HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 96/1/235    most recent 01094.2005v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Ivanov, A. I. Articles by Calabrese, R. L. Search for Related Content PubMed PubMed Citation Articles by Ivanov, A. I. Articles by Calabrese, R. L.

Spike-Mediated and Graded Inhibitory Synaptic Transmission Between Leech Interneurons: Evidence for Shared Release Sites

Department of Biology, Emory University, Atlanta, Georgia

Submitted 17 October 2005; accepted in final form 29 March 2006

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Inhibitory synaptic transmission between leech heart interneurons consist of two components: graded, gated by Ca²⁺ entering by low-threshold [low-voltage–activated (LVA)] Ca channels and spike-mediated, gated by Ca²⁺ entering by high-threshold [high-voltage–activated (HVA)] Ca channels. Changes in presynaptic background Ca²⁺ produced by Ca²⁺ influx through LVA channels modulate spike-mediated transmission, suggesting LVA channels have access to release sites controlled by HVA channels. Here we explore whether spike-mediated and graded transmission can use the same release sites and thus how Ca²⁺ influx by HVA and LVA Ca channels might interact to evoke neurotransmitter release. We recorded pre- and postsynaptic currents from voltage-clamped heart interneurons bathed in 0 mM Na⁺/5 mM Ca²⁺ saline. Using different stimulating paradigms and inorganic Ca channel blockers, we show that strong graded synaptic transmission can occlude high-threshold/spike-mediated synaptic transmission when evoked simultaneously. Suppression of LVA Ca currents diminishes graded release and concomitantly increases the ability of Ca²⁺ entering by HVA channels to release transmitter. Uncaging of Ca chelator corroborates that graded release occludes spike-mediated transmission. Our results indicate that both graded and spike-mediated synaptic transmission depend on the same readily releasable pool of synaptic vesicles. Thus Ca²⁺, entering cells through different Ca channels (LVA and HVA), acts to gate release of the same synaptic vesicles. The data argue for a closer location of HVA Ca channels to release sites than LVA Ca channels. The results are summarized in a conceptual model of a heart interneuron release site.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
At many synapses more than one kind of Ca channel is involved in neurotransmitter release and clusters of different Ca channel types may overlap and cooperate to trigger fusion (Meinrenken et al. 2003; Mintz et al. 1995; Reid et al. 1998; Wu and Shaggau 1994, 1997). On other hand, Ca²⁺ entering the cell through different types of Ca channels may have different effects on synaptic transmission. For example, in rat calyx-type synapses, P/Q-type Ca channels trigger release more effectively than N- and R-type Ca channels, which appear to be located more distantly from release sites (Wu et al. 1998, 1999). Mouse cortical CA3–CA1 synapses appear to be similarly organized (Qian and Noebels 2001).

Ca channel cooperativity in neurotransmitter release depends on the kind of Ca channels involved and the location of these channels with respect to active zones (Meinrenken et al. 2002, 2003; but see Gentile and Stanley 2005). Ca channels at calyx of Held synapses are clustered and located at different distances from the synaptic vesicle release trigger in different active zones, leading to variability in the release efficacy from one active zone to another (Meinrenken et al. 2002). Involvement of different Ca channels in neurotransmitter release may change during development (Fedchyshin and Wang 2005). Moreover, distant Ca channels that are not involved in Ca domain(s) that trigger release can nevertheless provide bulk (background) intracellular Ca²⁺ that may modulate neurotransmitter release (Bertram et al. 1999).

Synapses may be either phasic (spike-mediated and fast) or tonic (non-spike-mediated and slow, or graded) and involve synchronous and/or asynchronous release. At the same synapse, transmission may be both phasic (spike-mediated, synchronous) and tonic (graded, asynchronous) (Angstadt and Calabrese 1991; Ayali et al. 1998; Ivanov and Calabrese 2000, 2006; Pan et al. 2001; Warzecha et al. 2003), and in some cases the involvement of low-voltage–activated (LVA) Ca channels in transmission is well documented (Angstadt and Calabrese 1991; Ivanov and Calabrese 2000, 2006; Pan et al. 2001). Transmitter release in these neurons resembles catecholamine secretion in chromaffin cells (Artalejo et al. 1994), where activation of different Ca channels leads to a different mode of neurosecretion, and different Ca channels control catecholamine secretion with different efficacies as a result of their different proximity to release sites. At other neuronal synapses asynchronous (tonic, bulk Ca-dependent) release can often be recorded after spike-evoked (phasic, synchronous) release (Atluri and Regehr 1998) and be in a competitive relation with synchronous release (Otsu et al. 2004).

In leeches, reciprocally inhibitory synaptic transmission between heart interneurons is both spike-mediated [depends on high-voltage–activated (HVA) Ca channels] and graded (depends on two types of LVA Ca channels termed I_CaS and I_CaF) (Angstadt and Calabrese 1991; Ivanov and Calabrese 2000; Lu et al. 1997; Olsen and Calabrese 1996). Previously we showed that LVA Ca channels, which are widely distributed throughout heart interneurons, not only generate the plateau potential that drives the burst of action potential and mediate graded transmission, but also provide intracellular background Ca for modulation of spike-mediated transmission (Angstadt and Calabrese 1991; Arbas and Calabrese 1987; Ivanov and Calabrese 2000, 2003; Lu et al. 1997; Olsen and Calabrese 1996). Moreover, we demonstrated in the companion paper that both I_CaS and I_CaF mediate release from the same release sites (Ivanov and Calabrese 2006). We also showed that manipulating intracellular Ca²⁺ concentration with Ca chelators had a stronger effect on graded synaptic transmission than on spike-mediated transmission and its plasticity, indicating either different active zones for spike-mediated and graded synaptic transmission or a different location of HVA and LVA Ca channels in the same active zones (Ivanov and Calabrese 2003). These findings led us to propose two questions that we address here. Are the vesicle pools and release sites for high-threshold (thus spike-mediated) and low-threshold (graded) transmission shared? To the extent that these vesicle pools and release sites are shared, does Ca²⁺ entering by HVA and LVA Ca channels have equivalent access to the release trigger?

In this study, we recorded pre- and postsynaptic currents from voltage-clamped heart interneurons bathed in 0 mM Na⁺/5 mM Ca²⁺ saline. Through the use of different stimulating paradigms, inorganic Ca channels blockers, and photo-release of Ca²⁺ and Ca chelator, we were able to show that both graded and spike-mediated synaptic transmission depend on transmitter release from the same release sites using the same releasable pool of synaptic vesicles. Ca²⁺ ions entering cells through different Ca channels (low- and high-threshold) act to gate release the same synaptic vesicles. Our results further argue for a closer association of HVA Ca channels to release sites than LVA Ca channels and are summarized in a conceptual model of a heart interneuron release site.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Animals

Adult leeches (H. medicinalis) were obtained from Leeches USA and Biopharm and maintained in artificial pond water (Leeches USA) at approximately 15°C.

Preparation

Leeches were anesthetized in cold saline, after which individual ganglia (midbody ganglion 3 or 4) were dissected and pinned in clear, Sylgard-coated open bath recording/imaging chamber (RC-26, Warner Instrument) with a working volume of 150 µl. The sheath on the ventral surface of the ganglion was removed with 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 (HEPES) acid buffer, adjusted to pH 7.4 with NaOH or HCl. The preparation was mounted ventral side up on the stage of an Olympus BX50WI fluorescent microscope with an Olympus 40x/0.80W water immersion objective.

Electrophysiology

Heart interneurons were penetrated with thin-walled (1 mm OD, 0.75 mm ID) borosilicate microelectrodes (A-M Systems). and identified by the posterolateral position of their somata on the ventral surface of the ganglion and by their characteristic pattern of rhythmic bursting. In all experiments, the recording microelectrode, inserted into a postsynaptic cell, was filled with 4 M K-acetate, 20 mM KCl (unbuffered, pH 8.4). The "presynaptic" microelectrode was filled with 1 M K-acetate, 1.5 M tetraethyl ammonium acetate (TEA-acetate), and 1.5 Cs-acetate (unbuffered, pH 7.9) to block outward currents. Microelectrodes were coated along their shanks with Sylgard 186 (Dow-Corning) and had resistances of 20–45 M and time constants of 0.5–1.5 ms when capacity compensated.

Once the cells were penetrated with recording microelectrodes, for all experiments, the superfusate was immediately switched to Na⁺-free/5 mM Ca²⁺ saline containing (in mM) 110.0 N-methyl-D-glucamine (NMDG), 4.0 KCl, 5.0 CaCl₂, 10.0 glucose, and 10.0 HEPES acid buffer, adjusted to pH 7.4 with KOH or HCl. In a few of these experiments, Ca²⁺ in the saline was reduced to 2 mM with suitable osmotic adjustment of NMDG to 115.0 mM. In some cases, 150 µM Cd²⁺, 1 mM Ni²⁺, or both were added to the saline.

In all experiments, the activity of the pre- and postsynaptic cell was recorded in voltage-clamp mode. Voltage-clamp recordings were made with an Axoclamp-2A amplifier (Axon Instruments) in single-electrode voltage-clamp (SEVC) mode with a sampling rate of 5 kHz. In each case, the electrode potential was monitored on an oscilloscope to ensure that the potential settled between current injection cycles. All recordings were referenced to a chlorided silver wire used to ground the bath. All electrophysiological data were acquired, digitized, and stored on a Pentium IV (Intel) computer using pCLAMP 7.0/8.0 software with a Digidata 1200 or 1320A interface from Axon Instruments.

All stimulus protocols were 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, or from a different holding potential to a fixed depolarizing potential. Various approaches were used, from single voltage pulses/steps to combined pulses/steps, and to trains of "artificial spikes," which were copied from recordings of spontaneously active heart interneurons, recorded in separate experiments. Software-controlled leak subtraction was implemented as previously described (Ivanov and Calabrese 2003). More details on all stimulus protocols used are provided in RESULTS.

Ca imaging

In some experiments, we monitored changes in intracellular Ca²⁺ with the fluorescent indicator Calcium Orange (Molecular Probes). In these experiments, one cell (presynaptic) was iontophoretically filled with Calcium Orange [see Ivanov and Calabrese (2000, 2003) for details of methods and indicator properties] and then repenetrated after 5–15 min with a recording microelectrode. Changes of Calcium Orange fluorescence were continuously monitored and recorded with an ICCD-350f CCD camera (Video Scope International), connected to the fluorescent microscope mentioned above, equipped with an Olympus U-MNG (exciter filter BP 530–550 nm, dichroic mirror DM 570 nm, barrier filter BA 590 nm) filter set, 10% neutral density filter, and an Olympus 40x/0.80W water immersion objective and Axon Imaging Workbench 4.0 (AIW 4.0) software with a Digidata 2000 interface (Axon Instruments) on a Pentium III (Intel) computer. Intensifier gain and black (baseline) levels were adjusted to achieve minimal background fluorescence, convenient visualization of the filled neuron, and sufficient dynamic range for monitoring fluorescence changes.

Our setup permits the acquisition of full-frame images of 640 x 480 pixels size at a resolution of 0.379 µm² for 1 pixel (395 x 295 µm for full frame) with the Olympus 40x/0.80W water immersion objective. Changes of fluorescence were recorded from the approximate synaptic contact region of a heart interneuron (600–1,200 pixels, 235–470 µm²), as described by Ivanov and Calabrese (2003). In all experiments, the maximal available acquisition rate (video rate, 30 Hz) was used, yielding a time resolution of 33 ms. Video signals were accumulated for 33 ms per image, without any kind of gating, using the camera's DC mode.

To synchronize the acquisition of electrophysiological data and Ca fluorescence recording, the Digidata 2000 and Digidata 1200/1320A were connected using a DIO-3 cable interface (Axon Instruments) that permits one program to trigger the other. In our experiments, we used pCLAMP 7.0/8.0 protocols to trigger data acquisition by Axon Imaging Workbench, which in turn controlled the shutter for the imaging lamp.

UV photolysis of caged Ca²⁺ chelator

In some experiments Ca imaging/electrophysiology was combined with UV photolysis. For experiments of this kind, the optical system was modified and used as described by Ivanov and Calabrese (2003). Briefly, for UV photolysis we used a 100-W mercury lamp, equipped with a UV transmitting fused-silica condenser, an electronic shutter (Oriel Instruments), and a glass UV filter (U-360, Edmunds Industrial Optics). The lamp was connected by a UV transmitting fused-silica fiber (core diameter 1,000 µm, numerical aperture 0.22; Oriel Instruments) to an "ablation laser unit" (Photonic Instruments), attached to the microscope. The location and focusing of the spot of "uncaging" light were adjusted with controls on the ablation laser unit so the spot was centered in the image plane and as close to the estimated center of synaptic contact region as possible.

Diazo-2 (derived from BAPTA, "caged BAPTA"; Molecular Probes, diazo-2, tetrapotassium salt, "cell impermeant," MW 710.86, cat no. D-3034) is a photoactivatable Ca²⁺ scavenger; the nominal K_d of diazo-2 for Ca²⁺ changes on UV illumination from 2.2 µM to about 80 nM (Adams and Tsien 1993; Delaney 2000); diazo-2 photo-release Ca²⁺ chelator on UV illumination was <360 nm.

To fill cells with diazo-2, the same techniques as for filling with Calcium Orange but different microelectrode solutions were used. Filling electrodes for diazo-2, contained (in mM): 5 Calcium Orange, 40 diazo-2, and 40 KOH/HEPES, pH 7.2.

All protocols used for photo-release of caged Ca²⁺ chelator were generated using the pCLAMP program CLAMPEX, which controlled the shutter of the release lamp through the Digidata 1200/1320A connected to the shutter control unit. Typically, Ca²⁺/Ca²⁺ chelator were photo-released for 800–1,600 ms during electrophysiological and Ca fluorescence data acquisition.

Data analysis

All stored data were analyzed on the same computer using the pCLAMP program CLAMPFIT and Origin 7.5 (OriginLab) software. Calcium fluorescence data are presented mainly as changes in fluorescence (F/F), but in some cases as fluorescence (F); in this latter case, the data are presented in units of absolute fluorescence on a scale from 0 to 255 fu (fluorescence units). Statistical analyses used one-way ANOVA, factorial ANOVA, and repeated-measures ANOVA with post hoc comparisons made by Student's t-test with Bonferroni correction for multiple comparisons (Bonferroni test), linear regression with 95% confidence interval, and linear correlation analysis, all performed with Statistica 6 and Origin 7.5 software. Results are presented/plotted as means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
LVA channels are widely distributed throughout heart interneurons and serve to generate plateau potentials, support the formation of bursts of action potential during normal rhythmic activity, mediate graded synaptic transmission, and provide intracellular background Ca²⁺ for modulation of spike-mediated transmission (Angstadt and Calabrese 1991; Arbas and Calabrese 1987; Ivanov and Calabrese 2000, 2003; Lu et al. 1997; Olsen and Calabrese 1996). Spike-mediated transmission, on the other hand, is produced by HVA Ca channels of unknown distribution (Ivanov and Calabrese 2000, 2003; Simon et al. 1994). These findings left unanswered two questions addressed in the present study. Are the vesicle pools and release sites for high-threshold (thus spike-mediated) and low-threshold (graded) transmission shared? To the extent that these vesicle pools and release sites are shared, does Ca²⁺ entering by HVA and LVA Ca channels have equivalent access to the release trigger?

Do spike-mediated and graded synaptic transmission deplete the same pool of readily releasable vesicles in heart interneurons?

To determine whether vesicle pools and thus release sites are shared, we performed experiments designed to test whether depletion of releasable transmitter by low-threshold stimulation occluded release by high-threshold stimulation. In all experiments presented in this paper, ganglia were bathed in 5 mM Ca²⁺/0 mM Na⁺ saline, and both pre- and postsynaptic heart interneurons were voltage clamped with sharp microelectrodes. To test for occlusion (Fig. 1, n = 7), we elicited low-threshold (graded) transmission using depolarizing steps from a holding potential of –70 to –40 mV that evoked the two kinetically distinct LVA Ca currents presynaptically (I_CaF and I_CaS) (Angstadt and Calabrese 1991) and the corresponding postsynaptic responses (gIPSC_F and gIPSCs) (Fig. 1C). We also elicited high-threshold transmission, using a train of "artificial spikes" at 2.5 Hz from a holding potential of –40 mV that evoked HVA Ca current presynaptically and the corresponding postsynaptic (spike-mediated) responses (smIPSCs). The train of artificial spikes evoked robust smIPSCs with typical (compare with Ivanov and Calabrese 2003) stochastic variations in their peak amplitude (Fig. 1A). Artificial spikes elicited during a 2-s depolarizing step to –40 mV from a holding potential of –70 mV, however, failed to elicit smIPSCs during the intense graded transmission early in the step, but as the gIPSC waned (concomitantly with the inactivation of LVA Ca current), they began to elicit smIPSCs comparable in amplitude to those elicited from –40 mV (Fig. 1B). The "absolute" peak amplitude of these IPSCs (i.e., amplitude relative to holding current; see Fig. 1D) was almost constant (with some weak tendency to increase during the depolarizing step) and did not exceed the peak value of gIPSC [gIPSC (P)] (Fig. 1, A2 and B2). The peak amplitude of smIPSCs (i.e., amplitude relative to the baseline postsynaptic current recorded just before the onset of any given smIPSC; see Fig. 1D), however, increased concomitantly with the decrease in the gIPSC (P). Although changes in peak amplitudes of smIPSCs elicited from holding potential of –40 mV and "absolute" peak amplitude of IPSCs elicited during the depolarizing step to –40 mV did not depend on the time from the beginning of train/depolarization (P = 1.0 and 0.99989, respectively, one-way ANOVA), changes in peak amplitude of smIPSCs elicited during the depolarization to –40 mV and in gIPSC were strongly time dependent (P values of 0.000013 and 0.001, respectively, one-way ANOVA).

View larger version (19K): [in this window] [in a new window]   FIG. 1. Spike-mediated IPSCs (smIPSCs) are occluded by strong graded IPSCs (gIPSCs). A1: smIPSCs, evoked by a train of artificial spikes, applied from presynaptic holding potential of –40 mV. A2: data, as presented in A1, averaged over 7 similar experiments (different preparations) are plotted vs. the time. B1: smIPSCs and gIPSC, evoked by a train of artificial spikes generated during a depolarizing step to –40 mV; B2: data as presented in B1, averaged over 7 similar experiments (different preparations) are plotted vs. the time. In A2 and B2: black filled circles: peak amplitude of smIPSCs [smIPSCs (P)]; gray filled diamonds: absolute peak amplitude of IPSCs [absolute IPSCs (P)]; empty boxes: amplitude of gIPSCs evoked by depolarization to –40 mV measured immediately before presynaptic application of each artificial spike. C: typical example of a gIPSC evoked by step depolarization to –40 mV, for comparison. A1, B1, C1: recordings from different preparations. D: method used to calculate amplitudes of IPSCs for A2 and B2. Absolute peak amplitude of IPSC (P) = current at the peak of a given smIPSC (holding current); amplitude of smIPSC (P) = current at the peak of a given smIPSC (postsynaptic current just before the onset of the smIPSC); amplitude of gIPSC = current just before the onset of a given smIPSC (holding current). Here and in subsequent figures the following abbreviations are used: Pre: presynaptic cell, the cell that was stimulated and was thus functionally presynaptic. Post: postsynaptic cell, the opposite heart interneuron, where postsynaptic responses to the presynaptic cell stimulation were recorded. HVA Ca current: high-voltage–activated Ca current (high-threshold Ca current). LVA Ca current: low-voltage–activated Ca current (low-threshold Ca current). ICa: presynaptic Ca current; ICaF, ICaS, and ICaHT: fast and slow LVA calcium currents, and HVA calcium current, respectively. IPSC: inhibitory postsynaptic current; gIPSC and smIPSC: graded and spike-mediated inhibitory postsynaptic current; gIPSC and htIPSC: inhibitory postsynaptic current, evoked by ICaF + ICaS, and ICaHT, respectively. (P): peak signal; thus smIPSC (P) means peak spike-mediated postsynaptic current.

View larger version (16K): [in this window] [in a new window]   FIG. 5. Effects of 1 mM Ni2+ and 150 µM Cd2+ on postsynaptic responses to a train of brief depolarizing pulses to 10 mV superimposed on 2-s depolarization to –40 mV from presynaptic holding potential of –70 mV. Compared with Control saline (A), in 1 mM Ni2+-containing saline (B), LVA ICa and the corresponding gIPSC evoked by depolarization to –40 mV were substantially blocked, and the amplitudes of htIPSCs, elicited by a superimposed train of brief depolarizing pulses to 10 mV, were increased (relief of occlusion, see text). 150 µM Cd2+ and 1 mM Ni2+-containing saline blocked htIPSCs with a progressive decrease in ICa and gIPSC resulting from a lengthened exposure to 1 mM Ni2+ (C). D: gIPSC (P) and htIPSC (P), as illustrated in A, B, and C, averaged over 3 similar experiments (different preparations). In D data recorded in Control saline are presented in black, data recorded in 1 mM Ni2+-containing saline in white, and data recorded in 1 mM Ni2+ and 150 µM Cd2+-containing saline in gray. Black asterisk indicates gIPSC (P) recorded in 1 mM Ni2+-containing saline significantly different from that recorded in Control saline, and filled gray circle indicates gIPSC (P) recorded in 1 mM Ni2+ and 150 µM Cd2+-containing saline significantly different from that recorded in Control saline.

As a next step (Fig. 2, n = 3 different preparations), we used Ni²⁺ (1 mM), a blocker of LVA Ca currents in heart interneurons (Ivanov and Calabrese 2006), to selectively block low-threshold transmission and determine whether this relieved the above-observed occlusion of high-threshold transmission. We kept both the presynaptic and the postsynaptic heart interneurons voltage clamped. The presynaptic cell was held at –70 mV and a depolarizing step protocol, first to –40 mV for 5 s (to activate LVA Ca channels) and then to +10 mV for 1 s (to activate HVA Ca channels), was applied. Additionally, two brief pulses of 10 ms to +10 mV were applied during depolarizing step to –40 mV. The postsynaptic cell was held at –35 mV.

View larger version (8K): [in this window] [in a new window]   FIG. 2. High-threshold (depolarization-evoked) transmission (htIPSC) is occluded by strong low-threshold transmission (gIPSC) and blockade of LVA ICa, and thus low-threshold transmission with 1 mM Ni2+, relieves this occlusion. High-threshold depolarizations (to 10 mV) were applied during a 5-s depolarizing step to –40 mV from holding potential of –70 mV. A: presynaptic ICa and IPSCs recorded in Control saline (5 mM Ca2+/0 mM Na+). B: presynaptic ICa and IPSCs recorded 4 min after beginning superfusion with 1 mM Ni2+-containing saline. C: ICaF (P), gIPSC (P), and htIPSC (P), as illustrated in A and B, averaged over three similar experiments (different preparations). C: data recorded in Control saline are presented in black and data recorded in the 1 mM Ni2+-containing saline in white. Asterisks indicate that ICaF (P), gIPSC (P), and htIPSC (P) recorded in 1 mM Ni2+-containing saline are significantly different from those recorded in Control saline. Here and in subsequent figures, see the text for the probability value P.

For the next experiments (Fig. 3, n = 5, different preparations), we slightly modified the protocol to demonstrate that release by HVA and LVA Ca currents can occur independently, if they are separated temporally. Both the presynaptic and postsynaptic heart interneurons were voltage clamped. The presynaptic cell was held at –70 mV and the postsynaptic cell was held at –35 mV. The time between the beginning of the low-threshold depolarization and the high-threshold depolarization was increased to 12 s from 5 s in the previous protocol. The duration of high-threshold depolarization was 2 s, and the brief high-threshold depolarizations were excluded. Under these conditions (n = 5), whereas LVA Ca currents [I_CaF (P) in control: –1.109 (SD 0.434) nA, in 1 mM Ni²⁺-containing saline: –0.512 (SD 0.231) nA, P = 0.000332, one-way ANOVA, Bonferroni test] and corresponding gIPSCs [gIPSC (P), as above: 0.874 (SD 0.310) nA and 0.364 (SD 0.365) nA, P = 0.0014, one-way ANOVA, Bonferroni test] were significantly blocked by 1 mM Ni²⁺ as in previous experiments, the response to high-threshold depolarization remained unchanged (Fig. 3, A and B, compare with Fig. 2). Ni²⁺ (1 mM) blocked LVA Ca currents and the corresponding gIPSCs, but htIPSCs were not affected [htIPSC (P), as above: 0.582 (SD 0.250) nA and 0.572 (SD 0.379) nA, P = 1.0, one-way ANOVA, Bonferroni test] (Fig. 3C).

View larger version (9K): [in this window] [in a new window]   FIG. 3. Occlusion of high-threshold transmission (htIPSC) by strong low-threshold transmission (gIPSC) is relieved if low-threshold depolarization, which inactivates LVA ICa, is prolonged, indicating that with low-threshold release terminated for a sufficient period a shared vesicle pool can be restored. Under these conditions, 1 mM Ni2+ does not further relieve occlusion of htIPSCs by strong low-threshold transmission (gIPSC). A: in Control saline, a prolonged depolarizing step to –40 mV from holding potential of –70 mV for 12 s and subsequently to 10 mV for 2 s evokes a typical biphasic postsynaptic response with the low-threshold depolarization-evoked gIPSC decaying completely by the 4th second after the beginning of depolarizing step. B: 1 mM Ni2+ strongly reduced LVA ICa and the corresponding gIPSC, but the response to high-threshold depolarization (htIPSC) remained unchanged. C: ICaF (P), gIPSC (P), and htIPSC (P), as illustrated in A and B, averaged over 5 similar experiments (different preparations). In C data recorded in Control saline are presented in black; data recorded in 1 mM Ni2+-containing saline, in white. LVA ICa current and the corresponding gIPSP in response to low- and high-threshold depolarization in 1 mM Ni2+-containing saline were decreased compared with Control saline, whereas htIPSP was not changed. Asterisks indicate that ICaF (P) and gIPSC (P) recorded in 1 mM Ni2+-containing saline were significantly different from those recorded in Control saline.

The experiments illustrated in Fig. 4 (n = 5 different preparations) show that the changes in postsynaptic responses to high-threshold depolarization during low-threshold depolarization caused by 1 mM Ni²⁺ were not the result of some additional activation of LVA Ca channels but solely the result of Ca²⁺ entering the cell by HVA Ca channels. We used 1 mM Ni²⁺ and 150 µM Cd²⁺ to separate LVA and HVA Ca currents (see Figs. 10 and 11 in Ivanov and Calabrese 2006). Both presynaptic and postsynaptic heart interneurons were voltage clamped. The presynaptic cell was held at –70 mV and complex depolarizing step to –40 mV for 8 s and to +10 mV for 2 s was applied. The postsynaptic cell was held at –35 mV. In the presence of 1 mM Ni²⁺, the postsynaptic response to the high-threshold depolarization (which followed 8-s low-threshold depolarization) was significantly increased (Fig. 4B; compare with Fig. 4A), but subsequent addition of 150 µM Cd²⁺ that selectively blocks HVA Ca current (Ivanov and Calabrese 2000; Lu et al. 1997) significantly blocked this response [Fig. 4D: htIPSC (P) in control: 0.737 (SD 0.083) nA, in 1 mM Ni²⁺-containing saline: 1.322 (SD 0.304) nA (P = 0.016746, one-way ANOVA, Bonferroni test], in 1 mM Ni²⁺ and 150 µM Cd²⁺-containing saline: 0.386 (SD 0.356) nA (P = 0.000491, compared with 1 mM Ni²⁺-containing saline, one-way ANOVA, Bonferroni test). The peak amplitude of htIPSCs was significantly influenced by the experimental conditions used (control, 1 mM Ni²⁺-containing saline, 1 mM Ni²⁺ and 150 µM Cd²⁺-containing saline) (P = 0.0000001, one-way ANOVA). htIPSC time-to-peak in 1 mM Ni²⁺ compared with 150 µM Cd²⁺-containing saline was significantly increased [htIPSC (P) time-to-peak in control: 103.80 (SD 36.03) ms, in the saline containing 1 mM Ni²⁺ and 150 µM Cd²⁺: 698.40 (SD 435.73) ms, P = 0.030711, one-way ANOVA, Bonferroni test].

View larger version (9K): [in this window] [in a new window]   FIG. 4. Using the stimulus paradigm of Fig. 2, 150 µM Cd2+ blocks the Ni2+-augmented htIPSCs (relief of occlusion by low-threshold transmission), evoked by 2-s depolarization to 0 mV at the end of 8-s depolarization to –40 mV from holding potential of –70 mV. Compared with Control saline (A), 1 mM Ni2+-containing saline (B) blocked both ICa and the corresponding gIPSC evoked by low-threshold depolarization and increased the htIPSC after high-threshold depolarization. C: in 1 mM Ni2+ and 150 µM Cd2+-containing saline, the htIPSP was substantially blocked. D: ICaF (P), gIPSC (P), and htIPSC (P), as illustrated in A, B, and C, respectively, averaged over 5 similar experiments (different preparations). In D data recorded in Control saline are presented in black, data recorded in 1 mM Ni2+-containing saline in white, and data recorded in 1 mM Ni2+ and 150 µM Cd2+-containing saline in gray. Black asterisk indicates htIPSC (P) recorded in 1 mM Ni2+-containing saline significantly different from that recorded in Control saline, and gray filled circle indicates htIPSC (P) recorded in 1 mM Ni2+ and 150 µM Cd2+-containing saline was significantly different from that recorded in 1 mM Ni2+-containing saline.

View larger version (25K): [in this window] [in a new window]   FIG. 10. Effect of brief high-threshold and low-threshold presynaptic depolarizations of different duration on ICa and corresponding IPSCs. Presynaptic cell was depolarized from holding potential of –70 to –30 mV and to 10 mV with brief pulses of 5- to 50-ms duration, with a time increment of 5 ms. A1–A4: recordings of ICa and IPSCs evoked by presynaptic depolarizations of 5, 15, 25, and 35 ms, respectively. Recordings in response to depolarization to –30 mV are in blue and recordings in response to depolarization to 10 mV are in red. B: LVA ICa, htIPSCs, and gIPSCs evoked by presynaptic depolarizations of 5 to 50 ms, as partially illustrated in A, averaged over 8 similar experiments (different preparations) vs. duration of the applied depolarization. B1: IPSCs, htIPSCs (red), and gIPSCs (blue). B2: ICaF (P) (black) recorded in response to depolarizations to –30 mV.

View larger version (28K): [in this window] [in a new window]   FIG. 11. Schematic of proposed organization of a release site at synaptic connections between leech heart interneurons. A HVA Ca channel (or a small cluster of such channels) is localized in very close (nearest) proximity to a docked fusion competent vesicle and release trigger (10–40 nm). Associated with the vesicle-trigger complex is an enhancer site responsible for homosynaptic enhancement of spike-mediated transmission by background Ca2+ (see Ivanov and Calabrese 2003). HVA channel(s) provides enough Ca2+ to activate the release trigger but not sufficient to significantly activate the enhancer. Clusters of fast LVA Ca channels (ICaF channels) are localized more distant (up to several hundred nanometers) from the release trigger and cooperate to produce Ca2+ microdomains that provide enough Ca2+ to activate the release trigger and to activate strongly the enhancer. Widely distributed around the release site, nonclustered slow LVA Ca channels (ICaS channels) evoke release by radial diffusion of Ca2+ from distant Ca channels to release trigger, and thus much less effectively than fast low-threshold Ca channels. They also provide background Ca2+ that can activate the enhancer. Circle filled with red-to-blue color gradient is a fusion-competent synaptic vesicle. Yellow hexagon is the release trigger. Empty triangle is the enhancer. Light gray lines are the presynaptic membrane, and bright red, bright blue, and bright green doubled elongated ovals, respectively, are an HVA Ca channel, ICaF channels, and ICaS channels (shown in side view). Solid and dashed lines of the color corresponding to the Ca channel(s) are isolines of Ca2+ concentration triggering and enhancing release, respectively.

Effects of uncaging of caged Ca²⁺ chelator on the interaction between spike-mediated and graded synaptic transmission between heart interneurons

To understand further the interaction between LVA and HVA Ca currents in mediating synaptic transmission between heart interneurons, we used photo-uncaging of caged Ca²⁺ chelator (diazo-2). We reasoned that by reducing/blocking graded transmission by uncaging Ca²⁺ chelator, we should augment the response to high-threshold depolarization during concurrent low threshold and high-threshold depolarization.

In these experiments (n = 5 different preparations) (Fig. 6), the presynaptic cell was filled with diazo-2 (caged Ca²⁺ chelator) and Calcium Orange (to monitor changes in intracellular Ca²⁺ concentration) and held at –70 mV, whereas a train of artificial spikes was superimposed on a 5-s depolarizing step to –40 mV. The stimulus protocol evoked a robust gIPSC and greatly diminished htIPSCs that recovered somewhat at the end of the step (Fig. 6A). Diazo-2 itself is a weak Ca chelator (nominal K_d = 2.2 µM) (Adams and Tsien 1993; Delaney 2000), and thus intracellular injection of diazo-2 changed LVA Ca currents, apparently by slowing the inactivation of I_CaF, similarly to intracellularly injected BAPTA (see Ivanov and Calabrese 2003; Lu et al. 1997). Each of two subsequent releases of Ca chelator (Fig. 6, B and C) led to suppression of the gIPSC with a simultaneous increase in htIPSCs amplitudes. Two minutes after the last light flash, the gIPSC was strongly suppressed, whereas htIPSCs amplitudes were strongly increased (Fig. 6D). To quantify the effects of Ca²⁺ chelator uncaging on gIPSCs and htIPSCs, we averaged the amplitudes of gIPSCs (P) and htIPSCs (P) evoked before the first and after the last Ca²⁺ chelator uncaging (n = 1–3 uncagings in each preparation) over n = 5 different preparations (Fig. 6E). After Ca²⁺ chelator uncaging, the gIPSC (P) amplitudes were significantly smaller and htIPSC (P) amplitudes were significantly larger (P = 0.040792 and P = 0.0000001, respectively; repeated-measures ANOVA). The simultaneous increase in strength of spike-mediated (high-threshold) transmission and suppression of graded transmission by uncaged Ca²⁺ chelator further supports the existence of a common vesicle pool and release sites for both spike-mediated and graded synaptic transmission. The inability of diazo-2 (both caged and uncaged) to block htIPSCs argues that HVA Ca channels are closer to release sites than LVA Ca channels.

View larger version (21K): [in this window] [in a new window]   FIG. 6. Effects of uncaging (photo-release) of caged Ca2+ chelator (DM-Nitrophen) on spike-mediated (high-threshold) and graded synaptic transmission between heart interneurons. htIPSCs were elicited by a train of artificial spikes superimposed on a 5-s step depolarization to –40 mV from a holding potential of –70 mV that elicited a gIPSP (simulated burst), and the effect of uncaging of caged Ca2+ chelator on synaptic transmission was assessed. All recordings (A–D) are from the same preparation. A1: with no uncaging of caged Ca2+ chelator, evoked htIPSCs were occluded by strong graded transmission (gIPSP) but gradually began to recover toward the end of the simulated burst. B1: first uncaging of Ca2+ chelator prematurely terminated the gIPSP and promoted recovery of the htIPSPs. C1: in this run, the previous uncaging of Ca2+ chelator continued to suppress the gIPSP and relieved some occlusion of the htIPSPs. Second uncaging of caged Ca2+ chelator further promoted recovery of the htIPSPs. D1: in this run, no further uncaging of caged Ca2+ chelator was performed, but the 2 previous uncagings of Ca2+ chelator continued to suppress the gIPSP and substantially relieved occlusion of the htIPSPs. A2–D2: plots of htIPSCs (P) (empty circles) and gIPSCs (black line), extracted from A1–D1. Gray bar indicates the time and duration of the uncaging light flash. E: amplitudes of htIPSCs (P) before and after Ca2+ chelator uncaging (as illustrated in A1 and D1), averaged over 5 similar experiments (different preparations). htIPSCs recorded before Ca2+ chelator uncaging are shown in black; htIPSCs recorded after Ca2+ chelator uncaging are shown in white.

Incremental depolarization of presynaptic holding potential leads to inactivation of both LVA Ca currents evoked by low-threshold depolarization, and the corresponding gIPSCs are attributed to a progressive decrease in LVA Ca channel–mediated Ca²⁺ influx (Ivanov and Calabrese 2005). Because of an increase in presynaptic intracellular background Ca²⁺ associated with a more depolarized holding potential, however, the mean amplitude of htIPSCs, evoked from progressively depolarizing presynaptic holding potential increases (Ivanov and Calabrese 2003). In these experiments, we used presynaptic holding potential to affect the balance between high- and low threshold transmission (n = 6) (Fig. 7) with the aim of gaining further evidence in support of our hypothesis that LVA and HVA currents evoke release at shared release sites. We used the protocol used in our companion paper (Ivanov and Calabrese 2006) with some modifications. We held postsynaptic cell at –35 mV and applied 2-s depolarizing steps to –40 mV presynaptically from incrementally depolarized holding potentials (from –70 to –45 mV in 5-mV increments). Additionally, 15-ms depolarizing pulses to 10 mV were applied before, during, and after the 2-s depolarizing step.

View larger version (16K): [in this window] [in a new window]   FIG. 7. Effect of progressive incremental depolarization of the presynaptic holding potential on postsynaptic responses, elicited by brief high-threshold depolarizing pulses. Inactivation of LVA ICa prevents occlusion of high-threshold transmission by graded transmission. A1–A6: ICa and IPSCs evoked by 2-s step depolarization to –35 mV from different holding potentials (–45 to –70 mV, respectively) with brief depolarizing pulses to 10 mV applied before, during, and after the step depolarization. B: htIPSC (P) evoked before (htIPSC1, filled black circles) during (htIPSC2, empty circles), and after (htIPSC3, filled gray circles) low-threshold depolarization to –35 mV plotted vs. presynaptic holding potential. C, top: nonbinned nonnormalized htIPSC2 (P) plotted vs. nonbinned nonnormalized peak gIPSCF [gIPSCF (P)]; black line: linear fit. C, bottom: binned averaged nonnormalized htIPSC2 (P) plotted vs. binned averaged nonnormalized peak gIPSCF [gIPSCF (P)]. In B and C data were averaged over 6 similar experiments (different preparations).

These conclusions were further supported in experiments, similar to those illustrated in Fig. 7, in which after control recordings, we applied 1 mM Ni²⁺-containing saline to selectively block LVA Ca currents (Fig. 9) (n = 7). Control recordings were similar to Fig. 7 (Fig. 8A). As expected, 1 mM Ni²⁺ blocked (significantly reduced) both LVA Ca currents and the corresponding gIPSCs at all presynaptic holding potentials (Fig. 8B, 1a and 1b). [For the effect of presynaptic holding potential on I_CaF (P) and gIPSC_Fs (P), respectively, P = 0.005734 and P = 0.0000001, factorial ANOVA; for effect of 1 mM Ni²⁺ on I_CaF (P) and gIPSC_Fs (P), P = 0.0000001 and P = 0.048534, respectively, factorial ANOVA.] Amplitudes of the htIPSC₂s (Fig. 8B2) and the htIPSC₃s were greater in 1 mM Ni²⁺-containing saline than in Control saline at all presynaptic holding potentials tested [for the effect of presynaptic holding potential on htIPSC₂s (P) and htIPSC₃s (P), P = 0.000005 and P = 0.005734, respectively, factorial ANOVA; for the effect of 1 mM Ni²⁺ on htIPSC₂s (P) and htIPSC₃s (P), respectively, P = 0.011216 and P = 0.000676, factorial ANOVA], but at all holding potentials htIPSC₁s were not significantly different from those in Control saline (P = 0.505282, factorial ANOVA) (Fig. 8A). The negative correlations between the binned (according to holding potential) nonnormalized amplitude of both htIPSC₂s and htIPSC₃s with the binned (according to holding potential) nonnormalized gIPSC_Fs (P) in the presence of 1 mM Ni²⁺-containing saline were as strong as those in Control saline. In Control saline, R = –0.997 (P = 0.00282) for htIPSC₂s, and in 1 mM Ni²⁺-containing saline, R = –0.999 (P = 0.0007636). The increase of htIPSCs measured during and after low-threshold depolarization is consistent with a decreased depletion of readily releasable synaptic vesicles as a result of the decrease in low-threshold synaptic transmission brought on by Ni²⁺ blockade of LVA Ca channels.

View larger version (18K): [in this window] [in a new window]   FIG. 9. Effect of extracellular Ca2+ concentration on graded and high-threshold synaptic transmission between heart interneurons. A: IPSCs and fluorescent Ca signal evoked by a train of 15-ms depolarizing pulses to 10 mV from a presynaptic holding potential of –70 mV. B: IPSCs and fluorescent Ca signal evoked by a train of 15-ms depolarizing pulses to 10 mV superimposed on a 5-s depolarizing step to –40 mV from a presynaptic holding potential of –70 mV (simulated burst). A and B, left: recordings were made in 5 mM Ca2+/0 mM Na+ saline (control saline) and on the right in 2 mM Ca2+/0 mM Na+ saline. C: amplitudes of htIPSCs, evoked by a train of 15-ms depolarizing pulses to 10 mV from a presynaptic holding potential of –70 mV. D: amplitudes of htIPSCs evoked by a train of 15-ms depolarizing pulses to 10 mV, superimposed on a 5-s depolarizing step to –40 mV from a presynaptic holding potential of –70 mV. Data were averaged over 6 similar experiments (different preparations). In C and D, data recorded in 5 mM Ca2+/0 mM Na+ saline are shown in black and data recorded in 2 mM Ca2+/0 mM Na+ saline in white.

View larger version (14K): [in this window] [in a new window]   FIG. 8. Effect of progressive incremental depolarization of the presynaptic holding potential and addition of 1 mM Ni2+ on postsynaptic responses elicited by brief high-threshold depolarizing pulses. By blocking LVA ICa, 1 mM Ni2+ amplifies the effect of inactivation of LVA ICa in preventing occlusion of high-threshold transmission by graded transmission. A1–A3: ICa and IPSCs evoked by a 2-s step depolarization to –35 mV from different holding potentials (–45, –50, and –70 mV, respectively) with brief depolarizing pulses to 10 mV applied before, during, and after the step depolarization. Left: recordings in Control saline; right: recordings in 1 mM Ni2+-containing saline. B: data, as illustrated in A, averaged over 7 similar experiments (different preparations) plotted vs. presynaptic holding potential. B1a: ICaF (P). B1b: gIPSCF (P). B2: htIPSC2 (P). Data recorded in Control saline are shown in black and data recorded in 1 mM Ni2+-containing saline in white.

To further elucidate the interaction between high- and low threshold transmission between heart interneurons, we altered the driving force on Ca²⁺ entry into the neurons and compared both high-threshold transmission and a concomitant high- and low-threshold transmission in 5 (control) and 2 (reduced) mM Ca²⁺ saline (n = 6 different preparations). The presynaptic cell was filled with Ca Orange to monitor changes in intracellular Ca²⁺. Recordings were performed first in 2 mM Ca²⁺ saline and then in 5 mM Ca²⁺ saline or vice versa. The presynaptic cells were held at –70 mV and the postsynaptic cells were held at –40 mV. Trains of 15-ms pulses to 10 mV were applied either from the holding potential or during a 5-s depolarization to –40 mV. In 2 mM Ca²⁺ saline, amplitudes of htIPSCs evoked from –70 mV were smaller and the time course to attainment of maximal htIPSC amplitude delayed compared with those in 5 mM Ca²⁺ saline. Amplitudes of htIPSCs in 5 mM Ca²⁺ were significantly larger than those in 2 mM Ca²⁺ saline (P = 0.0000001, repeated-measures ANOVA) (Fig. 9, A and C). The time from the beginning of the first depolarizing pulse to maximal htIPSC in 5 mM Ca²⁺ was significantly shorter than that in 2 mM Ca²⁺ saline: 1,450.00 (SD 219.09) ms and 2,170.00 (SD 219.09) ms, P = 0.000826 (one-way ANOVA, Bonferroni test). The simultaneously recorded presynaptic Ca signal was smaller and its buildup delayed. [The time from the beginning of the first depolarizing pulse to the maximal F/F in 5 mM Ca²⁺ was significantly shorter than that in 2 mM Ca²⁺ saline: 2,450.80 (SD 1,334.62) ms and 3,966.80 (SD 596.16) ms, respectively, P = 0.048987 (one-way ANOVA, Bonferroni test).] In 2 mM Ca²⁺ saline, the LVA Ca currents evoked by the 5-s depolarization to –40 mV and the corresponding gIPSCs were decreased and delayed (Fig. 9B), whereas the presynaptic Ca signals were slightly diminished and strongly delayed (see Fig. 8 in Ivanov and Calabrese 2006). The htIPSCs evoked during the depolarization to –40 mV, however, were increased. Amplitudes of htIPSCs in 5 mM Ca²⁺ saline were significantly smaller than those in 2 mM Ca²⁺ saline (P = 0.000008, repeated-measures ANOVA) (Fig. 9, B and D). Factorial ANOVA indicates a significant cooperative effect (P = 0.0000001) of Ca²⁺ driving force and presynaptic depolarization on htIPSCs amplitudes. Thus although a decrease in Ca driving force reduced Ca²⁺ influx by LVA and presumably also HVA Ca channels, leading to decreased in gIPSCs and htIPSCs evoked from –70 mV, htIPSCs evoked during a depolarizing step to –40 mV (i.e., during low-threshold stimulation) increased. This increase in htIPSCs is consistent with the hypothesis that reduced low-threshold release results in greater availability (decreased depletion) of readily releasable synaptic vesicles for high-threshold release at shared release sites. The decreased depletion of readily releasable synaptic vesicles in 2 mM Ca²⁺ saline uncovers the enhancing effect of Ca²⁺ entering by LVA Ca channels on htIPSCs. In 5 mM Ca²⁺ saline, this effect was masked by severe depletion of the readily releasable pool.

Different latencies of brief high- and low-threshold depolarization-evoked release may reflect a differential localization of HVA and LVA Ca channels at shared release sites

The experiments described above support our main hypothesis that LVA and HVA Ca channels mediate transmitter release at common sites. To probe how these common sites might be organized, we compared the effectiveness of HVA and LVA Ca currents in mediating release. High-threshold depolarizations of different shapes and durations from hyperpolarized presynaptic holding potentials are widely used to mimic action-potential–mediated synaptic transmission (Borst and Sakmann 1999; Fedchyshin and Wang 2005; Gentile and Stanley 2005). We used a similar approach here: the presynaptic cell was depolarized from a holding potential of –70 to –30 mV (low-threshold depolarization) and to 10 mV (high-threshold depolarization) with pulses of increasing duration (5 to 50 ms in 5-ms increments), whereas the postsynaptic cell was held at –35 mV (n = 8 different preparations). Across the different pulse durations, htIPSCs were larger in amplitude (P = 0.0000001, repeated-measures ANOVA) and had on average an 11.87-ms shorter synaptic delay (measured from the beginning of presynaptic depolarization to the beginning of the postsynaptic response) [gIPSC latency was 22.23 (SD 4.29) ms and htIPSC latency was 10.36 (SD 3.53 ms); P = 0.0000001, repeated-measures ANOVA] than the corresponding gIPSCs (Fig. 11A). The latency between the pulse and the first recorded I_CaF was 4.1 (SD 1.8) ms (n = 8 different preparations). Although we had difficulty recording I_CaHT to determine its latency, it is unlikely that the large difference in synaptic delay between gIPSCs and htIPSCs can be accounted for solely by a difference in the latencies of I_CaF and I_CaHT. A small brief inward current was detected in seven of the eight preparations during high-threshold depolarization (data not shown) at a latency of 2.0 (SD 0.2) ms, significantly different from the latency of I_CaF to low-threshold depolarization (one-way ANOVA, P = 0.010686, Bonferroni test). htIPSCs were observed in some preparations in response to the 5-ms pulse and in all preparations in response to the 10-ms pulse, whereas gIPSCs were first recorded only in response to 20-ms pulses, at which point the simultaneously monitored LVA Ca current was nearly at its maximal peak value (90% of peak value for the 50-ms depolarization) (Fig. 10B). Subsequent, 3-min superfusion with 1 mM Ni²⁺-containing saline suppressed LVA Ca current and corresponding IPSCs, but had no effect on the amplitude of high-threshold depolarization-evoked IPSCs (data not shown). These data indicate that very brief high-threshold depolarizations (15 ms) from a hyperpolarized presynaptic holding potential evoke postsynaptic responses that depend mainly or exclusively on Ca²⁺ entry by I_CaHT, whereas depolarizations of longer duration lead to more complex mixed responses with low-threshold Ca currents contributing to release. The consistently shorter synaptic delay of htIPSCs suggests a closer localization of high-threshold Ca channels to the release trigger.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Widely distributed throughout neuritic tree of heart interneurons, LVA Ca channels are responsible for graded transmission and providing background Ca²⁺ for enhancement of spike-mediated transmission (Ivanov and Calabrese 2003), whereas HVA Ca channels are responsible for spike-mediated (high-threshold) synaptic transmission. Ultrastructural morphology suggests that there are 10 to 50 synaptic connections between heart interneurons in ganglia 3 and 4 (Calabrese et al. 1989). Taking into account different possible characteristics of synchronous and asynchronous synapses (Millar et al. 2005; Msghina et al. 1999; Quigley et al. 1999), these multiple synaptic contacts between heart interneurons allow for the possibility that some of them are specialized for graded and some for spike-mediated synaptic transmission. Thus there are two compatible alternatives: 1) HVA and LVA Ca channels share the same release sites or 2) their release sites are different, but some LVA Ca channels are present near (more distantly localized from release trigger) HVA Ca channel-operated release sites. The different degrees of effectiveness of fast and slow LVA Ca channels in mediating graded transmission that indicates different access of these channels to the same release trigger (Ivanov and Calabrese 2006) make the situation potentially more complex. We addressed two primary questions in the present paper: are the vesicle pools and release sites for high-threshold (spike-mediated) and low-threshold (graded) transmission shared and to the extent that these vesicle pools and release sites are shared, do LVA and HVA channels have equivalent access to the release trigger?

Through the use of different stimulating paradigms, inorganic Ca channels blockers, and photo-release Ca²⁺ chelator, we were able to support the hypothesis that both graded synaptic transmission and spike-mediated synaptic transmission use the same release sites by using the same readily releasable pool of synaptic vesicles. Ca²⁺ ions entering cells through different Ca channels (LVA and HVA) act to gate release the same synaptic vesicles. Moreover, our observations on the relative latency of htIPSCs versus gIPSCs further supports the hypothesis that HVA Ca channels are more closely associated than LVA Ca channels with the release trigger. In addition, as argued in Ivanov and Calabrese (2006), the location and distribution of fast and slow LVA Ca channels near release sites are also different.

Mitigating factors

Here we consider some potential pitfalls in our analysis that may mitigate the interpretations of our data. Imperfect space clamp (both pre- and postsynaptic) cannot be completely ruled out, although these voltage-clamp methods have been tested and used extensively in our laboratory (Angstadt and Calabrese 1991; Ivanov and Calabrese 2000, 2003). Imperfect space clamp could lead to different voltage control of different synaptic terminals with subsequent effects on transmitter release and postsynaptic responses. Nevertheless, the data presented (e.g., in Fig. 5) suggest that such potential problems cannot account for the results of our experiments. The relief from occlusion of high-threshold transmission by low-threshold transmission in the presence of 1 mm Ni²⁺, which blocks LVA Ca channels, with an almost uniform increase in amplitude of htIPSCs and a simultaneous decrease in the gIPSC is difficult to explain based on imperfect space clamp. The same may be said of the experiments illustrated in Fig. 6; block of gIPSCs by uncaged Ca²⁺ chelator immediately leads to an increase in amplitudes of htIPSCs.

Another potential concern is an uneven intracellular Ca²⁺ distribution throughout the neuritic tree, particularly release sites. Ivanov and Calabrese (2000) showed that the time course of changes in intracellular Ca fluorescence arising from activation of LVA Ca channels is uniform throughout the entire neuritic tree and synaptic contact region. This finding suggests a wide and more-or-less uniform distribution of LVA Ca channels in the neuritic tree and synaptic contact region, and thus it is likely that LVA Ca channels have access to most or all of the release sites. A Ca signal related to activation of HVA Ca channels could not be recorded with our current methods (Ivanov and Calabrese 2000), suggesting a more strict localization of a relatively small number of HVA channels to release sites and very spatially restricted Ca domains near the mouths of these channels (see also Ivanov and Calabrese 2003).

Postsynaptic effects, such as changes in number and sensitivity of postsynaptic receptors and diffusion and removal of transmitter from the cleft, might influence our recorded responses. Nevertheless, the results presented in the paper, especially the results of experiments that involved the block of HVA and LVA channels (e.g., Figs. 5 and 8), uncaging of caged Ca²⁺ chelator (Fig. 6), and changes in Ca²⁺ driving force (Fig. 9) (as well results of Ivanov and Calabrese 2000, 2003) indicate that the time course of postsynaptic responses is governed by transmitter release (i.e., by presynaptic Ca currents and vesicle availability) and that any potential influence of postsynaptic changes on synaptic transmission on the timescale used here is small.

Association of Ca channels and the release trigger

The structural organization of active zones and thus their functional properties vary from one synaptic connection to another (Msghina et al. 1999; Poage and Meriney 2002; Sätzler et al. 2002; Stanley 1997; Wachman et al. 2004). The spatial relations between Ca channels and the readily releasable synaptic vesicles and their associated release trigger within an active zone determine the release mode and reliability of a given synaptic connection. There are thought to be three main modes of relation between Ca channels and the release trigger (Augustine 2001; see also Neher 1998).

1. Ca nanodomains arise from local diffusion from single open Ca channel located in intimate proximity to a release trigger (few nanometers). Activation of just a single Ca channel can release acetylcholine at calyx-type presynaptic nerve terminal of the chick ciliary ganglion with the distance from Ca channel to release trigger being about 20 nm and the Ca²⁺ concentration near release trigger being about 10 µM (Stanley 1993). Secretion at such a release site is triggered by Ca²⁺ that enters through a discrete, specifically associated cluster of Ca channels. In such a cluster the members nearest the release trigger are the most effective. Such an organization has the advantage of allowing a molecular interaction between the release mechanism and its triggering Ca channels (Gentile and Stanley 2005; Stanley 1993, 1997