Potassium and Calcium Currents in Dissociated Muscle Fibres of the Mollusc Philine Aperta

Studies on the membrane properties of molluscan fibres have generally been restricted by their small size and inaccessibility, and our knowledge of their permeability is derived largely from investigations of the ionic mechanisms determining the resting and action potentials (Wilkens, 1972; Brezden and Gardner, 1984; Dorsett and Evans, 1989), although recently patch-clamp studies on isolated heart muscle fibres of Lymnaea have identified a potassium-selective channel sensitive to membrane stretch (Brezden et al. 1986; Sigurdson et al. 1987).

In other invertebrate preparations, these difficulties have been overcome by the use of single fibres obtained by enzymic dissociation of the connective tissue matrix of the muscle (Ishii and Takahashi, 1982; Hernandez Nicaise et al. 1980; Anderson, 1984). In this report the technique was used to isolate fibres from the buccal mass retractor muscles 4 and 5 of the opisthobranch mollusc Philine aperta (Evans and Dorsett, 1989). These are paired, anatomically distinct muscles with unstriated non-spiking fibres having a simple transverse tubule system (Dorsett and Roberts, 1980), which respond to motoneurone stimulation with summating excitatory junction potentials (Dorsett et al. 1989; Evans and Dorsett, 1989).

The present study was undertaken to characterise the ionic currents associated with depolarization of a non-spiking muscle fibre that initiates the development of tension in the muscle.

To isolate single muscle fibres, muscles were dissected free of their attachments to the buccal mass and body wall and rinsed in an artificial sea water (ASW) containing 0 mmol I⁻¹ Ca²⁺ to reduce the possibility of contraction during enzymic digestion of the sheath. The ASW contained (in mmoll⁻¹); Na⁺, 463; K⁺, 10; Ca²⁺, 9; Mg²⁺, 53; SO₄⁻², 17; Tris, 5; adjusted to pH8 with HC1. Measurements of the equilibrium potential involving elevated levels of K⁺ were made in salines where the [K⁺][C1⁻] product was kept constant (Dorsett and Evans, 1989).

The muscles were incubated in a 2mgml⁻¹ solution of Type XIV protease (Sigma) in zero-Ca²⁺ ASW until the outer connective tissue sheath was disrupted. At this point the muscles were carefully transferred to the experimental chamber containing sea water, and a number of single fibres were dissected free using electrolytically sharpened tungsten needles. Fibres separated in this way normally settled on the bottom of the chamber, retaining their characteristic shape and dimensions (see Fig. 1).

The experimental chambers were lined with sea water agar or with Sylgard (Dow Corning), the hydrophilic nature of the former offering an advantage in allowing the volume of fluid in the chamber to be reduced, thus minimising capacitance artefacts associated with the command pulse. The fibres were voltageclamped with a Dagan 8500 two-electrode amplifier. Both current and voltage electrodes had resistances in the range of 6–10 MΩ and were shielded to within a short distance of the tip with silver conducting paint. The driven shield of the voltage electrode was insulated and that of the current electrode grounded. The indifferent electrode consisted of an agar bridge to a KCl/AgCl junction connected to the virtual ground current monitor of the amplifier. Rectangular command pulses 70ms in duration, derived from a pulse generator (Farnell), resulted in membrane potential changes that were completed in 600 μs. No attempt was made to compensate for the series resistance as the cells were isolated and surrounded by a medium of high ionic strength. Membrane voltage and current traces were displayed and photographed on a Tetronix 5103N storage oscilloscope, digitised on a GTCO Digipad and stored for subsequent manipulation on a DEC VAX 8650. Leakage currents were measured for a series of hyperpolarising pulses and a straight line projection was used to estimate leakage currents in the depolarizing range. Temperatures ranged between 16 and 18°C.

The morphology of single fibres in the living muscle was studied by injecting them with 5-carboxyfluorescein (Rao et al. 1986). (Fig. 1A). They have a cylindrical form with tapering ends, and are between 1 and 2 mm long and about 20 μm in diameter. One end of each fibre is often forked, and the tapered end is commonly serrated, as if to provide a firm insertion into the connective tissue sheath. Dissociated fibres extracted from the muscle retained a similar morphology when relaxed (Fig. 1B), but became sigmoid or almost spherical when fully contracted, with the contractile apparatus consolidated as a central core.

In the present series of experiments 30 muscle fibres sampled in haemolymph were found to have a mean membrane potential of –74.9±1.35mV (±S.E.). The mean membrane potential of a further sample of 30 dissociated fibres in sea water was – 76.3±0.44mV and the mean membrane resistance was 42.2±3mΩ. It was concluded that the fibres were not unduly stressed by the isolation procedures.

When subjected to small depolarisations through the voltage-recording electrode, single fibres were observed to twist as they contracted and relaxed smoothly, the degree of contraction depending on the membrane potential. This observation is of particular interest as it establishes that contraction is initiated by and depends on depolarization, and supports a previous observation of small twitch-like contractions of the whole muscle in response to single excitatory junction potentials (EJPs) (Dorsett et al. 1989).

Isolated muscle fibres in sea water were subjected to a two-electrode voltageclamp. The voltage-recording electrode was inserted into the central region of the fibres and normally recorded potentials in excess of –70 mV, close to the normal resting potential of undissociated fibres in haemolymph or sea water. The current electrode was then inserted and the fibres clamped at –75 mV. The membrane potential was stepped to 0 mV by a 70 ms command pulse and the fibres responded with an outward current which rose to a delayed transient peak after approximately 25 ms before declining to a steady-state current that persisted until the end of the command pulse (Fig. 2). The complex profile of the outward current suggested the presence of more than one component.

A feature of the outward current is an inflection in the rising edge some 9–10 ms after the onset of the pulse when the current has attained a value of 25–30 nA (Figs 2, 4, inset i), marking a change in the rate at which the late transient current approaches its maximum. Stepping the membrane potential to zero from holding voltages of –60 and –80 mV did not substantially alter the profile of the total outward current other than to vary the value of the transient peak and the steadystate level.

Several lines of evidence support the conclusion that the outward current is mainly carried by potassium. The reversal potential of the tail currents was measured in three sets of muscle fibres following exposure to constant product [K⁺][C1⁻] ASW solutions containing 10, 20 or 40 mmol I⁻¹ potassium. The fibres were stepped to 0 mV for 70 ms and returned to a series of membrane potentials between –30 and –70mV to give the instantaneous current-voltage relationship for the steady-state current (Fig. 3). The regression of the equilibrium potential plotted against [K⁺]_o had a slope of 51 mV for a 10-fold change in [K⁺]_o, which compares well with the figure of 51.2 mV obtained for isolated radial fibres of Beroe ovata (Bilbaut et al. 1988). The mean equilibrium potential in 40 mmol I⁻¹ [K⁺]_o was –46.7 mV compared with a predicted value of 45.8 mV from the Nernst equation assuming a [K⁺]_i of 247 mmol I⁻¹ (Dorsett and Evans, 1989). Finally, application of saline containing the potassium channel blockers TEA-CI and 4-AP abolished all outward currents.

The current-voltage relationship of 10 muscle fibres was measured in sea water with an extended range of command pulses (+30 to +210 mV) from a membrane holding potential of –75 mV. At each step the current was measured after 25–35 ms to coincide with the late transient peak, and again at 70 ms to record the maintained current (Fig. 4, inset i). As the membrane was stepped to positive voltages the increase in the peak transient current became progressively smaller and eventually began to decline (Fig. 4, inset ii), whereas the steady-state current increase remained approximately linear until levelling out at about 90 mV.

The N-shaped current-voltage relationship shown by the transient current has been considered to indicate the presence of a calcium-dependent component of the total potassium current (Meech and Standen, 1975). The experiment was then repeated, first measuring the current at the time of the late transient peak in normal, and then in Ca²⁺-free, ASW (Fig. 5). Removal of calcium eliminated the rapid growth of the current associated with the transient peak, converting the N shape into a shallow curve that meets the former at the lowest point of its decline. Comparable results were obtained by adding the calcium current blockers 1 mmol I⁻¹ Cd²⁺ or 1 μmol I⁻¹ verapamil to the seawater bath. These treatments all abolished the transient current peak at 25–30 ms, giving an approximately linear relationship over the range, and supporting the interpretation that the late transient peak represents a calcium-dependent component of the potassium current.

The time course of the Ca²⁺-and Cd²⁺-sensitive potassium current was examined by comparing the total current in response to a 70 ms depolarising step from a holding voltage of –75 mV before and after the addition of 1 mmol I⁻¹ Cd²⁺ to the sea water bathing the fibre (Fig. 6A). This treatment abolished the transient current peak at 25–30 ms and reduced the steady-state component of the current by about 30 %. The profile of the calcium-dependent potassium current was then obtained from the total outward current by digital subtraction of the Cd²⁺-resistant component (see Fig. 11), after correction for an inward current thought to be carried by calcium (see below).

Although the most prominent feature of the Ca²⁺-dependent current is the transient peak, the reduction in the steady-state current following treatment with cadmium or lμmoll⁻¹ verapamil could have one of several explanations. There may be a residual component of the Ca²⁺-dependent current that inactivates very slowly or shows a voltage-dependency or, possibly, the delayed rectifier current I_K is partially affected by cadmium. Alternatively, it may be due to an entirely separate current that has yet to be isolated from the other three. At present we have not resolved this question.

The activation thresholds of the Ca²⁺-dependent transient peak and steadystate components of the outward current were determined from I/V plots of the current measured at 25 –30 ms and at 70 ms during a series of step depolarizations of a fibre from a holding potential of –75 mV (Fig. 7). As the membrane was stepped to potentials more depolarised than –45 mV a small outward current was activated which rose slowly to a maintained plateau. With larger depolarizations the current increased and developed a peak which reached its maximum amplitude between 25 and 30 ms, before inactivating to the steady-state level (Fig. 4, inset i).

Both the Ca²⁺-dependent transient and steady-state components of the outward current have activation thresholds at around –40 mV, although the transient current rapidly exceeds the maintained current with greater depolarization.

If the calcium-dependent K⁺ current is blocked with 1 mmol I⁻¹ Cd²⁺, the transient peak at 25–30 ms is replaced by a smaller early transient outward current which peaks 10–12 ms after the onset of the pulse (Fig. 6B). The activation kinetics of this early current were faster than those of the late calcium-dependent transient current and its inactivation coincided with the inflection in the leading edge of the total current profile previously noted (Figs 2, 4, inset i).

The addition of 50 mmoll ⁻¹ TEA⁺ to the bath containing sea water and 1 mmol I⁻¹ Cd²⁺ blocked both the residual Ca²⁺-dependent and steady-state components of the potassium current, leaving an early transient current which peaked at 12–15 ms and declined over 40–50 ms (Figs 6B, 9A). This current could also be isolated by blocking the late transient and steady-state currents with TEA⁺ in the absence of Cd²⁺, in which case its magnitude was reduced and its inactivation sometimes revealed a small inward current (see Fig. 9A). The early transient current was voltage dependent, resistant to 50 mmol I⁻¹ TEA⁺ and unaffected by either low-calcium saline or the presence of cadmium in the saline. It has many features associated with the potassium A-current described for molluscan neurones (Connor and Stevens, 1971a,b). At holding voltages of –75 mV, and in the presence of Cd²⁺ and TEA⁺, the activation threshold of the early transient current was –40 mV, a value close to that of the other two potassium currents (Fig. 7). The activation kinetics are faster than those of the calciumdependent current, maximum current amplitude occurring 12–15 ms after the onsets of depolarization (Figs 6B, 9A,B). Many neuronal A-currents are known to inactivate around the resting potential (Rudy, 1988), but measurement of the steady-state inactivation of I_A in these muscle fibres indicates that, although inactivation begins as the membrane is held depolarised to the resting potential, it is not complete until it reaches –30mV (Fig. 8).

If isolated muscle fibres are treated with both 50 mmol I⁻¹ TEA⁺ and 2 mmoll ⁻¹ 4-AP, the three potassium conductances are blocked, revealing a small inward current (Fig. 9A-D). The activation threshold of the inward current measured from a holding potential of –75 mV was around –35 mV (Fig. 10), the current increasing with greater depolarization to peak between 20 and 30 ms. Inward current amplitudes were generally in the range of 12–15 nA, but currents up to 25 nA were recorded from large (2 mm) fibres.

The inward current was blocked by the addition of calcium channel blockers such as 1mmoll⁻¹ Cd²⁺ or 1μmoll⁻¹ verapamil to the bath containing TEA⁺, with no sign of any residual inward current carried by other cations (Fig. 9C).

Profiles of the individual currents cannot be obtained directly from experiments on a single fibre, as the procedures necessary to isolate the Ca²⁺-dependent potassium current also block the inward calcium current. The total outward current recorded in response to the standard depolarising command pulse to 0 mV is opposed by the inwardly directed calcium current. The three potassium currents were derived by simple manipulation of digitized current records of the total outward current (I_t), the cadmium-resistant current (I_Cd), the early transient current (I_A) and the inward current (I_Ca) obtained from experiments on separate fibres (Fig. 11). In this instance the figure is derived from data on five separate fibres, in which the ratios of the component currents were similar, and were scaled to match on the initial total outward current profile. Subtraction of I_A from I_Cd leaves a TEA⁺-sensitive current which rises slowly to a maintained plateau and does not inactivate, having the characteristic profile of the well-known delayed rectifier current (I_K). I_K can also be demonstrated directly by depolarization of fibres in sea water containing 2 mmol I⁻¹ 4-AP, which blocks I_A, and 1 mmol I⁻¹ Cd²⁺, which blocks I_C (Fig. 9D).

When I_C is blocked with cadmium it has been noted that there is also some reduction in the steady-state current at the end of the 70 ms depolarising command pulse. This suggests that some part of I_C persists until the inward calcium current has inactivated. This did not always occur with the pulse duration used in these experiments. However, the results of experiments such as those illustrated in Fig. 9D indicate that, at times, I_K can account for almost all the steady-state current.

In spite of their normal non-spiking mode of operation, spike-like transients are occasionally observed in the fibres of these and other molluscan buccal muscles (Tattershall and Brace, 1987; Dorsett and Evans, 1989). As the two transient potassium currents I_A and Ic dominate the membrane conductance in the first 40 ms following a depolarization, the question of their possible role in preventing spike activity arises.

When isolated fibres in sea water are depolarised in a series of steps, the membrane voltage remains steady at each level. If the medium is exchanged for sea water containing 2 mmol I⁻¹ 4-AP to block I_A and the sequence repeated, as the membrane is stepped to potentials more depolarised than –40 mV, it continues to depolarise, eventually producing a series of regenerative overshooting spikes which may attain amplitudes of 40–50mV (see Fig. 12C). In this medium a spike train frequently occurs on penetration of a fibre. Individual spikes are preceded by a relatively slow depolarization and followed by a rapid repolarization, having a half-peak amplitude duration of around 40 ms. Each spike is accompanied by a twitch-like contraction of the fibre.

To investigate the possible role of the calcium-dependent I_C, a series of experiments was carried out in low-calcium ASW. In zero-calcium solutions no spiking was obtained, but in 2 mmol I⁻¹ Ca²⁺ ASW depolarising the membrane to –40 mV initiated slow oscillations that eventually developed 20 mV spike-like transients (Fig. 12A). Addition of 1 mmol I⁻¹ Cd²⁺ to the bath caused the oscillations to become slower and rather erratic, and it became impossible to induce spiking (not shown).

Isolated fibres were then depolarised in 2 mmol I⁻¹ Ca²⁺ saline containing 2 mmol I⁻¹ 4-AP (Fig. 12B). Under these conditions even small depolarizations became difficult to control, the membrane continuing to depolarise when stepped to around –60 mV, to produce large overshooting 50 mV spikes which attained a frequency of around 4 Hz. Further investigations into the mechanisms underlying the transition to a spiking state will be reported elsewhere.

Isolated fibres of a molluscan unstriated muscle have an array of ionic currents comparable to those found in molluscan neurones (Thompson, 1977). Depolarization of the fibre activates a voltage-dependent transient outward current which is similar in its pharmacology and kinetics to the A-current first described by Connor and Stevens (1971b), but, unlike many neuronal A-currents, its activation threshold is shifted to levels more depolarised than the resting potential (Rudy, 1988). Also, the steady-state inactivation, which in many neurones occurs close to the resting potential, begins as the membrane potential falls below –70 mV, but is only complete at membrane potentials around –35 mV. Comparable departures from what might be considered ‘typical’ A-current properties are also seen in Drosophila muscles (Salkoff and Wyman, 1983), and may be correlated with the more positive activation threshold of systems where the inward current is carried mostly by calcium.

There seems to be no generalised role for either I_A or I_C in excitable cells (Rogawski, 1985; Rudy, 1988). One function of the two transient currents in the muscle fibres of Philine seems to be to prevent spikes, which are readily induced if the early transient A-current is blocked with 4-AP or the fibres are immersed in a 2 mmol l^{− 1} Ca²⁺ saline, which may reduce the late calcium-dependent current I_C. A combination of these treatments both lowers spike threshold and induces regenerative large-amplitude spikes, which can be blocked by the addition of Cd²⁺ to the saline. In normal fibres, the faster activation kinetics and lower threshold of I_A, followed by the development of I_C, would be expected to shunt the inward calcium current and prevent its full expression. Indications that the muscle spikes are mediated by calcium are supported by preliminary experiments (not reported here) where spiking was observed in 2 mmol I⁻¹ Ca²⁺ ASW when choline chloride was substituted for sodium.

Normally non-spiking muscle fibres of crustaceans retain their capacity to develop calcium spikes following the application of TEA⁺ or when the internal calcium activity is reduced (Fatt and Ginsborg, 1958; Hagiwara and Naka, 1964). This implies a suppression of some component of the potassium current, but the details are unknown. In Drosophila melanogaster, the A-current is responsible for the rapid repolarization of the spike in flight muscle and in cervical giant axons (Salkoff and Wyman, 1983; Tanouye et al. 1981), while a similar function is performed by a transient calcium-dependent current in short Purkinje fibres of the calf (Siegelbaum and Tsien, 1980). In a preparation of Drosophila muscle fibres, a partial blockade of I_A with dendrotoxin can lead to explosive discharges of prolonged or giant EJPs when I_K is also blocked or reduced (Wu et al. 1989).

Long-term changes in excitability of some neurones have also been associated with reduction in the voltage-and calcium-dependent potassium currents. During conditioning trials of the mollusc Hermissenda crassicornis, a persistent increase in the excitability of the B photoreceptors results from a reduction in I_A and Ic brought on by the learning paradigm (Alkon, 1984).

The soma membranes of some apparently inexcitable neurones also retain the capacity for regenerative activity and are converted to spiking by perfusion with TEA⁺ or by reducing the calcium gradient (Pitman, 1979; Goodman and Heitler, 1979) but, again, details of the potassium currents involved in these preparations were not studied. However, in Procambarus clarkii, conversion of a non-spiking neurone to a calcium-dependent spiking one results from the voltage-dependent inactivation of the A-current (Czternasty et al. 1989). From the functional viewpoint, if the twitches which accompany the spikes in isolated fibres in low-Ca²⁺ and 4-AP saline were to occur during the normal feeding activity of Philine they would result in tetanic contractions of the entire muscle. In molluscs, where one motoneurone and all the fibres of a muscle may constitute a single motor unit, fine control of tension depends upon a graded depolarization of the fibre by summation of individual junction potentials, and could not be achieved with spiking fibres. In this instance, the two transient outward currents may constitute an important control mechanism in the development of tension by regulating the depolarization leading to calcium influx and preventing regenerative spiking.