The Journal of Neurophysiology Vol. 77 No. 4 April 1997, pp. 1779-1794
Copyright ©1997 by the American Physiological Society

Functional Role of Ca²⁺ Currents in Graded and Spike-Mediated Synaptic Transmission Between Leech Heart Interneurons

Jin Lu, John F. Dalton IV, Darrell R. Stokes, and Ronald L. Calabrese

Department of Biology, Emory University, Atlanta, Georgia 30322

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Lu, Jin, John F. Dalton, IV, Darrell R. Stokes, and Ronald L. Calabrese. Functional role of Ca²⁺ currents in graded and spike-mediated synaptic trnasmission between leech heart interneurons. J. Neurophysiol. 77: 1779-1794, 1997. We used intracellular recording and single electrode voltage-clamp techniques to explore Ca²⁺ currents and their relation to graded and spike-mediated synaptic transmissions in leech heart interneurons. Low-threshold Ca²⁺ currents (activation begins below 50 mV) consist of a rapidly inactivating component (I_CaF) and a slowly inactivating component (I_CaS). The apparent inactivation kinetics of I_CaF appears to be influenced by Ca²⁺; both the substitution of Ca²⁺ (5 mM) with Ba²⁺ (5 mM) in the saline and the intracellular injection of the rapid Ca²⁺ chelator, bis-(o-aminophenoxy)-N,N,N,N-tetraacetic acid (BAPTA), from the recording microelectrode, significantly increase its apparent inactivation time constant. The use of saline with a high concentration of Ba²⁺ (37.5 mM) permitted exploration of divalent ion currents over a broader activation range, by acting as an effective charge carrier and significantly blocking outward currents. Ramp and pulse voltage-clamp protocols both reveal a rapidly activating and inactivating Ba²⁺ current (I_BaF) and a less rapidly activating and slowly inactivating Ba²⁺ current with a broad activation range (I_BaS). Low concentrations of Cd²⁺ (100-150 µM) selectively block I_BaS, without significantly diminishing I_BaF. The current that remains in Cd²⁺ lacks the characteristic delayed activation peak of I_BaS and inactivates with two distinct time constants. I_BaF appears to correspond to a combination of I_CaF and I_CaS, i.e., to low-threshold Ca²⁺ currents, that can be described as T-like. I_BaS appears to correspond to a Ca²⁺ current with a broad activation range, which can be described as L-like. Cd²⁺ (100 µM) selectively blocks spike-mediated synaptic transmission between heart interneurons without significantly interfering with low-threshold Ca²⁺ currents and plateau formation in or graded synaptic transmission between heart interneurons. Blockade of spike-mediated synaptic transmission between reciprocally inhibitory heart interneurons with Cd²⁺ (150 µM), in otherwise normal saline, prevents the expression of normal oscillations (during which activity in the two neurons consists of alternating bursts), so that the neurons fire tonically. We conclude that graded and spike-mediated synaptic transmission may be relatively independent processes in heart interneurons that are controlled predominantly by different Ca²⁺ currents. Moreover, spike-mediated synaptic inhibition appears to be required for normal oscillation in these neurons.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Networks of central neurons contribute to the patterning of almost all rhythmic movements in animals (Delcomyn 1980). Oscillation within these motor-pattern-generating networks results from the combination of intrinsic electrical properties of the component neurons and their synaptic interactions (Dean and Cruse 1995; Harris-Warrick 1993; Jacklet 1989; Rossignol and Dubuc 1994). In several pattern-generating networks, especially those found in invertebrate ganglia, graded synaptic transmission (almost invariably inhibitory)release of transmitter by slow changes in membrane potentials and independent of action potentials per se (Burrows 1985; Siegler 1985)is thought to play an important, if not dominant role in network function (Büschges 1995; DiCaprio 1989; Heitler and Pearson 1980; Mendelson 1971; Paul and Mulloney 1985a,b; Pearson and Fourtner 1975; Wolf and Büschges 1995) even when the usual spike-mediated transmission is present (e.g., see Granzow et al. 1985; Graubard et al. 1983; Johnson and Harris-Warrick 1990; Johnson et al. 1991; Mangan et al. 1994; Raper 1979). It is not clear whether spike-mediated and graded transmissions represent separate processes or form a continuum (Hartline and Graubard 1992). Because graded and spike-mediated transmissions are active in different ranges of presynaptic membrane potential, one possibility is that they are controlled by Ca²⁺ entry through different populations of Ca²⁺ channels with different activation ranges.

The motor pattern generating network for heartbeat in the leech, Hirudo medicinalis, is paced by the rhythmic alternating activity of two pairs of reciprocally inhibitory heart interneurons, which are located in the third and fourth segmental ganglia (Fig. 1) (for a recent review, see Calabrese et al. 1995). These neurons inhibit one another via both graded and spike-mediated transmissions (Fig. 1C) (Arbas and Calabrese 1987a,b; Nicholls and Wallace 1978a,b). Spike-mediated inhibitory postsynaptic potentials (IPSPs) are large and prominent during normal oscillations, and graded transmission is particularly noticeable when hyperpolarizing perturbations elicit rebound plateaus in one of the neurons (Fig. 1B). Indeed, in high Ca²⁺ (5 mM) low Na⁺ (20 mM) salines, spikes are suppressed and oscillations based solely on graded inhibition occur (Arbas and Calabrese 1987b; Nadim et al. 1995).

View larger version (37K): [in this window] [in a new window]   FIG. 1. A: circuit diagram showing inhibitory connections among heart (HN) interneurons that pace heartbeat in leech. Cells are indexed by ganglion number and body side in this and subsequent figures. HN cells of 3rd and 4th segmental ganglia make reciprocal inhibitory connections across ganglionic midline, and each pair constitutes an independent neural oscillator. These oscillators are coordinated through their connections with HN(1) and HN(2) interneurons, which are here lumped together because they are functionally equivalent. B: typical alternating bursting activity of a pair of oscillator interneurons in an isolated ganglion in normal saline. HN(L,3) cell responds to a 6-s hyperpolarizing current pulse (DCC) with a pronounced sag potential and a robust plateau and intense impulse burst on release from the current pulse. The plateau is so intense that spikes are inactivated for a brief period, and associated with the plateau is strong graded inhibition of opposite HN neuron. C: transition to a burst in HN(L,3) cells and corresponding synaptic inhibition in the opposite HN neurons during normal oscillation (C1) and following a hyperpolarizing pulse (C2). Arrows indicate 50 mV.

We have presented evidence from experiments in which two reciprocally inhibitory interneurons were voltage clamped, that graded transmission between heart interneurons is mediated by low-threshold Ca²⁺ currents with two kinetic components, one rapidly inactivating (I_CaF) and one slowly inactivating (I_CaS) (Angstadt and Calabrese 1991). In these experiments, the low-threshold Ca²⁺ currents were well characterized, but contamination with outward currents prevented measurement of Ca²⁺ currents at presynaptic voltages above 30 mV, so that the possibility of Ca²⁺ currents activated at more depolarized potentials, that might underlie spike-mediated transmissions remained unexplored. However, the low-threshold currents we described were inactivated by holding the presynaptic cell at 40 mV or above (Angstadt and Calabrese 1991), yet even from a holding potential of 35 mV, depolarizing voltage pulses to 0 mV elicited strong synaptic inhibition of the postsynaptic cell (robust IPSCs) (Simon et al. 1994), thus providing indirect evidence for other Ca²⁺ currents that might underlie spike-mediated synaptic transmissions in these neurons.

Here we report on experiments in which we further explore low-threshold Ca²⁺ currents and use saline with a high concentration of Ba²⁺ to explore divalent cation currents over a broad voltage range. We then use selective block of a component of the Ca²⁺ currents with Cd²⁺ (100-150 mM) to provide direct evidence that graded and spike-mediated synaptic transmissions are separate processes, controlled by different components of the Ca²⁺ current. We also use selective block of spike-mediated transmission with Cd²⁺ to demonstrate the necessity of this transmission in sustaining oscillation in heart interneurons in normal saline.

	METHODS

Abstract Introduction Methods Results Discussion References

Animals

Adult leeches (H. medicinalis) were obtained from Leeches USA and Biopharm and stored at 15°C.

Preparation

Leeches were anesthetized in cold saline, after which individual ganglia (midbody ganglion 3 or 4) were dissected and pinned in small, clear, Sylgard-coated Petri dishes. The sheath on the ventral, surface of the ganglion was removed with fine scissors or microscalpels. Ganglia were superfused continually with normal leech saline (Nicholls and Baylor 1968) containing (in mM) 115 NaCl, 4 KCl, 1.8 CaCl₂, 10 glucose, and 10 N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid buffer, adjusted to pH 7.4 with NaOH or HCl. Heart interneurons were identified by the posterolateral position of their somata on the ventral surface of the ganglion and by their characteristic pattern of rhythmic bursting. Once a cell was identified, the superfusate usually was switched to a Na⁺-free (0 Na⁺) saline with altered divalent cation concentrations appropriate to a given experiment. The Na⁺ was replaced with the appropriate concentration of N-methyl-D-glucamine (NMDG) necessary to maintain the osmotic strength of normal saline, given the alterations in divalent cation concentrations. Anions were provided by neutralizing the NMDG with HCl. Zero Na⁺ salines containing 5 mM Ca²⁺, or 5 or 37.5 mM Ba²⁺ were used. Cd²⁺ (100-150 mM), other divalent ions and organic channel blockers were added to some salines to block Ca²⁺ currents. Stock solutions of nitrendipine (5 mM) and nifedipine (10 mM) were prepared as ethanol stocks and diluted to final concentrations with 0 Na⁺, 37.5 mM Ba²⁺ saline just before use. The ethanol concentration in saline was <1%, and no noticeable effects of this concentration of ethanol on electrical properties of neurons were detected. The stocks were kept in the dark at 4°C. Calcicludine and -conotoxin GVIA were dissolved in water and diluted to final concentrations with 0 Na⁺, 37.5 mM Ba²⁺ saline just before use. Stock solutions in water were stored in the dark at <20°C. All physiological salines were adjusted to pH 7.4.

Heart interneurons were penetrated with thin-walled (1 mm OD, 0.75 mm ID) borosilicate microelectrodes (A-M Systems). For the recording of normal activity and in current-clamp experiments, microelectrodes filled with 4 M K-acetate, 20 mM KCl (unbuffered, pH 8.4) were used. In voltage-clamp experiments, microelectrodes were filled with 2 M K-acetate, 2 M tetraethylammonium acetate (TEA-acetate) (unbuffered, pH 7.9), which blocks both I_A-like current (I_A) and delayed-rectifier-like outward currents (I_K1 and I_K2) (Simon et al. 1992), and 20 mM KCl. In some experiments, electrodes containing the Ca²⁺ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid (BAPTA; potassium salt; Sigma Chemicals) were used. These microelectrodes were filled with 0.2 M BAPTA, 2 M K-acetate, and 20 mM KCl, or with 0.2 M BAPTA, 2 M K-acetate, 2 M TEA-acetate, and 20 mM KCl. These BAPTA-containing solutions were stored at <20°C. Microelectrodes had resistances of 30-45 M and time constants of 0.2-0.5 ms.

Voltage-clamp recordings were made using an Axoclamp-2A (Axon Instruments) or a NPI SEC-05L (NPI) used in single-electrode voltage-clamp mode with a sampling rate of 2.5 and 20 kHz, respectively. Current clamp recordings were made with an Axoclamp-2A or a NPI SEC-05L used in discontinuous current-clamp (DCC) mode with a sampling rate of 2.5 and 20 kHz, respectively. In each case, the electrode potential was monitored on an oscilloscope to ensure that the potential settled between current injection cycles. Some current recordings were made with the same amplifiers in bridge mode. All recordings were referenced to a bath microelectrode of the same composition as the recording microelectrode; the bath itself was grounded through a chlorided silver wire. After a voltage-clamp experiment, microelectrodes were withdrawn from the cell, and only in experiments in which the electrode was within ±5 mV of bath potential were the data accepted. Thus all membrane potentials reported for voltage clamp experiments are accurate within a 10-mV range.

Data were digitized, stored, and analyzed on a 486 or Pentium (Intel) computer using Axotape or pCLAMP suite of software (Axon Instruments). All voltage-clamp data were acquired and voltage protocols generated using the pCLAMP program CLAMPEX. The usual voltage-clamp protocol consisted of voltage pulses from a holding potential of 70 mV to various depolarizing voltages. Each of these positive pulses was preceded by four negative prepulses of one-fourth magnitude and equal duration. The interval between the prepulses in the sequence, the delay between the prepulse sequence and the test pulse and the interval between prepulse-test pulse episodes were all adjusted for each different test pulse protocol so that the holding current returned to baseline between all pulses and/or prepulses by monitoring the holding current on-line with a chart recorder. The sum of the currents for these four prepulses was used as a measure leak current. All currents shown were leak subtracted automatically using this procedure in CLAMPEX unless otherwise noted. Although the raw (unsubtracted) currents were not digitized by CLAMPEX, they were monitored on-line with the chart recorder, so that we could verify that the estimated leak currents were time-invariant and approximately linear. In previous studies of low-threshold Ca²⁺ currents (Angstadt and Calabrese 1991), we digitized leak currents (using single negative voltage prepulses of equal magnitude) directly and subtracted them off-line. The results using the automatic procedure employed are similar to these previous results. Ramp protocols also were used, the height and duration of which were chosen to explore the currents optimally through reasonable voltage and time parameters.

Although the dendritic tree of the HN cell appears to be rather extensive, the cells are thought to be compact electrically, especially under the experimental conditions used here. When bathed in Na⁺-free saline the membrane resistance of the cells is relatively large (50-100 M), and many of the voltage-dependent currents were suppressed in our experiments. Use of Na⁺-free saline eliminates Na⁺ currents, and K⁺ currents (I_K and I_A) were suppressed partially with TEA, and/or BAPTA and/or Ba²⁺. These conditions support acceptable space-clamp. On the other hand, particularly for the fast currents (I_CaF and I_BaF), the steepness of the activation curve indicates that there may have been some problems holding the clamp potential in distal areas of the dendritic arbor where these currents reside. Still, these fast currents were graded with pulse potential. Moreover, the focus of these studies was on the function role of Ca²⁺ currents rather than on a quantitative description of their biophysical properties.

Voltage-clamp data were analyzed and filtered (RC filter with a user defined time constant of >4 ms) using the pCLAMP program CLAMPFIT. CLAMPFIT allows least-squares fits of exponential equations to current traces for the analysis of inactivation (or activation) time constants. Two different methods can be applied to make these fits in CLAMPFIT, the Simplex and Chebyshev methods. The former method is slower, requires a seed value, and is regarded as more accurate though subject to selecting local minima with certain seed values, while the later method is faster and does not require a seed value. Some of these data were fit with both methods and no differences were noted.

Current-clamp data were acquired using Axotape software or by recording on FM tape recorder (Vetter, Model 420C)

Experiments

Current-clamp experiments were performed in normal saline or 0 Na⁺, 5 mM Ca²⁺ saline, with or without added Cd²⁺ (100-150 µM). All voltage-clamp experiments were done on interneurons in ganglia initially superfused in 0 Na⁺, 5 mM Ca²⁺ saline; in this saline, cells ceased generating action potentials and hyperpolarized 10-15 mV. All cells initially were held in voltage clamp at 70 mV. For experiments done in 5 mM Ca²⁺ or 5 mM Ba²⁺, 0 Na⁺ salines, the protocol consisted of successive 1,500-ms voltage-pulses in 2.5-mV increments. For experiments done in 0 Na⁺, 37.5 mM Ba²⁺ saline, both ramp and pulse protocols were used. The ramp protocol steadily increased the voltage from 70 to +15 mV during 2,800 ms and quickly returned the voltage to 70 mV during 500 ms. The pulse protocol consisted of successive 1,500-ms voltage-pulses in 10-mV increments.

	RESULTS

Abstract Introduction Methods Results Discussion References

Inward currents recorded in 0 Na⁺ saline with 5 mM Ca²⁺ or 5 mM Ba²⁺

Our previous experiments on low-threshold Ca²⁺ currents demonstrated two kinetic components, one rapidly inactivating (I_CaF) and one slowly inactivating (I_CaS). Our data suggested that I_CaF showed Ca²⁺-mediated inactivation; inactivation was more rapid in elevated Ca²⁺ saline (5 mM) compared with normal Ca²⁺ saline (1.8 mM). To explore this possibility, we compared the inward currents of cells (penetrated with TEA microelectrodes2 M K-acetate, 2 M TEA-acetate, 20 mM KCl) recorded first in 0 Na⁺ elevated Ca²⁺ saline (5 mM) and then in 0 Na⁺ Ba²⁺ saline (5 mM), using a pulse protocol (V_hold of 70 mV, successive 1,500-ms pulses incremented by 2.5 mV) (Fig. 2, A and B). Currents were measured at peak amplitude and at 1 s from pulse onset to quantify both the rapidly and slowly inactivating components of the inward current (Fig. 2, C and D). Peak currents were greater with Ca²⁺ as the charge carrier than with Ba²⁺, but Ca²⁺ and Ba²⁺ currents were comparable in amplitude at 1 s. These results indicate that Ba²⁺ is not as effective as a charge carrier for the rapidly inactivating inward current (I_CaF) as is Ca²⁺.

View larger version (32K): [in this window] [in a new window]   FIG. 2. Comparison of low-threshold inward currents obtained from depolarizations of heart interneurons in single-electrode voltage-clamp (SEVC) from a holding potential of 70 mV with Ca2+ and with Ba2+ as charge carrier and tetraethylammonium (TEA; 2 M) in microelectrode. A: traces obtained in 0 Na+ saline containing 5 mM Ca2+ elicited by voltage pulses to potentials to right of each record. B: traces obtained in 0 Na+ saline containing 5 mM Ba2+ elicited by voltage pulses to potentials to left of each record. C: I-V plots of currents obtained in 0 Na+ saline containing 5 mM Ca2+. D: I-V plots of currents obtained in 0 Na+ saline containing 5 mM Ba2+ (same cell as in A). In both C and D, values were calculated by averaging measurements from 5 trials at peak amplitude (ICaF and IBaF) and at 1 s from pulse onset (ICaS and IBaS). In each case, Ca2+ and Ba2+ current was measured in same heart interneuron. Error bars indicate standard errors of all averaged values.

The apparent inactivation kinetics of the rapidly inactivating component of the inward current differed markedly in Ca²⁺ and Ba²⁺ salines. To quantify this difference, we determined the inactivation time constant of the rapidly inactivating component of the inward current by fitting a single exponential to the rapidly inactivating portion of the current trace (Fig. 3). The inactivation time constants in Ba²⁺ saline were on average about two times slower than those in Ca²⁺ saline at each voltage pulse, and the time constant values in the two salines corresponding to the same voltage pulse showed a statistically significant difference (paired t-test; P < 0.01). These results indicate that Ca²⁺ entry influences apparent inactivation kinetics of I_CaF. To explore this latter possibility further, we compared the inactivation time constant of the rapidly inactivating component in elevated Ca²⁺ saline (5 mM) of cells penetrated with TEA-filled microelectrodes and of cells penetrated with similar electrodes but with added BAPTA (0.2 M), a fast Ca²⁺ chelator. With BAPTA-containing microelectrodes, the outward currents appeared to be more effectively blocked in 0 Na⁺ 5 mM Ca²⁺ saline than with the normal TEA-filled electrodes. Current records began to be contaminated with outward current with depolarizing pulses above 37.5 mV with normal TEA-filled electrodes (Fig. 2), but they remained apparently uncontaminated with pulses up to 25 mV when BAPTA was added to the microelectrodes (Fig. 4) (n = 5). [This effect on outward currents was apparently due to BAPTA itself, because in separate experiments using microelectrodes containing BAPTA but not TEA (2 M K-acetate, 20 mM KCl, and 0.2 M BAPTA) comparable results were obtained (n = 5, data not shown).] Under these conditions I_CaF's activation appeared shifted to the right with the maximum current observed at about 35 mV (Fig. 4), as compared with 42.5 mV, when the current was recorded with normal TEA-filled microelectrodes. These results indicate that I_CaF is indeed "low-threshold" and transient. The late current, measured at 1 s after the onset of the pulse (n = 5), however, continued to increase (Fig. 4), indicating that I_CaF's inactivation was slowed sufficiently so that it contributed to the current at 1 s, or that the more slowly inactivating I_CaS's activation range might be broader than previously thought or that another different current was being activated. The inactivation time constant of I_CaF was increased dramatically by BAPTA throughout its activation range (Fig. 4). At 37.5 mV, the inactivation time constant of I_CaF was significantly greater when recorded with BAPTA containing electrodes [503 ± 162 ms (mean ± SE)] than when recorded with normal TEA electrodes (Fig. 3) (unpaired t-test P < 0.02). Thus intracellular Ca²⁺ does influence the apparent inactivation kinetics of I_CaF.

View larger version (24K): [in this window] [in a new window]   FIG. 3. Average values of inactivation time constants of rapidly inactivating component of inward currents obtained from depolarizations of heart interneurons in SEVC from a holding potential of 70 mV. Cells were bathed in 0 Na+ saline containing either 5 Ca2+ or 5 mM Ba2+. Recording electrodes all contained TEA (2 M) and in some cases1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid (BAPTA; 0.2 M). Data obtained without BAPTA were from 5 heart interneurons bathed successively in Ca2+ and Ba2+ saline (same cells as in Fig. 2), and data obtained with BAPTA were from 5 different heart interneurons. Error bars indicate standard errors.

View larger version (31K): [in this window] [in a new window]   FIG. 4. Effect of BAPTA on low-threshold Ca2+ currents in heart interneurons measured in SEVC. Currents, obtained in 0 Na+ saline containing 5 mM Ca2+ with TEA (2 M) and BAPTA (0.2 M) in recording microelectrode, were elicited from a holding potential of 70 mV by voltage pulses, which started at 55 mV and rose to 35 mV in 2.5-mV increments.

Inward currents recorded in 0 Na⁺ saline with 37.5 mM Ba²⁺

To explore the full activation range of Ca²⁺ currents in heart interneurons, we took advantage of two observations: Ba²⁺ is an effective charge carrier for Ca²⁺ channels in heart interneurons and high concentrations of Ba²⁺ block outward currents in leech cells (see, e.g., Thompson and Calabrese 1992). Initial tests were done in 0 Na⁺, 37.5 mM Ba²⁺ saline with voltage-ramp protocol that took advantage of the relatively low threshold and rapid inactivation of the rapidly inactivating component of the Ba²⁺current to separate it from other Ba²⁺ currents. From V_hold of 70 mV, the membrane potential was ramped to +15 mV during the course of 2.8 s (Fig. 5A). The currents obtained were leak subtracted based on extrapolation of I_leak obtained from the first 800 ms of the resulting current, which was linear. The leak subtracted currents were plotted versus membrane voltage to produce an I-V plot (Fig. 5B). This ramp protocol clearly separated the current into two distinct components one that activated at relatively low threshold, reaching its peak amplitude at about 35 mV and inactivated during the ramp, and one that activated over a broad range but predominantly above 20 mV and appeared to inactivate little (if at all) during the ramp. These currents were designated I_BaF and I_BaS, respectively, and their relationship to I_CaF and I_CaS will be considered in DISCUSSION. I_BaF had its maximum amplitude at approximately 35 mV on the ramp, whereas I_BaS reached its maximum amplitude at approximately 5 mV.

View larger version (14K): [in this window] [in a new window]   FIG. 5. Ba2+ currents obtained from ramp depolarizations of heart interneurons in SEVC from a holding potential of 70 mV to a peak potential of +15 mV in 0 Na+ saline containing 37.5 mM Ba2+. A: voltage protocol and current elicited by protocol. Current shown is average of 2 trials on same cell. B: I-V plot of above current. Current was leak subtracted by extrapolation of Ileak from 1st 600 ms of current elicited by the ramp. Down-ramp portion of currents (final 750 ms) was discarded to clarify plot.

In 0 Na⁺, 37.5 mM Ba²⁺ saline, with TEA microelectrodes, we used a conventional pulse protocol from V_hold of 70 mV to explore more fully I_BaF and I_BaS (Fig. 6A). To minimize the effects of run down that can occur in these experiments, the pulse protocol was performed on six heart interneurons in order of increasing pulse voltage and on five cells in order of decreasing pulse voltage and these data averaged (Fig. 6B). The pulses were 1,500 ms in duration and were produced in 10-mV increments with the smallest pulse to 60 mV and the largest pulse to +10 mV. I-V plots were constructed for the peak I_BaF and the peak I_BaS. These plots show that I_BaF reaches its maximum amplitude at approximately 20 mV, whereas I_BaS peaks at ~0 mV. During a depolarizing pulse, I_BaS reaches its peak amplitude after ~200-250 ms, well after I_BaF is inactivated substantially (Fig. 6A).

View larger version (21K): [in this window] [in a new window]   FIG. 6. Ba2+ currents in heart interneurons obtained from a holding potential of 70 mV by pulse depolarizations ranging from 60 mV to +10 mV (10-mV increments) of heart interneurons in SEVC in 0 Na+ saline containing 37.5 mM Ba2+. A: currents elicited by increasing voltage pulses to the potentials to left of each record. B: I-V plot of IBaF and IBaS. A total of 11 cells were recorded, 6 in which pulses were applied in increasing order and 5 in which pulses were applied in decreasing order. Values were calculated by averaging measurements from all 11 cells at peak amplitude of rapidly inactivating current [IBaF(peak)] and at peak amplitude of the slowly inactivating current [IBaS(peak)]. Error bars indicate standard errors.

Sensitivity of I_BaF to Cd²⁺

In 0 Na⁺, 37.5 mM Ba²⁺ saline, using TEA-filled microelectrodes and a pulse protocol, we tested various Ca²⁺ channels blockers (Table 1) to determine whether any selectively blocked one component of the Ba²⁺ current. The organic Ca²⁺ channel blockers nifedipine, nitrendipine (Janis and Triggle 1991), calcicludine (Schweitz et al. 1994), and -conotoxin GIVA (Kasai et al. 1987) had no detectable effect on I_Ba (Table 1). The inorganic channel blocker, Zn²⁺ (1 mM), nonselectively blocked both components of I_Ba. However, in five different heart interneurons, 100 µM Cd²⁺ effectively blocked I_BaS at all pulse potentials, while hardly interfering with I_BaF particularly at pulse potentials below 10 mV (Figs. 7 and 8; Table 1). The current that remains in Cd²⁺ (100 µM) appears to decay smoothly with two time constants, and there is no characteristic slow activation of I_BaS that appears to account for the late current (Fig. 7B). We fit two exponentials to this decay for the pulse potential to 10 mV for five cells from five different preparationsand found average fast time constants of 56.7 ± 4.7 ms(n = 5), and slow time constants of 1,876 ± 935 ms (n = 3for two cells slow time constants could not be fit because the decay was so gradual as to be almost linear), respectively. The difference current (Fig. 7C) in the presence and absence of Cd²⁺ (100 µM), albeit slightly contaminated with a small amount of I_BaF, apparently blocked by Cd²⁺, particularly at the higher pulse potentials, reveals the characteristic less rapid activation of I_BaS. The effects of Cd²⁺ on I_BaS were not reversible with persistent washing. Similar results to those obtained with Cd²⁺ (100 µM) were obtained with 200 µM Ni²⁺.

View larger version (25K): [in this window] [in a new window]   FIG. 7. Effect of Cd2+ (100 µM) on Ba2+ currents obtained from a holding potential of 70 mV by pulse depolarizations ranging from 40 to 0 mV of heart interneurons in SEVC in 0 Na+ saline containing 37.5 mM Ba2+. A: control current. B: current in presence of Cd2+. C: currents in B were subtracted from those in A to reveal current blocked by Cd2+. Records in A and B are from same cell.

View larger version (17K): [in this window] [in a new window]   FIG. 8. I-V plots showing effect of Cd2+ (100 µM) on Ba2+ currents obtained from a holding potential of 70 mV by pulse depolarizations ranging from 40 to 0 mV of heart interneurons in SEVC in 0 Na+ saline containing 37.5 mM Ba2+. Values were calculated by averaging measurements from 5 cells at peak amplitude of rapidly inactivating current (IBaF) and at peak of slowly inactivating current (IBaS; as in Fig. 6) in presence and absence of Cd2+. Error bars indicate standard errors.

Similarly Cd²⁺ (100 µM) appears to block I_BaS selectively, while leaving I_BaF relatively unaltered, when these currents are separated using a voltage ramp protocol (Fig. 9). There is some current that remains in the presence of Cd²⁺ at the more depolarized end of the ramp, which may correspond to the slowly inactivating late current observed with the pulse protocol in the presence of Cd²⁺ (Fig. 7B).

View larger version (12K): [in this window] [in a new window]   FIG. 9. Effect of Cd2+ (100 µM) on Ba2+ currents obtained from ramp depolarizations of heart interneurons in SEVC from a holding potential of 70 mV to a peak potential of +15 mV in 0 Na+ saline containing 37.5 mM Ba2+. I-V plots were constructed by procedure illustrated in Fig. 5. A: control currents. B: currents in the presence of Cd2+. Records in A and B are from same cell.

Effects of Cd²⁺ (100 µM) on I_CaF and I_CaS

We explored the effect of Cd²⁺ on I_CaF and I_CaS so that we might correllate these calcium currents with the barium currents, I_BaF and I_BaS, described above. We found that Cd²⁺ (100 µM) had very little effect on either I_CaF or I_CaS (Fig. 10). We obtained similar results with five heart interneurons from five different preparations.

View larger version (22K): [in this window] [in a new window]   FIG. 10. Effect of Cd2+ (100 µM) on low-threshold Ca2+ currents in heart interneurons measured in SEVC in 0 Na+ saline containing 5 mM Ca2+. Currents were elicited from a holding potential of 70 mV by voltage steps, which started at 55 mV and rose to 35 mV in 5-mV increments. A: control currents. B: currents in presence of Cd2+. C: currents in B were subtracted from those in A to reveal low-threshold currents blocked by Cd2+. Records in A and B are from same cell. Similar results were obtained in 5 preparations.

Effects of Cd²⁺ (100 µM) on graded and spike-mediated synaptic transmission between heart interneurons

We reasoned that because I_CaF or I_CaS both contribute to the production of plateau potentials and are associated with graded synaptic transmission between heart interneurons (Angstadt and Calabrese 1991) and that these currents are not sensitive to low concentrations of Cd²⁺ (100 µM), it might be possible to determine whether Ca²⁺ currents corresponding to I_BaS might underlie spike-mediated synaptic inhibition between these neurons while I_CaF or I_CaS underlie graded synaptic inhibition. First we determined whether Cd²⁺ (100 µM) had any effect on plateau potential production in or graded synaptic inhibition between heart interneurons. Pairs of reciprocally inhibitory heart interneurons were recorded in 0 Na⁺, 5 mM Ca²⁺saline, using DCC; one cell was held at 40 mV (presynaptic) and the other at 50 mV (postsynaptic) with injected current (Fig. 11). Plateau potentials were elicited in the presynaptic cell on termination of a hyperpolarizing current pulse (0.4 nA), and the graded IPSP recorded in the postsynaptic cell. Cd²⁺ (100 µM) had little effect on either the plateau potential in the presynaptic cell or the graded IPSP recorded in the postsynaptic cell. Next we tested whether Cd²⁺ could block selectively spike-mediated inhibition between heart interneurons while leaving graded inhibition intact. Two reciprocally inhibitory heart interneurons were recorded in normal saline (1.8 mM Ca²⁺) using DCC (Fig. 12). Such pairs oscillated in normal alternation (Fig. 12A1). On termination of a hyperpolarizing current pulse injected into one neuron, a plateau was elicited that caused strong graded inhibition of the other interneuron and normal spike-mediated IPSPs were observed (Fig. 12A2). Addition of Cd²⁺ (150 µM) to the saline severely disrupted normal oscillation, but did not noticeably alter the average "resting" potential of the neurons (Fig. 12B1). Robust plateaus and graded synaptic inhibition occurred on termination of hyperpolarizing current pulses, but no spike-mediated IPSPs were recorded in the presence of Cd²⁺ (Fig. 12B2). These results indicate that Cd²⁺-sensitive Ca²⁺ currents underlie spike-mediated inhibition between heart interneurons and that this spike-mediated inhibition is necessary for alternating oscillation in these neurons in normal saline. These results also suggest that graded synaptic inhibition between heart interneurons is not sufficient to sustain alternating oscillation in normal saline.

View larger version (13K): [in this window] [in a new window]   FIG. 11. Cd2+ (100 µM) does not block ability of heart interneurons to produce plateau potentials in 0 Na+ saline (5 mM Ca2+) or to inhibit their contralateral homolog by graded synaptic transmission. Recordings were made from a pair of reciprocally inhibitory heart interneurons in 0 Na+ saline, and plateaus were elicited by 1-s hyperpolarizing current pulses (0.4 nA DCC) in absence (A) and presence (B) of Cd2+. For each trace dashed line indicates baseline potential at which cell was held with steady current injection.

View larger version (29K): [in this window] [in a new window]   FIG. 12. Cd2+ (150 µM) blocks spike-mediated synaptic transmission between heart interneurons and prevents normal oscillation but does not block graded synaptic transmission between the neurons. Activity was recorded from a pair of reciprocally inhibitory heart interneurons in normal saline in absence (A1 and A2) and presence (B1 and B2) of Cd2+. Plateau potentials were elicited by 4-s hyperpolarizing current pulses (DCC). In each case, 2 shows a time expansion of record for period in 1 indicted by bar. Note in A2 that plateau elicits a graded inhibitory postsynaptic potential (IPSP) and associated spikes elicit distinct IPSPs in other heart interneuron, whereas in B2, plateau elicits a graded IPSP but associated spikes do not elicit IPSPs in other heart interneuron.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Ca²⁺ currents in heart interneurons

Our previous experiments implicated low-threshold Ca²⁺ currents in plateau production and graded synaptic transmission in heart interneurons (Angstadt and Calabrese 1991). In those studies, we demonstrated that Ca²⁺ currents with low threshold (below 50 mV) had two kinetic components, one rapidly inactivating and one slowly inactivating. These components were termed I_CaF and I_CaS, respectively. These currents were characterized over the range of membrane potential of 60 to 30 mV and both appeared to saturate below 30 mV. Because of contamination with outward currents, we could not explore more depolarized membrane potentials, and thus cannot state definitively that the currents truly saturated. Both currents were implicated in plateau production and graded synaptic transmission (Angstadt and Calabrese 1991): 1) the duration of the plateaus is too long to be accounted for by I_CaF alone. 2) voltage-clamp protocols that simulate the voltage excursions associated with normal plateaus elicit both I_CaF and I_CaS. 3) Graded transmission follows plateaus in duration and amplitude. 4) Brief voltage-clamp pulses that elicit predominantly I_CaF elicit correspondingly brief graded IPSP(C)s, while longer voltage pulses that elicit both I_CaF and I_CaS elicit correspondingly long graded IPSP(C)s. 5) Graded IPSP(C)s elicited by long voltage-clamp pulses show two decay time constants that correspond to the inactivation time constants of I_CaS and I_CaF. In those studies, no specific blocker for one or the other of the I_Ca components was found that would allow us to separate these components and assess their different roles in plateau production and graded synaptic transmission (Angstadt and Calabrese 1991).

Our modeling studies (De Schutter et al. 1993) indicated that I_CaF and I_CaSas characterized in our experiments (Angstadt and Calabrese 1991)could not account for spike-mediated synaptic transmission between heart interneurons. Moreover, in experiments (Simon et al. 1994) where one heart interneuron was held in voltage clamp at 35 mVa potential where both I_CaF and I_CaS had been shown to be inactivated completely (Angstadt and Calabrese 1991)while its postsynaptic partner was clamped at 35 mV to reveal IPSCs, presynaptic voltage pulses to 0 mV-elicited robust IPSCs. These results suggested to us a scenario in which heart interneurons possess two types of Ca²⁺ currents with distinctly different functions, one, a low-threshold type consisting of two kinetic components, I_CaF and I_CaS, that mediates plateau formation and graded synaptic transmission, and another, with a broad activation range that underlies spike-mediated synaptic transmission. A primary goal of the experiments reported here has been to characterize this other type of Ca²⁺ current and to determine its role in spike-mediated synaptic transmission. A secondary goal was to determine whether I_CaF inactivation is Ca²⁺ mediated.

Influence of intracellular Ca²⁺ on the apparent inactivation kinetics of I_CaF

Both our results comparing the rapidly inactivating component of inward current in the same neurons in 0 Na⁺, 5 mM Ca²⁺ saline and 0 Na⁺, 5 mM Ba²⁺ saline and our measurements of the inactivation time constant of I_CaF in cells recorded with BAPTA-filled microelectrodes, indicate that the time constant of the inactivation of I_CaF is Ca²⁺ sensitive. Several Ca²⁺ currents in both invertebrates and vertebrates share this characteristic (Byerly and Hagiwara 1988; Eckert and Chad 1984; Hille 1992). Still there is need for caution in interpreting our results, because despite our efforts to block outward currents with internal TEA, the possibility remains that Ca²⁺-mediated outward currents may contaminate our measurements. Both using Ba²⁺ as a charge carrier and BAPTA injection would be expected to block such contaminating currents and influence the apparent inactivation of I_CaF.

These considerations emphasize the need for caution in interpreting inactivation kinetics of fast Ca²⁺ currents in heart interneurons.

Ba²⁺ currents in heart interneurons

We were not able to measure directly Ca²⁺ currents in heart interneurons at more depolarized potentials, primarily because of our failure to block outward currents effectively. External TEA (Simon et al. 1992) and Cs⁺-filled microelectrodes (unpublished results) have proved ineffective. Our best efforts, before this study, at blocking outward currents have been with TEA-filled microelectrodes, which provide only a partial block so that no Ca²⁺ currents can be recorded at pulse potentials above 30 mV. Thus in this study, we were forced to measure Ca²⁺ currents indirectly by replacing Ca²⁺ with Ba²⁺ as a charge carrier. The high concentration (37.5 mM) of Ba²⁺ in the saline employed, in addition to providing robust Ba²⁺ currents, had the additional salubrious effect of blocking outward currents effectively when used in conjunction with TEA-filled microelectrodes. Under these experimental conditions, we were able to record largely uncontaminated Ba²⁺ currents over the range of membrane potential from 70 to +15 mV. Using a voltage-ramp protocol, we separated the Ba²⁺ currents into two typesrapidly inactivating and slowly inactivating. During the ramps, these currents had overlapping activation ranges, but clearly I_BaF reached its peak amplitude (around 40 mV) and appeared fully activated well below I_BaS's peak amplitude (around 10 mV) (Fig. 5). Thus I_BaF appears to be a low-threshold transient current, while I_BaS can be viewed as broadly activating and persistent. Studies with pulse protocols confirm these impressions but the range of overlap of the activation is shown to be more extensive. Because I_BaF appears to inactivate below the maximum activation of I_BaS during the ramp protocol, the ramp protocol provides better separation of the two currents. However, the pulse protocol clearly reveals the delayed activation of I_BaS, which does not reach peak amplitude until ~500 ms after the start of the pulse (particularly at pulse potentials 20 and above) during which time I_BaF is inactivated substantially (Figs. 6 and 7A).

Low concentrations of Cd²⁺ (100-150 µM) selectively block I_BaS, while leaving I_BaF relatively unaltered, except at more depolarized pulse potentials (Figs. 7 and 8). The currents, observed with the pulse protocol, that remain in the presence of Cd²⁺ (100 µM) show little of the delayed activation of I_BaS; they decay relatively smoothly with two time constants (~50 and 1,800 ms, respectively).

Relation between Ba²⁺ currents and Ca²⁺ currents in heart interneurons

A determination of relation between I_CaF and I_CaS, previously described (Angstadt and Calabrese 1991), the I_BaS and I_BaF described here, and ultimately the types of Ca²⁺ currents that heart interneurons possess requires careful consideration of the data. The experimental conditions under which Ba²⁺ currents were recorded were radically different from those under which Ca²⁺ currents were recorded; the high concentration of Ba²⁺(37.5 mM)a divalent cation which can dramatically alter membrane surface charge (Hille 1992)necessary to record inward currents, uncontaminated by outward currents at depolarized potentials, undoubtedly shifts the activation characteristics of these currents as well as altering their inactivation characteristics. It is tempting to equate the Ba²⁺ currents one-for-one with their Ca²⁺ counterparts, and this possibility is not excluded by our data, but results with Cd²⁺ block support a more complicated scenario in which I_BaF equates to I_CaF and I_CaS, and I_BaS reflects a qualitatively different Ca²⁺ current. Thus I_BaF reflects the low-threshold type and I_BaS reflects the broadly activating type Ca²⁺ current mentioned above. In the presence of Cd²⁺ (100 µM), the less rapidly developing inward current associated with I_BaS is blocked selectively, but the remaining current decays (inactivates) with two distinct time constants much as the two low-threshold Ca²⁺ currents, I_CaF and I_CaS, that we had previously described, decay with two distinct time constants. It is not possible to compare the time constants directly because the fast time constant is Ca²⁺ sensitive and because high concentrations of Ba²⁺ employed (37.5 mM) undoubtedly shift the voltage dependence of inactivation as well as activation, but nevertheless they are in congruence. Moreover, the late Ba²⁺ current observed with effective Cd²⁺ block appears to develop in conjunction with I_BaF and not in the delayed fashion associated with I_BaS. We cannot, however, rule out the possibility that Cd²⁺ blocks I_BaS incompletely and that it alters the activation characteristics of the unblocked current. Regardless of which scenario is correct, it is clear that Cd²⁺ (100 µM) selectively blocks a type of Ca²⁺ current with a relatively broad activation range and slow inactivation characteristics, referred to henceforth as broadly activatingCa²⁺ current, while leaving a low-threshold-type rapidly inactivating current intact. In the presence of Cd²⁺ the remaining slowly inactivating current may represent a low-threshold current or the dregs of the broadly activating current.

It is with considerable reserve that we compare our findings in leech heart interneurons with the vast literature onvertebrate Ca²⁺ currents, particularly because the organic blockers that are so important in defining channels types in vertebrate neurons (Hille 1992; Tsien et al. 1988) do not appear to work on leech Ca²⁺ currents. Although there are exceptions (Trudeau et al. 1993), other invertebrate Ca²⁺ currents likewise do not fit well into the vertebrate scheme (Charlton and Augustine 1990; Tsien et al. 1988) Nevertheless, I_CaF is similar to T-type currents found in vertebrate neurons that are low threshold and rapidly inactivating. I_CaF is also similar to T-type currents in that Ba²⁺ appears to be a less effective charge carrier than Ca²⁺ for this current. Unlike most T-type currents, I_CaF is relatively insensitive to block by low concentrations of Ni²⁺ (200 µM) and may show Ca²⁺-mediated inactivation. The broadly activating Ca²⁺ current identified here with I_BaS appears L-like; it activates predominantly above 20 mV (at least in 37.5 mM Ba²⁺ saline), it has less rapid activation and slow inactivation kinetics, and it is sensitive to block by low concentrations of Cd²⁺. The low-threshold Ca²⁺ current identified here with I_CaS does not fit well into any vertebrate classification scheme. It is interesting in this regard that Grolleau and Lapied (1996) have been able to demonstrate convincingly using selective block with Ni²⁺ and selective inactivation with holding potentials, that DUM neurons of cockroach possess three different types of Ca²⁺ currents two low-thresholdone rapidly and one more slowly inactivatingand a high-threshold current.

Separation of graded and spike-mediated synaptic transmission between heart interneurons

Inhibition between reciprocally inhibitory heart interneurons consists of both a spike-mediated component and a graded component associated with the slow wave of oscillation in these neurons and with plateau potential formation (Arbas and Calabrese 1987a,b; Olsen and Calabrese 1996). Synaptic inhibition is necessary for oscillation in these neurons, because when all synaptic inhibition is blocked with bicuculline (10⁴ M), oscillation stops and the neurons become tonically active (Schmidt and Calabrese 1992). We demonstrated here that plateau potentials in and graded synaptic transmission between heart interneurons is largely unaffected by Cd²⁺ (100-150 µM), indicating that the Ca²⁺ currents that control these processes, previously identified as I_CaF and I_CaS, were largely unaffected by these concentrations of Cd²⁺. This result and our results with Cd²⁺ block of I_BaS allowed us to test the role of broadly activating Ca²⁺ currents in spike-mediated synaptic transmission between heart interneurons. We found that spike-mediated IPSPs were blocked by Cd²⁺ (150 mM), but that plateau production and graded inhibition in the same preparation were largely unaffected. This result suggests that a broadly activating Ca²⁺ current, different from those that mediate plateau potentials and graded inhibition, underlies spike-mediated synaptic transmission. I_BaS described here may correspond to this current. T-like currents have been associated with neurosecretion in pituitary cells (Cohen et al. 1988; Tsien et al.1988), and they presumably play an important role in all graded synaptic transmission (see, e.g., Hartline and Graubard 1992). L-like currents are thought to participate in spike-mediated synaptic transmission and neurosecretion in several systems (Artalejo et al. 1994; Hille 1992; Tsien et al. 1988; Wang et al. 1993), while in others they are not associated with synaptic transmission but with activation of K⁺ currents (Fossier et al. 1993).

Relative roles of spike-mediated and graded synaptic inhibition in sustaining oscillation in reciprocally inhibitory heart interneuron pairs

Our observation that Cd²⁺ (150 µM) blocks alternating oscillation in reciprocally inhibitory pairs of heart interneurons in otherwise normal saline suggests that spike-mediated inhibition is necessary for and that graded inhibition is insufficient for oscillation in normal saline. This result fulfills the specific predictions of a detailed model of such an oscillatory pair of neurons that has been based on our previously obtained biophysical data on the intrinsic ionic and synaptic currents of heart interneurons (Nadim et al. 1995; Olsen et al. 1995). This model also predicts that spike-mediated synaptic transmission should be sufficient to sustain oscillation in heart interneurons, but unfortunately we have no selective blocker of graded inhibition in this system. In recent experiments designed to measure spike-mediated and graded transmissions, the graded inhibitory synaptic current was found to be relatively small when compared with the spike-mediated synaptic current (Olsen and Calabrese 1996). Graded inhibition was measured using voltage-clamp waveforms, which closely approximate the slow wave oscillation of heart interneurons, to drive the presynaptic interneuron (spikes were blocked by bathing the preparation in 0 Na⁺ saline), while the postsynaptic cell was held in voltage clamp. Spike-mediated inhibition was measured with the presynaptic heart interneuron oscillating normally in association with the greater heart interneuron network, while the postsynaptic cell was held in voltage clamp.

Selective block of spike-mediated synaptic transmission may be possible in other pattern generating networks, where spike-mediated and graded synaptic transmission work in conjunction, thus also allowing a direct assessment of their relative importance. In the pyloric network of the stomatogastric ganglion, spike-mediated transmission was blocked by blocking spikes with tetrodotoxin (TTX) and oscillation remained robust but only in the presence of the neuromodulator dopamine (Raper 1979). Such an experiment could not be performed in our system because TTX and similar compounds are not effective in blocking spikes (Kleinhaus and Angstadt 1995). More recent experiments in this system (Johnson et al. 1991, 1995) and in the leech swim system (Mangan et al. 1994) demonstrate the dramatic sensitivity of graded transmission to neuromodulators and other environmental factors. These findings point out the possibility that the proper modulatory environment may permit oscillation based solely on graded transmission in heart interneurons. Indeed in high Ca²⁺ (5 mM) low Na⁺ (20 mM) salines, spikes are suppressed and oscillations based solely on graded inhibition occur in heart interneurons (Arbas and Calabrese 1987b; Nadim et al. 1995).

	ACKNOWLEDGEMENTS

  This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-24072 to R. L. Calabrese and a Howard Hughes Undergraduate Summer Research Fellowship to J. F. Dalton.

	FOOTNOTES

  Address for reprint requests: R. L. Calabrese, Dept. of Biology, Emory University, 1510 Clifton Rd., Atlanta, GA 30322.

  Received 17 July 1996; accepted in final form 9 December 1996.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• ANGSTADT, J. D., CALABRESE, R. L. Calcium currents and graded synaptic transmission between heart interneurons of the leech. J. Neurosci. 11: 746-759, 1991.[Abstract]
• ARBAS, E. A., CALABRESE, R. L. Ionic conductances underlying the activity of interneurons that control heartbeat in the medicinal leech. J. Neurosci. 7: 3945-3952, 1987a.[Abstract]
• ARBAS, E. A., CALABRESE, R. L. Slow oscillations of membrane potential in interneurons that control heartbeat in the medicinal leech. J. Neurosci. 7: 3953-3960, 1987b.[Abstract]
• ARTALEJO, C. R., ADAMS, M. E., FOX, A. P. Three types of calcium channel trigger secretion with different efficacies in chromaffin cells. Nature Lond. 367: 72-76, 1994.[Medline]
• BURROWS, M. Nonspiking and spiking local interneurons in the locust. In: Model Neural Networks and Behavior, edited by and A. I. Selverston . New York: Plenum, 1985, p. 109-125
• BÜSCHGES, A. The role of local nonspiking interneurons in the generation of rhythmic motor activity in the stick insect. J. Neurobiol. 27: 488-512, 1995.[Medline]
• BYERLY, L., HAGIWARA, S. Calcium channel diversity. In: Calcium and Ion Channel Modulation, edited by A. D. Grinnel, D. Armstrong, and M. B. Jackson . New York: Plenum, 1988, p. 3-18
• CALABRESE, R. L., NADIM, F., OLSEN, Ø. H. Heartbeat control in the medicinal leech: a model system for understanding the origin, coordination, and modulation of rhythmic motor patterns. J. Neurobiol. 27: 390-402, 1995.[Medline]
• CHARLTON, M. P., AUGUSTINE, G. J. Classification of presynaptic calcium channels at the squid giant synapse: neither T-, L-, nor N-type. Brain Res. 525: 133-139, 1990.[Medline]
• COHEN, C. J., MCCARTHY, R. T., BARRETT, P. O., RASMUSSEN, H. Ca channels in adrenal glomerulosa cells: K⁺ and angiotensin II increase T-type Ca channel current. Proc. Natl. Acad. Sci. USA 85: 2412-2416, 1988.[Abstract/Free Full Text]
• DEAN, J., CRUSE, H. Motor pattern generation. In: The Handbook of Brain Theory and Neural Networks, edited by and M. A. Arbib . Cambridge, MA: MIT Press, 1995, p. 600-605
• DELCOMYN, F. Neural basis of rhythmic behavior in animals. Science Wash. DC 210: 492-498, 1980.[Abstract/Free Full Text]
• DE SCHUTTER, E., ANGSTADT, J. D., CALABRESE, R. L. A model of graded synaptic transmission for use in dynamic network simulations. J. Neurophysiol. 69: 1225-1235, 1993.[Abstract/Free Full Text]
• DICAPRIO, R. A. Nonspiking interneurons in the ventilatory central pattern generator of the shore crab, Carcinus maenas. J. Comp. Neurol. 285: 83-106, 1989.[Medline]
• ECKERT, R., CHAD, J. E. Inactivation of Ca channels. Prog. Biophys. Mol. Biol. 44: 215-267, 1984.[Medline]
• FOSSIER, P., BAUX, G., TAUC, L. Role of different types of Ca²⁺ channels and a reticulum-like Ca²⁺ pump in neurotransmitter release. J. Physiol. Paris 87: 3-14, 1993.[Medline]
• GRANZOW, B., FRIESEN, W. O., KRISTAN, W. B. JR Physiological and morphological analysis of synaptic transmission between leech motor neurons. J. Neurosci. 5: 2035-2050, 1985.[Abstract]
• GRAUBARD, K., RAPER, J. A., HARTLINE, D. K. Graded synaptic transmission between identified spiking neurons. J. Neurophysiol. 50: 508-521, 1983.[Abstract/Free Full Text]
• GROLLEAU, F., LAPIED, B. Two distinct low-voltage-activated Ca²⁺ currents contribute to the pacemaker mechanism in cockroach dorsal unpaired median neurons. J. Neurophysiol. 76: 936-976, 1996.
• HARRIS-WARRICK, R. M. Pattern generation. Curr. Opin. Neurobiol. 3: 982-988, 1993.[Medline]
• HARTLINE, D. K., GRAUBARD, K. Cellular and synaptic properties in the crustacean stomatogastric nervous system. In: Dynamic Biological Networks: The Stomatogastric Nervous System, edited by R. M. Harris-Warrick, E. Marder, A. I. Selverston, and M. Moulins . Cambridge, MA: MIT Press, 1992, p. 31-85
• HEITLER, W. J., PEARSON, K. G. Non-spiking interactions and local interneurons in the central pattern generator of the crayfish swimmeret system. Brain Res. 187: 206-211, 1980.[Medline]
• HILLE, B. In: Ionic Channels of Excitable Membranes., . Sunderland, MA: Sinauer, 1992
• JACKLET, J. W. (Editor). Neuronal and Cellular Oscillators. New York: Marcel Dekker, 1989.
• JANIS, R. A., TRIGGLE, D. J. Drugs acting on calcium channels. In: Calcium Channels: Their Properties, Functions, Regulation, and Clinical Relevance, edited by L. Hurwitz, L. D. Partridge, and J. K. Leach . Boca Raton, FL: CRC, 1991, p. 195-249
• JOHNSON, B. R., HARRIS-WARRICK, R. M. Aminergic modulation of graded synaptic transmission in the lobster stomatogastric ganglion. J. Neurosci. 10: 2066-2076, 1990.[Abstract]
• JOHNSON, B. R., PECK, J. H., HARRIS-WARRICK, R. M. Temperature sensitivity of graded synaptic transmission in the lobster stomatogastric ganglion. J. Exp. Biol. 156: 267-285, 1991.[Abstract/Free Full Text]
• JOHNSON, B. R., PECK, J. H., HARRIS-WARRICK, R. M. Distributed amine modulation of graded chemical transmission in the pyloric network of the lobster stomatogastric ganglion. J. Neurophysiol. 72: 437-452, 1995.
• KASAI, H., AOSAKI, T., FUKUDA, J. Presynaptic calcium antagonist -conotoxin GVIA irreversibly blocks N-type calcium channels in chick sensory neurons. Neurosci. Res. 4: 228-235, 1987.[Medline]
• KLEINHAUS, A. L., ANGSTADT, J. D. Diversity and modulation of ionic conductances in leech neurons. J. Neurobiol. 27: 419-433, 1995.[Medline]
• MANGAN, P. S., COMETA, A. K., FRIESEN, W. O. Modulation of swimming behavior in the medicinal leech. IV. Serotonin-induced alteration of synaptic interactions between neurons of the swim circuit. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 175: 723-736, 1994.[Medline]
• MENDELSON, M. Oscillator neurons in crustacean ganglia. Science Wash. DC 171: 1170-1173, 1971.[Abstract/Free Full Text]
• NADIM, F., OLSEN, Ø. H., DE SCHUTTER, E., CALABRESE, R. L. Modeling the leech heartbeat elemental oscillator. I. Interactions of intrinsic and synaptic currents. J. Comp. Neurosci. 2: 215-235, 1995.[Medline]
• NICHOLLS, J. G., BAYLOR, D. A. Specific modalities and receptive fields of sensory neurons in the central nervous system of the leech. J. Neurophysiol. 31: 740-756, 1968.[Free Full Text]
• NICHOLLS, J. G., WALLACE, B. G. Modulation of transmission at an inhibitory synapse in the central nervous system of the leech. J. Physiol. Lond. 281: 157-170, 1978a.[Abstract/Free Full Text]
• NICHOLLS, J. G., WALLACE, B. G. Quantal analysis of transmitter release at an inhibitory synapse in the CNS of the leech. J. Physiol. Lond. 281: 171-185, 1978b.[Abstract/Free Full Text]
• OLSEN, Ø. H., NADIM, F., CALABRESE, R. L. Modeling the leech heartbeat elemental oscillator. II. Exploring the Parameter Space. J. Comp. Neurosci. 2: 237-257, 1995.[Medline]
• OLSEN, Ø. H., CALABRESE, R. L. Activation of intrinsic and synaptic currents in leech heart interneurons by realistic waveforms. J. Neurosci. 16: 4958-4970, 1996.[Abstract/Free Full Text]
• PAUL, D. H., MULLONEY, B. Local interneurons in the swimmeret system of the crayfish. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 156: 489-502, 1985a.
• PAUL, D. H., MULLONEY, B. Nonspiking local interneuron in the motor pattern generator for the crayfish swimmeret. J. Neurophysiol. 54: 28-39, 1985b.[Abstract/Free Full Text]
• PEARSON, K. G., FOURTNER, C. R. Nonspiking interneurons in the walking system of the cockroach. J. Neurophysiol. 38: 33-52, 1975.[Abstract/Free Full Text]
• ROSSIGNOL, S., DUBUC, R. Spinal pattern generation. Curr. Opin. Neurobiol. 4: 894-902, 1994.[Medline]
• RAPER, J. A. Nonimpulse mediated synaptic transmission during the generation of a cyclic motor program. Science Wash. DC 205: 304-306, 1979.[Abstract/Free Full Text]
• SCHMIDT, J., CALABRESE, R. L. Evidence that acetylcholine is an inhibitory transmitter of heart interneurons in the leech. J. Exp. Biol. 171: 339-347, 1992.
• SCHWEITZ, H., HEURTEAUX, C., BOIS, P., MOINIER, D., ROMEY, G., LAZDUNSKI, M. Calcicludine, a venom peptide of the Kunitz-type protease inhibitor family, is a potent blocker of high-threshold Ca²⁺ channels with a high affinity for L-type channels in cerebellar granule neurons. Proc. Natl. Acad. Sci. USA 91: 878-882, 1994.[Abstract/Free Full Text]
• SIEGLER, M. V. S. Nonspiking interneurons and motor control in insects. Adv. Insect Physiol. 18: 249-304, 1985.
• SIMON, T. W., OPDYKE, C. A., CALABRESE, R. L. Modulatory effects of FMRFamide on outward currents and oscillatory activity in heart interneurons of the medicinal leech. J. Neurosci. 12: 525-537, 1992.[Abstract]
• SIMON, T. W., SCHMIDT, J., CALABRESE, R. L. Modulation of high-threshold transmission between heart interneurons of the medicinal leech by FMRF-NH₂. J. Neurophysiol. 71: 454-466, 1994.[Abstract/Free Full Text]
• THOMPSON, K. T., CALABRESE, R. L. FMRFAMIDE. Effects on membrane properties of Heart cells isolated from the leech, Hirudo medicinalis. J. Neurophysiol. 67: 280-291, 1992.[Abstract/Free Full Text]
• TRUDEAU, L.-E., BAUX, G., FOSSIER, P., TAUC, L. Transmitter release and calcium currents at an Aplysia buccal ganglion synapse. I. Characterization. Neuroscience 53: 571-580, 1993.[Medline]
• TSIEN, R. W., LIPSCOMBE, D., MADISON, D. V., BLEY, K. R., FOX, A. P. Multiple types of neuronal calcium channels and their selective modulation. Trends Neurosci. 111: 431-437, 1988.
• WANG, X., TREISTMAN, S. N., LEMOS, J. R. Single channel recordings of N_t- and L-type Ca²⁺ currents in rat neurohypophysial terminals. J. Neurophysiol. 70: 1617-1628, 1993.[Abstract/Free Full Text]
• WOLF, H., BÜSCHGES, A. Nonspiking local interneurons in insect leg motor control. II. Role of nonspiking local interneurons in the control of leg swing during walking. J. Neurophysiol. 73: 1861-1875, 1995.[Abstract/Free Full Text]

0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society