Sodium- and Calcium-Dependent Mechanisms in the Action Potential of the Secretory Epithelium of a Clam Mantle

The generation of action potentials has been shown in epithelia of several animal species from different and widely separated phyla (Anderson, 1980). In spite of this, much biophysical work is still required for the elucidation of the ionic mechanisms and for comparison with better-known excitable tissues.

One of the most interesting examples of epithelial excitability occurs in the clam mantle, whose outer (shell-facing) epithelium, which is responsible for the shell secretion, is excitable. This epithelium shows two types of action potentials,depending on its physiological state. During periods in which the animal is engaged in shell secretion, electrical stimulation elicits action potentials with a spike phase followed by a prominent plateau, resembling some cardiac action potentials. During the phases of the animal’s life in which it is resorbing the shell, the action potentials are spike-like, with no plateau (Beirão & Sorenson, 1986).

The present paper describes the ionic dependence of these action potentials. A Na⁺-dependent TTX-insensitive electrogenesis, with otherwise similar properties to the electrogenesis mediated by sodium channels, is responsible for the generation of the spike. Evidence for the occurrence of Ca²⁺ currents during the plateau is given. A preliminary account of the data has appeared elsewhere (Nascimento & Beirão, 1983).

Specimens of the clam Anomalocardia brasiliana were obtained from Rio de Janeiro or from São Vicente, Brazil. They were kept in an aquarium with aerated sea water containing plankton and algae as food sources. An animal was selected for use only if it actively pumped water when undisturbed and if it closed its valves quickly when handled. The largest possible area of the mantle (whole mantle, usually more than 1 cm²) was carefully removed and immediately assembled in the recording chamber. In some experiments, a narrow (2–3 mm) strip of the mantle was used, allowing the solution in the interstitial space to be easily exchanged with the bathing solution.

The impalements were restricted to the central part of the mantle, which consists of two single-layered epithelia separated by an interstitial space containing hemolymph. The outer (secretory) epithelium is formed by columnar cells, whose basolateral membranes are convoluted and communicate directly with the interstitial space (Beedham, 1958; Neff, 1972).

The mantle was pinned down with its secretory epithelium facing upwards. Supramaximal isolated current pulses were delivered through a suction electrode placed on top of the epithelium.

To study the effect of Na⁺ concentration on the spike overshoot, the mantle was stretched and assembled between the two compartments of a chamber, shutting off a 0·008cm² passageway that communicated them. Isolated current pulses were passed across this small area of the mantle, stimulating it almost simultaneously (see Beirão & Sorenson, 1986).

In either chamber, resting and action potentials were recorded with conventional intracellular (8–30 MΩ) glass microelectrodes connected to a capacity-compensated electrometer (BAK, Instruments for Life Sciences, Rockville, Maryland). Recordings were displayed and photographed on a storage oscilloscope (Tektronix, model 5111, Beaverton, Oregon). Resting potentials were recorded on a digital voltmeter (ECB, São Paulo, SP). Mantles with cardiac-like and spike-like action potentials were used but, for simplicity, only experiments with the latter were chosen to show the effects on the spike. The data shown are typical of three or more experiments.

Table 1 shows the composition of the solutions used. The normal solution resembled the extrapallial fluid (Crenshaw, 1972). Intermediate Na⁺ concentrations were obtained by the combination of normal solution with O-Na⁺ solution. All reagents were analytical grade. Tetrodotoxin, veratrine and lidocaine were from Sigma Chemical Co. All experiments were carried out at room temperature (21–26 °C).

The spike phase of the action potential of the whole mantle was very insensitive to ionic substitutions in the bathing solution. In the first experiment (Fig. 1), a narrow strip of the mantle was used. Na⁺ removal caused a reversible abolition of the action potential. Partial (50 %) removal of Na⁺ caused a decrease of the spike amplitude by 15 mV. This approach had the limitation that whenever Na⁺ was decreased to 100 mmol I⁻¹ or less, the Na⁺ current became too small to cause the spike to propagate and no action potential was seen. In order to use low Na⁺ concentrations, non-propagated (‘membrane’) action potentials were generated in mantles assembled between two compartments. The result of partial Na⁺ removal in both compartments is shown in Fig. 2, in which the peak value of the spike is plotted against the logarithm of the Na⁺ concentration. The straight Une obtained by linear regression has a slope of 54·5 mV for a 10-fold increase in Na⁺ concentration, which closely approaches the theoretical value for a Na⁺ electrode (59 mV).

To test if the increase in Na⁺ permeability was affected by sodium channel blockers, TTX was added to the medium bathing a strip of the mantle (Fig. 3A). No effect was observed at concentrations up to 36 μmol I⁻¹, even after 67 min of exposure to TTX. The results from one experiment (Fig. 3B) showed that it was unlikely that an unusually long delay in exchanging the solution in the interstitial space could have prevented TTX reaching the basolateral membrane. The strip of mantle was preincubated in 0-Na⁺ solution, resulting in the disappearance of action potentials, and then in normal solution containing 6μmoll⁻¹ TTX. After 10 min, normal action potentials were restored. They maintained the same amplitude for up to 40 min, showing that the solution in the interstitial space had been exchanged and that TTX had no effect. TTX from the same stock solution inhibited action potentials of frog skeletal muscle at a concentration of 30 nmol I⁻¹.

Local anesthetics block sodium channels by a mechanism different from that of TTX (Hille, 1977). Fig. 3C,D shows the reversible inhibition produced by 5mmoll⁻¹ lidocaine. Procaine (l·4mmoll⁻¹) and tetracaine (1 mmoll⁻¹) also blocked the spike electrogenesis. Veratrine, which specifically acts on sodium channels and inhibits their inactivation, caused long-lasting afterpotentials at low concentrations (data not shown).

Our results confirm the previous report by Beirão & Sorenson (1986), that the removal of Ca²⁺ markedly decreased the plateau. Fig. 4 shows that the increase in Ca²⁺ concentration caused by substitution of Ca²⁺ for Mg²⁺ increased the height and duration of the plateau.

The Ca²⁺-dependence of the plateau suggests that calcium channels are activated during its electrogenesis. This possibility was further investigated using inorganic calcium channel blockers. Mn²⁺ reversibly inhibited the plateau with a short delay, often less than 5 min (Fig. 5). The addition of Mn²⁺ did not transform a cardiac-like action potential into a spike-like action potential, because it slowed the rate of repolarization. Fig. 5C shows that the addition of Mn²⁺ to a preparation with a spike-like action potential reduced its repolarization rate. The same effects were seen with Cd²⁺.

Beirão & Sorenson (1986) described the occurrence of spike-like and cardiac like action potentials in the clam mantle, and proposed that the latter were made up of two distinct components: an earlier, faster one that originates the spike phase, and a later, slower one that forms its plateau. The data presented in this paper show that these components have distinct ionic dependences.

A remarkable observation was the absence of any effect of TTX at concentrations three orders of magnitude higher than the concentrations effective in nerve and skeletal muscle. This lack of effect cannot be attributed to a barrier for its diffusion to the site of the sodium channels, as shown by Fig. 3. Besides, there is no structure in the mantle that could provide a physical barrier specifically blocking TTX diffusion (Neff, 1972). We conclude that there are TTX-insensitive sodium channels in the basolateral membrane that generate the spike.

The existence of sodium channels with lower affinity for TTX has been reported in denervated mammalian muscle (Redfern & Thesleff, 1971), in cultured rat myoblasts and myotubes (Frelin et al. 1984), cultured rat heart cells (Catterall & Coppersmith, 1981) and in embryonic heart cells (Renaud et al. 1981). In these preparations, the concentration for half-maximal inhibition (K0·5) for TTX is between 0·1 and 1 μmol I⁻¹. Our results show that TTX had no effect on the spike electrogenesis, even at a concentration as high as 36μmoll⁻¹, showing an unusually high insensitivity of this preparation to TTX.

This property of the sodium channels of the clam mantle is probably related to its adaptation. It was shown that clams can accumulate saxitoxin (STX) at very high concentrations when dinoflagellates are present in the sea water. TTX and TTX bind to the same receptor and have the same inhibitory mechanism (Colquhoun et al. 1972). Twarog et al. (1972) have shown that many shellfish species show TTX-and TiX-resistant nerve excitability. Spalding (1980) showed that blocking a carboxyl group of the sodium channel with trimethyloxonium ion makes it insensitive to 90μmoll⁻¹ TTX without interfering with its sensitivity to local anesthetics and with its ionic selectivity.

Beirão & Sorenson (1986) showed that the substitution of Mg²⁻¹” for Ca²⁺ markedly decreased or blocked the plateau of cardiac-like action potentials. Fig. 4 shows that the increase in Ca²⁺ concentration increases the amplitude and duration of the plateau.

The addition of inorganic calcium channel blockers, such as Mn²⁺, inhibited the plateau in a few minutes. It is noteworthy that the repolarization rate in the presence of these cations was slow, producing a triangular action potential. This also occurred when Mn²⁺ was added to a preparation with spike-like action potentials, showing that it decreases the repolarization rate.

We suggest, therefore, that the presence of calcium channels, mostly in the apical membrane, accounted for the short time it took for the effects of calcium channel blockers and Ca²⁺ substitutions to take place, even when the whole mantle was used. The reason for the absence of the plateau in some mantles is under investigation.

The authors are indebted to Dr H. Masuda for his generous supply of clams and to Dr A. L. Sorenson for helpful discussions. This work was supported in part by grants 40.2912/82 and 30.0303/78-BF from the Conselho Nacional de Desenvolvi-mento Cientifico e Tecnológico-CNPq and by Convênio FINEP-Bioquimica. JHMN was a fellow from CAPES.