Pharmacological Properties of Axonal Sodium Channels in the Cockroach Periplaneta Americana L.: I. Selective Block by Synthetic Saxitoxin

Neurotoxins are essential tools in the analysis of the molecular events underlying axonal conduction and synaptic transmission (see Narahashi, 1974, 1975). Saxitoxin (STX) is one of the most toxic low molecular weight compounds known and in its natural form is extracted from the clam Saxidomus giganteus, the mussel Mytilus californianus and axenic cultures of the dinoflagellate Gonyaulax catenella (Ghazarossian et al. 1974). As is the case for the neuropoison tetrodotoxin (TTX) extracted from the puffer fish (Tetraodon lineatus), natural saxitoxin (STX) selectively blocks transient sodium currents in nerve membranes (Narahashi, Haas & Therrien, 1967; Evans, 1969). However, post-toxification recovery of sodium channel function is more rapid after exposure to STX than following TTX application (Narahashi et al. 1967; Hille, 1975). In addition, the dissociation of toxin from a component of solubilized garfish olfactory nerve membranes is almost four times faster for STX (Henderson, Ritchie & Strichartz, 1973) than for TTX (Henderson & Wang, 1972). Clearly, STX would be the chemical probe of choice for many experiments on membrane ionic permeability if it were more widely available. The structure of natural STX has now been established by X-ray crystallography (Schantz et al. 19751 Bordner et al. 1975) and the recent total synthesis of the toxin (Tanino et al. 1977) makes the more widespread use of this molecule in membrane studies a practical possibility. In an earlier, brief communication, giant axons of the cockroach (Periplaneta americana) were shown to be sensitive to synthetic STX (Pelhate & Sattelle, 1978). Here we report the first detailed account of the actions of synthetic STX on axonal membrane currents.

Giant axons of the cockroach (P. americana), which are used in the present study, closely resemble other unmyelinated axons in that they exhibit, under voltage-clamp conditions, transient sodium currents and slower, steady-state potassium currents (cf. Pichon, 1974). In this study, synthetic STX, natural STX and TTX are applied to isolated, voltage-clamped, cockroach axons. The aims of the investigation are as follows: (1) to assess the usefulness of synthetic STX as a specific inhibitor of sodium channel function; (2) to quantitatively compare the affinity for insect axonal sodium channels of various blocking agents (natural and synthetic STX and natural TTX).

Adult male cockroaches (Periplaneta americana) were used throughout these experiments. An isolated (2–3 mm) length of one giant axon was dissected from a connective linking the fourth and fifth abdominal ganglia and cleaned of adhering fibres. The preparation was transferred to an experimental chamber in which two lateral Ag-AgCl electrodes were in contact with the severed ends of the axon and a central Ag-AgCl electrode was in contact through the external bathing solution with a 100 /tm length of dissected axon. The preparation was immersed in paraffin oil and the ‘artificial node’ created by this non-electrolyte (see Pichon & Boistel, 1967) was voltage-clamped as described in detail earlier (Pelhate, Hue & Chanelet, 1978). This space-clamped region of the axon was superfused by either normal saline or test solutions. Analogue compensation for leakage and capacity currents, by the method of Hille & Campbell (1976), ensured that currents recorded during the depolarizing clamp pulses were specific ionic currents. Most experiments were performed using a step depolarization to a membrane potential (E_m) of – 10 mV from a holding potential (E_n) of – 60 mV. The leak correction was adjusted for this value of E_m. No leakage rectification is normally detected in cockroach axons until E_m values of ≃ +40 mV are reached (Pichon, 1969a).

In all experiments the preparation chamber was maintained at 12 ± 0·5 °C. Normal physiological saline had the following composition (mM): NaCl, 200; KC1, 3·1; CaCl₂, 5·4; MgCl₂, 5·0. The pH was maintained at 7·2 using a phosphate-carbonate buffer. Synthetic saxitoxin dihydrochloride was diluted in this saline from a stock solution (1·0 mg/ml toxin) generously provided by Prof. Y. Kishi (Harvard University). Natural saxitoxin was also diluted in saline from a stock solution (0·1 mg/ml toxin) kindly donated by Dr M. H. Evans (Agricultural Research Council Institute of Animal Physiology, Babraham, U.K.). Crystalline tetrodotoxin (natural) from Calbiochem was also prepared in physiological saline.

Membrane currents were measured under voltage-clamp conditions in response to a step depolarization from a holding potential (E_h) of –60 mV to a membrane potential (E_m) of –10 mV. When synthetic STX was externally applied to the axon at concentrations in the range 10⁻⁷ to 5 × 10⁻⁷ M the inward sodium current was rapidly blocked without affecting the steady-state outward potassium current (Fig. 1). Rebathing the axon in normal saline resulted in recovery of the inward sodium current (Fig. 1). An outward sodium current, generated by clamping the membrane at E_m =+60 mV, was also blocked by 10⁻⁷ M synthetic STX indicating that the toxin prevented ion movements through the sodium channels in both directions.

To measure directly the actions of synthetic saxitoxin on the inward sodium current alone, the axon was first pretreated with 4-aminopyridine (4-AP). It has been established that 4-AP specifically reduces potassium currents in the cockroach axon (Pelhate & Pichon, 1974), the squid axon (Meves & Pichon, 1977) and the frog node of Ranvier (Ulbricht & Wagner, 1976). Following suppression of the outward potassium current by 5 × 10⁻⁴ M 4-AP, synthetic STX at 10⁻⁷ M completely blocked the remaining inward (sodium) current (Fig. 2). We have therefore demonstrated total elimination of sodium and potassium currents using only synthetic, selective, channelblocking agents.

Axons were exposed to various concentrations of synthetic STX and the data plotted as the dose-dependence of the percentage inhibition of the peak transient inward sodium current (Fig. 3 a). STX binding to the axon membrane is considered as a simple case of reversible adsorption to surface sites (receptors), in which case we can write:

Saxitoxin (STX) and tetrodotoxin (TTX) from natural sources were applied to the axonal membrane of the cockroach and their effects on sodium currents compared to results obtained with synthetic STX. Both natural toxins were able to block axonal inward sodium currents completely without affecting outward potassium currents. Nevertheless to achieve the same percentage inhibition of the peak inward sodium current as that induced by synthetic STX it was necessary to elevate the concentration of natural STX by 2- to 5-fold and that of TTX by about an order of magnitude (Fig. 4). It was also noted that the actions of TTX on cockroach axons were much less readily reversed than the effects of saxitoxin (natural and synthetic). For sixteen axons which were blocked by 10⁻⁶ M synthetic STX, the mean time required for 50% recovery of the peak inward sodium current following re-exposure to normal saline was 7·38 (S.D. ±2·12) min. By contrast in two out of five axons blocked by 10⁻⁶ M TTX, recovery of 50% of the peak inward sodium current required rebathing in normal saline for 45 and 25 min respectively. In the three other TTX-treated axons only 25 % of the initial peak inward sodium current had returned after a 30 min wash in normal saline.

We have shown that chemically synthesized saxitoxin (STX) is a highly specific sodium channel blocking agent when applied to axonal membranes of the cockroach (Periplaneta americana). Since insect axonal sodium channels share many common properties with the sodium channels of other invertebrates and vertebrate unmyelinated axons (cf. Pichon, 1974, 1976), synthetic STX should prove to be widely applicable in studies of axonal ion permeability mechanisms.

Using voltage-clamp data on the concentration-dependence of the suppression of the peak inward sodium current by synthetic STX, it is estimated that one molecule of synthetic STX binds reversibly to each receptor site with a dissociation constant of 3·0 × 10⁻⁹ M. This value is close to the dissociation constant estimates for the binding of natural STX to other axonal membranes. For example, using voltage-clamp data, a value of 1·2 × 10⁻⁹ M has been estimated by Hille (1968) for the binding of natural STX, to the frog node of Ranvier. Also, from studies of the binding of tritiated natural STX to solubilized garfish olfactory nerve membranes, a dissociation constant of 6·7 × 10⁻⁹ has been obtained (Henderson et al. 1973). The simplest interpretation of the voltage-clamp, dose-response data is that each sodium channel in the cockroach axonal membrane contains a single receptor site for an STX molecule which when occupied by the toxin prevents the movement of sodium ions through the channel.

Synthetic STX has been shown to be identical to purified natural STX with respect to NMR spectrum, silica-gel TLC and whole animal toxicity (Tanino et al. 1977). When applied to the cockroach axonal membrane the chemically synthesized STX was always slightly more active than the natural toxin. A concentration of natural STX approximately 2–5 times greater than that used for synthetic STX was required to produce the same percentage suppression of the inward sodium current. Without detailed chemical analyses of the STX stock solutions it is not possible to account for these differences between the axonal actions of synthetic and natural STX, which may for example derive from trace impurities in the sample of natural toxin. More striking differences are noted between the concentrations of TTX and synthetic STX required to produce the equivalent reduction of the inward sodium current. Our observations on TTX are consistent with the results of earlier voltage-clamp experiments by Pichon (1969a, b) which showed that TTX at micromolar concentrations blocks cockroach axonal sodium channels. In the present study it has been shown that concentrations of TTX about 10-fold higher than those of synthetic STX are needed to produce a 50% suppression of the inward sodium current. Also, the actions of TTX are less readily reversed than the effects of STX (synthetic or natural). It therefore emerges that synthetic STX is the most useful of the sodium channel blocking agents tested on the cockroach axon. The development of a synthetic form of saxitoxin has provided the first chemically synthesized probe suitable for investigating the molecular pharmacology of sodium channels in cell membranes.

The support of The Royal Society European Exchange Programme and a travel award from the British Council are gratefully acknowledged. The authors are indebted to Prof. Y. Kishi (Harvard University) for helpful discussions and for supplying samples of synthetic saxitoxin. We thank M. Bedouet and Mrs M. Fuentes for technical assistance.