K ⁺ Currents in Cultured Neurones From a Polyclad Flatworm

The flatworms (Platyhelminthes) occupy a critical position in metazoan phylogeny as they are deeply rooted in the lineage that leads to most, if not all, phyla of bilateral animals (Valentine et al., 1999). Despite this, little is known about their cellular neurophysiology. According to most phylogenetic reconstructions, modern flatworms are expected to resemble the common ancestors of the protostomes and deuterostomes, and should retain some ancestral features that are common to both. The considerable phylogenetic distance between the flatworms and the eubilaterians (Field et al., 1988; Katayama et al., 1996; Turbeville et al., 1992) suggests that a comparison of the structural and functional plesiomorphic features of their nervous systems may expose fundamental structural motifs, with attendant functional properties, that are masked in higher taxa by accumulated selected characteristics. Identification of these fundamental motifs may provide novel insights into structural correlates of function in the more derived channels in the vertebrates.

Voltage-gated K⁺ channels play a central role in the control of neuronal excitability (Hille, 1992). What is known to date about voltage-gated K⁺ channels of basal phyla confirms that they display considerable diversity in fundamental design. Jellyfish K⁺ channels, for example, appear to possess novel features of voltage sensitivity and inactivation (Grigoriev et al., 1997, 1999a–c; Jegla et al., 1995).

Here, we describe the culture of central neurones from Notoplana atomata and the recording of voltage-gated K⁺ channels from these cells using the whole-cell, patch-clamp technique. Notoplana atomata is a polyclad genus for which detailed extracellular and intracellular data are available (Faisst et al., 1980; Keenan and Koopowitz, 1984; Keenan et al., 1979; Koopowitz et al., 1975, 1976, 1979a,b). No voltage-clamp recordings of ionic currents from the neurones of polyclads have been reported. The polyclads are free-living turbellarians and the most basal of the Platyhelminthes, except for the acoels (whose placement within the Platyhelmithes has been challenged; Ruiz-Trillo et al., 1999) and the catenulids, which form a sister group to the rest of the Bilateria (Carranza et al., 1997). The only K⁺ channel cloned from platyhelminths is from the parasitic trematode Schistosoma mansoni (Kim et al., 1995), and the only description of K⁺ currents in platyhelminths is for dispersed muscle cells of Schistosoma mansoni (Day et al., 1993) and for cultured neurones of the triclad Bdelloura candida (Blair and Anderson, 1993). We show that K⁺ currents of Notoplana atomata are extremely rightward-shifted, suggesting the extant occurrence of a class of comparatively voltage-insensitive, voltage-gated K⁺ channels in the Cnidaria, Platyhelminthes, Annelida and Mollusca, which functionally resemble the unusual Kv3 class of vertebrate Shaker-like channels.

Notoplana atomata (0.5–3 cm) were collected from the underside of rocks in the intertidal zone near Bamfield Marine Station, Bamfield, British Columbia, Canada, and were kept in a glass tank continually supplied with sea water (9–12 °C) from the system at the Station. Animals were used within 3 weeks of capture and fed on brine shrimp.

Individuals were anaesthetised by immersion in a 1:1 (v/v) mixture of sea water and 0.33 mol l⁻¹ MgCl₂ for at least 20 min. They were sterilised by brief immersion in 70 % ethanol, washed in filtered artificial sea water and then pinned out in a Petri dish lined with Sylgard resin. An incision was made in the dorsal epithelium, and the nerves were severed so as to leave fairly long stubs. The brain was removed and placed in a separate Petri dish containing sterile filtered (0.22 μm) artificial sea water. Following longitudinal bisection of the brain with fine scissors, the sheath was removed using two sharpened tungsten needles. The resulting hemiganglia were placed in a solution of 10 mg ml⁻¹ pronase in artificial sea water. Teasing the hemiganglia apart with sharpened tungsten needles assisted entry of the enzyme. After approximately 20 min, depending on the rate of dissociation, which varied from preparation to preparation, hemiganglia were rinsed well in several changes of culture medium and transferred to a Petri dish (30 mm, Falcon) containing filtered culture medium (see below) and gently triturated through fire-polished, silanized Pasteur pipettes or glass capillaries. Cultures were incubated at 12 °C for 1–7 days before use in electrophysiological experiments.

Cells with a cell body diameter of between 40 and 100 μm, and with processes extending for more than twice the cell body diameter, were selected for electrophysiological recording. Cultures were washed several times in artificial sea water (or modified artificial sea water if required by the experiment) and transferred to the stage of a Nikon inverted microscope. Glass pipettes were manufactured from borosilicate glass using a Sutter puller and fire-polished using a Narashige microforge. Pipettes had resistances in the range 300 kΩ to 2 MΩ when filled with the pipette solution described below. These pipettes readily formed seals in the 1–10 GΩ range, and the whole-cell configuration could be obtained by applying negative pressure or a combination of negative pressure and a brief (10–100 ms) pulse of depolarising current. Access resistances were in the range 2–4 MΩ. The cells were voltage-clamped using an Axopatch 1D amplifier, and currents were recorded using the pClamp 6.0 data-acquisition suite (Axon Instruments) running on a Dell Dimension P166v personal computer. Data were analysed using a combination of pClamp 6.02 (Axon Instruments) and SigmaPlot 4.0 (SPSS) analysis software. All recordings were compensated for series resistance by 50–80 % using the feedforward circuit built into the amplifier. Greater compensation could not be used without causing saturation of the feedback amplifier. Thus, a significant voltage error remained in some of our current recordings, which could amount to 20 mV for a 10 nA current after 50 % compensation. The estimated worst-case error is given in the figure legends and in the text where appropriate. Capacitative and leak currents were subtracted on-line using the P/N protocol provided with the data-acquisition software. N inverted copies of the test waveform (the amplitude of each being 1/N of the main trial) were applied before each trial. The sum of the currents applied during these prepulses was subtracted from the currents recorded during the main waveform.

Salines and pipette-filling solutions were prepared as in Table 1. The volume of 1 mol l⁻¹ KOH required to balance the pH was measured to determine the final concentration of K⁺. Culture medium consisted of 75 ml of artificial sea water mixed with 25 ml of Leibowitz L-15 culture medium (Sigma, USA) and 50 μg ml⁻¹ gentamycin sulphate to which 475 mg of NaCl was added to balance the osmotic pressure to a value approximating that of sea water (approximately 1 osmol l⁻¹).

Dissociated cells adhered readily to untreated plastic Petri dishes. Neurite outgrowth began within 3 h, although some cells had retained neurites through dissociation and trituration. A wide range of cell shapes was seen in culture, including monopolar, bipolar and multipolar cells (Fig. 1), with homopolar and heteropolar examples of these classes. We have assumed that most cells with elongated processes are neurones, although some may be non-neuronal, ensheathing cells.

Cell bodies were 5–100 μm in diameter. Many had aspect ratios near unity, whereas others were distinctly pear-shaped with the cell body merging continuously into the primary neurite. Most neurones had several obvious growth cones. Many neurites were flattened with veil-like lateral extensions (Fig. 1).

The whole-cell, recording configuration could be maintained for 10–30 min. To determine the current–voltage properties of cells in normal artificial sea water (ASW), a series of 100 ms voltage excursions from −90 mV to a range of positive potentials in 10 mV steps was imposed (Fig. 2). In response to a long (>100 ms) test pulse, most cells displayed outward currents that contained both a rapidly inactivating component and a slowly inactivating or non-inactivating component (Fig. 2A), while the remaining neurones expressed only a slowly or non-inactivating outward current (Fig. 2B). These currents were activated above approximately −50 mV. Although most cells produced little or no obvious inward current, 5–10 % of cells gave noticeable, rapid inward currents which reversed above +40 mV (Fig. 2C). These cells all had large cell bodies (capacitance >100 pF) and were difficult to space-clamp. Under current-clamp conditions, these large cells supported overshooting action potentials (Fig. 2D). Not all the cells with cell bodies of this size displayed action potentials.

Outward currents were activated at potentials above −10 mV (Fig. 3) when Na⁺ and Ca²⁺ currents were eliminated using ASW for the isolation of K⁺ currents (K-ASW; see Table 1). An inactivating component was present in 83 % of cells (Fig. 3A), whereas only a non-inactivating current was seen in the remaining cells (Fig. 3B). The mean ratio of the peak to steady-state current amplitudes was 2.88±0.47 (N=34) for a 100 ms depolarising step from −90 mV to +80 mV. The distribution of values falling in bins of width 0.5 were counted and found not to differ significantly from a normal distribution (Kolgorov–Smirnov test, P=0.29). Therefore, the ratio of peak to steady-state currents does not provide a criterion by which to classify cells. The difference between the initial and steady-state current amplitudes was not correlated with that of the steady-state current (Pearson product moment, P=0.637). Time to peak decreased slightly with increasing depolarisation from a value of 12.2±3.5 ms (N=18) for steps to 0 mV to 3.0±0.8 ms (N=18) for steps to +80 mV and to 1.37±0.13 ms (N=5) for steps to +130 mV (means ± S.E.M.). The latter two values differ significantly from the first, but not from each other (Kruskal–Wallis one-way analysis of variance on rank, P<0.05). The mean time constant for activation measured from the steep rising phase of responses to depolarising voltage steps to +80 mV was found to be 0.94±0.79 ms (mean ± S.E.M., N=15). When the rapidly inactivating component was isolated with a conditioning, 100 ms depolarising pre-pulse to +60 mV (Fig. 3C), the time constant of the steep rising phase of the remaining current was 1.10±0.2 ms (mean ± S.E.M., N=9).

where G(v) is conductance, G_max is maximum conductance, V is membrane potential, s is a slope factor, from which a half-activation potential (V₅₀) of 16.9±1.9 mV (mean ± S.E.M., N=6) and slope of 15.0±1.8 mV e⁻¹ (mean ± S.E.M., N=6) were estimated after correction for predicted series resistance errors (assuming a series resistance of 4 MΩ and 50 % compensation).

The time course of the tail currents was best fitted by a double-exponential function. Both components had similar activation properties, suggesting that they represent a single current decaying with two time constants or two currents with similar activation ranges.

The rapidly inactivating component was isolated by digitally subtracting currents obtained during a depolarising step to +80 mV, which were preceded by a +20 mV conditioning pulse, from those obtained without the conditioning pulse. In 22 out of 26 cells, the resulting current inactivated with a time constant of 26.1±1.9 ms. In four cells, however, a better fit was obtained using a double-exponential function with time constants of 20.5±7.0 ms and 98.7±50.8 ms (means ± S.E.M.).

The voltage-dependence of inactivation was measured using a paired-pulse protocol in ASW. From a holding potential of −90 mV, the membrane was stepped to a range of potentials ranging from −120 to +60 mV for 200 ms and then immediately to +100 mV or +70 mV. The amplitudes of the peak current in response to the second pulse were plotted against the membrane potential of the preceding pulse and normalised (Fig. 4). Individual inactivation curves were fitted by Boltzmann functions with a mean half-inactivation potential of −35.4±3.0 mV and a slope of −9.5±0.6 mV e⁻¹ (mean ± S.E.M., N=17). The range of values was −64.2 to −16.7 mV, although 14 of the 17 cells fell in the range −36.4 to −16.7 mV.

The time constant of inactivation of individual cells was measured during a depolarising pulse from −90 mV to +80 mV in ASW (12 mmol l⁻¹ K⁺), then again during perfusion with elevated [K⁺] (52 or 152 mmol l⁻¹). There was no significant difference between the time constants measured (12.6±5.2 ms for 152 mmol l⁻¹ K⁺ and 15.8±3.1 ms for 52 mmol l⁻¹ K⁺, t=0.921, d.f.=5, P=0.399, paired t-test). The rate of inactivation was also independent of the voltage of the conditioning pulse (data not shown).

To measure the time course of recovery from inactivation, two 100 ms depolarising steps to +100 mV from a holding potential of −70 mV were applied at successively diminishing intervals. The ratio of the peak amplitudes, after subtraction of the steady-state value, was plotted against the interval between the pulses (Fig. 5A, inset). Recovery from inactivation followed a single-exponential function, suggesting the existence of only one inactivated state. The rate of recovery was voltage-dependent (τ=50±2.5 ms at −90 mV, 111±4.9 ms at −70 mV and 167±8.3 ms at −50 mV, Fig. 5A) but was independent of the concentration of external K⁺ (τ=124.4±18.4, N=4, for 12 mmol l⁻¹ [K⁺]_out, 98.7±14.4, N=4, for 52 mmol l⁻¹ [K⁺]_out, values that are not significantly different (one-way analysis of variance, P=0.554).

Tail-current protocols were used to determine the dependency of the reversal potential upon the external concentration of K⁺. From a holding potential of −70 mV, the membrane potential was stepped to +100 mV for 10 ms to activate the currents and then to a series of potentials between −110 mV and +100 mV, at which point the tail currents were measured. This was performed in salines containing 12 mmol l⁻¹, 52 mmol l⁻¹ and 152 mmol l⁻¹ K⁺. The reversal potentials measured in this way (−68.0 mV, −35.7 mV and −10.5 mV respectively, Fig. 5B, inset) were dependent upon the concentration of extracellular K⁺ but differed from the equilibrium potentials for K⁺ predicted from the concentrations of K⁺ in the bathing medium and the pipette using the Nernst relationship (−78.3 mV, −45.9 mV and −21.8 mV, respectively, Fig. 5B). This suggests that these voltage-activated outward currents are mediated by K⁺ and by some other, unidentified ionic species.

Voltage-clamped cells were perfused for 3 min in K-ASW containing 10 mmol l⁻¹ 4-aminopyridine (4AP). The currents evoked by a series of depolarising steps were subtracted from those previously obtained from the same cell in normal saline to reveal the 4AP-sensitive currents (Fig. 6). These currents were rapidly and completely inactivating with a time constant of 20.6±5.4 ms (mean ± S.E.M., N=5). In contrast, bath-application of tetraethylammonium (25 mmol l⁻¹) had no measurable effect on either rapidly inactivating or non-inactivating currents (N=11, data not shown). Outward currents were still recorded in all cells in the presence of 10 mmol l⁻¹ Ba²⁺ in the bathing solution. Replacement of all the K⁺ in the saline and in the pipette-filling solution with Cs⁺ abolished most of the outward currents.

We have shown that there are at least two K⁺ currents present in cultured Notoplana atomata neurones: one rapidly activated (τ=0.9 ms) and inactivated (τ=26 ms), and the other also activated rapidly but showing no inactivation over 100 ms. Both currents activated at similar voltages (V₅₀=17 mV). The only characteristics that distinguished these two currents are the inactivation properties and sensitivity to 4AP. No other outward currents could be distinguished on the basis of the voltage-dependence of activation or inactivation, sensitivity to 4AP, barium or tetraethylammonium, or time to peak. Although we have not excluded the possibility that more than two current types exist in the cells from which we recorded, the simplest explanation of our results is that there are only two K⁺ currents present. This suggests that variation in excitability properties is brought about by the differential expression of these two channel types in each neuronal phenotype. However, we cannot exclude the possibility that there are other populations of cells displaying other types of K⁺ current that are not represented in our culture because of ‘loss’ during the culturing procedure.

The rate of inactivation of the rapidly inactivating current was independent of the concentration of external K⁺, in contrast to Polyorchis penicillatus K⁺ currents. This agrees with the suggestion of Grigoriev et al. (1999a) that K⁺-dependent regulation of inactivation is a compensatory mechanism in organisms that have a paucity of glial cells. The Notoplana atomata brain is rich in glial cells (Koopowitz and Chien, 1974), which might provide an effective K⁺ homeostatic mechanism.

Like most rapidly inactivating K⁺ currents, the activation and inactivation curves of the currents we describe overlap. However, the values of activation threshold and half-activation are much more positive than for most K⁺ currents reported to date and do not overlap with the resting potential of in situ neurones (Keenan and Koopowitz, 1984). In this respect, they resemble cloned K⁺ channels from the mollusc Aplysia californica (Pfaffinger et al., 1991), the platyhelminth Schistosoma mansoni (Kim et al., 1995) and two cnidarian channels (Jegla et al., 1995). Thus, a class of rapidly inactivating K⁺ channels with rightward-shifted activation curves is widespread among many invertebrates. This is probably not an adaptation of marine invertebrates to the high salt content of their extracellular fluids, since activation curves in muscle cells of Schistosoma mansoni (Day et al., 1993, 1995) and in lamina cells of the locust Schistocerca gregaria (Benkenstein et al., 1999) are also rightward-shifted, whereas those of some Shaker currents in Aplysia californica are not (Pfaffinger et al., 1991).

With a threshold of approximately −30 mV and a half-activation value of +17 mV, it is unlikely that either the rapidly inactivating or the slowly or non-inactivating component plays a role in subthreshold integration, as suggested for I_K(A) currents (Connor and Stevens, 1971). However, these currents are likely to be activated during action potentials since Notoplana atomata neurones in situ express overshooting action potentials whose peak in natural sea water may reach +50 mV with durations of 1–2 ms (Keenan and Koopowitz, 1984). Such depolarisations would be adequate to activate both K⁺ currents almost fully. However, with a time constant of inactivation of approximately 26 ms and a time constant of recovery from inactivation of approximately 50 ms, inactivation is only likely to accumulate significantly at interspike intervals of less than approximately 50 ms, corresponding to frequencies greater than approximately 20 s⁻¹, suggesting that inactivation is an adaptation to high-frequency discharge. Multimodal interneurones of Notoplana atomata are capable of sustaining discharges of the order of 100 s⁻¹ (Keenan and Koopowitz, 1984). Within the mammalian Shaker superfamily of genes cloned to date, only the Kv3.1 and Kv3.4 families have rightward-shifted activation curves. Interestingly, recent evidence suggests that Kv3 channels may play a specific role in adapting neurones to high-frequency firing (Liu and Kaczmarek, 1998; Wang et al., 1998). The selective block of the inactivating current by 4AP and the ineffectiveness of tetraethylammonium as a blocking agent resemble findings from other cultured flatworm (Bdelloura candida) neurones (Blair and Anderson, 1993) and recall the lack of action of extracellular tetraethylammonium on spikes recorded from Notoplana atomata neurones in situ. The slope factor (15 mV e⁻¹) for activation of the rapidly inactivating current in Notoplana atomata neurones is comparable with the 15 mV e⁻¹ reported for cloned and 12 mV e⁻¹ for in situ K⁺ channels from Schistosoma mansoni (Kim et al., 1995). It has been reported that the threshold and the slope factor are influenced by the total number of charges in the S4 segment of the channel (Liman et al., 1991; Logothetis et al., 1992, 1993; Papazian et al., 1991), although this is not always so (Smith-Maxwell et al., 1998). The gentle activation slope might therefore be explained by a reduction in the total nominal charge in the S4 segment of Notoplana atomata K⁺ channels. Resolution of these questions awaits the cloning and functional expression of Notoplana atomata K⁺ channels, which is in progress.

We thank the staff at Bamfield Marine Station for their assistance in providing ideal facilities for this study, which was supported by a NSERC research grant (A0419) to A.N.S.