Calcium Stores in Hippocampal Synaptic Boutons Mediate Short-Term Plasticity, Store-Operated Ca2+ Entry, and Spontaneous Transmitter Release

Neuron, Vol. 29, 197–208, January, 2001, Copyright 2001 by Cell Press

Calcium Stores in Hippocampal Synaptic Boutons Mediate Short-Term Plasticity, Store-Operated Ca2ⴙ Entry, and Spontaneous Transmitter Release Nigel J. Emptage,*§ Christopher A. Reid,* and Alan Fine*†‡ * Division of Neurophysiology National Institute for Medical Research London NW7 1AA United Kingdom † Department of Physiology and Biophysics Dalhousie University Faculty of Medicine Halifax, Nova Scotia B3H 4H7 Canada

Summary Evoked transmitter release depends upon calcium influx into synaptic boutons, but mechanisms regulating bouton calcium levels and spontaneous transmitter release are obscure. To understand these processes better, we monitored calcium transients in axons and presynaptic terminals of pyramidal neurons in hippocampal slice cultures. Action potentials reliably evoke calcium transients in axons and boutons. Calciuminduced calcium release (CICR) from internal stores contributes to the transients in boutons and to pairedpulse facilitation of EPSPs. Store depletion activates store-operated calcium channels, influencing the frequency of spontaneous transmitter release. Boutons display spontaneous Ca2ⴙ transients; blocking CICR reduces the frequency of these transients and of spontaneous miniature synaptic events. Thus, spontaneous transmitter release is largely calcium mediated, driven by Ca2ⴙ release from internal stores. Bouton store release is important for short-term synaptic plasticity and may also contribute to long-term plasticity. Introduction Stimulus-evoked neurotransmitter release is triggered by the local elevation of Ca2⫹ concentration that results when Ca2⫹ enters the synaptic terminal from the extracellular space through voltage-gated channels in the plasma membrane (Katz and Miledi, 1967; Miledi, 1973). Spontaneous transmitter release, in contrast, can occur in the absence of extracellular Ca2⫹ (Erulkar and Rahamimoff, 1978; Malgaroli and Tsien, 1992), and such release has generally been considered calcium independent. Calcium can, however, be made available from other sources than the extracellular space. Axons and nerve terminals contain intracellular calcium stores, including mitochondria and smooth endoplasmic reticulum– derived organelles (Blaustein et al., 1978a; Henkart et al., 1978), and also contain ryanodine receptors (Sharp et al., 1993) that could mediate calcium-induced calcium release (CICR). These stores have generally been con‡ To whom correspondence should be addressed (e-mail: afine@

nimr.mrc.ac.uk). § Present address: Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, United Kingdom.

sidered to function mainly in calcium sequestration and buffering (Blaustein et al., 1978b), although release of Ca2⫹ from stores has been implicated in the mediation of afterhyperpolarization following action potentials (Sah and McLachlan, 1991) and in generating glutamate receptor–gated postsynaptic calcium transients (Llano et al., 1991; Simpson et al., 1993; Kano et al., 1995; Korkotian and Segal, 1998; Emptage et al., 1999a; but see Mainen et al., 1999; Kovalchuk et al., 2000). Given the crucial role of residual calcium in various forms of shortterm presynaptically mediated synaptic plasticity (Zucker, 1999), the presence of potentially releasable Ca2⫹ stores in synaptic terminals raises important questions about their physiological functions. In this study, we describe the use of fast confocal imaging of fluorescent intracellular calcium indicators in hippocampal slice cultures to monitor Ca2⫹ transients in synaptic boutons and elsewhere along individual axons. We find that action potentials reliably trigger large Ca2⫹ transients in boutons, due both to Ca2⫹ influx and to CICR from internal stores. Depleting the stores or blocking CICR reduces the transients and also reduces a form of short-term synaptic plasticity, paired-pulse facilitation of excitatory postsynaptic potentials (EPSPs). Store depletion also activates store-operated calcium channels, resulting in strong modulation of the frequency of spontaneous transmitter release by extracellular Ca2⫹ concentration ([Ca2⫹]o). In addition, we find that the frequency of spontaneous miniature synaptic events in normal [Ca2⫹]o is reduced by CICR blockade. Together with the observation of spontaneous Ca2⫹ transients in boutons in the absence of action potentials, this suggests that spontaneous transmitter release can result from spontaneous calcium release from internal stores in boutons. The threshold and amplitude of CICR may be modulated by various factors such as the intracellular messenger cyclic ADP ribose and the calcium load of the store (Berridge, 1993; Koizumi et al., 1999a). Releasable calcium stores thus provide a mechanism by which diverse signaling pathways and the functional history of the cell may be integrated to modulate spontaneous and evoked synaptic transmission in the brain. Some of the results presented here have previously appeared in abstract form (Emptage et al., 1999b; Emptage and Fine, 2000). Results Action Potentials Evoke Reliable Calcium Transients in Individual Synaptic Boutons CA3 pyramidal cells in organotypic hippocampal cultures were impaled by sharp microelectrodes, filled with high-affinity Ca2⫹ indicator, and imaged by confocal laser scanning microscopy (Figure 1A). In most cells, a single axon, easily distinguished by its caliber, trajectory, and lack of spines, emerged from the soma or from the proximal segment of a basal dendrite. The main axon typically gave rise to fine branches with numerous small varicosities (arrows in Figure 1B). In preparations co-

Neuron 198

Figure 1. Visualization of Synaptic Boutons along Hippocampal Pyramidal Cell Axons (A) Confocal image of a living CA3 pyramidal neuron in a cultured hippocampal slice preparation, rendered fluorescent by intracellular injection of the Ca2⫹ indicator Oregon green 488 BAPTA-1 via the micropipette visible at top right.

Bouton Calcium Stores 199

Figure 2. Calcium Transients in Hippocampal Synaptic Boutons Three successive line scan images (B–D), 15 s apart, along the indicated trajectory (arrows in [A]) show an abrupt increase in Ca2⫹ signal throughout the axon, coincident with action potentials (black traces below; mV) evoked by intrasomatic depolarizing current injection (depolarization onset “time stamped” as white vertical line). In these “thermal” false-color images, colors from black through red to yellow and white denote increasing fluorescence intensities. Quantification of relative change in fluorescence intensity (red traces; %⌬F/F) at the bouton indicated by fiduciary lines in (A) reveals reliable Ca2⫹ transients associated with action potentials. The actual decay of the Ca2⫹ transient may be faster than the decay of the fluorescence transient, due to buffering by the high-affinity indicator.

injected with biocytin and subsequently labeled for the presynaptic marker synaptophysin, the overlapping pattern of labeling, illustrated in Figure 1C, indicates that these axonal varicosities correspond to synaptic boutons. Rapid repetitive scanning of a single line along the axon, as illustrated in Figure 2, allowed us to determine the time course and amplitude of action potential– evoked Ca2⫹ transients in the boutons. Figures 2B–2D show the responses of one axon segment to action potentials evoked by three successive intrasomatic depolarizing current injections at 15 s intervals. Each action potential, evident in the simultaneous somatic electrical recordings (bottom traces), elicited a Ca2⫹ transient throughout the axon with minimal latency (⬍2 ms) and rapid rise (peaking within 8 ms). The Ca2⫹ levels returned to baseline relatively rapidly; the mean decay time constant, 262 ⫾ 23 ms (n ⫽ 10), must be considered an upper limit because of the high affinity of our calcium indicator (Regehr and Atluri, 1995; Helmchen et al., 1997;

Sinha et al., 1997). The amplitudes of calcium transients evoked by single action potentials were not limited by indicator saturation, since in all cases higher peak amplitudes were evoked by second spikes when longer depolarizing pulses elicited multiple action potentials (e.g., Figures 2C and 2D). Somatic action potentials reliably evoked axonal Ca2⫹ transients, with no failures detected at boutons (n ⫽ 45) separated from the axon initial segment by several branchpoints; reliability was sustained even with multiple action potentials, in agreement with other demonstrations of the absence of branchpoint failure in axons of pyramidal neurons (Cox et al., 2000). Ca2⫹ transient amplitudes at boutons were significantly larger than those along intervening nonvaricose segments of axon (%⌬F/F ⫽ 111.6% ⫾ 6.4% versus 40.5% ⫾ 4.2%; p ⬍ 0.001, n ⫽ 8), but Ca2⫹ transients at branchpoints were not significantly larger than those along intervening axon segments (%⌬F/F ⫽ 56.3% ⫾ 8.1% versus 48.7% ⫾ 11.8%; n ⫽ 4), in contrast to observations on branchpoints of cerebellar basket cell axons (Llano et al., 1997).

(B) Boxed area in (A) (50 ␮m wide) shown at higher magnification; the vertically oriented axon segment bears several varicosities (arrows) and is easily distinguished from spiny dendritic segments (e.g., lower right). (C) Synaptophysin immunoreactivity colocalizes with axon varicosities. Red Cy3-streptavidin fluorescence labels varicosities (middle) along the axon of a cell injected with biocytin, while green Cy2-anti-mouse-IgG labels synaptophysin-immunoreactive structures (left) in the same preparation. The varicosities are yellow in the double-labeled image (right), indicating that they correspond to synaptic boutons. Scale bar, 10 ␮m.

Neuron 200

Adenosine (50 ␮m) reduced Ca2⫹ transients at boutons by 55.3% (p ⬍ 0.01; n ⫽ 4), in contrast to its lack of effect on synaptically evoked Ca2⫹ transients in dendritic spines (Emptage et al., 1999a). Calcium-Induced Calcium Release Contributes to Bouton Calcium Transients The bouton Ca2⫹ transients were blocked by addition of 10 ␮M cadmium to the superfusate (%⌬F/F ⫽ 72.4% ⫾ 13.3% in artificial cerebrospinal fluid [ACSF] versus 5.4% ⫾ 0.9% in cadmium; p ⬍ 0.03, n ⫽ 2), indicating that they result from influx of Ca2⫹ from the extracellular space via voltage-gated Ca2⫹ channels (VGCCs). Earlier indirect evidence (Reyes and Stanton, 1996) and recent demonstrations that glutamate receptor activation and transmembrane Ca2⫹ influx can trigger CICR from internal stores in hippocampal dendritic spines (Korkotian and Segal, 1998; Emptage et al., 1999a) led us to investigate whether a similar process occurs in boutons. Indeed, we found that addition of 20 ␮M ryanodine (a concentration that irreversibly blocks the ryanodine receptors responsible for CICR [McPherson et al., 1991]) to the ACSF reduced the amplitude of action potential– evoked Ca2⫹ transients in boutons (Figures 3A and 3B) by 29% (%⌬F/F ⫽ 95.6% ⫾ 8.0% in ACSF versus 67.7% ⫾ 10.8% in ryanodine; p ⬍ 0.05, n ⫽ 5). Cyclopiazonic acid (CPA; 30 ␮M), which reversibly depletes intracellular Ca2⫹ stores by selectively blocking sarco/ endoplasmic reticulum Ca2⫹/ATPases (SERCA pumps; Seidler et al., 1989), reversibly reduced Ca2⫹ transient amplitudes (Figures 3C and 3D) by a similar 32% (%⌬F/F ⫽ 80.4% ⫾ 14.3% in ACSF versus 55.0% ⫾ 9.1% in CPA; p ⬍ 0.025, n ⫽ 4; and 86.7% ⫾ 26.0% after washout; n ⫽ 3). La3⫹ (400 ␮M) was present in the ACSF throughout this experiment to prevent CPA-induced changes in basal fluorescence (see below; Figure 5). La3⫹ itself had no effect on bouton Ca2⫹ transients (%⌬F/F ⫽ 80.0% ⫾ 12.6; Figure 3D). Thapsigargin (3.8 ␮M), an irreversible SERCA pump inhibitor (Smith and Reed, 1996; Treiman et al., 1998), similarly reduced Ca2⫹ transient amplitudes by 34% (%⌬F/F ⫽ 87.8% ⫾ 17.9% in ACSF versus 57.9% ⫾ 17.9% in thapsigargin; p ⬍ 0.003, n ⫽ 4). CPA and thapsigargin also increased the mean decay time constant of the Ca2⫹ transient from 262 ms in ACSF (see above) to 334.0 ⫾ 53.3 ms in CPA (p ⫽ 0.10, n ⫽ 5) and 421.0 ⫾ 68.3 ms in thapsigargin (p ⬍ 0.015, n ⫽ 4). Thus, internal Ca2⫹ stores are present and functional in synaptic boutons, contributing substantially to the amplitude (via CICR) as well as to the decay (via Ca2⫹ uptake) of action potential–evoked Ca2⫹ transients in the boutons. CICR Contributes to Short-Term Synaptic Plasticity Does this CICR in boutons have physiological consequences? Our previous work (Emptage et al., 1999a) indicated that ryanodine and CPA had no effect on the amplitudes of EPSPs evoked by single stimuli. Transmitter release, however, occurs extremely rapidly and may be finished before CICR has peaked. The released Ca2⫹ could nevertheless persist long enough to influence the probability of transmitter release by a closely following action potential. To test this possibility, we examined

the effects of ryanodine and CPA on the phenomenon of paired-pulse facilitation (PPF), a form of short-term synaptic plasticity in which the second of two closely following EPSPs is larger than the first (Zucker, 1989). PPF is generally believed to reflect an increase in transmitter release probability due to residual Ca2⫹ from the first impulse summing with the Ca2⫹ influx evoked by a second action potential (Wu and Saggau, 1994; Zucker, 1999), although other mechanisms such as prolongation of the presynaptic action potential (Sabatini and Regehr, 1997; Borst and Sakmann, 1998; Cuttle et al., 1998; Qian and Saggau, 1999), facilitation of presynaptic Ca2⫹ currents (Borst and Sakmann, 1998; Cuttle et al., 1998), or transient calcium-dependent phosphorylation of proteins involved in transmitter release (Llinas et al., 1991) may also be involved. It has previously been assumed that such residual Ca2⫹ results from influx through VGCCs. Although neither ryanodine nor CPA significantly affected the amplitude of the first EPSP of a pair, both drugs greatly reduced the enhancement of response to the second stimulus 75 ms later (Figure 4; percent enhancement of the second pulse with respect to the first: 28.8% ⫾ 6.9% in ACSF versus 8.9% ⫾ 4.1% in ryanodine [p ⬍ 0.03; n ⫽ 5] and 12.3% ⫾ 3.9% in CPA [p ⬍ 0.05; n ⫽ 4]), indicating that the crucial residual calcium for PPF is store derived. Store Depletion Opens Store-Operated Channels in Boutons In the course of these experiments, we noticed that depletion of Ca2⫹ stores by 30 ␮M CPA generally increased the basal fluorescence of Ca2⫹ indicator–filled boutons (to 142.1% ⫾ 11.6% of fluorescence in ACSF; n ⫽ 4, p ⬍ 0.01; Figure 5). This increase appeared not to be a simple consequence of leakage of Ca2⫹ from the stores, since it persisted as long as CPA was present. When such preparations were subsequently transferred to EGTA-buffered Ca2⫹-free ACSF with CPA (0 Ca ⫹ CPA), bouton basal fluorescence decreased even below the level in ACSF alone. Transfer from ACSF to 0 Ca2⫹ in the absence of CPA had no effect on basal fluorescence of boutons (fluorescence in 0 Ca2⫹ ⫽ 103.3% ⫾ 8.4% of fluorescence in ACSF; n ⫽ 4, n.s.); this may reflect an intensity floor due to noise or the relatively short (15–20 min) exposure to 0 Ca. Thus, by depleting internal calcium stores, CPA appears to activate store-operated channels (SOCs) (a process also known as capacitative Ca2⫹ entry) that render the bouton membrane permeable to Ca2⫹. Lanthanum (400 ␮M), which is known to block SOCs (Toescu et al., 1998), had no effect on basal fluorescence in ACSF (bouton basal fluorescence after addition of La3⫹ ⫽ 107% ⫾ 12% of control level in ACSF) but completely suppressed the CPA-mediated increase in basal fluorescence after further addition of CPA (70% ⫾ 10% of control; p ⫽ 0.20, n ⫽ 4; Figure 5). SOCs are common in nonexcitable tissue (Parekh and Penner, 1997; Putney and McKay, 1999), and evidence of store-operated Ca2⫹ influx has been obtained in neuroblastoma (Takemura et al., 1991; Mathes and Thompson, 1995), phaeochromocytoma (Clementi et al., 1992; Bennett et al., 1998), and dorsal root ganglion (Usachev and Thayer, 1999) cells, but its presence in synaptic boutons in the brain has not been

Bouton Calcium Stores 201

Figure 3. Calcium-Induced Calcium Release from Internal Stores Contributes to Action Potential–Evoked Calcium Transients in Boutons (A and B) Ca2⫹ fluorescence transients evoked by single action potentials are reduced in the presence of ryanodine (RYAN; 20 ␮M) compared to normal ACSF. (A) Representative traces from one bouton in a single experiment, obtained as in Figure 2. (B) Summary histogram from five experiments. (C and D) Bouton Ca2⫹ transients are similarly and reversibly reduced by addition of cyclopiazonic acid (CPA; 30 ␮M); the effect is most clearly seen when CPA-mediated changes in basal fluorescence are prevented by the presence of La3⫹ (see text and Figure 5). (C) Representative bouton fluorescence traces from a single experiment. (D) Summary histogram from four experiments. La3⫹ alone has no effect on transients.

previously reported, and, indeed, it has been believed not to be functional in hippocampal cells (Shmigol et al., 1994; Koizumi et al., 1999b). Store-Operated Channels Can Modulate the Rate of Spontaneous Transmitter Release Activation of SOCs is normally considered to serve in refilling the depleted Ca2⫹ stores, but it may play other roles as well. The regulation of exocytosis by SOCs has been observed in nonexcitable cells (Parekh and Penner, 1997); recently, SOCs have been implicated in transmitter release from adrenal chromaffin cells (Fomina and Nowycky, 1999) and from inhibitory interneurons in the hippocampus (Savic and Sciancalepore, 1998).

As noted above, the amplitude of single EPSPs was not changed by conditions that activated SOCs, but the dominant control of such evoked transmitter release by voltage-gated Ca2⫹ channels (Augustine et al., 1987) could mask any effect of SOCs. To determine whether SOC activation can influence spontaneous glutamate release in the hippocampus, we monitored miniature excitatory postsynaptic currents (mEPSCs) in voltageclamped CA3 pyramidal neurons in the presence of 2.5 ␮M tetrodotoxin (TTX). In keeping with the previously observed increase in basal Ca2⫹ fluorescence in boutons, the frequency of miniature events was greatly increased (from 1.84 ⫾ 1.46 to 28.17 ⫾ 7.30 events/s; p ⬍ 0.001, n ⫽ 4) by addition of 30 ␮M CPA to the ACSF. In

Neuron 202

Figure 4. CICR Is Responsible for the Majority of Paired-Pulse Facilitation (A) Individual intracellular recordings illustrating that 20 ␮M RYAN (i) and 30 ␮M CPA (ii) have no effect on the amplitude of first EPSPs of a pair but greatly reduce enhancement of the second EPSPs 75 ms later. The effect of CPA is reversed by 45 min washout in ACSF. (B) Summary histogram of reduction in PPF after exposure to ryanodine (n ⫽ 5) or CPA (n ⫽ 4) or after washout from CPA (WASH), compared to normal ACSF.

contrast, whereas transfer of cultures from normal ACSF to 0 Ca2⫹ medium has little effect on mEPSC frequency (1.27 ⫾ 0.27 events/s), addition of 30 ␮M CPA to the 0 Ca2⫹ medium (Figure 6) caused a significant decrease in mEPSC frequency (to 0.36 ⫾ 0.12 events/s); restora-

tion of normal Ca2⫹ to the medium now increased mEPSC frequency (to 10.70 ⫾ 0.20 events/s; p ⬍ 0.001, n ⫽ 4). Thus, SOC activation by Ca2⫹ store depletion greatly enhances the influence of extracellular [Ca2⫹] on spontaneous transmitter release.

Bouton Calcium Stores 203

Figure 5. Calcium Store Depletion Increases Sensitivity of Presynaptic Resting Ca2⫹ Levels to Extracellular Ca2⫹ Mean basal Ca2⫹ indicator fluorescence from boutons in ACSF is increased 10 min after addition of 30 ␮M CPA (ACSF ⫹ CPA) but decreased 15 min after subsequent transfer to calcium-free ACSF with 30 ␮M CPA (0 Ca ⫹ CPA) (n ⫽ 4). The effect of CPA is blocked by the presence of 400 ␮M La3⫹ (ACSF ⫹ La ⫹ CPA) (n ⫽ 4). Neither addition of La3⫹ alone nor transfer to calcium-free ACSF significantly affects basal fluorescence over the time span of these experiments (see text).

Spontaneous Transmitter Release under Normal Conditions Results from Spontaneous CICR Do nondepleted Ca2⫹ stores influence spontaneous release? Application of 30 ␮M ryanodine blocks ryanodine receptor channels in the closed state (McPherson et al., 1991); unlike CPA, ryanodine does not deplete Ca2⫹

stores and does not alter basal Ca2⫹ indicator fluorescence in boutons (Figure 7A). Nevertheless, ryanodine significantly reduces mEPSC frequency (Figures 7B and 7C; 0.97 ⫾ 0.22 events/s in ACSF versus 0.52 ⫾ 0.13 events/s with 30 ␮M ryanodine; p ⬍ 0.03, n ⫽ 6). The mEPSC amplitude frequency distributions were not significantly different with versus without ryanodine (p ⬍ 0.63, Kolmogorov–Smirnov test; data not shown), indicating that the effect of ryanodine on observed mEPSC frequency was not an artifact of changes in detectability. Thus, roughly half of all spontaneous transmitter release under normal conditions appears to be due to spontaneous release of Ca2⫹ from internal stores via ryanodine receptors. If this is the case, it should be possible to visualize the spontaneous Ca2⫹ release as spontaneous Ca2⫹ transients in individual boutons. Indeed, with fast confocal imaging at video rates, Ca2⫹ transients similar to but smaller than those normally evoked by action potentials can be observed, at low frequency, in boutons in the presence of TTX (Figure 7D). The spatial coherence of these changes and the absence of changes in fluorescence elsewhere in the axon indicate that the transients are not artifacts of instrument noise, movement, or dye redistribution. Their isolated occurrence differs from the periodic Ca2⫹ oscillations previously observed in synaptic boutons of neuromuscular junctions (Melamed et al., 1993). Rather, the Ca2⫹ transients we observe have amplitudes and spatial and temporal extents that closely resemble those of spontaneous elementary calcium release events (Ca2⫹ sparks) seen in cardiac muscle cells (Cheng et al., 1993). Very recently, similar Ca2⫹ sparks have been described

Figure 6. The Effect of Calcium Store Depletion on Spontaneous Transmitter Release Is Dependent on Extracellular Ca2⫹ Level Sequential records illustrate that spontaneous mEPSCs in 0 Ca2⫹ (A) occur with reduced frequency 15 min after addition of 30 ␮M CPA (B); restoration of Ca2⫹ to the medium now greatly increases mEPSC frequency (C). (D) Summary histogram of results from four experiments.

Neuron 204

Figure 7. Spontaneous CICR Mediates Spontaneous Transmitter Release (A) Addition of ryanodine (30 ␮M) to the superfusate does not affect basal Ca2⫹ indicator fluorescence in boutons. The frequency of mEPSCs in ACSF ([B]; upper sequential records), however, is reduced by such addition of ryanodine ([B]; lower sequential records). (C) Summary histogram of results from six experiments. (D) High-speed imaging of spontaneous Ca2⫹ transients in a bouton in the presence of 1 ␮M tetrodotoxin to block action potentials. Seventytwo sequential frames of the axon segment, collected at 84 ms intervals, before (i) and after (ii) the addition of 20 ␮M ryanodine. Mean Ca2⫹ indicator fluorescence in the bouton (region outlined by the small square in the first frame of each series) over the duration of the image sequence shown (600 ms) is plotted in red beneath each sequence. Spontaneous Ca2⫹ transients, restricted to the bouton, are abolished by ryanodine. False-color lookup table as in Figure 2.

Bouton Calcium Stores 205

in somata and neurites of dissociated hippocampal neurons and PC12 cells and in presynaptic boutons of lizard neuromuscular junctions (Koizumi et al., 1999b; Melamed-Book et al., 1999). In order to characterize the modulation of these events, their frequency was pharmacologically augmented by application of nicotine (see Experimental Procedures) from ⬍0.01 per s to 0.66 ⫾ 0.24 per s (n ⫽ 4). Under these conditions, addition of 20 ␮M ryanodine significantly reduced Ca2⫹ transient frequency (to 0.02 ⫾ 0.02 events/s; n ⫽ 4, p ⬍ 0.03) at the same time as reducing spontaneous miniature transmitter release events (from 3.12 ⫾ 0.41 to 1.21 ⫾ 0.44 events/s; n ⫽ 4, p ⬍ 0.05), providing further support for the conclusion that the ryanodine-mediated reduction in spontaneous transmitter release is due to a reduction in the frequency of calcium store release events. Discussion Our results demonstrate the presence of functional calcium stores in synaptic boutons in the hippocampus. The stored Ca2⫹ can be released either spontaneously or by action potential–evoked Ca2⫹ influx through VGCCs. The substantial contribution of stores to the evoked Ca2⫹ transient is in contrast with observations on cerebellar parallel fibers (Sabatini and Regehr, 1995); this may reflect differences in the characteristics of the parallel fiber boutons or the fact that bouton-associated signals would only have constituted a small proportion of the cerebellar signals, which were monitored in populations of axons rather than in individual boutons as here. Our observation of the absence of failures in action potential–evoked bouton Ca2⫹ transients is in agreement with observations on boutons of dissociated cultured neurons and of pyramidal cells in brain slice (Mackenzie et al., 1996; Mackenzie and Murphy, 1998; Cox et al., 2000) and indicates that the probabilistic nature of synaptic transmission at synapses made by these boutons must be due to variance in the evoked bouton Ca2⫹ flux or to elements of the transmitter release machinery per se. The precise subcellular identity of the presynaptic Ca2⫹ stores participating in the CICR observed here is not known. Calcium-containing cisternae, presumably extensions of smooth endoplasmic reticulum, have been identified in synaptic boutons (McGraw et al., 1980; Hartter et al., 1987), but possible contribution of other membranous structures such as vesicles (transmitter containing or otherwise) (Hartter et al., 1987) or mitochondria (Blaustein et al., 1980; Jonas et al., 1999) cannot be excluded. Similarly, the SOCs described here are only functionally defined; the molecular identity of the channels responsible for our observed Ca2⫹ fluxes is unknown. However, TRP4 and TRP5, mammalian homologs of the Drosophila transient receptor potential protein (TRP) implicated in capacitative calcium entry (Petersen et al., 1995), are expressed in hippocampal pyramidal neurons (Philipp et al., 1998) and could be responsible for the effects we report. Previous reports have suggested a role for presynaptic CICR in modulating synaptic transmission from motor nerve terminals (Onodera, 1973), photoreceptors (Krizaj et al., 1999), and sympathetic nerve terminals (Peng, 1996; Smith and Cunnane, 1996; but cf. Brain and Ben-

nett, 1997), as well as in certain forms of synaptic plasticity (Lee et al., 1987). Interestingly, caffeine, which may act both as a CICR agonist and as an antagonist of IP3 receptors, has been reported to inhibit hippocampal PPF (Lee et al., 1987). The present results demonstrate that PPF depends upon the functional integrity of internal Ca2⫹ stores; therefore, a major source of residual Ca2⫹ in the bouton is CICR. Although not addressed in our studies, it is therefore likely that CICR contributes to other forms of short-term synaptic enhancement believed to be dependent upon residual presynaptic Ca2⫹, such as facilitation (Katz and Miledi, 1968) and posttetanic potentiation (Erulkar and Rahamimoff, 1978; Zucker, 1989; Frenguelli et al., 1996). The release of calcium from internal stores may also contribute to the slow, asynchronous component of stimulus-associated transmitter release that is exaggerated by replacement of extracellular Ca2⫹ with Sr2⫹ (Dodge et al., 1969; Narita et al., 1998; Chen and Regehr, 1999). Our observations indicate that spontaneous CICR triggers approximately half of all spontaneous transmitter release events under normal conditions. The larger reduction in mEPSC frequency (from 0.97 ⫾ 0.22 to 0.36 ⫾ 0.12 events/s) when stores are depleted by CPA in 0 Ca2⫹ suggests that the majority of spontaneous transmitter release is a Ca2⫹-dependent process. Nicotine can increase the frequency of spontaneous glutamate release in the hippocampus via a TTX-insensitive mechanism, presumably Ca2⫹ influx mediated by presynaptic ␣7-containing nicotinic acetylcholine receptors (Radcliffe and Dani, 1998). Our observations indicate that such Ca2⫹ influx serves to increase the frequency of CICR in the bouton and that it is this CICR, rather than the transmembrane Ca2⫹ influx per se, that is responsible for triggering the increased spontaneous transmitter release. Where boutons contain multiple release sites, triggering of spontaneous transmitter release by store Ca2⫹ release could account for multiquantal spontaneous EPSCs or for the temporal clustering of mEPSCs (Bornstein, 1978; Voronin et al., 1999). Proof of these hypotheses will require simultaneous postsynaptic electrophysiology and presynaptic calcium imaging. Spontaneous transmitter release is not merely a curiosity. There is evidence that spontaneous and evoked transmitter release are regulated by shared mechanisms and that long-term synaptic plasticity is reflected in corresponding changes in both (Malgaroli et al., 1995). In addition, spontaneous transmitter release appears to have important functions, including the maintenance of synaptic structure (McKinney et al., 1999). The observation that spontaneous transmitter release is driven by calcium release from stores suggests that the store release mechanisms may be an important locus for longterm synaptic plasticity. Experimental Procedures Preparation of Hippocampal Slice Cultures Transverse 350 ␮m slices of hippocampus from 8d male Wistar rats were prepared as described previously (Stoppini et al., 1991; Emptage et al., 1999a) and cultured for 10–21 days before use. Electrophysiological and Optical Recording Cultures were transferred to a recording chamber beneath a Zeiss Axioskop upright microscope with a BioRad MRC-1000 confocal

Neuron 206

scanhead and superfused with oxygenated (95% O2/5% CO2) ACSF (composition in mM: NaCl, 120; KCl, 3; MgCl2, 1; CaCl2, 3; NaH2PO4, 1.2; NaHCO3, 23; glucose, 11; and Trolox, 1) at 30⬚C. 0 Ca2⫹ ACSF was similar but with 1 mM EGTA in place of CaCl2. Intracellular recording and injection of the calcium indicator Oregon green 488 BAPTA-1 (Molecular Probes) in CA3 pyramidal neurons was carried out via sharp microelectrodes as described previously (Emptage et al., 1999a). For some experiments, microelectrode tips also contained 2% Cascade blue–biocytin (Molecular Probes). Indicatorfilled axons in the upper 50 ␮m of the preparation were visualized by confocal laser scanning microscopy (488 nm excitation) using a 60⫻, NA 0.9 water immersion objective (Olympus), beginning ⵑ10 min after initial cell penetration. The scan axis was aligned parallel to a region of axon bearing conspicuous boutons (at least twice the axon diameter and located 75–300 ␮m and at least two branchpoints distal to the axon initial segment), using a scan rotator (Scientific Systems Design, Inc., Mercerville, NJ), prior to collecting line scans (2 ms/line) or fast area scans (39–84 ms/frame) with an in-house modification of COMOS (BioRad) software. A minimum of 18 action potentials were sampled per bouton for each condition. Experiments were rejected if morphological changes in the axon were detected at any time. EPSPs were recorded via intracellular electrodes in response to brief (40 ␮s) stimuli delivered to CA3 stratum radiatum via a sharpened tungsten microelectrode (AM Systems) at least 100 ␮m from the recorded cell. For paired-pulse facilitation experiments, pairs of stimuli were delivered 75 ms apart; facilitation was measured as the percent increase in the amplitude of the second EPSP compared to the first. Data were collected using A/DVANCE software (McKellar Designs, Vancouver). mEPSCs were recorded in wholecell mode from cells maintained at a holding potential of ⫺65mV with 1–2.5 ␮M TTX added to the ACSF, using A/DVANCE with a standard patch-clamp amplifier (Axopatch 1D, Axon Instruments) and pipettes (3–5 M⍀) filled with (in mM) K gluconate, 110; KCl, 5; HEPES, 50; Mg SO4, 4; and Na2ATP, 4 (pH adjusted to 7.3 with KOH; osmolarity 305 mOsm). Series resistance was ⬍20 M⍀; experiments were rejected if the series resistance was found to change by ⬎10%. Spontaneous mEPSCs were detected and analyzed offline using unbiased automated routines (optimally scaled template; Clements and Bekkers [1997]) in Axograph software (Axon Instruments). For imaging spontaneous Ca2⫹ transients in boutons, spontaneous event frequency was augmented by the presence of 1 ␮M nicotine (Radcliffe and Dani, 1998) and 10 mM Ca2⫹ in the ACSF. Images were analyzed using NIH Image, Axograph, and custom-written software. Bouton Ca2⫹ transients (measured over their initial 10 ms) were expressed as fractional change in fluorescence, %⌬F/F ⫽ 100 ⫻ (F ⫺ Finitial)/(Finitial ⫺ Fbackground). Ryanodine, thapsigargin (Alamone), and cyclopiazonic acid (Tocris) were dissolved in dimethyl sulfoxide as 1000⫻ stock solutions, from which they were added directly to the ACSF. All reagents were from Sigma, except as noted. Data are expressed as mean ⫾ SEM. Statistical significance was evaluated by two-tailed t tests or, in one case, by Kolmogorov–Smirnov test.

Immunohistochemistry At the conclusion of experiments with biocytin-containing intracellular electrodes, slice cultures were fixed by immersion in 3% paraformaldehyde/0.1% glutaraldehyde in phosphate-buffered saline (PBS [pH 7.4]) for 20 min. After washing in PBS, cultures were permeabilized with 0.1% saponin in PBS for 20 min and then preincubated for 30 min in PBS containing 1% bovine serum albumin and 10% (v/v) fetal calf serum (dilution and blocking medium [DBM]). Cultures were then incubated for 24 hr with monoclonal anti-synaptophysin antibody (SY38; Boehringer Mannheim; 1:100 in DBM), washed three times for 15 min in PBS, and incubated with Cy3-labeled streptavidin (Sigma; 1:200 in DBM) to detect biocytin-filled axons and Cy2-labeled donkey anti-mouse IgG (Jackson ImmunoResearch; 1:200 in DBM) to detect synaptophysin immunoreactivity. All procedures were carried out at 4⬚C, except washes, which were at room temperature. After further washing (two times for 15 min in PBS), preparations were embedded in Mowiol (Hoechst) and imaged by simultaneous two-channel confocal microscopy, using BioRad A1 and A2 filter blocks for standard fluorescein and rhodamine double labeling with bleed-through correction.

Acknowledgments We thank Sukhvinder Dhanjal, Gian-Luca Raimondi, and the staff of the Mechanical Engineering Section, National Institute for Medical Research, for technical assistance; Tim Bliss for support and advice; and Ruth Fabian-Fine, Dimitri Kullmann, and Dmitri Rusakov for advice and criticism. This work was funded by the Medical Research Council (United Kingdom) and by grants to A. F. from the Human Frontier Science Program and the Natural Sciences and Engineering Research Council Canada. C. A. R. is a Howard Florey Fellow of the Royal Society. Received April 17, 2000; revised November 28, 2000. References Augustine, G.J., Charlton, M.P., and Smith, S.J. (1987). Calcium action in synaptic transmitter release. Annu. Rev. Neurosci. 10, 633–693. Bennett, D.L., Bootman, M.D., Berridge, M.J., and Cheek, T.R. (1998). Ca2⫹ entry into PC12 cells initiated by ryanodine receptors or inositol 1,4,5-trisphosphate receptors. Biochem. J. 329, 349–357. Berridge, M.J. (1993). Inositol trisphosphate and calcium signaling. Nature 361, 315–325. Blaustein, M.P., Ratzlaff, R.W., Kendrick, N.C., and Schweitzer, E.S. (1978a). Calcium buffering in presynaptic nerve terminals. I. Evidence for involvement of a nonmitochondrial ATP-dependent sequestration mechanism. J. Gen. Physiol. 72, 15–41. Blaustein, M.P., Ratzlaff, R.W., and Schweitzer, E.S. (1978b). Calcium buffering in presynaptic nerve terminals. II. Kinetic properties of the nonmitochondrial Ca2⫹ sequestration mechanism. J. Gen. Physiol. 72, 43–66. Blaustein, M.P., Ratzlaff, R.W., and Schweitzer, E.S. (1980). Control of intracellular calcium in presynaptic nerve terminals. Fed. Proc. 39, 2790–2795. Bornstein, J.C. (1978). Spontaneous multiquantal release at synapses in guinea-pig hypogastric ganglia: evidence that release can occur in bursts. J. Physiol. 282, 375–398. Borst, J.G.G., and Sakmann, B. (1998). Facilitation of presynaptic calcium currents in the rat brainstem. J. Physiol. 513, 149–155. Brain, K.L., and Bennett, M.R. (1997). Calcium in sympathetic varicosities of mouse vas deferens during facilitation, augmentation and autoinhibition. J. Physiol. 502, 521–536. Chen, C., and Regehr, W.G. (1999). Contributions of residual calcium to fast synaptic transmission. J. Neurosci. 19, 6257–6266. Cheng, H., Lederer, W.J., and Cannell, M.B. (1993). Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science 262, 740–744. Clementi, E., Scheer, H., Zacchetti, D., Fasolato, C., Pozzan, T., and Meldolesi, J. (1992). Receptor-activated Ca2⫹ influx. Two independently regulated mechanisms of influx stimulation coexist in neurosecretory PC12 cells. J. Biol. Chem. 267, 2164–2172. Clements, J.D., and Bekkers, J.M. (1997). Detection of spontaneous synaptic events with an optimally scaled template. Biophys. J. 73, 220–229. Cox, C.L., Denk, W., Tank, D.W., and Svoboda, K. (2000). Action potentials reliably invade axonal arbors of rat neocortical neurons. Proc. Natl. Acad. Sci. USA 97, 9724–9728. Cuttle, M.F., Tsujimoto, T., Forsythe, I.D., and Takahashi, T. (1998). Facilitation of the presynaptic calcium current at an auditory synapse in rat brainstem. J. Physiol. 512, 723–729. Dodge, F.A., Jr., Miledi, R., and Rahamimoff, R. (1969). Strontium and quantal release of transmitter at the neuromuscular junction. J. Physiol. 200, 267–283. Emptage, N.J., and Fine, A. (2000). Intracellular calcium stores in hippocampal boutons modulate short term plasticity, capacitative calcium entry and spontaneous transmitter release. Eur. J. Neurosci. Suppl. 11, 366. Emptage, N., Bliss, T.V.P., and Fine, A. (1999a). Single synaptic

Bouton Calcium Stores 207

events evoke NMDA receptor-mediated release of calcium from internal stores in hippocampal dendritic spines. Neuron 22, 115–124. Emptage, N.J., Bliss, T.V.P., and Fine, A. (1999b). Confocal imaging of calcium transients in axonal boutons of hippocampal pyramidal cells. Soc. Neurosci. Abstr. 25, 1255. Erulkar, S.D., and Rahamimoff, R. (1978). The role of calcium ions in tetanic and post-tetanic increase of miniature end-plate potential frequency. J. Physiol. 278, 501–511. Fomina, A.F., and Nowycky, M.C. (1999). A current activated on depletion of intracellular Ca2⫹ stores can regulate exocytosis in adrenal chromaffin cells. J. Neurosci. 19, 3711–3722. Frenguelli, B.G., Irving, A.J., and Collingridge, G.L. (1996). Ca2⫹ stores and hippocampal synaptic plasticity. Seminars in the Neurosciences 8, 301–309. Hartter, D.E., Burton, P.R., and Laveri, L.A. (1987). Distribution and calcium-sequestering ability of smooth endoplasmic reticulum in olfactory axon terminals of frog brain. Neuroscience 23, 371–386. Helmchen, F., Borst, J.G.G., and Sakmann, B. (1997). Calcium dynamics associated with a single action potential in a CNS presynaptic terminal. Biophys. J. 72, 1458–1471. Henkart, M.P., Reese, T.S., and Brinley, F.J., Jr. (1978). Endoplasmic reticulum sequesters calcium in the squid giant axon. Science 202, 1300–1303. Jonas, E.A., Buchanan, J., and Kaczmarek, L.K. (1999). Prolonged activation of mitochondrial conductances during synaptic transmission. Science 286, 1347–1350. Kano, M., Garaschuk, O., Verkhratsky, A., and Konnerth, A. (1995). Ryanodine receptor-mediated intracellular calcium release in rat cerebellar Purkinje neurones. J. Physiol. 487, 1–16. Katz, B., and Miledi, R. (1967). A study of synaptic transmission in the absence of nerve impulses. J. Physiol. 192, 407–436. Katz, B., and Miledi, R. (1968). The role of calcium in neuromuscular facilitation. J. Physiol. 195, 481–492. Koizumi, S., Lipp, P., Berridge, M.J., and Bootman, M.D. (1999a). Regulation of ryanodine receptor opening by lumenal Ca(2⫹) underlies quantal Ca(2⫹) release in PC12 cells. J. Biol. Chem. 274, 33327– 33333. Koizumi, S., Bootman, M.D., Bobanovic, L.K., Schell, M.J., Berridge, M.J., and Lipp, P. (1999b). Characterization of elementary Ca2⫹ release signals in NGF-differentiated PC12 cells and hippocampal neurons. Neuron 22, 125–137. Korkotian, E., and Segal, M. (1998). Fast confocal imaging of calcium released from stores in dendritic spines. Eur. J. Neurosci. 10, 2076– 2084. Kovalchuk, Y., Eilers, J., Lisman, J., and Konnerth, A. (2000). NMDA receptor-mediated subthreshold Ca(2⫹) signals in spines of hippocampal neurons. J. Neurosci. 20, 1791–1799. Krizaj, D., Bao, J.X., Schmitz, Y., Witkovsky, P., and Copenhagen, D.R. (1999). Caffeine-sensitive calcium stores regulate synaptic transmission from retinal rod photoreceptors. J. Neurosci. 19, 7249– 7261. Lee, W.L., Anwyl, R., and Rowan, M. (1987). Caffeine inhibits posttetanic potentiation but does not alter long-term potentiation in the rat hippocampal slice. Brain Res. 426, 250–256. Llano, I., Dreessen, J., Kano, M., and Konnerth, A. (1991). Intradendritic release of calcium induced by glutamate in cerebellar Purkinje cells. Neuron 7, 577–583. Llano, I., Tan, Y.P., and Caputo, C. (1997). Spatial heterogeneity of intracellular Ca2⫹ signals in axons of basket cells from rat cerebellar slices. J. Physiol. 502, 509–519. Llinas, R., Gruner, J.A., Sugimori, M., McGuiness, T.L., and Greengard, P. (1991). Regulation by synapsin I and Ca2⫹-calmodulin-dependent protein kinase II of transmitter release in squid giant synapse. J. Physiol. 436, 257–282. Mackenzie, P.J., and Murphy, T.H. (1998). High safety factor for action potential conduction along axons but not dendrites of cultured hippocampal and cortical neurons. J. Neurophysiol. 80, 2089– 2101.

Mackenzie, P.J., Umemiya, M., and Murphy, T.H. (1996). Ca2⫹ imaging of CNS axons in culture indicates reliable coupling between single action potentials and distal functional release sites. Neuron 16, 783–795. Mainen, Z.F., Malinow, R., and Svoboda, K. (1999). Synaptic calcium transients in single spines indicate that NMDA receptors are not saturated. Nature 399, 151–155. Malgaroli, A., and Tsien, R.W. (1992). Glutamate-induced long-term potentiation of the frequency of miniature synaptic currents in cultured hippocampal neurons. Nature 357, 134–139. Malgaroli, A., Ting, A.E., Wendland, B., Bergamaschi, A., Villa, A., Tsien, R.W., and Scheller, R.H. (1995). Presynaptic component of long-term potentiation visualized at individual hippocampal synapses. Science 268, 1624–1628. Mathes, C., and Thompson, S.H. (1995). The relationship between depletion of intracellular Ca2⫹ stores and activation of Ca2⫹ current by muscarinic receptors in neuroblastoma cells. J. Gen. Physiol. 106, 975–993. McGraw, C.F., Somlyo, A.V., and Blaustein, M.P. (1980). Localization of calcium in presynaptic nerve terminals. An ultrastructural and electron microprobe analysis. J. Cell Biol. 85, 228–241. McKinney, R.A., Capogna, M., Durr, R., Gahwiler, B.H., and Thompson, S.M. (1999). Miniature synaptic events maintain dendritic spines via AMPA receptor activation. Nat. Neurosci. 2, 44–49. McPherson, P.S., Kim, Y.K., Valdivia, H., Knudson, C.M., Takekura, H., Franzini-Armstrong, C., Coronado, R., and Campbell, K.P. (1991). The brain ryanodine receptor: a caffeine-sensitive calcium release channel. Neuron 7, 17–25. Melamed, N., Helm, P.J., and Rahamimoff, R. (1993). Confocal microscopy reveals coordinated calcium fluctuations and oscillations in synaptic boutons. J. Neurosci. 13, 632–649. Melamed-Book, N., Kachalsky, S.G., Kaiserman, I., and Rahamimoff, R. (1999). Neuronal calcium sparks and intracellular calcium “noise.” Proc. Natl. Acad. Sci. USA 96, 15217–15221. Miledi, R. (1973). Transmitter release induced by injection of calcium ions into nerve terminals. Proc. R. Soc. Lond. B 183, 421–425. Narita, K., Akita, T., Osanai, M., Shirasaki, T., Kijima, H., and Kuba, K. (1998). A Ca2⫹-induced Ca2⫹ release mechanism involved in asynchronous exocytosis at frog motor nerve terminals. J. Gen. Physiol. 112, 593–609. Onodera, K. (1973). Effect of caffeine on the neuromuscular junction of the frog, and its relation to external calcium concentration. Jap. J. Physiol. 23, 587–597. Parekh, A.B., and Penner, R. (1997). Store depletion and calcium influx. Physiol. Rev. 77, 901–930. Peng, Y. (1996). Ryanodine-sensitive component of calcium transients evoked by nerve firing at presynaptic nerve terminals. J. Neurosci. 16, 6703–6712. Petersen, C.C.H., Berridge, M.J., Borgese, M.F., and Bennett, D.L. (1995). Putative capacitative calcium entry channels: expression of Drosophila trp and evidence for existence of vertebrate homologues. Biochem. J. 311, 41–44. Philipp, S., Hambrecht, J., Braslavski, L., Schroth, G., Freichel, M., Murakami, M., Cavalie, A., and Flockerzi, V. (1998). A novel capacitative calcium entry channel expressed in excitable cells. EMBO J. 17, 4274–4282. Putney, J.W., Jr., and McKay, R.R. (1999). Capacitative calcium entry channels. Bioessays 21, 38–46. Qian, J., and Saggau, P. (1999). Modulation of transmitter release by action potential duration at the hippocampal CA3-CA1 synapse. J. Neurophysiol. 81, 288–298. Radcliffe, K.A., and Dani, J.A. (1998). Nicotinic stimulation produces multiple forms of increased glutamatergic synaptic transmission. J. Neurosci. 18, 7075–7083. Regehr, W.G., and Atluri, P.P. (1995). Calcium transients in cerebellar granule cell presynaptic terminals. Biophys. J. 68, 2156–2170. Reyes, M., and Stanton, P.K. (1996). Induction of hippocampal longterm depression requires release of calcium from separate presyn-

Neuron 208

aptic and postsynaptic intracellular stores. J. Neurosci. 16, 5951– 5960. Sabatini, B.L., and Regehr, W.G. (1995). Detecting changes in calcium influx which contribute to synaptic modulation in mammalian brain slice. Neuropharmacology 34, 1453–1467. Sabatini, B.L., and Regehr, W.G. (1997). Control of neurotransmitter release by presynaptic waveform at the granule cell to Purkinje cell synapse. J. Neurosci. 17, 3425–3435. Sah, P., and McLachlan, E.M. (1991). Ca(2⫹)-activated K⫹ currents underlying the afterhyperpolarization in guinea pig vagal neurons: a role for Ca(2⫹)-activated Ca(2⫹) release. Neuron 7, 257–264. Savic, N., and Sciancalepore, M. (1998). Intracellular calcium stores modulate miniature GABA-mediated synaptic currents in neonatal rat hippocampal neurons. Eur. J. Neurosci. 10, 3379–3386. Seidler, N.W., Jona, I., Vegh, M., and Martonosi, A. (1989). Cyclopiazonic acid is a specific inhibitor of the Ca2⫹-ATPase of sarcoplasmic reticulum. J. Biol. Chem. 264, 17816–17823. Sharp, A.H., McPherson, P.S., Dawson, T.M., Aoki, C., Campbell, K.P., and Snyder, S.H. (1993). Differential immunohistological localization of inositol 1,4,5-trisphospahate- and ryanodine-sensitive Ca2⫹ release channels in rat brain. J. Neurophysiol. 13, 3051–3063. Shmigol, A., Kirischuk, S., Kostyuk, P., and Verkhratsky, A. (1994). Different properties of caffeine-sensitive Ca2⫹ stores in peripheral and central mammalian neurons. Pflugers Arch. 426, 174–176. Simpson, P.B., Challis, R.A.J., and Nahorski, S.R. (1993). Involvement of intracellular stores in the Ca2⫹ responses to N-methyl-Daspartate and depolarization in cerebellar granule cells. J. Neurochem. 61, 760–763. Sinha, S.R., Wu, L.-G., and Saggau, P. (1997). Presynaptic calcium dynamics and transmitter release evoked by single action potentials at mammalian central synapses. Biophys. J. 72, 637–651. Smith, A.B., and Cunnane, T.C. (1996). Ryanodine-sensitive calcium stores involved in neurotransmitter release from sympathetic nerve terminals of the guinea-pig. J. Physiol. 497, 657–664. Smith, P.M., and Reed, H.E. (1996). Amplification of the thapsigarginevoked increase in the cytosolic free Ca2⫹ concentration by acetylcholine in acutely isolated mouse submandibular acinar cells. Biochem. J. 317, 779–783. Stoppini, L., Buchs, P.A., and Muller, D. (1991). A simple method for organotypic cultures of nervous tissue. J. Neurosci. Meth. 37, 173–182. Takemura, H., Ohshika, H., Yolosawa, N., Oguma, K., and Thastrup, O. (1991). The thapsigargin-sensitive intracellular Ca2⫹ pool is more important in plasma membrane Ca2⫹ entry than the IP3-sensitive intracellular Ca2⫹ pool in neuronal cell lines. Biochem. Biophys. Res. Comm. 180, 1518–1526. Toescu, E.C., Moller, T., Kettenmann, H., and Verkhratsky, A. (1998). Long-term activation of capacitative Ca2⫹ entry in mouse microglial cells. Neuroscience 86, 925–935. Treiman, M., Caspersen, C., and Brogger Christensen, S. (1998). A tool coming of age: thapsigargin as an inhibitor of sarco-endoplasmic reticulum Ca(2⫹)-ATPases. Trends Pharmacol. 19, 131–135. Usachev, Y.M., and Thayer, S.A. (1999). Ca2⫹ influx in resting rat sensory neurones that regulates and is regulated by ryanodinesensitive Ca2⫹ stores. J. Physiol. 519, 115–130. Voronin, L.L., Volgushev, M., Sokolov, M., Kasyanov, A., Chistiakova, M., and Reymann, K.G. (1999). Evidence for an ephaptic feedback in cortical synapses: postsynaptic hyperpolarization alters the number of response failures and quantal content. Neurosci 92, 399–405. Wu, L., and Saggau, P. (1994). Presynaptic calcium is increased during normal synaptic transmission and paired-pulse facilitation, but not in long-term potentiation in area CA1 of hippocampus. J. Neurosci. 14, 645–654. Zucker, R.S. (1989). Short-term synaptic plasticity. Annu. Rev. Neurosci. 122, 13–31. Zucker, R.S. (1999). Calcium- and activity-dependent synaptic plasticity. Curr. Opin. Neurobiol. 9, 305–313.