Cholinergic Modulation of the Swimmeret Motor System in

JOURNALOF NEUROPHYSIOLOGY Vol. 70, No. 6, December 1993. Printed in U.S.A.

Cholinergic Modulation of the Swimmeret Motor System in Crayfish

GijTZ BRAUN AND BRIAN MULLONEY Section of Neurobiology, Physiology and Behavior, and Neuroscience Doctoral Program, University of California, Davis, California 95616

SUMMARY AND CONCLUSIONS

1. The muscarinic agonist pilocarpine induced the swimmeret motor pattern in resting isolated preparations of the crayfish abdominal nerve cord and modulated the burst frequency in a dose- dependent manner.

2. Nicotine did not elicit rhythmic activity in resting isolated preparations but increased the burst frequency in active preparations. Nicotine produced higher burst frequencies than pilocarpine.

3. The acetylcholine ( ACh) analogue carbachol combined the effects of pilocarpine and nicotine. It activated isolated resting preparations and increased the burst frequency as effectively as nicotine. The ACh-esterase inhibitor eserine also increased the burst frequency in active preparations.

4. Neither muscarinic nor nicotinic antagonists disrupted the proctolin-induced motor pattern, suggesting that proctolin and cholinergic agonists affect two different pathways for the activation of the swimmeret system.

5. We conclude that cholinergic interneurons participate in initiation of the swimmeret motor pattern and can modulate its burst frequency.

INTRODUCTION

The swimmerets are paired appendages on the abdominal segments of crayfish and other decapod crustaceans. In the intact animal the swimmerets beat in a bilaterally synchronous metachronal wave that propagates caudally to rostrally. These behavioral features are paralleled by the activity of motor neurons innervating the swimmeret mus- cles. In the isolated abdominal nerve cord there are alternat- ing bursts of return-stroke (RS) motor neurons and power- stroke (PS) motor neurons in the anterior and posterior branches of the nerve Nl that innervates the swimmeret. These bursts are bilaterally synchronous in each pair of N 1 s and are coordinated in a metachronal sequence from cau- da1 to rostra1 ganglia (Hughes and Wiersma 1960). This pattern can occur spontaneously, or can be induced by stimulation of axons in the connectives between the last thoracic and first abdominal ganglia (Wiersma and Ikeda 1964). Pharmacologically, the swimmeret motor pattern can be induced by application of the peptide proctolin ( Mulloney et al. 1987 ) . However, proctolin-induced patterns display only a narrow range of burst frequencies (Mulloney et al. 1987)) whereas the stimulation of axons in the connective between the first and second abdominal ganglion in the lobster induces a broader range of burst frequencies (Davis and Kennedy 1972).

Acetylcholine (ACh) occurs as a neurotransmitter in arthropods, including crustaceans (Barker et al. 1972; Florey 1973; Marder 1976), and receptors similar to verte- brate nicotinic and muscarinic receptors have been de-

scribed pharmacologically (Barker et al. 1986; Florey and Rathmayer 1980; Freschi 199 1; Marder and Paupardin- Tritsch 1978; Pfeiffer-Lynn and Glantz 1990). Recently it has been shown that the muscarinic agonist oxotremorine also induces expression of the swimmeret motor pattern in isolated crayfish abdominal nerve cords (Chrachri and Neil 1993). In contrast to proctolin, oxotremorine modulated the burst frequency in a dose-dependent manner (Chrachri and Neil 1993).

The ability of oxotremorine to modulate burst frequency motivated us to investigate the roles of muscarinic and nicotinic receptors in initiation and modulation of the swimmeret rhythm.

METHODS

Crayfish (Pac$zstacus leniusculus) of both sexes were obtained from local fisherman and kept in freshwater tanks at 15OC.

Preparation

Before dissection, animals were cooled on ice. The ventral nerve cord with the abdominal ganglia l-6 (A 1 -A6) was dissected free and pinned out in a Sylgard-coated dish containing modified van Harreveld’s crayfish saline (in mM: 195 NaCl, 5.36 KCl, 2.6 MgCl, , 13.5 CaCl, , and 10 tris ( hydroxymethyl ) aminomethane- maleate buffer, pH 7.4). The sheath around the ganglia was re- moved to facilitate the diffusion of solutions into the tissue.

Extracellular recordings

Each swimmeret is innervated by one nerve, Nl, which splits into an anterior and a posterior branch. The axons of RS motor neurons exit through the anterior branch; those of the PS motor neurons exit through the posterior branch (Mulloney et al. 1990; Stein 197 1). Action potentials of both motor neuron populations were recorded separately with extracellular Vaseline-insulated pin electrodes. Signals were amplified and stored on videotape for later analysis.

Intracellular recordings

Intracellular recordings were made using conventional tech- niques and equipment. Glass microelectrodes were pulled (Sutter Instruments) and filled with 3 M KCl. Their resistances were 15- 20 MQ. Signals were amplified (Axoclamp 2A, Axon Instru- ments) and stored on tape for later analysis.

Drugs and solutions

All drugs were purchased from Sigma and dissolved in-saline. Drugs were added to the bath solution and washed out by superfusion with saline. When a series of concentration steps was performed, increments in concentration were separated by 5 min. Between applications of different drugs all preparations were washed in saline for r 1 h.

0022-3077/93 $2.00 Copyright 0 1993 The American Physiological Society 2391

2392

1. saline

es3_.... ~ I

PS4

2. IOpM pilocarpine (1.44 s-l) I

3. 3OpM pilocarpine (2.14 s-l)

II 1~1 I/ ( 1’

’ /““i i ’ ‘/ ! 1’

4. IOOpM pilocarpine (2.31 s-1)

G. BRAUN AND B. MULLONEY

5. 1 h in saline

Is

6. 4pM proctolin (I .41 s-1)

7. 1 OJIM proctolin (I .82 s-1)

8. 5OpM proctolin (1.82 s-1) I ,

9. IOpM proctolin + 1 OpM pilocarpine (I .82 s-1)

10. + 3OpM pilocarpine (2.09 s-1)

11. + 1 OOpM pilocarpine (I .89 s-1)

FIG. 1. Activity patterns induced by pilocarpine and by proctolin in 1 isolated nerve cord. l-4: initiation of the swimmeret motor pattern by pilocarpine. Burst frequency increased with the concentration of pilocarpine. 5: after l-h wash in saline, the preparation returned to a resting state. 6-8: different concentrations of proctolin induced a smaller increase in burst frequency and generated bursts that differed in shape from the pilocarpine-induced burst. 9- 11: after 1 -h wash in saline (not shown), the preparation was exposed to different pilocarpine concentrations in the presence of 10 ,uM proctolin. The increase in burst frequency was smaller than in pilocarpine alone and bursts had a different shape than in either pilocarpine or proctolin alone. Arrowheads: set of large spikes that were absent from the pilocarpine- and proctolin-induced bursts [ PS3, PS4: extracellular recordings of power-stroke motor neurons in ganglia A3 and A4. The burst frequency of the respective pattern is shown in parentheses].

Data analysis

Burst frequency and phase were calculated from 20-s intervals where the pattern showed little variation. The motor pattern was normally very regular, so that in single preparations the SE of the mean of the periods were 4% of the mean. Recording traces were digitized and analyzed on a PC with software described elsewhere (Mulloney and Hall 1987). Data were evaluated with t test statis- tics. For multigroup comparisons an analysis of variance and Stu- dent-Newman-Keuls test were performed.

RESULTS

Pilocarpine initiated the swimmeret motor pattern in isolated nerve cords and modulated the burst frequency

In resting preparations (i.e., there was either no or only a little tonic firing of motor neurons observed in extracellular recordings), pilocarpine elicited a swimmeret motor pattern at concentrations as low as 10 PM within 0.2- 1 min after application to the bath (Fig. 1, l-2). Resting preparations either began to express intermittent bouts of coordinated activity and then settled down to produce a continu-

ous pattern or began at once to express a continuous pattern. The periods and phases during intermittent bouts were in the same ranges as those observed during subsequent continuous activity. Pilocarpine concentrations < 10 ,uM did not activate resting preparations (n = 3). With increasing doses of pilocarpine the burst frequency increased (Fig. 1, 2-4; for a quantitative description on the modulation of burst frequency, see below). Superfusion with saline stopped the rhythmic activity within a few minutes (not shown).

Subsequent application of proctolin to the same preparations again activated the motor pattern, but the structure of motor neuron bursts appeared to be different from those during the pilocarpine-induced activity (Fig. 1, 6). An increase in the proctolin concentration increased the burst frequency less than pilocarpine (Fig. 1, 6-8 ) . When different pilocarpine concentrations were applied in the presence of proctolin the burst frequency was lower than in pilocarpine alone, and the structure of bursts was different from both the pilocarpine- and the proctolin-activated patterns (Fig. 1, 9-11).

CHOLINERGIC MODULATION OF MOTOR ACTIVITY 2393

Differences in the burst structure of motor patterns induced by different agonists and combinations of agonists were a consistent feature throughout this study. Basically, the appearance of a burst could change 1) when one or more units changed their intraburst firing frequency or 2) when different or additional motor units were recruited. We did not try systematically to discriminate between these two possibilities. However, Fig. 1, 9-l 1, provides a clear example of the additional recruitment of one or more motor units that were seen only in the presence of pilocarpine plus proctolin but not with either drug alone.

Nicotine modulated the burst frequency in isolated nerve cords

Nicotine, when applied to a resting preparation, did not elicit a coordinated swimmeret motor pattern but caused only tonic activity in some motor neurons (not shown). However, when added to an already active preparation, it increased the burst frequency in a dose-dependent manner (Fig. 2, A I-&). The threshold for detectable effects of nicotine on the burst frequency in active preparations was -0.1 FM (not shown). Nicotine also changed the structure of single bursts (Fig. 2A) but these changes were not consistent over the whole range of concentrations tested. Lower concentrations (52 ,uM, Fig. 2, A2 and A3) increased the number of larger motor neuron spikes, whereas higher concentration decreased the number of smaller spikes (Fig. 2, n4 and A.5). At concentrations >3 ,uM the motor pattern became unstable (Fig. 2A5) and finally deteriorated into tonic activity (not shown).

Carbachol initiated the swimmeret motor pattern and modulated the burstfiequency in isolated nerve cords

In a few experiments we applied ACh. However, doses > 100 ,uM were necessary to activate resting preparations, and the resulting changes were quite unstable over time (not shown). Also, in a previous study, Mulloney et al. ( 1987) were unable to observe consistent effects of ACh superfusion. We attribute this instability to the breakdown of ACh by ACh esterases ( AChE; Maynard 197 1). There- fore we employed the cholinergic analogue carbachol, which acts on both nicotinic and muscarinic receptors and is not affected by enzymatic breakdown.

Carbachol increased the burst frequency in preparations already expressing the swimmeret motor pattern (Fig. 2, Bl and B2). Carbachol also could initiate the swimmeret motor pattern in resting preparations with a threshold concentration - 1 ,uM (Fig. 2 B4). Higher doses led to an increase in burst frequency and to changes in burst shape (Fig. 2 B.5). At concentrations > 10 PM the pattern became more likely to deteriorate (not shown).

Eficts ofnicotine on individual motor neurons

Nicotine not only modulated the burst frequency but of all cholinergic agonists had the most variable effects on burst structure. A small set of experiments (n = 4) with intracellular recordings revealed several nicotine-induced effects on PS motor neurons (Fig. 3).

In the first example, a PS motor neuron had displayed large oscillations and generated phase-locked spikes during

A 1. 1 OpM proctolin

I PS 3 * I, II

1. 1 O!M proctolin

PS 4

2. + 0.25pM nicotine 2. + IpM carbachol

3. + IpM nicotine

4. + 3pM nicotine

5. + ~,YM nicotine

3. 1 h in saline

4. IpM carbachol

tt -,

5. IOpM carbachol

FIG. 2. Activity patterns induced by nicotine, carbachol, and proctolin. A : burst frequency of a motor pattern activated by 10 PM proctolin (I) was increased by addition (+) of nicotine. At 0.25- 1 PM (2 and 3) the burst shape indicated the activation of additional motor units; at higher concentrations the number of active units decreased (4). With 4 PM nicotine, the pattern deteriorated (5). B: burst frequency of a motor pattern activated by 10 FM proctolin (I) was increased by addition of 1 PM carbachol(2). After 1 -h wash in saline (3) the preparation returned to a resting state and could be activated by 1 PM carbachol (4). Increasing the carbachol concentration also increased the burst frequency (5). (PS 3, PS 4: extracellular recordings from PS motor neurons in A3 and A4. Different preparations in A and B).

rhythmic activity induced by proctolin alone (Fig. 3A). Application of 1 ,uM nicotine increased the frequency of the oscillations but caused only a small change in the average membrane potential (Fig. 3A). Increasing the nicotine concentration to 2 ,uM further increased the frequency but decreased the number of spikes (Fig. 3A).

Another PS motor neuron displayed only small oscillations with no spikes in the presence of proctolin alone (Fig. 3 B). Adding 1 ,uM nicotine depolarized the cell by - 15 mV, increased oscillation amplitude and frequency, and induced phase-locked action potentials (Fig. 3 B). After a l-h wash, application of 10 FM proctolin together with 1.5 FM nicotine caused phase-locked oscillations at a membrane

2394 G. BRAUN AND B ‘. MULLONEY

A RS 3i

+ IpM nicotine + 2pM nicotine (1 .96s-1) (2.64s-1) IS

R II

RS 4J

PS MN

-66 mV’ l%%%L,L 3 ,F& 1 7, ,y , L

+ 1.5pM nicotine -2.2nA (3.49s~1)

proctolin + IpM nicotine (1 .76s-1) (2.33s-1)

FIG. 3. Effects of nicotine on individual motor neurons. A: PS motor neuron in A3 showed rhythmic oscillation with phase-locked spikes during proctolin-induced activity. With the addition of 1 PM nicotine, the membrane potential depolarized slightly and phase-locked spikes were still present. Increasing the nicotine concentration to 2 FM depolarized more and reduced the number of spikes per burst. B: PS motor neuron in A4 showed only small oscillations during proctolin-induced activity. Adding 1 ,uM nicotine depolarized the cell by 15 mV and produced phase-locked bursts of spikes. After washing, 10 PM proctolin plus 1.5 PM nicotine depolarized the cell by 20 mV. The membrane potential still oscillated but no spikes occurred. Hyperpolarizing pulses generated rebound spikes that were reduced in number when release from hyperpolariza- tion was out-of-phase with the PS bursts. (PS 3, PS 4, RS 3, RS4, extracellular recordings; PS MN, intracellular recording from PS motor neuron in the same ganglion. Burst frequency of the respective pattern is shown in parentheses.)

potential more depolarized than during previous nicotine application, and action potentials were absent (Fig. 3B). The absence of spikes was probably due to adaptation caused by the strong depolarization; after pulses of hyperpolarizing current the neuron generated rebound spikes, which were reduced in number when the release from hyper- polarization occurred out-of-phase with the other PS motor neurons (Fig. 3 B).

The accompanying extracellular recordings ( Fig. 3, A and B) again illustrate the changes in burst structure already mentioned.

Eserine increased the burst frequency of the swimmeret motor pattern in isolated nerve cords

The AChE inhibitor eserine also increased burst frequency when applied to preparations already activated by proctolin (n = 4, Fig. 4). The onset of the response was later than with cholinergic receptor agonists (> 1 min). Eserine also prolonged the time required to terminate rhythmic activity using saline wash (Fig. 4), presumably because of the different kinetics involved in enzyme-inhibitor interaction as opposed to receptor-ligand interaction.

In resting preparations, eserine had no effect (not shown). This suggests that eserine enhanced ongoing cholinergic neurotransmission during the proctolin-activated

state. A detailed dose-response relationship was not established but doses from 1 to 10 ,uM increased the burst frequency, whereas higher doses disrupted the rhythm, as with nicotine and carbachol (not shown).

Eficts ofcholinergic agonists on burst frequency

During the present study, proctolin activated the motor pattern only within a small frequency range. In 24 preparations activated by 10 PM proctolin, frequencies ranged from 0.9 to 2.22 Hz, distributed around 1.58 per second (Fig. 5A). We did not test other concentrations because previous experiments ( Mulloney et al. 1987 ) indicated that

1 OpM proctolin + 5,uM eserine 20 min Is

1 .92 s-1 3.57 s-1 i.n saline 2.23 s-1

FIG. 4. Eserine increased the frequency of motor neuron bursts induced by 10 FM proctolin. After 20-min wash in saline the motor pattern persisted but the burst frequency had decreased.


I

1 .o

1 OuM proctolin

-I 1.2 1.5

I

1.8

burst frequency

[pilocarpine] uM r c I 0 20406080100

0

0246810 ‘g n= 4 4 6 5 L

[nicotine] uM [] pilocarpine l proctolin + nicotine 0 pilocarpine

F7A proctolin + pilocarpine m proctolin + nicotine

0 proctolin + pilocarpine m carbachol

FIG. 5. Effects of proctolin and cholinergic drugs on burst frequency in isolated nerve cords. A : distribution of mean burst frequencies induced by proctolin in 24 preparations. Bin width is 0.26 Hz, lowest frequency was 0.90 Hz, highest was 2.22 Hz. B: dose-frequency relations. For pilocarpine and proctolin plus pilocarpine n = 4; for proctolin plus nicotine n is indicated in the figure because not all animals were tested at every concentration. Error bars: mean + SE. C: maximum frequencies. Error bars: mean + SE. Dots: highest burst frequency ( * : P < 0.05, Student-Newman-Keuls test).

concentrations > 10 ,uM did not further increase the burst frequency (see also Fig. 1, 7 and 8).

In contrast, all cholinergic agonists seemed to have dose- dependent effects on the burst frequency. We established dose-response curves for the burst frequency activated by 1) pilocarpine alone, 2) pilocarpine in the presence of 10 PM proctolin, and 3) nicotine in the presence of 10 FM proctolin. The first and second experiments were performed with the same groups of animals; the third was obtained from a separate set. Concentrations were increased every 5 min; changes in frequency were usually observed within 1 min after application of a higher concentration. Only preparations showing a stable swimmeret pattern were evaluated.

With pilocarpine alone, the burst frequency increased up to a concentration of 50 PM (~2 = 4, maximum burst frequency = 2.42 t 0.16 (SE) Hz; Fig. 5, B and C). The highest burst frequency reached in one preparation was 2.89 Hz. When the same concentrations of pilocarpine were applied in the presence of 10 FM proctolin, the increase in burst frequency was attenuated (n = 4, maximum burst frequency = 2.19 t 0.06 Hz; Fig. 5, B and C).

The effect of nicotine was examined in six preparations activated with 10 PM proctolin. With increasing nicotine doses between 0.25 and 3 PM the burst frequency increased rapidly (n = 6, maximum burst frequency = 2.97 t 0.15 Hz; Fig. 5, B and C) . At higher concentrations the number of preparations still showing patterned activity decreased. The highest burst frequency reached was 3.49 Hz. Not all preparations were tested with every concentration.

Because the maximum frequency, especially in the case of nicotine, was not always reached at the highest concentration, we compared the maximum burst frequency evoked by the different drugs irrespective of the drug concentration (Fig. 5C). Pilocarpine alone was as effective as pilocarpine plus proctolin (Fig. 5C), whereas nicotine in the presence of proctolin was more effective than pilocarpine (P < 0.05, Student-Newman-Keuls test; Fig. 5C).

We also compared the results obtained from five preparations that were exposed to 10 FM carbachol. The pattern elicited by 10 PM carbachol had a frequency as high as that

elicited by proctolin plus nicotine, significantly greater than the pilocarpine-induced frequency (maximum burst frequency = 3.12 t 0.16 Hz; P < 0.05, Student-Newman- Keuls test; Fig. 5C). In carbachol the highest burst frequency obtained was 3.64 Hz.

Independent of the pharmacological treatment and burst frequency, the phase relation between ganglia remained rel- atively constant. Figure 6 shows the phase of PS bursts in ganglion A4 relative to A5 in three experiments where different burst frequencies were obtained by varying the concentration of either pilocarpine, nicotine in the presence of proctolin, or carbachol. The phases were within the range observed in spontaneous, “command-driven,” and proctolin-induced motor patterns described in a previous study ( Acevedo 1990).

Muscarinic antagonists did not disrupt proctolin-induced activity

How does the pattern induced by cholinergic drugs relate to that induced by proctolin? Can the proctolin-induced pattern persist in the presence of cholinergic antagonists?

The muscarinic antagonists scopolamine and atropine reliably blocked activity induced by carbachol or pilocarpine (n = 5). Whereas scopolamine was effective at doses as low as 1 PM, > 10 PM atropine was necessary to block rhythmic activity. Figure 7, 1 and 2, shows the blocking of activity in 2 PM carbachol by 1 PM scopolamine. Washes of

0 pilocarpine 0 nicotine + proctolin

carbachol

zi 0.4 - g 0.2 -

V 0

00

00’ . 0.5 1.0 1.5

@ ov V

2.0 2.5 3.0

frequency-s -1

FIG. 6. Phase of PS motor neuron bursts in A4 relative to A5 did not change with frequency or drug. Three different experiments with carbachol ( 1, 2, 4, 5, 7, and 10 PM), .pilocarpine ( 10, 20, and 70 PM), and 10 PM proctolin with nicotine (0.25, 0.5, an.d 1 PM).

G. BRAUN AND B.MULLONEY 2396

1. 2pM carbachol I

ps3 --w++++**+* i I

PS4 m

2. + 1pM scopolamine

-P: :; : : :;

3. 2pM carbachol

4. 1OpM proctolin

5. + 1OOpM scopolamine

FIG. 7. Muscarinic antagonists did not block proctolin-induced activity. Activity induced by 2 PM carbachol ( 1) was blocked by 1 PM scopolamine within 2 min (2). After 2-h wash, activity was restored with carbachol ( 3 ) . Activity induced by 10 ,uM proctolin (4) was not blocked by 100 PM scopolamine ( 5). The burst frequency increased in the presence of scopolamine.

l-2 h were required to restore the pattern in carbachol again (Fig. 7, 3). Subsequent activity induced by 10 PM proctolin persisted in the presence of 100 ,uM scopolamine (Fig. 7,4 and 5). Also the simultaneous application of scopolamine (50 PM) and the nicotinic antagonist mecamylamine in concentrations as high as 50 PM did not interrupt the proctolin-induced activity (n = 2, not shown). There- fore the expression of a proctolin-induced swimmeret motor pattern does not depend on cholinergic neurotransmission. However, during application of atropine or scopolamine the proctolin-induced burst frequency increased from 1.26&0.13Hzto 1.51 &0.17Hz(n=4;P<O.Q5,pairedt test). After washing the frequency decreased to the initial level [ 1.3 t 0.15 Hz]. Application of mecamylamine in concentrations 550 PM had no effect on the burst frequency (n = 3, not shown).

DISCUSSION

Initiation of the motor pattern and modulation of burst frequency

Our results demonstrate that cholinergic agonists can activate the swimmeret motor pattern and modulate the burst frequency in a dose-dependent manner (Figs. l-5 ). In the swimmeret system, activation of the motor pattern by the muscarinic agonist oxotremorine has been reported recently (Chrachri and Neil 1993). We confirmed and ex- tended these findings with the use of a different muscarinic agonist (Figs. 1 and 5 ). Muscarinic induction of patterned activity has also been found in other arthropod locomotor systems including crayfish thoracic ganglia (Chrachri and Clarac 1990), lobster stomatogastric ganglion ( Elson and Selverston 1992; Marder and Eisen 1984), and locust thoracic ganglia (Ryckebusch and Laurent 199 1) , indicating a widespread role for muscarinic receptors in activation of motor systems throughout the arthropods.

An important finding in the present study is that nicotinic receptors also seem to be involved in the frequency modulation of the motor pattern. Nicotine enhanced the burst frequency more effectively than pilocarpine (Fig. SC). Carbachol, which affects both muscarinic and nicotinic receptors, both activated the motor pattern and modulated its burst frequency as effectively as nicotine (Figs. 2B and 5C). With maximum burst frequencies > 3 Hz (Fig. 5C), nicotine and carbachol came close to the frequency range observed in intact crayfish (unpublished results) and the burst frequencies obtained by stimulation of axons in the connectives between the first and second abdominal ganglia in the nerve cord of the lobster (Davis and Kennedy 1972). The motor neuron bursts in different ganglia occurred at constant phase at different frequencies, a prerequi- site for the maintenance of coordinated swimmeret beating (Fig. 6).

Relation ofmuscarinic and nicotinic receptor activation

Pilocarpine elicited and maintained a stable motor pattern in a wide range of concentrations (Figs. 1 and 5 B). In contrast, the modulatory effect of nicotine was limited to a narrow concentration range (Figs. 2A and 5 B). At higher concentrations nicotine disrupted the rhythm, and it produced only tonic activity in resting preparations (see RE- SULTS). A similar dualism of muscarinic and nicotinic effects has been reported for the pyloric rhythm in the foregut of crabs (Marder and Meyrand 1989). Marder and Mey- rand ( 1989) explain this by the different mode of action of muscarinic and nicotinic agonists: pilocarpine activates slow, voltage-dependent conductances (Freschi and Liven- good 1989 ) , whereas nicotine activates “classical” conductances that depolarize the membrane irrespective of the state of individual neurons and thus tends to disrupt rhythmic activity, especially at high concentrations. This explana- tion might also apply to the results reported here. However, the application of agonists to an in vitro preparation only poorly reflects the spatial and temporal patterning of nervous activity in vivo. It seems probable that ACh released by neurons in the crayfish swimmeret system acts on muscarinic and nicotinic receptors and thus can initiate the motor program and modulate its frequency.

Intracellular recordings showed that different swimmeret motor neurons can respond differently to similar doses of nicotine (Fig. 3, A and B). Given the small numbers of neurons tested, we cannot say whether the motor neuron pool can be divided into distinct groups with a different sensitivity to nicotine. However, the example in Fig. 3 B, which shows strong depolarization and changes in the number of spikes per cycle, suggests that this motor neuron could receive cholinergic inputs other than from the central pattern-generating circuit. In motor neurons like this, cholinergic sensory feedback or descending pathways (Davis 1968, 1969) could exert correctional action that would still be phase-locked with the ongoing motor pattern.

Although we cannot rule out the idea that some of the observed effects of cholinergic drugs might reflect the influ- ence of sensory feedback, the activation of resting preparations by muscarinic drugs is not what would be expected of sensory afferents providing information to central circuits. -+


Sensory afferents are commonly modulated by the pattern- -renerating network itself to provide phase-locked feedback kithout disruption of the ongoing activity (Clarac et al. 1992) and central modulation of sensory afferents does occur in the crayfish swimmeret system (Paul 1989). More- over, eserine increased the burst frequency in active preparations, probably by enhancing cholinergic transmission through inhibition of AChE (Fig. 4). Because there was no sensory feedback present in isolated nerve cords and the terminals of sensory afferents were not active, this also suggests that the observed effects of cholinergic drugs reflect the activity of central cholinergic interneurons.

Relation ofcholinergic and proctolinergic excitation of the motor pattern

Previously, application of the peptide proctolin was the only pharmacological way to activate the swimmeret motor pattern in vitro (Mulloney et al. 1987 ) . At least three pairs of descending interneurons, probably originating in the sub- esophageal ganglion, were shown to be immunoreactive for proctolin, and proctolin is released when some of these axons are stimulated ( Acevedo 1990). Thus the application of proctolin in vitro is likely to mimic the activation of the abdominal motor system by higher centers. Neither muscarinic nor nicotinic antagonists blocked the proctolin-induced activity (Fig. 7, and see RESULTS). This suggests that proctolinergic and cholinergic activation of the swimmeret motor system mimic the effects of two different pathways capable of activating the system. Studies on cardiac motor neurons in lobster showed that proctolin and muscarinic agonists elicit quite similar sodium inward currents, although they act on different receptors (Freschi 1989; Fres- chi and Livengood 1989). Therefore it is conceivable that proctolinergic and cholinergic pathways might to some ex- tent converge on common targets in the neural circuitry that produces the swimmeret motor pattern.

Although cholinergic neurotransmission is not necessary for the expression of the proctolin-induced motor pattern, it seems to be involved since the AChE-inhibitor eserine increased the burst frequency in proctolin-activated preparations (Fig. 4). Eserine had no effect on resting preparations (see REsuLTs), so its effects on active preparations imply the activity of cholinergic neurons in the pattern-generating or coordinating circuits (Mulloney et al. 1993; Mur- chison et al. 1993 ) , whose effects were enhanced by inhibit- ing AChE.

A second inference of cholinergic activity in the swimmeret circuits follows from the observations that muscarinic antagonists increased the proctolin-induced burst frequency (Fig. 7) and that the effects of simultaneous application of proctolin and pilocarpine were not additive; rather, proctolin slightly decreased the burst frequency caused by pilocarpine alone (Fig. 5, B and C) . It is conceivable that in the proctolin-activated state the application of muscarinic antagonists abolishes hidden cholinergic transmission and thus increases the burst frequency. The nicotinic antagonist mecamylamine had no detectable effect on the frequency of the proctolin-induced pattern (see RE-

SULTS), which suggests that nicotinic receptors either play a negligible role during low levels of activitv or that the appli-

cation of nicotine activated elements in the circuitry that are not active in the presence of proctolin alone.

Because phasic sensory feedback was absent from the deafferented preparations, we think that the muscarinic pharmacology reflects the presence of cholinergic interneurons in the swimmeret system that can initiate and modulate the expression of the swimmeret motor pattern. Be- cause our knowledge about central cholinergic neurons in arthropods is scarce, a future investigation of the physiology of cholinergic neurotransmission will be of broad in- terest.

We thank D. Murchison, C. Sherff, and K. Sigvardt for reviewing earlier drafts of this manuscript and W. Hall for help with the illustrations.

This work was supported by National Science Foundation Grants BNS 87-19397 and IBN 92-22470 to B. Mulloney.

Address for reprint requests: B. Mulloney, 2320 Storer Hall, UCD, Davis CA 956 16.

Received 5 March 1993; accepted in final form 13 July 1993.

REFERENCES

ACEVEDO, L. D. Control ofthe Expression ofa Simple Behavior: Activation and Inhibition of’the Crayjsh Swimmeret Rhythm by Command Neu- rons and Their Transmitters ( PhD thesis). Davis, CA: Univ. of Califor- nia, 1990.

BARKER, D. L., HERBERT, E., HILDEBRAND, J. G., AND KRAVITZ, E. A. Acetylcholine and lobster sensory neurones. J. Physiol. Lond. 226: 205- 229, 1972.

BARKER, D. L., MURRAY, T. F., SIEBENALLER, J. F., AND MPITSOS, G. J. Characterization of muscarinic cholinergic receptors in the crab nervous system. J. Neurochem. 46: 583-588, 1986.

CHRACHRI, A. AND CLARAC, F. Fictive locomotion in the fourth thoracic ganglion of the crayfish, Procambarus clarkii. J. Neurosci. 10: 707-7 19, 1990.

CHRACHRI, A. AND NEIL, D. M. Interaction and synchronization between two abdominal motor systems in the crayfish. J. Neurophysiol. 69: 1373-1383,1993.

CLARAC, F., EL MANIRA, A., AND CATTAERT, D. Presynaptic control as a mechanism of sensory-motor integration. Curr. Opin. Neurobiol. 2: 764-769,1992.

DAVIS, W. J. Lobster righting responses and their neural control. Proc. R. Sot. Land. B Biol. Sci. 70: 435-456, 1968.

DAVIS, W. J. Reflex organization in the swimmeret system of the lobster. I. Intrasegmental reflexes. J. Exp. Biol. 5 1: 547-563, 1969.

DAVIS, W. J. AND KENNEDY, D. Command interneurons controlling swimmeret movements in the lobster. I. Types of effects on motoneurons. J. Neurophysiol. 35: 1-12, 1972.

ELSON, R. C. AND SELVERSTON, A. I. Mechanisms of gastric rhythm gener- ation in the isolated stomatogastric ganglion of spiny lobsters: bursting pacemaker potentials, synaptic interactions, and muscarinic modulation. J. Neurophysiol. 68: 890-907, 1992.

FLOREY, E. Acetylcholine as sensory transmitter in crustacea. J. Comp. Physiol. 83: 1-16, 1973.

FLOREY, E. AND RATHMAYER, M. Pharmacological characterization of cholinoreceptors of cardiac ganglion cells of crustaceans. Gun. Pharma- col. 11: 47-53, 1980.

FRESCHI, J. E. Proctolin activates a slow, voltage-dependent sodium current in motoneurons of the lobster cardiac ganglion. Neurosci. Lett. 106:188-192, 1989.

FRESCHI, J. E. The effect of subtype-selective muscarinic receptor antagonists on the cholinergic current in motoneurons of the lobster cardiac ganglion. Brain Res. 552: 87-92, 199 1.

FRESCHI, J. E. AND LIVENGOOD, D. R. Membrane current underlying muscarinic cholinergic excitation of motoneurons in lobster cardiac ganglion. J. Neurophysiol. 62: 984-995, 1989.

HUGHES, G. M. AND WIERSMA, C. A. Theco-ordination of swimmeret movement in the crayfish, Procambarus clarkii (Girard). J. Exp. Biol. 37:657-670, 1960.

MARDER, E. Cholinergic motoneurons in the stomatogastric system of the lobster. J. Physiol. Lond. 257: 63-86, 1976.

2398 G. BRAUN AND B. MULLONEY

MARDER, E. AND EISEN, J. S. Electrically coupled pacemaker neurons respond differently to the same physiological inputs and neurotransmit- ters. J. Neurophysiol. 5 1: 1363- 1374, 1984.

MARDER, E. AND MEYRAND, P. Chemical modulation of an oscillatory neural circuit. In Neuronal and Cellular Oscillators, edited by J. W. Jacklet. New York: Dekker, 1989, p. 317-337.

MARDER, E. AND PAUPARDIN-TRITSCH, D.The pharmacologicalproper- ties of some crustacean neuronal acetylcholine, T-aminobutyric acid, and L-glutamate responses. J. Physiol. Lond. 280: 2 13-236, 1978.

MAYNARD, E. A. Microscopic localization of cholinesterases in the nervous system of the lobsters Panulirus argus and Homarus americanus. Tissue Cell 3: 2 15-250, 197 1.

MULLONEY, B., ACEVEDO, L.D., AND BRADBURY, A.G. Modulation of the crayfish swimmeret rhythm by octopamine and the neuropeptide proctolin. J. Neurophysiol. 58: 584-597, 1987.

MULLONEY, B., ACEVEDO, L. D., CHRACHRT, A., HALL, W. M., AND SHERFF, C. M. A confederation of neural circuits: control of swimmeret movements by a modular system of pattern generators. In: Frontiers in Crustacean Neurobiology, edited by K. Wiese et al. Basel, Birkhauser, 1990, p. 439-447.

MULLONEY, B. AND HALL, W. The PD programs: a method for the quanti-

tative description of motor patterns. J. Neurosci. Methods 19: 47-59, 1987.

MULLONEY, B., MURCHISON, D., AND CHRACHRI, A. Modularorganiza- tion of pattern-generating circuits in a segmental motor system: the b swimmerets of crayfish. Semin. Neurosci. 5: 49-57, 1993.

MURCHISON, D. CHRACHRI, A, AND MULLONEY, B. A separatelocalpat- tern-generating circuit controls the movements of each swimmeret in crayfish. J. Neurophysiol. 70: 2620-263 1, 1993.

PAUL, D. H. Nonspiking stretch receptors of the crayfish swimmeret receive an efference copy of the central motor pattern for the swimmeret. J. Exp. Biol. 14 1: 257-264, 1989.

PFEIFFER-LYNN, C. AND GLANTZ, R. A neuronal nicotinic acetylcholine = receptor in crayfish neurons. Neurosci. Lett. 120: 234-236, 1990.

RYCKEBUSCH, S. AND LAURENT, G. Pilocarpine induces rhythmic activity in leg motoneurons in isolated thoracic ganglia of locusts. Sot. Neurusci. Abstr. 17: 1581, 1991.

STEIN, P. S. Intersegmental coordination of the swimmeret motoneuron activity in crayfish. J. Neurophysiol. 34: 3 1 O-3 18, 197 1.

WIERSMA, C. A. G. AND IKEDA, K. Interneurons commanding swimmeret movements in the crayfish. Comp. Biochem. Physiol. 12: 509-525, 1964.

Documents

Cholinergic Modulation of the Swimmeret Motor System in