Pharmacological Characterization of Ionic Currents that Regulate the Pacemaker Rhythm in a Weakly Electric Fish G. Troy Smith and Harold H. Zakon Section of Neurobiology, School of Biological Sciences, Patterson Laboratories (C0920), University of Texas, Austin
Received 6 July 1999; accepted 10 September 1999
Electric organ discharge (EOD) frequency in the brown ghost knifefish (Apteronotus leptorhynchus) is sexually dimorphic, steroid-regulated, and determined by the discharge rates of neurons in the medullary pacemaker nucleus (Pn). We pharmacologically characterized ionic currents that regulate the firing frequency of Pn neurons to determine which currents contribute to spontaneous oscillations of these neurons and to identify putative targets of steroid action in regulating sexually dimorphic EOD frequency. Tetrodotoxin (TTX) initially reduced spike frequency, and then reduced spike amplitude and stopped pacemaker activity. The sodium channel blocker O-conotoxin MrVIA also reduced spike frequency, but did not affect spike amplitude or production. Two potassium channel blockers, 4-aminopyridine (4AP) and A-conotoxin SIVA, increased pacemaker firing rates by approximately 20% and then stopped pacemaker firing. Other potassium
channel blockers (tetraethylammonium, cesium, ␣-dendrotoxin, and agitoxin-2) did not affect the pacemaker rhythm. The nonspecific calcium channel blockers nickel and cadmium reduced pacemaker firing rates by approximately 15–20%. Specific blockers of L-, N-, P-, and Q-type calcium currents, however, were ineffective. These results indicate that at least three ionic currents—a TTX- and O-conotoxin MrVIA-sensitive sodium current; a 4AP- and A-conotoxin SIVA-sensitive potassium current; and a T- or R-type calcium current— contribute to the pacemaker rhythm. The pharmacological profiles of these currents are similar to those of currents that are known to regulate firing rates in other spontaneously oscillating neural circuits.
Like other weakly electric fish species, the brown ghost knifefish (Apteronotus leptorhynchus) produces an electric organ discharge (EOD) that functions in electrolocation and communication (reviewed in Hopkins, 1988). Objects near the fish distort the electric field generated by the EOD, and the fish use specialized electroreceptors in their skin to detect these distortions. In addition, the frequency and waveform of
the EOD is species specific, sexually dimorphic, and steroid-sensitive, enabling the EOD to be used as a social signal communicating species, sex, and breeding status. The neural circuitry that controls this behavior is remarkably simple. The frequency of the EOD is determined by the medullary pacemaker nucleus (Pn), which contains three neuron types: (1) pacemaker cells, which generate the EOD rhythm; (2) relay cells, which convey this rhythm to the electromotor neurons (EMNs) in the spinal cord; and (3) small interneurons, which form chemical synapses on relay and pacemaker cells, but whose function has not yet been well
ABSTRACT:
Correspondence to: G. Troy Smith (
[email protected]). Contract grant sponsor: National Institutes of Health; contract grant number: F32 NS10150 (GTS) and R01 MH56535 (HHZ). © 2000 John Wiley & Sons, Inc. CCC 0022-3034/00/020270-17
270
© 2000 John Wiley & Sons, Inc. J Neurobiol 42: 270 –286, 2000
Keywords: pacemaker rhythm; conotoxin; weakly electric fish; sodium current; potassium current; calcium current
Ionic Currents in Pacemaker Nucleus
characterized (Ellis and Szabo, 1980; Tokunaga et al., 1980; Elekes and Szabo, 1981, 1985; Turner and Moroz, 1995; Smith et al., 1999). Pacemaker cells are extensively electrotonically coupled with each other and with relay cells, and the connections between relay cells and EMNs are also electrotonic (Tokunaga et al., 1980; Elekes and Szabo, 1981). In brown ghosts and other apteronotids, the peripheral portions of the EMN axons form the electric organ. Thus, the circuit that controls the EOD consists of only four identifiable cell types and their relatively simple connections. The neurons that control the EOD all fire action potentials in a 1:1 correspondence with the EOD; e.g., in a fish that has an EOD frequency of 800 Hz, the pacemaker, relay, and electromotor neurons continuously produce 800 action potentials per second. The EOD is also a highly precise behavior; indeed, the neural circuit that generates the EOD rhythm is reported to be the most precise biological oscillator known (Bullock, 1970; Moortgat et al., 1998). Furthermore in brown ghosts EOD frequencies are frequently as high as 1000 Hz, and the EODs of other apteronotids can approach 1800 Hz (Hopkins, 1974; Kramer et al., 1980). Thus, the apteronotid electromotor circuit is also one of the most rapid neural oscillators. The electromotor system is therefore wellsuited for investigating the mechanisms by which neural circuits generate rapid and precise rhythmic behavior. Neurons of in vitro preparations of the Pn and EMNs continue to fire at frequencies that are highly correlated with the fish’s EOD frequency (Meyer, 1984; Schaefer and Zakon, 1996). These preparations provide an opportunity to investigate the biophysical mechanisms by which neurons generate spontaneous rhythms. Furthermore, because the EOD and the firing frequencies of the Pn and EMNs are sexually dimorphic and regulated by gonadal steroids, these preparations are also excellent models in which to study steroid action on the biophysical properties of neurons that control a reproductive behavior. A previous study (Dye, 1991) used ionic manipulations and channel-blocking drugs to qualitatively characterize the ionic currents important for Pn activity. Our study extends these results by quantifying the effects of ion channel blockers on the firing frequency of Pn neurons and by using blockers of specific subclasses of ion channels to further characterize the ionic currents that underlie EOD rhythm generation and that are thus putative targets for steroid action on EOD frequency.
271
METHODS Subjects Brown ghost knifefish were obtained from commercial suppliers and were housed in communal tanks at room temperature (approximately 24 –27°C). All experiments were performed within the guidelines outlined by the National Institutes of Health and by using protocols approved by the University of Texas Animal Care Committee.
Pharmacological Agents Toxins were obtained from the following suppliers: 4-aminopyridine (4AP), Aldrich (Milwaukee, WI); tetraethylammonium chloride (TEA), nifedipine, and -conotoxins (CTXs) MVIIC and GVIA, Sigma (St. Louis, MO); Oconotoxin MrVIA and A-conotoxin SIVA, generous gift of Dr. Baldomero Olivera, University of Utah (see McIntosh et al., 1995; Craig et al., 1998 for purification and characterization); -agatoxin IVA (AGA IVA), Alexis Biochemicals (San Diego, CA); tetrodotoxin (TTX) and agitoxin-2 (AGI-2), Alomone Laboratories (Jerusalem, Israel); and ␣-dendrotoxin (␣DTX), Calbiochem (La Jolla, CA).
Tissue Preparation Brown ghosts were deeply anesthetized with 0.75% 2-phenoxyethanol (Sigma) in deionized water. The fish were placed on ice, and their brains were quickly removed and placed in cold artificial cerebrospinal fluid (ACSF, in mM: NaCl, 124; KCl, 2; KH2PO4, 1.25; MgSO4, 1.1; CaCl2, 1.1; NaHCO3, 16, D-glucose, 10; pH 7.4). The Pn was clearly visible as a protrusion on the ventral surface of the medulla. The dura mater overlying the Pn was removed with fine forceps, and the Pn was excised from the brain with iridectomy scissors. The Pn was placed in a recording chamber and continuously superfused (1–2 mL/min) with ACSF that was bubbled with 95% O2/5% CO2. The preparation was secured to the Sylgard bottom of the recording chamber by pinning a nylon mesh over it. We used a thermistor probe placed within 2 mm of the preparation and a temperature controller (model TCU-2, Fine Science Tools, Foster City, CA or model TC2BIP, Cell Micro Controls, Virginia Beach, VA) to monitor and control the temperature of the recording chamber.
Electrophysiological Recordings Pn recordings were made with borosilicate microelectrodes (A-M Systems, Everett, WA, 1.2 mm od, 0.68 mm id, resistance 30 – 80 M⍀) filled with 3 M KCl. Signals were amplified with an Intra 767 amplifier (World Precision Instruments, Sarasota, FL), filtered at 10 kHz, and stored on a data recorder (model 400, Vetter, Rebersburg, PA) for off-line analysis. Intra- and extracellular recordings were made from neurons in the in vitro Pn. A total of 255 cells
272
Smith and Zakon
were exposed to a channel-blocking toxin during recordings, and 185 cells were exposed to vehicle alone. Of the 255 cells exposed to channel blockers, 44 were recorded intracellularly during the entire drug application. In a few cases, we could distinguish intracellular recordings of pacemaker versus relay cells based on action potential waveform. For example, gradually depolarizing pacemaker potentials, characteristic of pacemaker neurons, were observed in a few intracellular recordings, particularly when firing frequency was reduced by certain drugs [Fig. 1(c)]. For most intracellular recordings and all extracellular recordings, however, we could not unambiguously determine whether we were recording from a pacemaker or relay cell, nor could we determine whether the recordings were from axons versus somata. Dye and Heiligenberg (1987) similarly reported that brown ghost relay and pacemaker neurons were difficult to unambiguously distinguish from each other by spike waveform alone. All pacemaker and relay neurons are electrotonically coupled and fire synchronously, however, and any change in the firing rates of pacemaker neurons would therefore be reflected in recordings made from relay neurons and vice versa.
Drug Perfusion The baseline activity of each neuron was recorded for 3–5 min. The entire preparation was then exposed to either a channel-blocking drug or vehicle control. For experiments with some drugs (4AP, cesium, nickel, cadmium, and nifedipine), the preparation was continuously perfused with artificial cerebrospinal fluid (ACSF) containing the drug. In these cases, the drugs reached the bath within 30 – 60 s of starting perfusion. For other drugs (conotoxins, AGA IVA, AGI-2, ␣DTX, TTX), the perfusion of the bath was stopped before baseline recording, and the volume of ACSF in the bath was fixed at 200 L. After baseline recordings, drugs were directly added to the bath to yield an appropriate final concentration, and the drugs were thoroughly mixed into the bath solution by repetitively applying gentle positive and negative pressure to the bath through a 20L pipette. A 95%/5% mixture of O2 and CO2 was blown over the surface of the bath to oxygenate, maintain the pH, and gently mix the bath solution. For drug concentrations above 1 mM (cesium and TEA experiments), the concentration of NaCl in the ACSF was reduced to maintain the same osmolarity. Some drugs required special vehicles. Several peptide drugs (CTXs, AGA IVA, and ␣DTX) were dissolved in a stock solution of ACSF containing 0.1% cytochrome C (Sigma) to prevent the peptides from adhering to the containers, pipettes, or recording chamber. AGI-2, OCTX MrVIA, and ACTX SIVA were dissolved in stock solutions containing 0.1% bovine serum albumin (BSA, Grade V, Sigma), and the bath to which these drugs were applied contained 0.01% BSA. Stock concentrations of bath-applied drugs were 10 – 40 times the final concentrations to which the preparation was exposed. Nifedipine was dissolved in
ethanol or DMSO before being diluted 1:1000 in ACSF. Cadmium could not be dissolved in the bicarbonate-buffered ACSF because it would precipitate as the insoluble CdCO3. Cadmium experiments were therefore conducted in HEPES-buffered ACSF (same as normal ACSF, except NaHCO3 was replaced by 5 mM HEPES and pH was adjusted to 7.4 with NaOH). For all of these drugs, the same preparations were exposed to the vehicle alone to control for effects of the agents in the vehicle. Recordings were made in drugs for time periods ranging from 2 to 60 min, depending on the effect of the drug and the stability of the recording. The drugs were then washed out of the bath for 20 – 60 min. In cases in which multiple cells were recorded from the same preparation, drugs were washed out for at least 1 h between recordings.
Data Analysis Recordings were digitized with a USM-100 signal manifold (World Precision Instruments) and an analog– digital converter (model DT2821, Data Translation, Marlborough, MA, or Digidata 1200, Axon Instruments, Foster City, CA), and analyzed using data analysis software (Spike, Hillal Associates, Englewood, NJ, or Axoscope 1.1, Axon Instruments) on a microcomputer. Peak-to-peak firing frequency was measured every 30 s from 25-ms windows. EOD frequency and the firing rate of the Pn varies with temperature (Enger and Szabo, 1968; Meyer, 1984). As part of a separate study (K. D. Dunlap, G. T. Smith, and A. Yekta, unpublished observations), the firing rates of cells from several Pn preparations were measured over a temperature range of 23–27°C. From these data, we calculated a Q10°C of 1.78 for the firing rate of neurons in the isolated Pn. This Q10°C was used to temperature-compensate the measured firing frequencies to those expected at the temperature at which the fish’s EOD was measured. The Q10°C we used in this study was somewhat higher than that previously reported for EOD frequency (1.5; Enger and Szabo, 1968) or firing frequency of neurons in an in vitro Pn preparation (1.58; Meyer, 1984). Using the lower Q10°Cs of these previous studies, however, did not change the direction or statistical significance of any of the results reported here.
Statistical Analyses We compared the percent change from baseline firing frequency of vehicle control versus drug-treated cells in the Pn preparations by using one-factor analyses of variance (ANOVAs). In cases in which the data did not meet the ANOVA’s assumption of homogeneity of variance between treatment groups (TTX, nickel, and cadmium comparisons), ANOVAs were performed on logarithmically transformed data. When the results of the ANOVAs were significant, comparisons between different doses of drugs and control treatments were made by using Fisher’s Protected Least Significant Difference (PLSD) tests. The effects of TTX and
Ionic Currents in Pacemaker Nucleus 4AP on action potential waveform parameters were determined by using paired t tests to compare values before and after drug treatment. An alpha level of .05 was used for all statistical tests.
RESULTS Sodium Channel Blockers The sodium channel blocker TTX affected the firing of Pn neurons in two ways: (1) TTX decreased Pn firing rate; and (2) TTX decreased spike amplitude (Fig. 1). Five nanomolar TTX reduced spike frequency by approximately 20% within 60 min [Fig. 1(a)] None of the cells exposed to this dose of TTX, however, stopped firing, and the effects of 5 nM TTX were largely reversible (70 – 80% recovery in 30 min, not shown). At higher doses (25 nM–1 M), TTX had a more profound effect on spike frequency [Fig. 1(a)]. Furthermore, spike amplitude declined dramatically in these cells, and all of them stopped firing completely between 5 and 30 min after TTX exposure. Some cells resumed firing when TTX was washed out, but only at low rates. The effects of 25 nM versus 1 M TTX on pacemaker firing did not differ significantly from each other, which indicates that the TTX dose response saturates by at most 25 nM. In intracellular recordings of TTX-treated neurons, the reduction in firing frequency of Pn neurons was accompanied by a significant hyperpolarization of the minimum membrane potential [Fig. 1(c), Table 1, n ⫽ 5]. Before stopping pacemaker activity, TTX also significantly altered action potential waveform by decreasing the slopes of both the rising and decay phases (Table 1). The effects of TTX on spike frequency and amplitude were temporally separable. In the cell shown in Figure 1(b,c), TTX decreased spike frequency between 200 and 400 s after TTX application. During this time, spike amplitude remained constant, or even increased slightly. Approximately 403 s after TTX application, the cell stopped firing, though subthreshold membrane potential oscillations persisted for several more seconds [bottom panel of Fig. 1(c)]. In other cells, a more gradual decline in spike amplitude preceded the cessation of firing. In no case, however, did the decline in spike amplitude precede the decline in spike frequency. In the cell shown in Figure 1(c), gradually depolarizing pacemaker potentials preceded each action potential, and these pacemaker potentials were more prominent after firing frequency slowed in response to
273
TTX. The presence of such pacemaker potentials is characteristic of pacemaker neurons, although the pacemaker potentials can be difficult to unambiguously detect in rapidly firing apteronotid pacemaker neurons (Dye and Heiligenberg, 1987). A pronounced afterhyperpolarization was present in this cell after TTX treatment and was also observed in the only other cell exposed to TTX that had unambiguous pacemaker potentials (but was not observed in other intracellularly recorded neurons exposed to TTX). The reduction in action potential width seen in the post-TTX recordings in Figure 1(c) was not observed in other intracellularly recorded neurons exposed to TTX (Table 1). Pacemaker firing frequency, but not the production of action potentials, was also affected by the sodium channel-blocking peptide, O-conotoxin (OCTX) MrVIA (Fig. 2). OCTX MrVIA, a toxin isolated from the venom of the marine snail Conus marmoreus, blocks rat brain type II sodium channels with an IC50 of ⬃200 nM, but does not affect peripheral mammalian sodium channels (McIntosh et al., 1995; Terlau et al., 1996). At doses of 250 nM, OCTX MrVIA significantly reduced pacemaker firing frequency by 11.82 ⫾ 0.61% within 30 min. This effect was not reversed by up to 30 min of drug washout. No additional effect of OCTX MrVIA was found at concentrations of 1 M, which suggests that 250 nM was a saturating dose. Unlike TTX, OCTX MrVIA at these saturating concentrations did not cause the pacemaker to stop firing, even after 1 h of exposure. Furthermore, in two intracellularly recorded neurons, OCTX MrVIA did not affect action potential amplitude, but did reduce spike frequency by 15.8 –16.5% and hyperpolarize the minimum membrane potential by 6 –7 mV.
Potassium Channel Blockers The potassium channel blocker 4AP also profoundly affected the firing of Pn neurons. At concentrations of 100 M, 4AP caused Pn neurons to fire ⬃20% faster within 2–3 min and then caused these neurons to stop firing, presumably because of depolarization of the cell and inactivation of voltage-gated conductances [Fig. 3(a,b)]. In intracellular recordings (n ⫽ 5), the increase in firing frequency was accompanied by several changes in action potential waveform: (1) a depolarization of the minimum membrane potential; (2) decreases in the slopes of both the rising and falling phases of the action potential; (3) an increase in the width of the action potentials; and (4) a reduction in spike amplitude that culminated in the complete ab-
274
Smith and Zakon
Figure 1 Effects of tetrodotoxin (TTX) on pacemaker firing frequency. (a) Percent change from baseline firing frequency of Pn neurons exposed to TTX at 5 nM (n ⫽ 8), 25 nM (n ⫽ 5), and 0.1-1.0 M (n ⫽ 13) compared to control preparations exposed to vehicle alone (n ⫽ 11). *25 nM and 0.1–1.0 M TTX versus controls (one-factor ANOVA, PLSD, p ⬍ .05). † 5 nM TTX versus controls (one-factor ANOVA, PLSD, p ⬍ .05). Too few control preparations recorded at 60 min for statistical comparison. (b) Spike amplitude (filled squares) and frequency (filled circles) for a single intracellularly recorded pacemaker neuron exposed to 25 nM TTX at time 0. Spike frequency began to decline approximately 200 s after TTX application. Spike amplitude remained relatively constant until approximately 403 s after TTX application, when spikes failed entirely. (c) Representative voltage traces from the same cell as in (b), taken at time points indicated by numbered arrows. Pacemaker potentials were evident between spikes (solid arrowheads), indicting that this cell was a pacemaker cell. Note the decrease in firing frequency and baseline membrane potential following TTX treatment. The narrowing of spikes in traces 2 and 3 was not seen in other cells, but the pronounced afterhyperpolarization following TTX treatment (open arrowhead) did occur in another cell that had apparent pacemaker potentials. Trace 3 shows last two spikes produced by this cell. Subthreshold membrane potential oscillations persisted for several seconds after the last action potential. The peak of the first subthreshold oscillation is indicated by the arrow. Scale bar ⫽ 1 ms, 10 mV. Spike frequency and amplitude for the three traces: (1) 598 Hz, 32 mV; (2) 345 Hz, 33 mV; (3) 141 Hz (frequency of last 5 spikes), 34 mV.
sence of activity [Fig. 3(b); Table 2]. Most of the neurons exposed to 4AP stopped firing 3–5 min after 4AP exposure. These effects were largely reversible in most cases. Firing resumed and returned to baseline rates within 20 –30 min of 4AP washout (not shown). Another potassium channel blocker, A-conotoxin
(ACTX) SIVA, similarly disrupted the firing of Pn neurons. ACTX SIVA, a toxin recently isolated from the venom of the marine cone snail Conus striatus, blocks potassium channels in frog neuromuscular preparations and Drosophila Shaker potassium channels expressed in Xenopus oocytes (Craig et al.,
Ionic Currents in Pacemaker Nucleus Table 1
Effects of Tetrodotoxin on Action Potential Waveform Parameters (n ⴝ 5, mean ⴞ SEM)
Parameter
Before TTX
Baseline membrane potential (mV) Amplitude (mV) Maximum rising slope2 (mV/ms) Maximum decay slope2 (mV/ms) Half-width (ms)
⫺65.5 ⫾ 3.2
1 2
275
After TTX1 ⫺75.6 ⫾ 1.8
30.0 ⫾ 3.0 179.5 ⫾ 33.2
21.1 ⫾ 3.2 97.1 ⫾ 29.6
⫺102.7 ⫾ 19.2
⫺61.1 ⫾ 18.9
0.331 ⫾ 0.080
0.520 ⫾ 0.298
Difference
p-value (paired t-test)
⫺10.1 ⫾ 2.0
0.0071
⫺9.0 ⫾ 0.8 ⫺82.4 ⫾ 17.4
0.0003 0.0091
41.6 ⫾ 5.1
0.0012
0.189 ⫾ 0.101
0.14
4 –5 min after TTX application. Measurement based on maximum change in membrane potential between samples at 10 kHz sampling frequency.
1998). Like 4AP, ACTX SIVA (250 nM) increased the firing rate of Pn neurons by 15–20%, depolarized the minimum membrane potential by approximately 15 mV, and decreased action potential amplitude within 5 min [Fig. 3(c)]. Neurons exposed to ACTX SIVA then stopped firing completely after 5–20 min. The effects of ACTX SIVA were similar at 1 M concentrations, which indicates that 250 nM is a saturating dose. The latencies of the ACTX SIVA effects were somewhat longer than those of 4AP [cf. Fig. 3, (a) and(c)], possibly because it may have taken longer for the larger peptide toxin to penetrate the tissue. Unlike 4AP, the effects of ACTX SIVA were not reversed by a 30-min washout. Other potassium channel blockers did not influence
Figure 2 Effects of OCTX MrVIA on pacemaker firing frequency. Percent change from baseline firing frequency of Pn neurons treated with 250 nM (n ⫽ 5) or 1 M (n ⫽ 13) OCTX MrVIA compared to control preparations treated with vehicle alone (n ⫽ 13). *250 nM OCTX MrVIA versus controls (one-factor ANOVA, PLSD, p ⬍ .05); † 1 M OCTX MrVIA versus controls (one-factor ANOVA, PLSD, p ⬍ .05).
the pacemaker rhythm. TEA (5–10 mM) did not affect pacemaker firing rate relative to controls [Fig. 4(a)]. A slight (2–3%) decrease in pacemaker firing frequency was observed in both TEA- and control-treated preparations, but this effect was possibly due to a decrease in sodium ion concentrations, which were lowered in both the TEA and control salines to maintain osmolarity. We investigated the possible role of hyperpolarization-activated cation currents (Ih) in regulating Pn firing rate by treating the Pn with external cesium, which blocks Ih at doses of 0.1–3.0 mM (DiFrancesco, 1982; Pape, 1996). Because external cesium also blocks inward rectifier, delayed rectifier, and calciumactivated K⫹ currents (Hille, 1992), this experiment also addressed the role of cesium-sensitive K⫹ currents in regulating the rate at which Pn neurons fire. Cesium did not significantly affect the firing frequency of Pn neurons at concentrations up to 5 mM [Fig. 4(b)], suggesting that neither Ih nor cesiumsensitive potassium currents play a role in Pn oscillations. Pacemaker firing frequency was sensitive to 4AP and ACTX SIVA, but resistant to cesium or TEA. Many members of the Kv1 family of mammalian potassium channels are also sensitive to 4AP, but resistant to TEA (Chandy and Gutman, 1995). Furthermore, ACTX SIVA blocks Drosophila Shaker channels (Craig et al., 1998), which are homologous to the vertebrate Kv1 channels. We therefore used two specific blockers of several Kv1 channel subtypes to test the hypothesis that the effects of 4AP on Pn firing rate were mediated by blocking channels that are pharmacologically similar to Kv1 channels. Two specific blockers of Kv1 channels had no effect on Pn firing rate. ␣-Dendrotoxin (␣DTX), which blocks Kv1.1, 1.2, 1.3, and 1.6 channels with EC50s ranging from 2.8 to 250 nM, but is relatively
276
Smith and Zakon
Figure 3 Effects of two potassium channel blockers on pacemaker firing frequency. Both 4AP (a, 100 M, n ⫽ 13) and ACTX SIVA [c, n ⫽ 18 (250 nM); and 4 (1 M)] rapidly increased pacemaker firing rate relative to control preparations [n ⫽ 15 (4AP) and 9 (ACTX SIVA)], and then stopped pacemaker oscillations completely. *4AP or ACTX SIVA (both doses) differed significantly from controls (one-factor ANOVA, PLSD, p ⬍ .05). (b) Intracellular recording of Pn cell exposed to 100 M 4AP. Within two min, there was a marked increase in firing frequency, decrease in action potential amplitude, and depolarization of baseline membrane potential. By 3 min, only high-frequency, low-amplitude membrane potential oscillations were still present. Most cells exposed to 4AP stopped firing completely 3–5 min after 4AP exposure, and most cells exposed to ACTX SIVA stopped firing completely 5–10 min after drug application. Scale bars ⫽ 10 mV, 2 ms.
ineffective at blocking Kv1.4, Kv1.5, and channels outside the Kv1 family (Chandy and Gutman, 1995; Harvey, 1997) had no significant effect on the firing rate of Pn neurons at concentrations up to 2 M [Fig. 4(c)]. Similarly, agitoxin-2 (AGI-2), which blocks Kv1.1, 1.3, and 1.6 channels with EC50s of 4 –37 pM, but does not block Kv2.1 channels (Garcia et al., 1994), had no effect on Pn firing rate at concentrations of 200 nM [Fig. 4(d)]. These data indicate that the effects of 4AP and ACTX SIVA on Pn firing were not mediated by blocking channels with pharmacological sensitivities similar to mammalian Kv1.1, 1.2, 1.3, or 1.6 channels.
Calcium Channel Blockers Two broad-spectrum calcium channel blockers reversibly decreased Pn firing frequency. Nickel had no significant effect on Pn firing rate at concentrations of 100 M; but 500 M nickel significantly decreased Pn firing rate by 7.1 ⫾ 2.2% after 20 min, and 1 mM nickel significantly decreased Pn firing rate by 11.3 ⫾ 3.6% within 3 min, and by 14.7 ⫾ 1.9% after 20 min [Fig. 5(a,b)]. Pn firing frequency was slightly more sensitive to cadmium than to nickel. A 10-min exposure to cadmium reduced Pn firing rate by 7.1 ⫾ 0.7% at 100 M and by 18.3 ⫾ 2.3% at 500 M
Ionic Currents in Pacemaker Nucleus Table 2
Effects of 4-aminopyridine (4AP) on Action Potential Waveform Parameters (n ⴝ 5, mean ⴞ SEM)
Parameter Baseline membrane potential (mV) Amplitude (mV) Maximum rising slope2 (mV/ms) Maximum decay slope2 (mV/ms) Half-width (ms) 1 2
277
Before 4AP ⫺61.9 ⫾ 1.3 27.9 ⫾ 2.1
After 4AP1 ⫺53.3 ⫾ 2.3 12.6 ⫾ 4.3
141.3 ⫾ 17.1
39.1 ⫾ 16.0
⫺75.4 ⫾ 7.4 0.400 ⫾ 0.019
⫺29.5 ⫾ 11.5 0.682 ⫾ 0.102
Difference
p-value (paired t-test)
8.7 ⫾ 1.5 ⫺15.3 ⫾ 2.9
0.0048 0.0059
⫺102.2 ⫾ 22.9
0.0112
45.9 ⫾ 10.5 0.282 ⫾ 0.088
0.0118 0.032
1.5–2 min after 4AP application. Measurement based on maximum change in membrane potential between samples at 10 kHz sampling frequency.
[Fig. 5(c)]. The effects of nickel and cadmium were reversed by wash-out (⬃75% recovery in 10 min, not shown). The effects of nickel and cadmium on Pn firing rate suggest that a calcium current may regulate the firing frequency of Pn neurons. In order to more specifically characterize the calcium current that underlies this
effect, we used specific blockers of four calcium channel subtypes. These specific calcium channel blockers did not significantly affect Pn firing frequency (Fig. 6). Nifedipine, which blocks L-type calcium channels in micromolar concentrations, did not significantly affect Pn firing rate compared to vehicle controls.
Figure 4 Effects of four other potassium channel blockers on pacemaker firing frequency. Neither TEA (a, 5–10 mM, n ⫽ 11) nor 5 mM cesium (b, n ⫽ 13) significantly affected pacemaker firing rates relative to control treatments [n ⫽ 9 (TEA) and 16 (cesium)]. Similarly, two specific blockers of some members of the Kv1 potassium channel gene family, ␣DTX (c, 2 M, n ⫽ 17) and agitoxin-2 (d, 200 nM, n ⫽ 12), failed to significantly affect pacemaker firing rates compared to vehicle controls (n ⫽ 12 and 10, respectively).
278
Smith and Zakon
Figure 5 Effects of divalent calcium channel blockers on pacemaker firing frequency. Percent change in firing frequency of neurons in pacemaker preparations exposed to nickel [a, n ⫽ 14 (100 M), 11 (500 M), and 7 (1 mM)] or cadmium [c, n ⫽ 17 (100 M) and 5 (500 M) ] compared to vehicle controls [n ⫽ 20 (nickel) and 10 (cadmium) ]. *1 mM dose versus control (one-factor ANOVA, PLSD, p ⬍ .05); † 500 M dose versus control (one-factor ANOVA, PLSD, p ⬍ .05); ‡ 100 M dose versus control (one-factor ANOVA, PLSD, p ⬍ .05). (b) Traces from an intracellular recording of Pn neuron exposed to 1 mM NiCl2. Nickel caused a 23% decrease in firing frequency over 20 min. Initially, nickel hyperpolarized the baseline membrane potential of the cell, and increased the amplitude of the action potentials. By 20 min, however, the baseline membrane potential was depolarized relative to before nickel treatment. A similar biphasic effect of nickel on membrane potential was observed in two other neurons that were intracellularly recorded during nickel treatment. Scale bars ⫽ 10 mV, 2 ms.
CTX GVIA, which specifically blocks mammalian N-type calcium currents at concentrations of 1 M or less (Olivera et al., 1994; Randall and Tsien, 1995), had no significant effect on Pn firing rate at concentrations up to 5 M. CTX MVIIC, which blocks mammalian N-, P-, and Q-type calcium currents at concentrations of 0.3–5.0 M (Hillyard et al., 1992; Olivera et al., 1994; Randall and Tsien, 1995; Nooney and Lodge, 1996), did not affect Pn firing rate at concentrations of 2 M. -Agatoxin (AGA) IVA, which completely blocks mammalian P- and Q-type calcium currents at concentrations of 100 –200 nM (Mintz et al., 1992), failed to significantly influence Pn firing rate at concentrations of up to 5 M. These results suggest that the pacemaker rhythm is not reg-
ulated by L-, N-, P-, or Q-type calcium currents and implicate T- or R-type calcium currents, both of which are sensitive to nickel and cadmium, but resistant to the specific blockers used in this study.
DISCUSSION Our results indicate that at least three ionic currents regulate the firing frequency of neurons in the brown ghost Pn: (1) a sodium current that is sensitive to TTX and OCTX MrVIA; (2) a potassium current that is sensitive to 4AP and ACTX SIVA, but resistant to cesium, TEA, ␣DTX, and AGI-2; and (3) a calcium
Ionic Currents in Pacemaker Nucleus
279
Figure 6 Effects of specific blockers of L-, N-, P-, and Q-type calcium channels on pacemaker firing frequency. Percent change in firing frequency of pacemaker preparations treated with the L-type calcium channel blocker nifedipine [a, n ⫽ 16 (1 M) and 11 (10 M)]; the N-type calcium channel blocker -conotoxin GVIA [b, n ⫽ 15 (1 M) and 6 (5 M)]; the N-, P-, and Q-type calcium channel blocker -conotoxin MVIIC (c, n ⫽ 9); and the P- and Q-type calcium channel blocker, -agatoxin IVA [d, n ⫽ 6 (1 M) and 6 (5 M)]. None of these blockers significantly changed pacemaker frequency compared to vehicle controls [n ⫽ 13 (a), 25 (b), 6 (c), 16 (d); one-factor ANOVAs, p ⬎ .05].
current that is sensitive to nickel and cadmium, but resistant to specific blockers of L-, N-, P-, and Q-type calcium currents.
Sodium Currents TTX reduced spike frequency, hyperpolarized baseline membrane potential, reduced spike amplitude, and eventually stopped pacemaker firing altogether. These results indicate that one or more TTX-sensitive sodium currents are necessary for spike generation and also regulate the membrane potential and firing frequency of Pn neurons. The effects of TTX on the spike amplitude and
spike frequency of Pn neurons were temporally separable. TTX reduced spike frequency before affecting spike amplitude [Fig. 1(b,c)]. Furthermore, saturating doses of the sodium channel blocker OCTX MrVIA reduced pacemaker frequency without affecting spike generation. These results may be explained by two possible scenarios. One possibility is that Pn neurons express two TTX-sensitive sodium currents: a TTX- and OCTX MrVIA-sensitive current that regulates firing frequency; and a TTX-sensitive, but OCTX MrVIA-resistant, current that generates the depolarizing phase of action potentials. Many neuron types possess multiple sodium currents (Honmou et al., 1994; Crill, 1996;
280
Smith and Zakon
Baker and Bostock, 1997). In several rhythmically firing neurons and neural circuits, transient sodium currents generate action potentials, while persistent sodium currents provide tonic depolarization that influences the timing between spikes (Opdyke and Calabrese, 1994; Ju et al., 1996; Oka, 1996; Elson and Selverston, 1997; Onimaru et al., 1997; Takakusaki and Kitai, 1997). Although some persistent sodium currents are resistant to TTX (Oka, 1996), many others are TTX-sensitive (Crill, 1996; Ju et al., 1996; Baker and Bostock, 1997; Elson and Selverston, 1997). Indeed, in some neurons, persistent sodium currents are more sensitive to TTX than transient sodium currents (Hammarstrom and Gage, 1998), which is consistent with the fact that TTX hyperpolarized the Pn neurons and slowed their firing rates before reducing spike amplitude. The sensitivity of persistent sodium currents to OCTX MrVIA has not been investigated. It is possible that like many other rhythm-generating neurons, the neurons of the Pn express transient sodium currents that generate spikes, and persistent sodium currents that regulate spike frequency. An alternative possibility is that Pn neurons possess a single TTX-sensitive sodium current that is partially sensitive to OCTX MrVIA, and that spike frequency and amplitude are differentially affected by blocking this current. Partial blockade, which occurs initially on TTX treatment or on treatment with saturating doses of OCTX MrVIA, may hyperpolarize the cell and gradually slow Pn firing rate without reducing spike amplitude. More complete blockade, which occurs after longer exposure to TTX, may reduce spike amplitude and eventually stop spike production. Analysis of sodium currents of Pn neurons under voltage clamp is necessary to distinguish between these possibilities.
Table 3
Potassium Currents Consistent with the previous report of Dye (1991), we found that 4AP rapidly increased pacemaker firing frequency, and then stopped pacemaker oscillations entirely. A similar effect was observed with ACTX SIVA, which was previously reported to block potassium channels in frog neuromuscular preparations and Drosophila Shaker channels (Craig et al., 1998). Other potassium channel blockers, including TEA, cesium, ␣DTX, and AGI-2, had little or no effect on pacemaker rhythmicity. The effects of the potassium channel blockers used in this study on pacemaker neurons and on mammalian potassium channel types are summarized in Table 3. There may be differences between the potassium channels expressed in mammals and fish, and some caution should therefore be used in attempting to classify potassium currents in fish based on their pharmacological similarities to mammalian potassium currents. Nevertheless, potassium channel gene families tend to be highly conserved across taxa (reviewed in Salkoff et al., 1992; Chandy and Gutman, 1995). The sequences of 19 recently identified potassium channel gene fragments from brown ghost knifefish could each be classified as homologs of one of the mammalian voltage-gated potassium channel gene families (Kv1-4) (Rashid and Dunn, 1998), and clear homologs of mammalian potassium channel genes also occur in Drosophila (Shaker, Shab, Shal, and Shaw; reviewed in Salkoff et al., 1992; Chandy and Gutman, 1995). Furthermore, the Drosophila potassium channels often have similar pharmacological profiles and kinetics to their mammalian homologs (Wei et al., 1990; Chandy and Gutman, 1995), which indicates that the pharmacology of these channels may also be largely conserved across taxa. The results of the present study suggest that one or more potassium currents that are sensitive to low
Effects of Potassium Channel Blockers on Pacemaker Rhythm and Voltage-Gated Potassium Channels
Channel Blocker
Effect on Pn Firing Frequency
Kv1.1
4-Aminopyridine (⬍1 mM) Tetraethylammonium ␣-Dendrotoxin Agitoxin-2
1* ⫺ ⫺ ⫺
⫹ ⫹ ⫹ ⫹
Potassium Channel(s) Blocked Kv1.2, Kv1.4, Kv2 Kv3 Kv1.3, Kv1.6 Kv1.5 Family Family ⫹ ⫺ ⫹ ⫹
⫹ ⫺ ⫺ ⫺
⫺ ⫹ ⫺ ⫺
⫹ ⫹ ⫺ ND
Kv4 Family ⫹?† ⫺ ND ND
1, increased firing frequency. ⫹, channel blocked by toxin. ⫺, no effect on firing frequency or channel. ND, not determined. Channels with pharmacologically sensitivities consistent with effects on Pn firing frequency indicated in bold. *4-aminopyridine transiently increased frequency, then stopped firing (see Results, fig. 3). † IC50 typically 1–10 mM (see Discussion).
Ionic Currents in Pacemaker Nucleus
concentrations of 4AP and ACTX SIVA, but resistant to TEA, AGI-2, and ␣DTX, contribute to the pacemaker rhythm and regulate its frequency (Table 3). Known potassium channels that are consistent with this pharmacological profile include Kv1.4 and Kv1.5 channels. It is also possible that some members of the Kv4 family of potassium channels fit this profile. The extreme sensitivity of the pacemaker rhythm to 100 M 4AP, however, argues against this possibility, because Kv4 channels are typically blocked by 4AP with IC50s in the range of 1–10 mM (Chandy and Gutman, 1995; Seroˆdio et al., 1996; Song et al., 1998). Some of the kinetic properties of the currents generated by Kv1.4 and Kv1.5 channels are consistent with a role in pacemaking. Kv1.4 channels produce currents with A-type (transient) kinetics, and Kv1.5 channels also produce A-type currents when coexpressed with the Kv1.1 beta subunit (Chandy and Gutman, 1995; Heinemann et al., 1996). A-type potassium currents regulate firing rates in other rhythmic neural circuits (Byrne, 1982; Getting, 1989; Tierney and Harris-Warrick, 1992). Their rapid kinetics and the overlap of their activation and inactivation curves allow them to influence both the rate of spike repolarization and the interval between spikes in repetitively firing neurons (Connor and Stevens, 1971; Hille, 1992). These features may enable Kv1.4 and/or Kv1.5-like channels to generate the potassium currents that contribute to the oscillations of the brown ghost Pn neurons. It is possible, however, that the potassium currents in the Pn neurons have some unique kinetic features. The Pn in brown ghosts and other apteronotid electric fish generates rhythms with extremely high frequencies (e.g., up to 1800 Hz in Sternarchella schotti; Kramer et al., 1980). These firing rates far exceed those of other known neural oscillators, including those in which A-type potassium currents are known to play important roles in regulating interspike intervals. Although the kinetics of currents generated by mammalian Kv1.4 and Kv1.5 channels are rapid, particularly when they are coexpressed with certain beta subunits (Snyders et al., 1993; Fiset et al., 1997; McIntosh et al., 1997), it is unclear how these currents would function in response to membrane potential oscillations of 1 kHz or more. One possibility is that Pn neurons express potassium channels that are pharmacologically similar to Kv1.4 and Kv1.5 channels, but which produce currents with kinetics that are more rapid than those generated by mammalian potassium channels. Such kinetic differences could result from differences in the potassium channel alpha subunits or
281
from interactions with beta subunits, which dramatically affect the kinetics of mammalian Kv1.4- and Kv1.5-generated currents (Morales et al., 1995; Sasaki et al., 1995; Heinemann et al., 1996; McIntosh et al., 1997). Voltage-clamp studies of pacemaker neurons are needed to accurately describe the kinetics of the potassium currents that contribute to the EOD rhythm.
Calcium Currents Dye (1991) investigated the role of calcium currents in the Pn rhythm by treating the Pn with ACSF lacking calcium and containing the calcium chelator EGTA. This zero calcium treatment reversibly extinguished pacemaker activity. In the present study, the broad-spectrum calcium channel blockers nickel and cadmium significantly reduced pacemaker firing frequency, but did not suppress its oscillations. This difference may be explained by two alternative hypotheses: (1) Pn neurons may express a nickel- and cadmium-insensitive calcium current that is required for spontaneous activity; or (2) the zero calcium/ EGTA treatment used by Dye (1991) may have extinguished pacemaker activity by interfering with intracellular calcium metabolism. Examining the calcium currents of Pn neurons under voltage clamp and comparing the effects of nickel and cadmium versus zero calcium on these currents may help to distinguish between these possibilities. Six calcium current types have been characterized in mammals, and all are sensitive to nickel and cadmium (Table 4). Although nickel and cadmium both reduced pacemaker firing rates by 15–20%, specific blockers of four calcium current subtypes (L, N, P, and Q) had no effect on the pacemaker rhythm. This suggests that the rhythm of the pacemaker may be regulated by one of the two remaining calcium current subtypes, i.e., T- or R-type currents. Although T- and R-type calcium currents were initially believed to be relatively sensitive to nickel and resistant to cadmium (Hille, 1992), more recent studies have reported Tand R- type currents that are equally sensitive to nickel and cadmium, or even slightly more sensitive to cadmium (Soong et al., 1993; Kwiecien et al., 1998). T-type calcium currents are found in other endogenously oscillating neurons (Soltesz et al., 1991; Erickson et al., 1993; Kwiecien et al., 1998). Their low threshold of activation and the large overlap of their activation and inactivation curves enables them to provide a sustained calcium influx that contributes to subthreshold oscillations in many types of pace-
282 Table 4
Smith and Zakon Effects of Calcium Channel Blockers on Pacemaker Rhythm and Calcium Current Types
Channel Blocker
Effect on Pn Firing Frequency
L
N
Nickel Cadmium Nifedipine -Conotoxin GVIA -Agatoxin IVA -Conotoxin MVIIC
2 2 ⫺ ⫺ ⫺ ⫺
⫹ ⫹ ⫹ ⫺ ⫺ ⫺
⫹ ⫹ ⫺ ⫹ ⫺ ⫹
Calcium Current Type(s) Blocked P Q R ⫹ ⫹ ⫺ ⫺ ⫹ ⫹
⫹ ⫹ ⫺ ⫺ ⫹ ⫹
⫹ ⫹ ⫺ ⫺ ⫺ ⫺
T ⫹ ⫹ ⫺ ⫺ ⫺ ⫺
2, decreased firing frequency. ⫹, current type blocked by toxin. ⫺, no effect on firing frequency or calcium current type.
making neurons (Randall and Tsien, 1997). Unlike T-type currents, R-type calcium currents have a high threshold of activation, and they are therefore unlikely to provide sustained calcium influx during interspike intervals, when membrane potential is hyperpolarized. It is possible, however, that R-type calcium currents could influence the pacemaker rhythm by altering the time course of action potentials. Indeed, R-type currents function in pacemaking in rat tumor somatotroph cells (Kwiecien et al., 1998), and their kinetics may be rapid enough to allow them to participate in pacemaking on a cycle-by-cycle basis in the very rapid (600 –1000 Hz) brown ghost Pn rhythm. Another possibility is that T- and/or R-type calcium currents do not directly influence the Pn rhythm as charge carriers, but rather provide a source of calcium influx that indirectly influences the Pn rhythm (e.g., by affecting other ionic currents through second-messenger cascades).
cies) of 800 –1000 Hz, while females typically have EODs with frequencies of 600 – 800 Hz (Meyer et al., 1987; Dunlap et al., 1998). Female EODs are thus 20 –25% lower in frequency than male EODs in this species. The effects of various channel-blocking drugs on Pn firing frequency are summarized in Table 5. The size of these drug effects is only a crude approximation of the abilities of their respective target currents to regulate the Pn rhythm. The fact that blocking these ionic currents produced changes in Pn firing rates of similar or greater magnitude than the sexual dimorphism in EOD frequency, however, indicates that these currents have the capacity to mediate changes in the EOD rhythm similar to those produced by gonadal steroids. If blocking these currents had produced much smaller effects on Pn firing rate, it would be unlikely that steroidal regulation of these currents would be able to produce the sexually dimorphic EOD frequencies observed in this species.
Ionic Currents and Sexual Dimorphisms in EOD Frequency
Proposed Ionic Mechanisms Underlying Pn Rhythm
The effects of ion channel blockers on the Pn rhythm suggest that the currents blocked by these drugs exert sufficient control over Pn firing rate to mediate sexual dimorphisms in EOD frequency. Male brown ghosts have EOD frequencies (and thus Pn firing frequen-
We pharmacologically characterized three ionic currents that regulate the frequency of the pacemaker rhythm. Although it is not possible to quantitatively model the interactions of these currents without using voltage-clamp experiments to determine their kinet-
Table 5 Comparison of Effects of Channel Blockers on Pacemaker Firing Frequency with Sexual Dimorphism in EOD Frequency Channel Blockers
Channel Type Blocked
Tetrodotoxin, O-conotoxin MrVIA 4-Aminopyridine, A-conotoxin SIVA Nickel, cadmium % difference between average brown ghost male (⬃900 Hz) and female (⬃700 Hz) EOD frequencies:
Na⫹ K⫹ Ca2⫹
Maximum Effect on Pn Firing Frequency 2 10–40% 1 ⬃20% 2 20–30% 25%
Ionic Currents in Pacemaker Nucleus
Figure 7 Model of how pharmacologically characterized ionic currents in the Pn may contribute to the pacemaker rhythm. Depolarizing phase of action potential is produced by voltage-gated, inactivating sodium current (INa(i)), that is sensitive to TTX, but resistant to OCTX MrVIA. The repolarizing phase of the action potential results from inactivation of this current, and activation of a potassium current (IK(A)) that is sensitive to ACTX SIVA and 4AP. The ramped pacemaker potential between spikes (portion of trace between open arrowheads) results from the interaction of IK(A) with low-threshold inward currents, such as a persistent, TTX- and OCTX MrVIA-sensitive sodium current (INa(p)) and/or a low-threshold calcium current (ICa(T)) that is blocked by nickel or cadmium. The slope of the pacemaker potential influences the interspike interval, and thus firing rate. The currents that contribute to the pacemaker potential (INa(p), IK(A), and/or ICa(T)) are thus potential control points for regulating the frequency of the pacemaker rhythm.
ics, it is interesting to speculate, based on the functions of pharmacologically similar currents in other neuronal oscillators, on how these currents may interact to generate the pacemaker rhythm (Fig. 7). The pharmacological profiles of the ionic currents in the brown ghost Pn are similar to those of currents that generate rhythms in other spontaneously oscillating neural circuits (Fig. 7). In most of these oscillators, a sustained, low-threshold, depolarizing current maintains the spontaneous activity of the oscillator and regulates membrane potential between spikes, thus influencing firing rates. This role may be fulfilled in the Pn by a persistent, TTX- and OCTX MrVIAsensitive sodium current (INa(p)) and/or by a lowthreshold, T-type calcium current (ICa(T)). When these currents were blocked in the in vitro Pn, the baseline membrane potential hyperpolarized, and the pacemaker rhythm was dramatically slowed. The lack of
283
an effect of cesium on the Pn rhythm suggests that, unlike in many other neural oscillators, a hyperpolarization-activated cation current (Ih) does not regulate the interspike interval in the Pn. This is similar to what occurs in cerebellar Purkinje neurons, where persistent and/or resurgent sodium currents, rather than Ih, drive spontaneous activity (Raman and Bean, 1999). In addition to its effect on spike frequency, TTX also disrupted the ability of the Pn neurons to generate action potentials, which suggests that an inactivating sodium current (INa(i)) contributes to spike generation. Transient (A-type) potassium currents frequently regulate the interval between successive spikes in rhythmically firing neurons. The effects of potassium channel blockers on the Pn rhythm are consistent with a transient, Kv1.4- or Kv1.5-like, potassium current (IK(A)), playing a similar role in this system. When this current was blocked with 4AP or ACTX SIVA, baseline membrane potential of the Pn neurons depolarized, and firing frequency increased. This result suggests that the 4AP- and ACTX SIVAsensitive potassium current may interact with lowthreshold inward currents during the interspike interval to regulate rates of membrane depolarization, and thus spike frequency. 4AP also affected action potential waveform (Table 2). Thus, a 4AP-sensitive potassium current also influences spike formation and waveform, either by directly participating in spike repolarization or indirectly via its effects on baseline membrane potential. Furthermore, both 4AP and ACTX SIVA eventually stopped pacemaker firing completely, which indicates that the potassium current(s) blocked by these toxins are also necessary for the Pn to produce its rhythm. The localization of the ionic currents that were blocked by the toxins used in this study is uncertain. One likely possibility is that these ionic currents are present in pacemaker neurons and that they allow the pacemaker neurons to spontaneously generate the EOD rhythm via a model similar to that proposed in Figure 7. Other scenarios are, however, possible. Our recordings were made from neurons (some identified as pacemaker neurons, but others likely to be relay neurons) in intact in vitro Pn preparations. In these preparations, pacemaker and relay neurons were electrotonically coupled and fired synchronously. Our measurements of changes in the firing rates of individual Pn neurons therefore reflect changes in the firing rates of all of the neurons in the Pn. The presence of electrotonic coupling between Pn neurons, however, means that effects of toxins on the firing patterns of one type of Pn neuron could affect the firing patterns of other neurons in the Pn that do
284
Smith and Zakon
not necessarily have channels that are blocked by those toxins. In addition, although the somata of the descending prepacemaker and sublemniscal prepacemaker projections were not present in the in vitro Pn preparation, the Pn does contain small interneurons, which form chemical synapses on the relay and pacemaker neurons (Turner and Moroz, 1995; Smith et al., 1999). The drugs used in this study were bath-applied, and thus all cells in the preparation were exposed to them. It is therefore possible that at least some of the toxins’ effects on the firing rates of the recorded cells may have indirectly resulted from their effects on ionic currents in other cells. In addition, many of the channel blockers in this study affected the membrane potential of the Pn neurons, and the conductance of some types of gap junctions can be sensitive to membrane potential (reviewed in Bennett, 1997). It is therefore possible that some of the toxins may have indirectly affected the firing patterns of the recorded neurons by altering the electrotonic coupling of different cells in the Pn network. Future studies that locally apply channel-blocking drugs to isolated relay and/or pacemaker cells may help to address this question. Ultimately, a complete model of how the Pn generates the EOD rhythm must incorporate an understanding of both how ionic conductances, such as those identified in this study, contribute to the spontaneous, rhythmic excitability of pacemaker (and possibly relay) neurons, and how the different cell types in the Pn interact to generate a synchronous, rhythmic command signal. An important next step will be to characterize the currents of the Pn neurons under voltage clamp. It will then be possible to quantitatively model the currents of Pn neurons to determine how they contribute to the high-frequency rhythm that underlies the brown ghost EOD. Further studies investigating the actions of gonadal steroid hormones on the density and/or kinetics of these currents may then allow us to understand the biophysical mechanisms that underlie sexual dimorphism in the EOD. The authors thank Dr. Baldomero Olivera for the kind donation of the O-conotoxin MrVIA and A-conotoxin SIVA used in this study.
REFERENCES Baker MD, Bostock H. 1997. Low-threshold, persistent sodium current in rat large dorsal root ganglion neurons in culture. J Neurophys 77:1503–1513.
Bennett MVL. 1997. Gap junctions as electrical synapses. J Neurocytol 26:349 –366. Bullock TH. 1970. The reliability of neurons. J Gen Physiol 55:565–584. Byrne JH. 1982. Cellular and biophysical mechanisms contributing to regulation of reflex excitability of inking behavior in Aplysia. Fed Proc 41:2147–2152. Chandy KG, Gutman GA. 1995. Voltage-gated potassium channel genes. In: North RA, editor. Handbook of receptors and channels: ligand- and voltage-gated ion channels. Boca Raton, FL: CRC Press, p 1–71. Connor JA, Stevens CF. 1971. Voltage clamp studies of a transient outward membrane current in gastropod neural somata. J Physiol (Lond) 213:21–30. Craig AG, Zafaralla G, Cruz LJ, Santos AD, Hillyard DR, Dykert J, Rivier JE, Gray WR, Imperial J, DelaCruz RG, Sporning A, Terlau H, West PJ, Yoshikami D, Olivera BM. 1998. An O-glycosylated neuroexcitatory peptide. Biochemistry 37:16019 –16025. Crill WE. 1996. Persistent sodium current in mammalian central neurons. Annu Rev Physiol 58:349 –362. DiFrancesco D. 1982. Block and activation of the pacemaker channel in calf Purkinje fibres: effects of potassium, caesium, and rubidium. J Physiol (Lond) 329:482– 507. Dunlap KD, Thomas P, Zakon HH. 1998. Diversity of sexual dimorphism in electrocommunication signals and its androgen regulation in a genus of electric fish, Apteronotus. J Comp Physiol A 183:77– 86. Dye J. 1991. Ionic and synaptic mechanisms underlying a brainstem oscillator: an in vitro study of the pacemaker nucleus of Apteronotus. J Comp Physiol A 168:521–532. Dye J, Heiligenberg W. 1987. Intracellular recording in the medullary pacemaker nucleus of the weakly electric fish, Apteronotus, during modulatory behaviors. J Comp Physiol A 161:187–200. Elekes K, Szabo T. 1981. Comparative synaptology of the pacemaker nucleus in the brain of weakly electric fish (Gymnotidae). Adv Physiol Sci 31:107–127. Elekes K, Szabo T. 1985. Synaptology of the medullary command (pacemaker) nucleus of the weakly electric fish (Apteronotus leptorhynchus) with particular reference to comparative aspects. Exp Brain Res 60:509 –520. Ellis DB, Szabo T. 1980. Identification of different cell types in the command (pacemaker) nucleus of several gymnotiform species by retrograde transport of horseradish peroxidase. Neuroscience 5:1917–1929. Elson RC, Selverston AI. 1997. Evidence for a persistent Na⫹ conductance in neurons of the gastric mill rhythm generator of spiny lobsters. J Exp Biol 200:1795–1807. Enger PS, Szabo T. 1968. Effects of temperature on discharge rate of the electric organ of some gymnotids. Comp Biochem 27:625– 627. Erickson KR, Ronnekleiv OK, Kelly MJ. 1993. Role of T-type calcium current in supporting a depolarizing potential, damped oscillations, and phasic firing in vaso-
Ionic Currents in Pacemaker Nucleus pressinergic guinea pig supraoptic neurons. Neuroendocrinology 57:789 – 800. Fiset C, Clark RB, Larsen TS, Giles WR. 1997. A rapidly activating sustained K⫹ current modulates repolarization and excitation-contraction coupling in adult mouse ventricle. J Physiol (Lond) 504:557–563. Garcia ML, Garcia-Calvo M, Hidalgo P, Lee A, MacKinnon R. 1994. Purification and characterization of three inhibitors of voltage-dependent K⫹ channels from Leiurus quinquestriatus var. hebraeus venom. Biochemistry 33: 6834 – 6839. Getting PA. 1989. A network oscillator underlying swimming in Tritonia. In: Jacklet J, editor. Neuronal and cellular oscillators. New York: Marcel Dekker, p 215– 236. Hammarstrom AK, Gage PW. 1998. Inhibition of oxidative metabolism increases persistent sodium current in rat CA1 hippocampal neurons. J Physiol (London) 510:735– 741. Harvey AL. 1997. Recent studies on dendrotoxins and potassium ion channels. Gen Pharmacol 28:7–12. Heinemann SH, Rettig J, Graack HR, Pongs O. 1996. Functional characterization of Kv channel beta subunits from rat brain. J Physiol (Lond) 493:625– 633. Hille B. 1992. Ionic channels of excitable membranes. Sunderland, MA: Sinauer. Hillyard DR, Monje VD, Mintz IM, Bean BP, Nadasdi L, Ramachandran J, Miljanich G, Azimi-Zoonooz A, McIntosh JM, Cruz LJ, Imperial JS, Olivera BM. 1992. A new conus peptide ligand for mammalian presynaptic Ca2⫹ channels. Neuron 9:69 –77. Honmou O, Utzschneider DA, Rizzo MA, Bowe CM, Waxman SG, Kocsis JD. 1994. Delayed depolarization and slow sodium currents in cutaneous afferents. J Neurophysiol 71:1627–1637. Hopkins CD. 1974. Electric communication in fish. Am Sci 62:426 – 437. Hopkins CD. 1988. Neuroethology of electric communication. Annu Rev Neurosci 11:497–535. Ju Y, Gage PW, Saint DA. 1996. Tetrodotoxin-sensitive inactivation-resistant sodium channels in pacemaker cells influence heart rate. Pfluegers Archiv Eur J Physiol 431: 868 – 875. Kramer B, Kirschbaum F, Markl H. 1980. Species specificity of electric organ discharges in a sympatric group of gymnotoid fish from Manaus (Amazonas). In: Szabo´ T, Cze´ h G, editors. Sensory physiology of aquatic lower vertebrates. Budapest: Akade´ miai Kiado´ , p 195–219. Kwiecien R, Robert C, Cannon R, Vigues S, Arnoux A, Kordon C, Hammond C. 1998. Endogenous pacemaker activity of rat tumour somatotrophs. J Physiol (Lond) 508:883–905. McIntosh JM, Hasson A, Spira ME, Gray WR, Li W, Marsh M, Hillyard DR, Olivera BM. 1995. A new family of conotoxins that blocks voltage-gated sodium channels. J Biol Chem 270:16796 –16802. McIntosh P, Southan AP, Akhtar S, Sidera C, Ushkaryov Y,
285
Dolly JO, Robertson B. 1997. Modification of rat brain Kv1.4 channel gating by association with accessory Kv1.1 and 1.2 subunits. Pflugers Archiv Eur J Physiol 435:43–54. Meyer JH. 1984. Steroid influences upon discharge frequencies of intact and isolated pacemakers of weakly electric fish. J Comp Physiol A 154:659 – 668. Meyer JH, Leong M, Keller CH. 1987. Hormone-induced and maturational changes in electric organ discharges and electroreceptor tuning in the weakly electric fish Apteronotus. J Comp Physiol A 160:385–394. Mintz IM, Adams ME, Bean BP. 1992. P-type calcium channels in rat central and peripheral neurons. Neuron 9:85–95. Moortgat KT, Keller CH, Bullock TH, Sejnowski TJ. 1998. Submicrosecond pacemaker precision is behaviorally modulated: the gymnotiform electromotor pathway. Proc Natl Acad Sci USA 95:4684 – 4689. Morales MJ, Castellino RC, Crews AL, Rasmusson RL, Strauss HC. 1995. A novel beta subunit increases rate of inactivation of specific voltage-gated potassium channel alpha subunits. J Biol Chem 270:6272– 6277. Nooney JM, Lodge D. 1996. The use of invertebrate peptide toxins to establish Ca2⫹ channel identity of CA3-CA1 neurotransmission in rat hippocampal slices. Eur J Pharmacol 306:41–50. Oka Y. 1996. Characterization of TTX-resistant persistent Na⫹ current underlying pacemaker potentials of fish gonadotropin-releasing hormone (GnRH) neurons. J Neurophysiol 75:2397–2404. Olivera BM, Miljanich GP, Ramachandran J, Adams ME. 1994. Calcium channel diversity and neurotransmitter release: The -conotoxins and -agatoxins. Annu Rev Biochem 63:823– 867. Onimaru H, Arata A, Homma I. 1997. Neuronal mechanisms of respiratory rhythm generation: an approach using in vitro preparation. Jpn J Physiol 47:385– 403. Opdyke CA, Calabrese RL. 1994. A persistent sodium current contributes to oscillatory activity in heart interneurons of the medicinal leech. J Comp Physiol A 175:781– 789. Pape H-C. 1996. Queer current and pacemaker: The hyperpolarization-activated cation current of neurons. Annu Rev Neurosci 58:299 –327. Raman IM, Bean BP. 1999. Ionic currents underlying spontaneous action potentials in isolated cerebellar Purkinje neurons. J Neurosci 19:1663–1674. Randall A, Tsien RW. 1995. Pharmacological dissection of multiple types of Ca2⫹ channel currents in rat cerebellar granule neurons. J Neurosci 15:2995–3012. Randall AD, Tsien RW. 1997. Contrasting biophysical and pharmacological properties of T-type and R-type calcium channels. Neuropharmacol 36:879 – 893. Rashid AJ, Dunn RJ. 1998. Sequence diversity of voltagegated potassium channels in an electric fish. Brain Res Mol Brain Res 54:101–107. Salkoff L, Baker K, Butler A, Covarrubias M, Pak MD, Wei
286
Smith and Zakon
A. 1992. An essential ’set’ of K⫹ channels conserved in flies, mice, and humans. Trends Neurosci 15:161–166. Sasaki Y, Ishii K, Nunoki K, Yamagishi T, Taira N. 1995. The voltage-dependent K⫹ channel (Kv1.5) cloned from rabbit heart and facilitation of inactivation of the delayed rectifier current by the rat  subunit. FEBS Lett 372:20 – 24. Schaefer JE, Zakon HH. 1996. Opposing actions of androgen and estrogen on in vitro firing frequency of neuronal oscillators in the electromotor system. J Neurosci 16: 2860 –2868. Seroˆ dio P, Vega-Saenz de Miera E, Rudy B. 1996. Cloning of a novel component of A-type K⫹ channels operating at subthreshold potentials with unique expression in heart and brain. J Neurophysiol 75:2174 –2179. Smith GT, Lu Y, Zakon HH. 1999. Parvocells: a recentlyidentified neuron type in the pacemaker nucleus of the weakly electric fish Apteronotus leptorhynchus. Soc Neurosci Abstr 25:1363. Snyders DJ, Tamkon MM, Bennett PB. 1993. A rapidly activating and slowly inactivating potassium channel cloned from human heart. Functional analysis after stable mammalian cell culture expression. J Gen Physiol 101: 513–543. Soltesz I, Lightowler S, Leresche N, Jassik-Gerschenfeld D, Pollard CE, Crunelli V. 1991. Two inward currents and the transformation of low-frequency oscillations of rat and cat thalamocortical cells. J Physiol (London) 441: 175–197. Song W-J, Tkatch T, Baranauskas G, Ichinohe N, Kitai ST, Surmeier DJ. 1998. Somatodendritic depolarization activated potassium currents in rat neostriatal cholinergic
interneurons are predominantly of the A type and attributable to coexpression of Kv4.2 and Kv4.1. J Neurosci 18:3124 –3137. Soong TW, Stea A, Hudson CD, Dubel SJ, Vincent SR, Snutch TP. 1993. Structure and functional expression of a member of the low voltage-activated calcium channel family. Science 260:1133–1136. Takakusaki K, Kitai ST. 1997. Ionic mechanisms involved in the spontaneous firing of tegmental pendunculopontine nucleus neurons of the rat. Neuroscience 78:771-794. Terlau H, Stocker M, Shon K-J, McIntosh JM, Olivera BM. 1996. O-conotoxin MrVIA inhibits mammalian sodium channels, but not through site I. J Neurophysiol 76:1423– 1429. Tierney AJ, Harris-Warrick RM. 1992. Physiological role of the transient potassium current in the pyloric circuit of the lobster stomatogastric ganglion. J Neurophysiol 67:599 – 609. Tokunaga A, Akert K, Sandri C, Bennett MVL. 1980. Cell types and synaptic organization of the medullary electromotor nucleus in a constant frequency weakly electric fish Sternarchus albifrons. J Comp Neurol 192:407– 426. Turner RW, Moroz LL. 1995. Localization of nicotinamide adenine dinucleotide phosphate-diaphorase activity in electrosensory and electromotor systems of a gymnotiform teleost, Apteronotus leptorhynchus. J Comp Neurol 356:261–274. Wei A, Covarrubias M, Butler A, Baker K, Pak M, Salkoff L. 1990. K⫹ current diversity is produced by an extended gene family conserved in Drosophila and mouse. Science 248:599 – 603.
