Ionic Basis of Action Potentials and Bursting Activity in the Hydromedusan Jellyfish Polyorchis Penicillatus

Intracellular recordings from the motor neurones which control swimming, the swimming neurones, were obtained in the manner described previously (Anderson & Mackie, 1977). Experiments were carried out in normal sea water or artificial media (Wilkins, 1970). Choline chloride was used to substitute for NaCl in media with reduced concentrations of sodium. Manganese, cobalt and other divalent cations were added as their chloride salts.

The swimming neurones of Polyorchis have resting potentials of −60 ± 5 mV and conduct action potentials with amplitudes in the range 80–100 mV (Anderson & Mackie, 1977). These action potentials occur spontaneously but they can, if desired, be evoked by extracellular stimulation of the inner nerve ring.

All spontaneous and evoked action potentials were abolished when the preparation was bathed in sodium-free sea water. If sodium were reintroduced and the concentration raised to 20% of that of normal sea water, small (10 mV) transient depolarizations could be evoked by stimuli to the nerve ring (Fig. 1 A). If the sodium concentration were further increased, the action potentials became progressively larger and eventually attained their normal amplitudes (Fig. 1 A).

When the waveforms of the action potentials were monitored through a descending series of sodium concentrations, it became clear that the preparation was acting as an ion store and delaying displacement of sodium ions from around the cells. Since this effect was absent in experiments involving ascending series of concentrations, it indicated that the ionic concentration of the bath could not necessarily be taken as truly representative of that surrounding the cells.

Tetrodotoxin (TTX) in concentrations of up to 4 × 10⁻⁵ g/ml was applied for up to 4 h with no observable effect even when, in an attempt to optimize penetration of the drug, the TTX solution was perfused through the circulatory canal system of the animal.

The possible role of calcium in the action potential was studied using calcium-free and high calcium sea waters and known calcium antagonists such as manganese (10 mm), cobalt (10–55 mm) and barium (10–55 mm) The results of these experiments were ambiguous, and while changes in action potential waveform were observed, these changes were difficult to distinguish from the changes in waveform that occur spontaneously (Anderson & Mackie, 1977).

Spontaneous changes in waveform were originally attributed to some effect of electrotonic coupling between neurones. It appears that while strong coupling is present the waveform of the action potential is, instead, determined by the level of depolarization of the neurone at the time the action potential is triggered. During experiments in which action potentials were evoked by stimulation of the nerve ring, it was observed that high intensity stimuli evoked relatively short duration action potentials which appeared after a negligible delay (Fig. 1B). In contrast, just threshold or low intensity stimuli depolarized the neurones for varying intervals before the action potential and as the delay increased so did the duration of the action potential (Fig. 1B).

The most notable effect of calcium-free sea water was on the firing pattern of the swimming neurones. Normally, the cells fire irregularly, and there is evidence of much synaptic activity (Fig. 1 C and Anderson & Mackie, 1977). However, in calcium-free sea water the cells fired very regular bursts of 3–5 action potentials interrupted by a pronounced 40 mV hyperpolarization lasting for approximately 5–7 s (Fig. 1D, E). If cobalt (7·4 mm) were then added, the firing pattern changed further, and the burst and interburst durations increased dramatically (Fig. 1E). Eventually, the cell would begin firing almost incessantly. Continuous firing also occurred if barium or tetra ethylammonium (50 mm) were added to the sea water bathing a preparation.

The dependence of the action potential on the sodium concentration of the bathing medium implies that the inward current of the action potential is carried by sodium. The effect of calcium-free sea water on the firing pattern of the neurones suggests, for reasons which will be discussed later, that there is also a small, but undetected, calcium component to the inward current.

The apparent insensitivity of these cells to TTX is somewhat surprising. TTX acts by blocking sodium currents in excitable cells and this action is relatively consistent throughout the animal kingdom. The swimming muscles of Polyorchis are sensitive to TTX but here prolonged exposure to high concentration is necessary (Spencer, unpublished). It is conceivable that the swimming neurones of Polyorchis are also TTX sensitive but that the sensitivity was not detected in these experiments. This, however, appears unlikely since high concentrations and long exposure times were used and care was taken to optimize penetration of the drug.

Neurones which fire in a bursting or oscillatory manner are common throughout the animal kingdom and in many cases mechanisms to explain their behaviour have been described (Moffatt, 1977; Chalazonitis, 1977). The cell designated R15 in Aplysia fires in very regular bursts which appear to be controlled by a calcium-activated potassium efflux mechanism (Meech, 1974) in which calcium, which accumulates intracellularly during the burst, triggers a potassium efflux which hyperpolarizes the cell and terminates the burst. The potassium conductance diminishes as the calcium is sequestered intracellularly and the cell repolarizes and another burst occurs. While there are many theories to account for bursting behaviour in neurones, this calcium-based mechanism best explains the results obtained from Polyorchis.

Calcium-free sea water normally contains some calcium and in coelenterates, where there is an extensive acellular mesogloea, these levels may be quite high. If the calcium concentration in calcium-free sea water were sufficient to support a calcium-based bursting mechanism but insufficient to support synaptic activity, records such as those of Fig. 1 D would be expected. The strongest evidence to suggest that a calcium based mechanism is present comes from the results of adding divalent cations when a cell is undergoing bursting activity.

If the extra cations were to compete with calcium during the inward current of the action potential, fewer calcium ions would enter the cell during the action potential and, consequently, more action potential would be required to elevate the intracellular calcium concentration to the level needed to trigger the potassium permeability increase. Assuming that there is competition between calcium and other divalent cations for the intracellular uptake mechanisms, the presence of many other divalent cations could result in a delay in the removal of calcium and hence a prolongation of the interburst interval.

The action of TEA, a recognized potassium current blocker, in producing continuous firing is also compatible with this theory since it suggests that the burst-terminating hyperpolarization is indeed produced by a potassium efflux.

In such a complex multicellular preparation, it is difficult to separate the intrinsic properties of spontaneous repetitive activity in a cell from external influences of other cells. Nevertheless, the available evidence favours a model of a calcium-based bursting mechanism. Whatever the mechanism, however, the presence of underlying bursting activity in these neurones holds important consequences for the behaviour of Polyorchis since it has been shown that the firing pattern of the swimming neurones directly determines the swimming rhythm of the animal (Anderson & Mackie, 1977).

These various results reaffirm the previous view (Anderson & Mackie, 1977) that the primitive nervous system of the coelenterate as typified by that of Polyorchis shows many similarities with those of higher organisms.

This study was supported by an N.R.C. Operating Grant to Dr G. O. Mackie. I would like to thank him for his assistance and support during this work.