin (10 mmol l21) does not affect this response (C). The mosthyperpolarized potentials of PD are 260 mV (A), 257 mV (B and C) and 254 mV (D).

A

Control

Pilocarpine

Pilocarpine+cinanserin

10 mV

1 s

B

C

Wash

D

LP became tonically active in the absence of AB/PD bursting. Under these conditions,the 5-HT1 agonist 5-CT depolarized LP and increased its tonic spiking in a dose-dependent manner with a maximal effect at 10 mmol l21 (Table 2; Fig. 10B). This isdifferent from the actions of 5-CT on AB/PD bursting, which were maximal at 5 mmol l21

and declined at higher concentrations (Table 2). In a weakly oscillating preparation, 5-CTstrongly excited LP and enhanced plateau potentials in LP (Fig. 10C). Other 5-HTagonists we tested, including a-me-5HT, had no effect on LP activity (Table 4).


B

C

LP

LP

LP

LP

A

GPR(20 Hz)

5-HT

5-HTGPR(20 Hz)

15 mV1 s

15 mV5 s

Control

Cinanserin

Control 5-CT Wash

Control 5-CT 5-CT+cinanserin Wash

Fig. 10. 5-HT and 5-CT excitation of LP and the blockade by cinanserin. (A) GPR stimulation(20 Hz) and a brief puff of 5-HT (0.1 mmol l21) induce tonic excitation (depolarization andincreased action potentials). Cinanserin (20 mmol l21) blocks these responses. The restingpotential of LP is 250 mV. (B) 5-CT mimics the effect of 5-HT. In the absence of AB/PDbursting, 5-CT (10 mmol l21) depolarizes (from 240 mV to 235 mV) and enhances tonicspiking in LP. (C) In a slowly oscillating preparation, 5-CT (10 mmol l21) enhances theplateau ability of LP and increases the number of spikes per burst. This effect is blocked bycinanserin (10 mmol l21). The most hyperpolarized potentials of LP are 256 mV (control and5-CT), 259 mV (5-CT+cinanserin) and 257 mV (wash).

Cinanserin also blocks LP tonic excitation evoked by GPR, 5-HT and 5-CT

Cinanserin (1–20 mmol l21) abolished both GPR-evoked and 5-HT-induced LP activity(Fig. 10A). Cinanserin also blocked tonic LP excitation evoked by the 5-HT agonist 5-CT(Fig. 10 B,C). None of the other antagonists, including atropine (Figs 4 and 5), blockedthe response of the LP neuron to GPR, 5-HT or 5-CT (Table 5).

Discussion

GPR cells use 5-HT as a slow neuromodulator

Motor patterns generated by the crustacean STG can be altered by a variety ofneuromodulators that are released by identified neurons (for reviews, see Katz and Harris-Warrick, 1990b; Harris-Warrick et al. 1992). In the crab, the GPR cells aremechanosensory neurons that appear to use both ACh and 5-HT as co-transmitters. GPRstimulation evokes rapid EPSPs with a typical nicotinic cholinergic pharmacologicalprofile as well as a variety of slow modulatory responses (Katz and Harris-Warrick, 1989,1990a). We have studied three distinct slow modulatory responses to GPR stimulationobserved in different pyloric neurons: (1) tonic inhibition in the PY neurons; (2) inductionand enhancement of rhythmic bursting in the AB/PD neurons; and (3) prolonged tonicexcitation in the LP neuron (Katz and Harris-Warrick, 1989, 1990a). A number of otherSTG neurons also show slow modulatory responses to GPR stimulation; these responsesinclude induction of plateau potentials (Katz and Harris-Warrick, 1989; Kiehn andHarris-Warrick, 1992) as well as responses that appear to be similar to one of these threeclasses of response (Katz and Harris-Warrick, 1989), but they were not examined in thisstudy. It has been proposed (Katz and Harris-Warrick, 1989) that these slow modulatoryresponses are induced by the release of 5-HT from the GPR cells. In support of this, theslow responses are not blocked by a combination of muscarinic (except atropine; seebelow) and nicotinic cholinergic antagonists (Katz and Harris-Warrick, 1989, 1990a;Katz, 1989). Further, the slow effects of GPR stimulation are mimicked by either bathapplication (Katz and Harris-Warrick, 1989, 1990a) or pressure ejection (this paper;Kiehn and Harris-Warrick, 1992) of 5-HT. In this paper, we show that a set of 5-HTagonists selectively mimics the different effects of GPR and 5-HT on different neurons.Moreover, the specific responses to GPR stimulation and application of 5-HT or itsagonists have the same profile of sensitivity to a set of pharmacological 5-HT antagonists.This provides further support for our hypothesis that the GPR cells use 5-HT to induceslow modulatory responses in STG neurons.

Evidence for serotonergic modulation via multiple 5-HT receptors

Our experiments used pharmacological agonists and antagonists of 5-HT to determinewhether there is more than one type of 5-HT receptor evoking the different responses ofidentified pyloric neurons to 5-HT and GPR stimulation. Our results show that 5-HTexerts its modulatory effects via multiple receptors that are pharmacologically distinct onthe different STG neurons.

GPR stimulation or 5-HT application elicits a transient inhibition of the PY neurons.Katz and Harris-Warrick (1989, 1990a) showed that this inhibition is a direct effect of the


GPR cell onto PY, rather than an indirect effect mediated by excitation of other STGneurons. The inhibition of PY was selectively mimicked by 2-me-5HT, but not by anyother agonists that we tested (Fig. 3; Table 4). Gramine can block or reduce the PYinhibition induced by GPR stimulation as well as by application of 5-HT or 2-me-5HT(Figs 2 and 3). Gramine itself had no effect on the baseline activity of the PY neuron, andit had no effect on the responses of the other neurons to 5-HT and its agonists or to GPRstimulation. None of the other drugs we tested antagonized the inhibitory response of PYto 5-HT (Table 5). Thus, we conclude that 2-me-5HT is a selective agonist and gramine aselective antagonist for the 5-HT receptors that cause PY inhibition in the STG.

Serotonin and GPR stimulation evoke or enhance rhythmic bursting in the electricallycoupled AB/PD neurons (Figs 4, 5 and 6A). Two 5-HT agonists, 5-CT and a-me-5HT,mimic the effects of 5-HT and GPR on the AB/PD neurons (Fig. 6B,C). Cinanserin andatropine antagonized burst enhancement of the AB/PD neurons by 5-HT and GPR(Figs 5, 6A and 7). Cinanserin (but not atropine) also blocked the effects of 5-CT and a-me-5HT on AB/PD bursting (Fig. 6B,C). Moreover, cinanserin did not affect AB/PDbursting induced by dopamine or the muscarinic agonist pilocarpine (Fig. 9), suggestingthat it is selective for 5-HT receptors.

Atropine is typically considered as a muscarinic cholinergic antagonist. However, inthe crustacean STG, atropine is a weak antagonist at muscarinic receptors, causingcomplete block only at 1 mmol l21 (Marder and Eisen, 1984; Katz, 1989). In contrast,atropine blocked GPR-and 5-HT-induced AB/PD bursting at 10–100 mmol l21 (Figs 4and 5; Katz and Harris-Warrick, 1989), at which concentration it only has a weak effecton muscarinic receptors activated by the muscarinic agonist pilocarpine (Fig. 8). Inaddition, atropine did not antagonize any of the other 5-HT- or GPR-induced responses inthe pyloric neurons (Table 5). Thus, in the STG, atropine at low concentrations appears toact as a relatively selective antagonist of a subset of the 5-HT receptors inducing burstingin the AB/PD neurons. Atropine also blocks 5-HT receptors in guinea pig ileum (Gaddumand Picarelli, 1957), in molluscan neurons (Gerschenfeld and Stefani, 1966) and in leechneurons (Smith and Walker, 1975), so there are precedents for this effect. However,atropine at the concentrations we used failed to block AB/PD bursting induced by the 5-HT agonists 5-CT or a-me-5HT. We do not understand this result. It is unlikely thatatropine is acting in some non-selective manner, since it abolished GPR- and 5-HT-evoked AB/PD bursting but had little or no effect on pilocarpine- and dopamine-inducedbursting. It is possible that the AB/PD neurons possess multiple 5-HT receptors withdifferential selectivity for 5-CT, a-me-5HT, cinanserin and atropine, but theirinteractions would have to be complex to explain this result.

Katz and Harris-Warrick (1989, 1990a) demonstrated that GPR stimulation directlyexcites the LP neuron. We have shown that a brief puff of 5-HT and bath-applied 5-CTmimicked the effect of GPR stimulation on LP activity (Fig. 10). Cinanserin(20 mmol l21) blocked the tonic excitation induced by GPR stimulation, by puffed 5-HTor by bath-applied 5-CT. However, atropine, which blocked the AB/PD response, did notblock the LP response (Fig. 5; Table 5). Further, a-me-5HT, which mimics the action of5-HT on the AB/PD neurons, had no effect on the LP neuron (Table 4). Thus, the 5-HTreceptor(s) mediating tonic excitation in the LP neuron appear not to be identical to those


mediating rhythmic burst induction in the AB/PD neurons. Either these neurons expresscompletely different 5-HT receptors that are all capable of being blocked by cinanserin orthe LP neuron shares a cinanserin-sensitive receptor with the AB/PD neurons, but differsin its other 5-HT receptors.

Comparative views of vertebrate and invertebrate 5-HT receptors

The characterization of 5-HT receptor types in vertebrate nervous systems is quiteadvanced as a result of the large number of selective agonists and antagonists that exist(Peroutka, 1991). However, little is known about 5-HT receptor types in invertebrates.Using available 5-HT agonists and antagonists developed from vertebrate studies, wehoped to begin to identify 5-HT receptor subtypes in a well-defined invertebrate neuralnetwork in which cells can be readily identified and the functions of multiple receptortypes can be studied.

Several recent studies have shown that certain vertebrate 5-HT antagonists block 5-HTreceptors in crustaceans and insects. 5-HT stimulates the Na+/K+-ATPase that is involvedin the retraction of the proximal pigment in crayfish photoreceptor cells. Highconcentrations of methysergide and cyproheptadine (1 mmol l21) block this effect of 5-HT, presumably through antagonism of specific 5-HT receptors (Aréchiga et al. 1990;Frixione and Hernández, 1989). In Rhodnius prolixus, mianserin, methysergide andcyproheptadine block a 5-HT-induced increase in intracellular cyclic AMP concentrationin adult anterior midgut and Malpighian tubules (Barrett and Orchard, 1990). The 5-HT2

antagonists ketanserin and spiperone also block 5-HT-induced fluid secretion in theMalpighian tubules (Maddrell et al. 1991). In the mollusc Aplysia californica, 5-HTincreases the efficacy of synaptic transmission between siphon sensory cells and motorneurons. The 5-HT antagonist cyproheptadine blocks spike-broadening and theenhancement of synaptic facilitation induced by 5-HT or sensitizing stimulation of thetail (Mercer et al. 1991). In central neurons of the snail Helix pomatia, several 5-HTagonists (including 5-CT and a-me-5HT) and antagonists (including cinanserin andMDL 72222) were found to act on specific 5-HT receptors (Vehovszky and Walker,1991; Vehovszky et al. 1992). We have demonstrated that the vertebrate 5-HT1c/2

antagonist cinanserin blocks 5-HT receptors and that three 5-HT agonists selectivelymimic 5-HT-induced responses in crab STG neurons.

Despite the fact that some vertebrate agonists and antagonists work well oninvertebrate receptors, there is increasing evidence that 5-HT receptors in invertebratenervous systems are strikingly different from the well-characterized 5-HT receptorsubtypes of vertebrates. These differences are both pharmacological and functional(Smith and Walker, 1975; Vehovszky and Walker, 1991). Tables 4 and 5 show thatclassical vertebrate receptor subtype agonists and antagonists cannot define the crab 5-HTreceptors. For example, the 5-HT3 agonist 2-me-5HT inhibits the PY neurons, but another5-HT3 agonist, 1-phenylbiguanide, does not. Moreover, two 5-HT3 antagonists (MDL72222 and zacopride) fail to block the effects of 5-HT or 2-me-5HT on the PY neurons,while the non-classical antagonist gramine is very effective. The 5-HT1a,1b agonist 5-CTexcites the AB/PD and LP neurons, and this effect is antagonized by the 5-HT1c/2

antagonist cinanserin, but not by other 5-HT1c/2 antagonists (methysergide and


mianserin). In fact, the great majority of vertebrate 5-HT antagonists we tested had nodetectable effect on the crab 5-HT receptors (Table 5).

In addition, the vertebrate receptor subtypes evoke typical membrane electricalresponses in their neurons. In vertebrates, 5-HT1 agonists typically evokehyperpolarization via a G-protein-coupled second-messenger mechanism (Andrade andChaput, 1991b). In contrast, the 5-HT1 agonist 5-CT excites the crab AB/PD and LPneurons. The vertebrate 5-HT3 receptor is a ligand-gated ion channel mediating rapiddepolarization (Yankel and Jackson, 1988; Derkach et al. 1989), while the 5-HT3 agonist2-me-5HT slowly inhibits the crab PY neurons.

These results show that, although certain vertebrate 5-HT agonists and antagonists can beused to identify multiple 5-HT receptor subtypes in the crab STG neurons, these receptorsare sufficiently different that the vertebrate subtype classifications cannot be used to namethe crab receptors and other invertebrate receptors (Walker, 1985). Further research will beneeded before a rational invertebrate 5-HT receptor classification can be made.

We wish to thank Dr O. Kiehn for very helpful discussions of the data and forcomments on the manuscript. We also thank Drs D. Baro, B. Johnson, P. Katz, J. Peckand two anonymous reviewers for their comments on the manuscript, Drs E. Kravitz andM. Goy for discussions of their unpublished work on gramine and Drs G. W. G. Sharpand B. D. Schultz for their generous supply of spiperone and for many stimulatingdiscussions. We are grateful to Dr R. Hoy for the use of computer and printing facilitiesand to P. Faure and S. Singer for their assistance in preparing some of the figures. Thiswork was supported by NIH no. NS 17323 and USDA Hatch Act grant NYC-191410.

ReferencesAGHAJANIAN, G. K. (1981). The modulatory role of serotonin at multiple receptors in brain. In Serotonin

Neurotransmission and Behavior (ed. B. L. Jacobs and A. Gelperin), pp. 156–185. Cambridge, MA,USA: The MIT Press.

AMIN, A. H., CRAWFORD, T. B. B. AND GADDUM, J. H. (1954). Distribution of substance P and 5-hydroxytryptamine in the central nervous system of the dog. J. Physiol., Lond. 126, 596–618.

ANDRADE, R. AND CHAPUT, Y. (1991a). 5-Hydroxytryptamine4-like receptors mediate the slowexcitatory response to serotonin in the rat hippocampus. J. Pharmac. exp. Ther. 257, 930–937.

ANDRADE, R. AND CHAPUT, Y. (1991b). The electrophysiology of serotonin receptor subtypes. InSerotonin Receptor Subtypes: Basic and Clinical Aspects (ed. S. J. Peroutka), pp. 103–124. NewYork: Wiley-Liss, Inc.

ARÉCHIGA, H., BAÑUELOS, E., FRIXIONE, E., PICONES, A. AND RODRIGUEZ-SOSA, L. (1990). Modulation ofcrayfish retinal sensitivity by 5-hydroxytryptamine. J. exp. Biol. 150, 123–143.

BARBEAU, H. AND ROSSIGNOL, S. (1990). The effects of serotonergic drug on the locomotor pattern andon cutaneous reflexes of the adult chronic spinal cat. Brain Res. 514, 55–67.

BARRETT, M. AND ORCHARD, I. (1990). Serotonin-induced elevation of cyclic AMP levels in theepidermis of the blood-sucking bug, Rhodnius prolixus. J. Insect Physiol. 36, 625–633.

BELTZ, B., EISEN, J. S., FLAMM, R. E., HARRIS-WARRICK, R. M., HOOPER, S. L. AND MARDER, E. (1984).Serotonergic innervation and modulation of the stomatogastric ganglion of three decapod crustaceans(Panulirus interruptus, Homarus americanus and Cancer irroratus). J exp. Biol. 109, 35–54.

BIDAUT, M. (1980). Pharmacological dissection of the pyloric network of the lobster stomatogastricganglion using picrotoxin. J. Neurophysiol. 44, 1089–1101.

BOBKER, D. H. AND WILLIAMS, J. T. (1990). Ion conductances affected by 5-HT receptor subtypes inmammalian neurons. Trends Neurosci. 13, 169–173.


BRADLEY, P. B., ENGEL, G., FENIUK, W., FOZARD, J. R., HUMPHREY, P. P. A., MIDDLEMISS, D. N.,MYLECHARANE, E. J., RICHARDSON, B. P. AND SAXENA, P. R. (1986). Proposals for the classification andnomenclature of functional receptors for 5-hydroxytryptamine. Neuropharmacology 25, 563–576.

BRODIE, B. B. AND SHORE, P. A. (1957). A concept for a role of serotonin and epinephrine as chemicalmediators in the brain. Ann. N. Y. Acad. Sci. 66, 631–642.

COLPAERT, F. C., NIEMEGEERS, C. J. E. AND ANSSEN, P. A. J. J. (1982). A drug discrimination analysis oflysergic acid diethylamine (LSD): in vivo agonist and antagonist effects of purported 5-hydroxytryptamine antagonists and of pirenperone, a LSD antagonist. J. Pharmac. exp. Ther. 221,206–214.

DERKACH, V., SURPRENANT, A. AND NORTH, R. A. (1989). 5-HT3 receptors are membrane ion channels.Nature 339, 706–709.

EISEN, J. S. AND MARDER, E. (1982). Mechanisms underlying pattern generation in lobsterstomatogastric ganglion as determined by selective inactivation of identified neurons. III. Synapticconnections of electrically coupled pyloric neurons. J. Neurophysiol. 48, 1392–1451.

FELDER, C. C., KANTERMAN, R. Y., MA, A. L. AND AXELROD, J. (1990). Serotonin stimulatesphospholipase A2 and the release of arachidonic acid in hippocampal neurons by a type 2 serotoinreceptor that is independent of inositophospholipid hydrolysis. Proc. natn. Acad. Sci. U.S.A. 87,2187–2191.

FRIXIONE, E. AND HERNANDEZ, J. (1989). Modulation of screening-pigment position in crayfishphotoreceptors by serotonin: possible involvement of Na+/K+-ATPase activity. J. exp. Biol. 143,459–473.

GADDUM, J. H. AND PICARELLI, Z. P. (1957). Two kinds of tryptamine receptor. Br. J. Pharmac. 12,323–328.

GERSCHENFELD, H. M. AND PAUPARDIN-TRITSCH, D. (1974a). Ionic mechanisms and receptor propertiesunderlying the responses of molluscan neurones to 5-hydroxytryptamine. J. Physiol, Lond. 243,427–456.

GERSCHENFELD, H. M. AND PAUPARDIN-TRITSCH, D. (1974b). On the transmitter function of 5-hydroxytryptamine at excitatory and inhibitory monosynaptic junctions. J. Physiol., Lond. 243,457–481.

GERSCHENFELD, H. M. AND STEFANI, E. (1966). An electrophysiological study of 5-hydroxytryptaminereceptors of neurons in the molluscan nervous system. J. Physiol, Lond. 185, 684–700.

HARRIS-WARRICK, R. M. AND FLAMM, R. E. (1987). Multiple mechanisms of bursting in a conditionalbursting neuron. J. Neurosci. 7, 2113–2128.

HARRIS-WARRICK, R. M., NAGY, F. AND NUSBAUM, M. P. (1992). Neuromodulation of stomatogastricnetworks by identified neurons and transmitters. In Dynamic Biological Networks: TheStomatogastric Nervous System (ed. R. M. Harris-Warrick, E. Marder, A. I. Selverston and M.Moulins), pp. 87–137. Cambridge, MA, USA: The MIT Press.

HOLOHEAN, A. M., HACKMAN, J. C. AND DAVIDOFF, R. A. (1990). Changes in membrane potential of frogmotoneurons induced by activation of serotonin receptor subtypes. Neurosci. 34, 555–564.

IRELAND, S. J. AND TYERS, M. B. (1987). Pharmacological characterization of 5-hydroxytryptamine-induced depolarization of the rat isolated vagus nerve. Br. J. Pharmac. 90, 229–238.

KATZ, P. S. (1989). Motor pattern modulation by serotonergic sensory cells. In The StomatogastricNervous System. PhD thesis, Cornell University.

KATZ, P. S., EIGG, M. H. AND HARRIS-WARRICK, R. M. (1989). Serotonergic/cholinergic muscle receptorcells in the crab stomatogastric nervous system. I. Identification and characterization of thegastropyloric receptor cells. J. Neurophysiol. 62, 558–570.

KATZ, P. S. AND HARRIS-WARRICK, R. M. (1989). Serotonergic/cholinergic muscle receptor cells in thecrab stomatogastric nervous system. II. Rapid nicotinic and prolonged modulatory effects on neuronsin the stomatogastric ganglion. J. Neurophysiol. 62, 571–581.

KATZ, P. S. AND HARRIS-WARRICK, R. M. (1990a). Neuromodulation of the crab pyloric central patterngenerator by serotonergic/cholinergic proprioceptive afferents. J. Neurosci. 10, 1495–1512.

KATZ, P. S. AND HARRIS-WARRICK, R. M. (1990b). Actions of identified neuromodulatory neurons in asimple motor system. Trends Neurosci. 13, 367–373.

KATZ, P. S. AND HARRIS-WARRICK, R. M. (1991). Recruitment of crab gastric mill neurons into thepyloric motor pattern by mechanosensory afferent stimulation. J. Neurophysiol. 65, 1442–1451.

KIEHN, O. AND HARRIS-WARRICK, R. M. (1992). Serotonergic stretch receptors induce plateau


properties in a crustacean motor neuron by a dual-conductance mechanism. J. Neurophysiol. 68,485–495.

LUCKI, I., WARD, H. R. AND FRAZER, A. (1989). Effects of 1-(m-chlorophenyl)piperazine and 1-(m-trifluoromethylphenyl)piperazine on locomotor activity. J. Pharmac. exp. Ther. 249, 155–164.

MADDRELL, S. H. P., HERMAN, W. S., MOONEY, R. L. AND OVERTON, J. A. (1991). 5-Hydroxytryptamine:a second diuretic hormone in Rhodnius prolixus. J. exp. Biol. 156, 557–566.

MARDER, E. (1987). Neurotransmitters and neuromodulators. In The Crustacean Stomatogastric System(ed. A. I. Selverston and M. Moulins), pp. 263–300. New York: Springer-Verlag.

MARDER, E. AND EISEN, J. S. (1984). Transmitter identification of pyloric neurons: electrically coupledneurons use different transmitters. J. Neurophysiol. 51, 1345–1361.

MAWE, G. H., BRANCHEK, T. A. AND GERSHON, M. D. (1986). Peripheral neural serotonin receptors:identification and charaterization with specific antagonists and agonists. Proc. natn. Acad. Sci. U.S.A.83, 9799–9803.

MAYNERT, E. W., LEVI, R. AND DE LORENZO, A. J. D. (1964). The presence of norepinephrine and 5-hydroxytryptamine in vescicles from disrupted nerve endings particles. J. Pharmac. exp. Ther. 144,385–392.

MCCALL, R. B. AND AGHAJANIAN, G. K. (1979). Serotonergic facilitation of facial motoneuronexcitation. Brain Res. 169, 11–27.

MCCALL, R. B. AND AGHAJANIAN, G. K. (1980). Pharmacological charaterization of serotonin receptorsin the facial motor nucleus: A microiontophoretic study. Eur. J. Pharmac. 65, 175–183.

MERCER, A. R., EMPTAGE, N. J. AND CAREW, T. J. (1991). Pharmacological dissociation of modulatoryeffects of serotonin in Aplysia sensory neurons. Science 254, 1811–1813.

MILLER, J. P. AND SELVERSTON, A. I. (1979). Rapid killing of single neurons by irradiation ofintracellularly injected dye. Science 206, 702–704.

NEALE, R. F., FALLONS, S. L., BOYAZ, W. C., WASLEY, J. W. F., MARTIN, L. L., STONE, G. A., GLAESERB,B. S., SINTON, C. M. AND WILLIAMS, M. (1987). Biochemical and pharmacological characterization ofCGS-12066B, a selective serotonin-1B agonist. Eur. J. Pharmac. 136, 1–9.

PEROUTKA, S. J. (ED.) (1991). Serotonin Receptor Subtypes: Basic and Clinical Aspects. New York:Wiley-Liss, Inc.

PROUDFIT, H. K. AND ANDERSON, E. G. (1973). Influence of serotonin antagonists on bulbospinalsystems. Brain Res. 61, 331–341.

RASMUSSEN, K. AND AGHAJANIAN, G. K. (1990). Serotonin excitation of facial motoneurons: Receptorsubtype characterization. Synapse 5, 324–332.

RICHARDSON, B. P., ENGEL, G., DONATSCH, P. AND STADLER, P. A. (1985). Identification of serotonin M-receptor subtypes and their specific blockade by a new class of drugs. Nature 316, 126–131.

ROBERTS, M. H. T., DAVIES, M., GIRDLESTONE, D. AND FOSTER, G. A. (1988). Effects of 5-hydroxytryptamine agonists and antagonists on the responses of rat spinal motoneurones to rapheobscurus stimulation. Br. J. Pharmac. 95, 437–448.

RUSSELL, D. F. (1979). CNS control of pattern generators in the lobster stomatogastric ganglion. BrainRes. 177, 598–602.

SCHMIDT, A. W. AND PEROUTKA, S. J. (1989). 5-Hydroxytryptamine receptor ‘families’. FASEB J. 3,2242–2249.

SELVERSTON, A. I. AND MOULINS, M. (1987). The Crustacean Stomatogastric System. Berlin: Springer-Verlag.

SELVERSTON, A. I., RUSSELL, D. F., MILLER, J. P. AND KING, D. G. (1976). The stomatogastric nervoussystem: Structure and function of a small neuronal network. Prog. Neurobiol. 7, 215–290.

SMITH, P. A. AND WALKER, R. J. (1975). Further studies on the action of various 5-hydroxytryptamineagonists and antagonists on the receptors of neurons from the leeches, Hirudo medicinalis andHaemopis sanguisuga. Comp. Biochem. Physiol. 51C, 195–203.

VEHOVSZKY, A., KEMENES, G. AND S.-ROZSA, K. (1992). The monosynaptic connections between theserotonin-containing LP3 and RPas neurons in Helix are serotonergic. J. exp. Biol. 173, 109–122.

VEHOVSZKY, A. AND WALKER, R. J. (1991). An analysis of the 5-hydroxytryptamine (serotonin) receptorsubtypes of central neurons of Helix aspersa. Comp. Biochem. Physiol. 100C, 463–476.

WALKER, R. J. (1985). The pharmacology of serotonin receptors in invertebrates. InNeuropharmacology of Serotonin (ed. A. R. Green), pp. 366–408. Oxford, UK: Oxford UniversityPress.


WELSH, J. H. AND MOOREHEAD, M. (1960). The quantitative distribution of 5-hydroxytryptamine in theinvertebrates, specially in their nervous systems. J. Neurochem. 6, 146–169.

WHITE, S. R. AND NEUMAN, R. S. (1983). Pharmacological antagonism of facilitatory but not inhibitoryeffects of serotonin and norepinephrine on excitability of spinal motoneurons. Neuropharmac. 22,489–494.

YANKEL, J. L. AND JACKSON, M. B. (1988). 5-HT3 receptors mediate rapid responses in culturedhippocampus and a clonal cell line. Neuron 16, 615–621.

YANKEL, J. L., TRUSSELL, L. O. AND JACKSON, M. B. (1988). Three serotonin responses in cultured mousehippocampal and striatal neurons. J. Neurosci. 8, 1273–1285.

ZHANG, B. AND HARRIS-WARRICK, R. M. (1991). Multiple receptors mediate the modulatory effects ofserotonergic neurons in a small neural network. Soc. Neurosci. Abstr. 17, 200.


190, 55–77 (1994) 55 printed in great britain © the ...(lp) neuron. these effects were blocked by...

Documents