Voltage-Independent Block of a Neuronal Nicotinic Acetylcholine Receptor by N-Methyl Lycaconitine

The principal toxin of the plant Delphinium brownii is the alkaloid /V-methyl lycaconitine (Fig. 1) and its blocking actions on vertebrate cholinergic pathways have been described (Nambi Aiyar et al. 1979). By means of an external-electrode recording method based on the sucrose-gap technique of Callec & Sattelle (1973), it has recently been shown that micromolar concentrations of N-methyl lycaconitine block synaptic transmission at cholinergic synapses of the cockroach Periplaneta americana (Jennings et al. 1987). The same authors showed the alkaloid to be an inhibitor of [³H]-αa-bungarotoxin binding sites in housefly (Musca domestica) head extracts. However, it is not known if synaptic blockade at cockroach nicotinic cholinergic synapses (Sattelle, 1978; Sattelle et al. 1983) is due to pre-or postsynaptic actions of the toxin. Though an action on a putative nicotinic cholinergic receptor can be inferred, there is no direct evidence for either an agonist action or an antagonist effect at a functional insect nicotinic acetylcholine receptor.

To test for a receptor action of N-methyl lycaconitine, its actions are examined on the nicotinic acetylcholine response (David & Sattelle, 1984) recorded from the cell body membrane of the cockroach fast coxal depressor motor neurone (D_f). A population of nicotinic acetylcholine receptors that are blocked by α-bungarotoxin and κ-bungarotoxin account for the cholinergic chemosensitivity of the cell body membrane of this motor neurone (David & Sattelle, 1984; Pinnock et al. 1988). The cell body receives no direct synaptic input, thereby facilitating interpretation of pharmacological data. Voltage-clamp techniques can be applied, and this method, in combination with radioligand binding experiments on the same insect tissue, permits a detailed analysis of neuronal nicotinic acetylcholine receptor actions of N-methyl lycaconitine.

Adult male cockroaches (Periplaneta americana) were used throughout the investigation. They were obtained from cultures maintained at 27°C with access to unlimited food and water.

The isolated, desheathed metathoracic ganglion of the cockroach Periplaneta americana was mounted under saline (composition: 214 mmol I⁻¹ NaCl; 3 mmol I⁻¹ KCl; 5 mmol I⁻¹ CaC12; 4mmoll⁻¹MgCl₂; 50 mmol I⁻¹ sucrose; 10mmoll⁻¹ TES; pH7·4) in a Perspex experimental chamber (total volume 0·3 ml). The cell body of the fast coxal depressor motor neurone D_f (Pearson & Iles, 1970, 1971) was located visually and then impaled with two microelectrodes (filled with 0·5moll⁻¹ K₂SO₄, 5·0mmoll¹ KCl) each of resistance 20–25 MΩ. Changes in membrane potential, current and resistance were monitored as described in detail elsewhere (David & Sattelle, 1984). Acetylcholine was applied ionophoretically from a micropipette (resistance 5–15 MΩ) filled with 1·0 mol I⁻¹ acetylcholine chloride (pH 4·0). A retaining current of 50 nA was used to prevent leakage of the drug from the ionophoretic pipette.

N-Methyl lycaconitine citrate was the generous gift of Dr M. H. Benn of the Chemistry Department, University of Calgary, Alberta, Canada.

Results

N-methyl lycaconitine citrate caused a dose-dependent inhibition of specific [³H]-α-bungarotoxin binding to cockroach ventral nerve cord homogenates (Fig. 2). The IC₅₀ value of 2·0×10⁻⁹mol⁻¹, determined graphically, yields a K_i value of l·4×10⁻⁹moll⁻¹. Inhibition curves for cholinergic ligands (α-bungarotoxin and d-tubocurarine) known to block functional cockroach nicotinic acetylcholine receptors (Sattelle, 1986) are also illustrated. For comparison, the actions of a muscarinic antagonist (atropine) and a nicotinic agonist (nicotine) are also included (Fig. 2).

α-Bungarotoxin, the most effective of all ligands tested at insect neuronal Nicotinic receptors (Sattelle et al. 1980, 1983; David & Sattelle, 1984), is the only ligand of comparable effectiveness to N-methyl lycaconitine. Both ligands are much more effective than d-tubocurarine in inhibiting the binding of [³H]-α-bungarotoxin to cockroach nerve cord extracts.

At resting potential, ionophoretic doses of acetylcholine resulted in depolarization of the cell body membrane of motor neurone D_f, accompanied by a drop in input resistance. Repeated applications (400 nA) at intervals of 2 min or more yielded responses of constant amplitude over a period of several hours in control conditions when the cells were continuously perfused (2·0 ml min⁻¹) with normal saline. Bath-application of 1·0×10⁻⁶moll⁻¹N-methyl lycaconitine resulted in a progressive block of the depolarization induced by ionophoretic application of acetylcholine (Fig. 3). After 15min, the acetylcholine-induced response was completely blocked. Increasing the ionophoretic dose to 600 nA and 1000 nA, or prolonged washing of the preparation (60 min) in normal saline, did not restore the response to acetylcholine. Using a submaximal response to acetylcholine (30-60% of the maximum depolarization), the percentage reduction in amplitude of the original response was determined after 20 min of bath-application of various concentrations of N-methyl lycaconitine. In this way a dose-dependent block of the insect neuronal nicotinic receptor was demonstrated in the range 1·0×10⁻⁸ to 1·0×moll⁻¹ (Fig. 4).

To characterize the interaction between N-methyl lycaconitine and the nicotinic receptor, acetylcholine-induced currents were recorded in the presence and absence of the alkaloid (Fig. 5). Non-competitive antagonists often cause a voltage-dependent block of insect neuronal nicotinic receptors, whereas the blockade produced by competitive nicotinic antagonists is voltage-independent (Sattelle & David, 1983; David & Sattelle, 1984). To examine whether the blockade produced by N-methyl lycaconitine was dependent on membrane potential, the motor neurone cell body was voltage-clamped at –60mV and a series of square-wave voltage-excursions was imposed over the range –30 to –100mV in the presence or absence of N-methyl lycaconitine. The voltage steps were imposed before and during the acetylcholine-induced responses.

Examples of the current traces obtained in this way are illustrated in Fig. 5 which shows records from a single neurone. Fig. 6A shows, for a different neurone, the acetylcholine-induced current as a function of membrane potential before and after application of 1·0×10⁻⁷moll⁻¹N-methyl lycaconitine. The percentage block of acetylcholine-induced currents resulting from this exposure to N-methyl lycaconitine was determined for each command potential for three separate cells and the data combined (Fig. 6B). N-Methyl lycaconitine block of the nicotinic receptor of motor neurone D_f was independent of membrane potential over the range examined (–100 to –30mV).

Though the insecticidal properties of Delphinium seed extract were noted in AD 77 by Pliny the Elder (Book XXIII, Chapter XII), who recorded that ‘pounded, they rid the head of lice’ (see Jones, 1961), it is only in the present century that this has been confirmed (Nambi Aiyar et al. 1979; Grina et al. 1986). Recently, Jennings et al. (1986, 1987) have demonstrated a potent synaptic blocking action of N-methyl lycaconitine, a major active fraction of the seed extract. Here, we provide direct evidence that V-methyl lycaconitine is active at an insect neuronal nicotinic acetylcholine receptor. A K_i value of 4·0×10⁻⁹moll⁻¹ for inhibition of [³H]-α-bungarotoxin binding to cockroach (Periplaneta americana) nerve cord extracts is higher than the value (2·5×10⁻¹⁰moll⁻¹) obtained in binding studies on housefly (Musca domestica) head extracts, a preparation which contains a wide variety of tissues in addition to nervous tissue. Differences of this order have been seen for some other ligands when comparing putative receptors from different insect tissues (see Sattelle, 1986). One possibility is that there may be some pharmacological differences between fly and cockroach nicotinic acetylcholine receptors. Alternatively, differences in preparative procedures may account for these discrepancies. Nevertheless, it is clear that N-methyl lycaconitine is the most potent low molecular weight blocking agent so far tested on the insect nervous system [³H]-α-bungaro toxin binding site.

Interpretation of receptor-ligand interactions is facilitated when radioligand binding and electrophysiological data are available for the same tissue. The finding that the alkaloid blocks acetylcholine-induced currents in the cell body membrane of an identified cockroach neurone located in the same tissue on which the binding studies have been performed is the first direct evidence for an antagonist action of N-methyl lycaconitine. The fast coxal depressor motor neurone (D_f) nicotinic receptors are blocked by α-bungarotoxin (Sattelle et al. 1980; David & Sattelle, 1984), κ-bungarotoxin (Pinnock et al. 1988) and a wide range of nicotinic ligands (David & Sattelle, 1984). Hitherto, the most effective low molecular weight blocking agent was dihydro-β-erythroidine. A-Methyl lycaconitine reduces the acetylcholine-induced current at concentrations above 1·0×10⁻⁸moll⁻¹, completely blocking the response following brief exposures to a 1·0×10⁻⁶moll⁻¹ concentration. Although the IC₅₀ value determined from binding studies (2·0×10⁻⁹moll⁻¹) is much lower than that estimated from physiological experiments (≈1·0×10⁻⁷ mol l⁻¹), factors such as accessibility to receptors of the in vivo preparation may account for these differences. A blocking action at an insect neuronal nicotinic acetylcholine receptor is the simplest explanation of these findings.

The blocking action of N-methyl lycaconitine is voltage-independent over the membrane potential range –100 to –30 mV. Other ligands which block the insect nicotinic receptor in a voltage-independent manner include α-bungarotoxin (David & Sattelle, 1984), ≈-bungarotoxin (Pinnock et al. 1988), dihydro-β-erythroidine and benzoquinonium (David & Sattelle, 1984). By contrast, d-tubocurarine, atropine (David & Sattelle, 1984), histrionicotoxin (Sattelle & David, 1983) and amantadine (Artola et al. 1984) are all strongly voltagedependent in their blocking actions over a similar range of membrane potentials. The precise site of action of N-methyl lycaconitine remains to be determined. An irreversible blocking action at the acetylcholine recognition site remains a possibility. A noncompetitive, voltage-independent action cannot be ruled out, though the radioligand binding data presented here suggest a competitive action. With the recent detection of acetylcholine-induced single channels in dissociated cockroach neurones (Sattelle et al. 1986), it will be of interest to examine the actions of N-methyl lycaconitine on the unitary conductance and lifetime properties of this insect nicotinic acetylcholine receptor.

When physiological data from the present study are compared with data obtained on the rat phrenic diaphragm preparation (Nambi Aiyar et al. 1979), it appears that N-methyl lycaconitine is more effective on the insect neuronal nicotinic receptor than on the vertebrate peripheral nicotinic receptor. When the high affinity of N-methyl lycaconitine for the insect nicotinic receptor is considered together with the finding that the insect nervous system is one of the richest sources of neuronal nicotinic receptors in the animal kingdom (see Sattelle, 1986,Breer & Sattelle, 1987), it is possible that the role of the alkaloid in plants is related to protection from insect damage. All insects so far tested, including such diverse forms as cockroaches (Lummis & Sattelle, 1985), moths, locusts and flies, are rich in neuronal nicotinic receptors (Breer & Sattelle, 1987) which may account for the observation that N-methyl lycaconitine exhibits insecticidal activity against a range of insect species.