The Pharmacology of Cholinoceptors on the Somatic Muscle Cells of the Parasitic Nematode Ascaris Suum

The presence of acetylcholine (ACh) (Mellanby, 1955), the excitatory action of ACh on the muscle cells (Del Castillo et al. 1963) and the presence of cholinesterase in the nervous system (Lee, 1962; Knowles and Casida, 1966) have led to the suggestion that ACh is the excitatory transmitter at the neuromuscular junction in Ascaris suum.

The nematodes are probably the lowest phylum of the evolutionary tree to use ACh as a transmitter (see Venter et al. 1988, for a review). The receptor for ACh on Ascaris muscle cells is, therefore, of particular interest from a phylogenetic standpoint. It has also been demonstrated that the primary site of action of the anthelmintics pyrantel, morantel and levamisole is the Ascaris ACh receptor (Aubry et al. 1970; Harrow and Gration, 1985). Thus, an understanding of the relationships between structure and function at this receptor is important for the development of anthelmintics.

Most of the early studies on the Ascaris ACh receptor have used a muscle strip preparation. The receptor has not been characterised by electrophysiological means. Previous studies on the pharmacology of this receptor using the muscle strip preparation have given a preliminary basis for its classification. The nicotinic agonists nicotine and dimethylphenylpiperazinium (DMPP) are good at eliciting muscle contraction, whereas muscarinic agonists are weak or ineffective (Baldwin and Moyle, 1949; Natoff, 1969; Rozhkova et al. 1980) and, on the basis of these studies, it has been proposed that the receptor in Ascaris is of the nicotinic type. Tubocurare is an antagonist, though of varying potency (Del Castillo et al. 1963; Baldwin and Moyle, 1949; Natoff, 1969; Rhozkova et al. 1980). The selective muscarinic antagonist atropine (Baldwin and Moyle, 1949; Natoff, 1969; Rozhkova et al. 1980) and the ganglion blocker hexamethonium are weak antagonists of the ACh-elicited contraction (Natoff, 1969; Rhozhkova et al. 1980).

In this study we have used intracellular recordings of Ascaris muscle cells to look at the direct effect of various compounds and to provide a more complete picture of the pharmacology of the cholinoceptor in this preparation. We highlight some important differences between mammalian, insect and Ascaris nicotinic receptors.

Ascaris suum were obtained from a local abattoir on a weekly basis. They were transported to the laboratory in a flask containing artificial perienteric fluid (APF) the composition of which was (in mmol 1⁻¹) NaCl, 67; sodium acetate, 67; KC1, 3; MgCl₂, 15.7; CaCl₂, 3; Tris base, 5; adjusted to pH 7.6 with glacial acetic acid. In the laboratory they were maintained in APF with 3 mmol 1⁻¹ glucose at 37°C in a water bath. The worms remained healthy for up to 1 week, as indicated by their appearance and the resting membrane potentials (–30 mV) and input conductances (2–3μS) of the somatic muscle cells.

An anterior section (approximately 1 cm) of the worm was excised and slit along one lateral line. The section was pinned out, cuticle side down, in a Perspen perfusion chamber (volume 2 ml). The intestine was carefully removed using fine forceps, revealing the muscle bag cells underneath. The preparation was continuously perfused with APF at 10 ml min⁻¹. The temperature in the bath was 32–33°C. The muscle bag cells (diameter 50–200 pu) were impaled with two glass microelectrodes (10–30 MΩ), filled with 4 mol 1⁻¹ potassium acetate and 10 mmol 1⁻¹ KC1 and connected to the headstage of an Axoclamp 2A (Axon Instruments) by Ag/AgCl wire. The reference electrode was a 3 mol 1⁻¹ KCl/agar bridge. Membrane potential was routinely monitored with the Axoclamp in bridge mode and current pulses (10–40 nA, 500ms pulse width, 0.1 Hz) were passed through the second microelectrode using a Grass stimulator and Grass stimulus isolation unit. The true current being passed through the current electrode was constantly monitored and permanent records of both the current stimulus and the muscle cell membrane potential were obtained on a Gould two-channel chart recorder. Membrane potential was monitored on a Gould digital storage oscilloscope. Smaller cells near the nerve cord were chosen for voltage-clamp experiments (approx, diameter 50μm). These experiments were performed with the Axoclamp in two-electrode voltage-clamp mode.

Drugs were applied directly to the area of the cell from which a recording was being made, via a fine-bore tube, at a rate of 10 ml min⁻¹. Switching from APF to the drug (by moving the tubing from one beaker to the other) introduced a bubble into the tubing, which separated the two solutions. To minimise mechanical disturbance, the bubble was allowed to escape from the tubing via a small slit just prior to entry to the bath. Temperature was maintained at a constant 32–33°C during drug application by keeping the drug solution in the same water bath as the APF. In addition, temperature was monitored by a fine temperature probe placed near to the cell under study. This method of drug application enabled the concentration of drug around the cell to be changed rapidly and completely, allowing fast on and off switching of responses. Agonists were applied for up to 45 s, with the duration of application being consistent for each individual experiment. Antagonists were applied for 2 min prior to the application of the agonist and concurrent with the application of the agonist.

Application of agonists to the muscle preparation causes muscle cell depolarization, which in many cases results in contraction and subsequent displacement of the recording electrode from the cell. To minimise this problem and to enable a full dose-response relationship for ACh to be studied, we investigated the action of higher concentrations of ACh in Ca²⁺-free APF in some experiments. (The composition of this was identical to that of normal APF except that CaCl₂ was omitted and 1 mmol 1⁻¹ EGTA was added.) The protocol of these experiments was as follows. The cell was impaled and the effect of ACh (1–10μmoll⁻¹) in the Ca²⁺-containing APF was determined. The perfusing medium was then changed to Ca²⁺-free APF. The cell was bathed in this medium for 15 min, which was probably sufficient to deplete extracellular Ca²⁺ levels (as indicated by the inhibition of the Ca²⁺-mediated spontaneous activity; see Fig. 2). The doseresponse relationship for 10⁻⁶–10⁻³moll⁻¹ ACh was then determined. The relative potency of the agonist nicotine compared to ACh was also studied in Ca²⁺-free APF.

The potency of agonists was expressed relative to that of ACh in the following manner: the response, either conductance increase or depolarization, to increasing concentrations (1–30 μmol 1⁻¹) of ACh was determined and plotted as a dose-response curve. This was repeated for the compound being assessed. Relative potency of the agonist compared to ACh was determined by taking the ratio of the concentration of ACh to the concentration of agonist that produced equivalent responses from approximately parallel portions of the dose-response curves. The sensitivity of the cells to ACh did not change appreciably during the course of the experiment, as tested by the application of ACh at the end of each experiment, though the depolarization sometimes diminished slightly.

The potency of the antagonists was determined as an IC₅₀ value. The IC₅₀ value was the concentration of antagonist that reduced the response to a submaximal concentration of ACh (5–10μmoll⁻¹) by 50% and was determined using at least four concentrations of antagonist.

In order to pool the data for agonists the results were normalized with respect to the response to 5μmoll⁻¹ ACh for each cell. All results are expressed as the mean±s.E.M. for N experiments.

Drugs were obtained from the following sources; acetylcholine bromide, nicotine, trimethylammonium, bethanechol, methacholine, pilocarpine, quina- crine, curare, atropine, hexamethonium, neostigmine and physostigmine from Sigma; DMPP, TPP and Decamethonium from Aldrich; carbachol from KochLight; muscarone and muscarine from Ciba-Geigy; arecoline from BDH; McN-A-343 from Research Biochemicals. We are grateful to the following for their generous gifts of compounds: Smith Kline & French for furtrethonium; Merck Sharp & Dohme for mecamylamine; Ciba-Geigy for chlorisondamine; Organon for pancuronium; Roche Products Ltd for trimethaphan; Dr M. Caulfield (University College London) for HPPT (metahydroxyphenylpropyltrimethylammonium); Bayer for benzoquinonium; and Merck & Co. for dihydro-β-erythroidine.

The typical resting membrane potential of the somatic muscle cells was −30 mV Experiments were only performed on those cells with resting input conductances of less than 3 μS and a resting membrane potential greater than −25 mV, as this was taken to indicate that a reasonably healthy impalement of the cell had been achieved. Some cells, particularly those near the nerve cord, exhibited spontaneous activity, which took the form of either ‘slow’ waves or action potentials of an amplitude up to 40 mV. Acetylcholine depolarized the cells with a threshold of around 1 μmoll⁻¹. The response to ACh did not show any signs of desensitization at the concentrations used in these studies.

Cells varied in the amount of rectification that they exhibited. Generally, the cells showed little rectification in the hyperpolarizing direction, particularly if the applied hyperpolarizing current was below 30 nA (Fig. 1A,B). However, current voltage plots indicate rectification in the depolarizing direction in some cells and, as we have not been able routinely to clamp the membrane potential of these cells, it is important to bear this in mind when using the increase in input conductance as a measure of the activation of the ACh receptor. For example, in one cell ACh (10 μmol 1⁻¹) elicited a depolarization of 10 mV and an increase in conductance of 0.82 μS. When this cell was stepped by current injection alone to the same potential to which it depolarized in the presence of ACh, the input conductance increased by 0.27 μS. Thus, 33% of the increase in input conductance apparently elicited by ACh could have been secondary to the ACh-induced depolarization. For this reason, data on the effect of ACh on membrane potential have been included in this paper.

Typical responses to ACh and their dose-response relationships are shown in Fig. 2. ACh produced a dose-dependent and reproducible depolarization and increase in input conductance of the muscle cells. In Ca²⁺-free APF, the EC₅₀ (the concentration that produced a half-maximal response) for ACh to elicit a depolarization was 10±2μmoll⁻¹ (N=8; Fig. 2B). A value could not be obtained in Ca²⁺-containing APF as the recording was not stable when high concentrations of ACh were applied in this medium. The EC₅₀ for the ACh-elicited conductance increase could not be estimated either, as it continued to increase even at 1 mmol 1⁻¹ ACh (Fig. 2C). At the lower concentrations of ACh, the dose response curves in the presence and absence of Ca²⁺ were not significantly different (Fig. 2B,C).

Voltage-clamp experiments indicated that ACh elicited an inward current of between 7 and 29 nA (Figs 2D, 11) when the cells were voltage-clamped at resting membrane potential.

The relative potencies of the agonists studied are given in Table 1 and Figs 3 and 4. For all compounds except nicotine these were calculated in Ca²⁺-containing APF only. Nicotine mimicked the effect of ACh, though with a lower potency. The relative potency of nicotine compared to ACh was studied in the presence and absence of Ca²⁺. The values in the presence of Ca²⁺ are given in Table 1. The values in the absence of Ca²⁺ were 0.08 for the depolarization and 0.008 for the conductance increase. Both these values are lower than the relative potency for nicotine obtained in the presence of Ca²⁺. In all seven cells studied (four in the presence and three in the absence of Ca²⁺) it was found that the response to ACh was reduced after the cell had been exposed to nicotine (Fig. 5).

For the other compounds studied, the relative potencies determined from the depolarization and from the conductance increase are comparable (Table 1). HPPT and DMPP were the only compounds that were more potent than ACh (Fig. 4). TMA⁺ was less potent than ACh. The duration of the response to carbachol was greater than that to ACh (Fig. 6B). Muscarone was 100 times less potent than ACh. Arecoline and furtrethonium both elicited depolarizations and conductance increases with about one-thousandth of the potency of ACh. McN-A-343 was virtually inactive, requiring a concentration of greater than 1 mmol 1⁻¹ to elicit a depolarization of about 2 mV. Bethanechol and methacholine were without effect. Pilocarpine and muscarine caused a slight (up to 3 mV) hyperpolarization with a threshold for the response of 500 μmol 1⁻¹. The muscarinic agonist oxotremorine (1 mmol 1⁻¹) had no effect on two cells. NMTHT was without effect at concentrations up to 1 mmol 1⁻¹ on three cells.

The effects of two anticholinesterases were investigated. Neither physostigmine nor neostigmine had any significant effect on resting membrane potential or input conductance when applied on their own at concentrations up to 10μmoll⁻¹. Physostigmine had no effect on the response to ACh at low concentrations; however, at concentrations greater than 10μmoll⁻¹ it blocked the response to ACh in a reversible manner (V=3). Neostigmine potentiated the response to ACh at concentrations of 1–10μmoll⁻¹ (Fig. 6A,C).

The potencies of the antagonists at blocking the depolarization and the conductance increase in response to ACh are given in Table 2. For most compounds these are comparable; however, for both atropine and chlorisondamine the IC₅₀ values calculated from the conductance change are lower than the IC₅₀ values calculated from the block of the depolarization. For example, the IC₅₀ jjetermined for atropine from the conductance response was 6.7±2.1μmol 1⁻¹ (N=4, determined by extrapolation) compared to 52±4 μmoll⁻¹ (N=4) determined from the ability to block the depolarization (Fig. 7).

Pancuronium and curare antagonised the ACh response. The antagonism by curare was readily reversible, with 100% recovery in 5–15min (Fig. 8A). The block by atropine and pancuronium lasted longer, requiring a wash of 20 min for a recovery of greater than 50%.

Mecamylamine was a potent antagonist of the ACh response (Table 2; Fig. 8B). The block of the ACh response was reversed slowly. After a 5–15 min wash it reversed by about 30% in six cells (Fig. 8B). The block reversed completely in one cell washed for longer than 30min. The other ganglion blockers tested in this study, trimethaphan and chlorisondamine, also antagonised the ACh response (Table 2), though less potently than mecamylamine.

The relative potencies of hexamethonium and Decamethonium as blockers of the ACh response were studied in three cells. Decamethonium was found to be more potent than hexamethonium (Fig. 9; Table 2). It also had a direct, depolarizing effect on membrane potential (N=3, Fig. 10).

Quinacrine blocked the inward current elicited by ACh with an IC₅₀ of 2.05±0.60μmoll⁻¹ (N=3, Fig. 11). TPP also blocked the current elicited by ACh with an IC₅₀ of 1.9±0.3 μmol1⁻¹ (N=3; Fig. 12).

Dihydro-β-erythroidine (1 mmol 1⁻¹) did not block the action of ACh on the muscle cells. However, the other neuromuscular blocking agent, benzoquinonium, produced a potent, reversible block of the ACh response (N=3; Table 2).

These results confirm the excitatory nature of ACh on Ascaris muscle cells. It is suspected that ACh is the excitatory transmitter at the Ascaris neuromuscular junction. However, it should be noted that the responses to bath-applied ACh in the preparation described in this paper will involve both the physiologically important synaptic ACh receptors and the extrasynaptic ACh receptors on the muscle bag cells. Pharmacological differences may exist between the synaptic and extrasynaptic ACh receptors.

The EC₅₀ for ACh at this receptor could not be determined exactly in all cells because the recordings made after application of concentrations of ACh capable of eliciting a maximal response in the presence of Ca²⁺ were unstable. However, it can be estimated to be in the low micromolar range. From eight studies in the absence of Ca²⁺, the ECw for the depolarization was determined to be 10μmoll⁻¹. The effect of ACh on the cell input conductance was not maximal even at 1 mmol 1⁻¹. Harrow and Gration (1985) estimated that the EC₅₀ of ACh for eliciting a conductance increase was 110μmoll⁻¹. This may indicate that there are a large number of ‘spare’ receptors for ACh in this preparation, i.e. the maximal depolarization occurs at a concentration of ACh at which all the available receptors have not been activated. The methodological problems of measuring the conductance increase in a cell that is not voltage-clamped were dealt with in the Materials and methods section. We measured both depolarization and conductance increase in response to ACh application.

We have not reported here on the ionic mechanism of this response. However, ACh elicits an inward current in voltage-clamped cells (Figs 2D, 11) and we have also found that the response is decreased when Na⁺ is partly replaced by glucosamine in the external medium (L. Colquhoun, L. Holden-Dye and R. J. Walker, unpublished observations). Harrow and Gration (1985) have estimated that the reversal potential for the ACh current is about +10 mV, which is consistent with the receptor gating a cation channel through which Na⁺ is probably the major charge carrier. Therefore, by analogy with other nicotinic receptors, the Ascaris receptor seems to be a ligand-gated cation channel. The nicotinic nature of this receptor has been confirmed by the effects of agonists, as described below.

In this study the magnitude and duration of the ACh response were increased by the anticholinesterase neostigmine, but not by physostigmine. Physostigmine (100μmol 1⁻¹) inhibited the maximum response to ACh by more than 50%. The different effects of these two anticholinesterases may be explained by their differential affinity for the cholinesterase binding site and the nicotinic receptor. It has been shown in other preparations that these anticholinesterases can have a direct action at the ACh receptor site (Slater et al. 1986), and in Aplysia neurones the blockade of the receptor by physostigmine can occur at quite low concentrations (Oyama et al. 1989). It would seem, therefore, that the duration of action of ACh in the Ascaris muscle preparation is regulated by the presence of a cholinesterase susceptible to blockade by neostigmine. The precise mechanism of action of physotigmine in this tissue will require further study. It is likely that it acts both as a cholinesterase inhibitor and as a receptor antagonist. The physiological relevance of the cholinesterase in terminating the action of synaptic ACh will have to be elucidated by looking at the possible effect of neostigmine on the excitatory junction potentials.

We did not routinely employ a Ca²⁺-free medium to study the relative potencies of agonists. The potency of nicotine on the Ascaris muscle preparation was apparently lower in the absence of Ca²⁺. The significance of this has not been investigated. The kinetics of the nicotine response also seemed to be much slower than that for ACh (Fig. 5). Nicotine was the only agonist that seemed to elicit a form of desensitization of the Ascaris muscle, in that the response to ACh was reduced after exposure of the cell to nicotine (Fig. 5). Therefore, the action of nicotine at the ACh receptor seems to possess some interesting features that would be best investigated at the single-channel level using the patch-clamp technique.

The most effective cholinomimetic agents in Ascaris muscle were nicotinic agonists (Table 1). However, nicotine itself was not as potent as might be expected. A similar low potency of nicotine, at an otherwise ‘nicotinic’ receptor, in an invertebrate preparation has been shown in Manduca sexta (Trimmer and Weeks, 1989). In the central nervous system of Manduca the potency of nicotine is low, with an EC50 of around 20μmoll⁻¹. This may be due to the presence in Manduca of a nicotine-resistant form of the receptor.

Muscarone, a compound with slight nicotinic activity, was a very weak agonist. The muscarinic agonists bethanechol and methanechol were completely inactive. This agrees well with past results obtained from the muscle strip preparation showing that the nicotinic ganglionic agonist DMPP was potent at eliciting Ascaris muscle strip contraction (Natoff, 1969; Rozhkova et al. 1980; Hayashi et al. 1980). In our preparation too, DMPP was one of the most potent agonists. This would suggest that the Ascaris nicotinic receptor is like the nicotinic receptor at mammalian autonomic ganglia. However, it should also be noted that we found that Decamethonium had a slight agonist action on Ascaris muscle (Fig. 10). Decamethonium is not an agonist at mammalian ganglia (Paton and Perry, 1953; Ascher et al. 1979; Gurney and Rang, 1984), though it has been shown to be an agonist at frog ganglia (Lipscombe and Rang, 1988). Also, the slight stimulatory action of furtrethonium was surprising, as this compound is generally regarded as a muscarinic agonist. Thus, even from an agonist profile, the Ascaris ACh receptor does not readily fit into a mammalian classification scheme.

HPPT is one of a series of compounds synthesized by Barlow and Thompson (1969) and shown to be 50 times more potent than nicotine as an agonist in the frog rectus muscle. In rat superior cervical ganglion it is 15 times more potent than DMPP (Caulfield et al. 1990). In our preparation this compound was the most potent cholinomimetic tested, being at least twice as potent as DMPP.

The observation that pilocarpine and muscarine produced a weak hyperpolarization is interesting, particularly in view of the fact that muscarinic receptors, of lower sensitivity than the nicotinic ACh receptors, are known to coexist with nicotinic receptors on other invertebrate cell types (Woodruff et al. 1971; Benson, 1988). We have not pursued this possibility further in this study; however, the potent muscarinic agonist, oxotremorine, was without effect on Ascaris muscle cells.

The nitromethylene insecticide NMTHT had no effect on the Ascaris muscle cells. It has been shown to be cholinomimetic on cultured cockroach neurons where it opens channels with the characteristics of nicotinic receptor-gated channels at a low micromolar concentration (Buckingham et al. 1989).

None of the antagonists that blocked the response to ACh was very potent. Of the ganglion blockers studied, mecamylamine was the most potent compound and both chlorisondamine and trimethaphan were relatively weak antagonists. Although mecamylamine is widely regarded as a ganglion blocker (see Martin et al. 1989, for a review) at the concentrations that were effective in Ascaris, it has also been shown to block transmission at the frog neuromuscular junction in a manner characteristic of open-channel block (Varanda et al. 1985). Mecamylamine is one of the most potent antagonists of the nicotinic ACh response on the cockroach D_f motoneurone and acts in a voltage-independent manner (David and Sattelle, 1984). It has also recently been shown to be a potent antagonist of the ACh response in Xenopus oocytes expressing the α-4 and non-anicotinic receptor subunits cloned from rat brain (Bertrand et al. 1990). Therefore, it can be seen that our results with mecamylamine do not help us to classify the Ascaris muscle nicotinic receptor as a neuromuscular junction type, ganglionic type or brain type of nicotinic receptor. In future experiments it would be interesting to test the voltage-dependency of mecamylamine action at the Ascaris ACh receptor, as it would seem that a voltage-dependent action may be typical of the nicotinic neuromuscular receptor.

Two of the antagonists, chlorisondamine and atropine, show a marked discrepancy in their ability to block the conductance increase and the depolarization elicited by ACh. Both compounds block the conductance increase more potently than the depolarization (Table 2; Fig. 7). It has been shown that 50 μmol 1⁻¹ atropine is required to block the ACh-stimulated contraction in Ascaris muscle strip (Baldwin and Moyle, 1949; Rozhkova et al. 1980; Natoff, 1969; Onuaguluchi, 1989; Hayashi et al. 1980). The relatively weak antagonism of the contractile response and the muscle cell depolarization by atropine is compatible with the receptor being of a nicotinic rather than a muscarinic subtype. However, the receptor shows little discrimination between curare and atropine, with curare blocking the ACh-induced depolarization with an IC₅₀ of about 5μmoll⁻¹. A similar lack of discrimination by a nicotinic receptor between these two compounds has also been noted in leech neuropile glial cells (Ballanyani and Schlue, 1989) and cockroach D_f motoneurone (David and Sattelle, 1984).

The nicotinic receptor in Ascaris also fails to distinguish between the neuromuscular blocking agents curare and pancuronium. Pancuronium is a steroidal non-depolarising antagonist at the vertebrate neuromuscular junction of 10-fold greater potency than curare (Buckett et al. 1968), but it has the same potency as curare at the Ascaris neuromuscular junction.

Of the other neuromuscular junction blockers tested, benzoquinonium and dihydro β-erythroidine, only benzoquinonium was a potent antagonist. Dihydroβ-erythroidine had no discernible effect even at 100 μmol 1⁻¹. The lack of action of dihydro-β erythroidine is of interest as it is a nondepolarising neuromuscular junction blocker, similar in mechanism of action to curare. However, unlike curare and other neuromuscular junction blockers, it is a tertiary rather than a quaternary compound.

Tetraphenylphosphonium (TPP) was tested against the ACh response in Ascaris because of its known potency in blocking nicotinic gated cation channels (reviewed in Heidmann et al. 1983; Spivak & Alburquerque, 1985). Trimethyl phenylphosphonium, belonging to the same group of compounds as TPP, binds to the extracellular beginning of the M2 transmembrane domain of the Torpedo nicotinic receptor (Hucho et al. 1986). In Ascaris muscle cells voltage-clamped at the resting membrane potential, TPP blocked the inward current elicited by ACh in a dose-dependent manner (Fig. 12). Thus, the Ascaris nicotinic receptor channel may have similar recognition sites to those that have been extensively characterised on the Torpedo receptor at the molecular level.

Quinacrine (mepacrine) is another noncompetitive inhibitor of nicotinic receptors, although it is also known to have an action at the ACh recognition site (Tsai et al. 1979). This compound was a potent inhibitor of the ACh response in Ascaris. As well as having a clinical use as an antimalarial, this drug is also used as a vermifuge, mainly for tapeworm infections. In this context, its potent blockade of the Ascaris nicotinic receptor is of some interest.

The bisquaternary methonium compounds hexamethonium and Decamethonium were both weak antagonists of the ACh response in Ascaris. However, Decamethonium was more potent than hexamethonium (Fig. 9). At mammalian ganglia, hexamethonium blocks transmission by entering the nicotinic channel (Gurney and Rang, 1984), whereas hexamethonium has little effect on transmission at the vertebrate neuromuscular junction (Milne and Byrne, 1981). Thus, the low potency of hexamethonium implies that the nicotinic channel in Ascaris lacks a recognition site that is present in the channel of mammalian ganglionic nicotinic receptors. As mentioned previously, the weak agonist action of Decamethonium in Ascaris is similar qualitatively, if not quantitiatively, to its action at the neuromuscular junction.

Little is known of the pharmacology of the nicotinic receptors in other nematodes. Radioligand binding studies in the free-living nematode Caenorhabditis elegans identified a binding site with a high affinity for the levamisole group of compounds. This site also had a micromolar affinity for DMPP (Lewis et al. 1987).

In one species where electrophysiological studies on the pharmacology of the muscle cell receptors have been performed, Dipetalonema viteae, the ACh response seems to be very different from that in Ascaris: both arecoline and pilocarpine were cholinomimetic (Rohrer et al. 1988) and the response was not blocked by curare (10μmol 1⁻¹). It did resemble the Ascaris receptor, however,that the response was not blocked by hexamethonium. In the trematode Schistosoma mansoni there is believed to be a cholinoceptor with similar pharmacology to that of the receptors in autonomic ganglia (Barker et al. 1966).

Nicotinic receptors are present on leech muscle (Walker et al. 1970), neurones (Woodruff et al. 1971) and glial cells (Ballanyi & Schlue, 1989). All three types of receptor are insensitive to hexamethonium, as in Ascaris. Decamethonium had a partial agonist action at the receptor on leech glia, but no direct action on leech muscle. Thus, the receptor on the glia may most closely resemble the Ascaris receptor in terms of recognition of the bisquatemary methonium compounds.

The nicotinic receptor in Ascaris differs from those of certain insect preparations in a number of ways. The Df motoneurone of the cockroach is insensitive to DMPP (David and Sattelle, 1984), whereas it was the most potent agonist in Ascaris. A further difference is indicated by the lack of sensitivity of the Ascaris ACh receptor to dihydro-β-erythroidine, whereas this is one of the most potent antagonists on the D_f neurone and is also a potent displacer of [¹²⁵I]-a-bungarotoxin binding to locust ganglia (Macallan et al. 1988). Levamisole, pyrantel and morantel (Harrow and Gration, 1985) are more potent than ACh at the Ascaris ACh receptor but are only weakly effective on the Df motoneurone (Pinnock et al. 1988). However, the receptors in insects and Ascaris share a common sensitivity to mecamylamine and benzoquinonium, though, as mentioned earlier, neither of these antagonists was very potent on either preparation. Decamethonium is an agonist in Ascaris, whereas it is a weak antagonist in the D_f motoneurone (David and Sattelle, 1984). Both preparations are insensitive to hexamethonium. Dorsal unpaired median (DUM) cells that respond to ACh with a depolarization have been identified in the metathoracic ganglia of grasshoppers (Goodman & Spitzer, 1979). The depolarization was blocked by curare and by high (millimolar) concentrations of hexamethonium. The nicotinic receptor at an identified synapse in Manduca sexta has some of the characteristics of the Ascaris nicotinic receptor (Trimmer and Weeks, 1989), as Decamethonium was an agonist and hexamethonium did not block the ACh response. Mecamylamine and curare were equipotent, though not very good, antagonists.

ACh receptors have also been described in molluscs. Three types of response to ACh have been studied in Aplysia (Kehoe, 1972). The excitatory response is blocked by 50–100 μmol 1⁻¹ curare (Kehoe, 1972; Ascher et al. 1978) and 10–100μmoll⁻¹ hexamethonium. The studies on these receptors have concentrated on the mechanism of action of antagonists rather than the pharmacological profile of the receptor. It is interesting, however, that the receptor mediating the excitatory response to ACh in Aplysia is blocked by hexamethonium with moderate potency, as is the excitatory response to ACh in Limulus polyphemus (Walker and Roberts, 1982). Yavari et al. (1979) estimated pA₂ values for compounds that block the inhibitory and excitatory response to ACh in Helix aspersa. The excitatory response was weakly antagonised by hexamethonium, whereas the inhibitory response was completely unaffected. Curare was a weak antagonist of both responses.

In conclusion, common features of the nicotinic receptors in invertebrates would seem to be a poor recognition of the potent ganglion blocker hexamethonium and an inability to discriminate well between curare and atropine. There also seem to be marked interspecies differences. In this study we have characterised the somatic muscle cell ACh receptor in Ascaris. As with the receptors in other invertebrates, it cannot readily be classified according to mammalian nomenclature. The true identity of these receptors, and their relationship to those in higher groups such as mammals, will not be elucidated until the nicotinic receptor subunits for these species have been cloned and sequenced. The information from the molecular biological approach, together with the pharmacological data outlined above, will enable the relationships between structure and function to be determined for the invertebrate nicotinic receptors.

We gratefully acknowledge financial support from the SERC and Jersey Government. We are indebted to Chris Willis for a regular supply of healthy Ascaris.