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

This Article Abstract Full Text (PDF) Supplemental data All Versions of this Article: 580/3/835    most recent jphysiol.2006.125062v1 Services Email this article to a friend Similar articles in this journal 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 Google Scholar Google Scholar Articles by Power, J. M. Articles by Sah, P. Search for Related Content PubMed PubMed Citation Articles by Power, J. M. Articles by Sah, P. Related Collections Neuroscience

¹ Queensland Brain Institute, University of Queensland, St Lucia, Queensland 4072, Australia

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Metabotropic receptor activation is important for learning, memory and synaptic plasticity in the amygdala and other brain regions. Synaptic stimulation of metabotropic receptors in basolateral amygdala (BLA) projection neurons evokes a focal rise in free Ca²⁺ in the dendrites that propagate as waves into the soma and nucleus. These Ca²⁺ waves initiate in the proximal dendrites and show limited propagation centrifugally away from the soma. In other cell types, Ca²⁺ waves have been shown to be mediated by either metabotropic glutamate receptor (mGluR) or muscarinic receptor (mAChR) activation. Here we show that mGluRs and mAChRs act cooperatively to release Ca²⁺ from inositol 1,4,5-trisphosphate (IP₃)-sensitive intracellular Ca²⁺ stores. Whereas action potentials (APs) alone were relatively ineffective in raising nuclear Ca²⁺, their pairing with metabotropic receptor activation evoked an IP₃-receptor-mediated Ca²⁺-induced Ca²⁺ release, raising nuclear Ca²⁺ into the micromolar range. Metabotropic-receptor-mediated Ca²⁺-store release was highly compartmentalized. When coupled with metabotropic receptor stimulation, large robust Ca²⁺ rises and AP-induced amplification were observed in the soma, nucleus and sparsely spiny dendritic segments with metabotropic stimulation. In contrast, no significant amplification of the Ca²⁺ transient was detected in spine-dense high-order dendritic segments. Ca²⁺ rises evoked by photolytic uncaging of IP₃ showed the same distribution, suggesting that IP₃-sensitive Ca²⁺ stores are preferentially located in the soma and proximal dendrites. This distribution of metabotropic-mediated store release suggests that the neuromodulatory role of metabotropic receptor stimulation in BLA-dependent learning may result from enhanced nuclear signalling.

(Received 19 November 2006; accepted after revision 9 February 2007; first published online 15 February 2007)
Corresponding author P. Sah: Queensland Brain Institute, The University of Queensland, St Lucia, Queensland 4072, Australia. Email: pankaj.sah{at}uq.edu.au

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
The basolateral amygdala (BLA) is an important locus in assigning affective value to sensory stimuli. The cellular mechanism that underlies this association is thought to be long-term synaptic plasticity within the BLA (LeDoux, 2000; Davis & Whalen, 2001). It is generally believed that these changes that underlie emotional learning are largely due to NMDA-receptor-dependent synaptic plasticity. However, neurons in the amygdala also express a variety of metabotropic receptors (Sah et al. 2003) that have been implicated in BLA-dependent learning. Thus, activation of metabotropic glutamate receptors (mGluRs) within the BLA has been found to be necessary both for fear conditioning and for some forms of synaptic plasticity (Fendt & Schmid, 2002; Rodrigues et al. 2002). The BLA also receives a dense cholinergic innervation (Ben-Ari et al. 1977), with cholinergic activity in the BLA being thought to enhance consolidation of hippocampal-dependent memories that evoke an emotional response (Power et al. 2003). Blockade of muscarinic receptors (mAChRs) attenuates long-term potentiation in the amygdala (Watanabe et al. 1995) and impairs Pavlovian fear conditioning, as well as amygdala-dependent instrumental conditioning (McGaugh, 2004; Tinsley et al. 2004). However, while glutamatergic and cholinergic metabotropic receptors have been clearly implicated in BLA-dependent learning, the underlying cellular mechanisms for these actions are not well understood.

In BLA projection neurons, activation of either cholinergic or glutamatergic metabotropic receptors suppresses a number of K⁺ conductances that control neuronal excitability (Washburn & Moises, 1992b; Womble & Moises, 1992, 1993, 1994; Yajeya et al. 1999). In addition, these receptors also stimulate the formation of inositol 1,4,5-trisphosphate (IP₃) which can release Ca²⁺ from intracellular stores (Berridge, 1998). In recent years, activation of metabotropic receptors linked to the generation of IP₃ has been shown to lead to release of Ca²⁺ from internal stores in a variety of neuronal cell types. This release can be focal, as in cerebellar Purkinje neuron spines (Finch & Augustine, 1998), or can propagate as waves through the dendritic tree as in hippocampal and cortical pyramidal neurons and midbrain dopamine neurons (Jaffe & Brown, 1994; Pozzo-Miller et al. 1996; Nakamura et al. 1999; Kapur et al. 2001; Power & Sah, 2002; Larkum et al. 2003; Morikawa et al. 2003). Of these, Ca²⁺ waves have only been described in detail in hippocampal and cortical pyramidal neurons (Ross et al. 2005). While the exact physiological role of these propagating Ca²⁺ rises is not known, release of Ca²⁺ from intracellular Ca²⁺ stores has been suggested to play a role in the induction of synaptic plasticity (Barbara, 2002; Fitzjohn & Collingridge, 2002), providing a possible link between synaptic activity and changes in gene transcription (Berridge, 1998). Consistent with this proposal, exogenous application of muscarinic agonists in hippocampal pyramidal neurons evokes Ca²⁺ waves that propagate from the dendritic tree and invade the nucleus (Nakamura et al. 1999; Kapur et al. 2001; Power & Sah, 2002; Larkum et al. 2003). Such rises in nuclear Ca²⁺ have been shown to activate CREB-mediated gene transcription (Dolmetsch et al. 1998; Li et al. 1998; Hardingham et al. 2001) in isolated cell systems. However, while dendritic Ca²⁺ waves can be evoked by repetitive synaptic stimulation (Jaffe & Brown, 1994; Pozzo-Miller et al. 1996; Nakamura et al. 1999; Kapur et al. 2001; Power & Sah, 2002; Larkum et al. 2003), synaptically evoked rises in nuclear Ca²⁺ have not been shown.

We have recently reported that synaptic stimulation of BLA projection neurons can evoke Ca²⁺ waves that initiate in the dendritic tree and propagate to the soma (Power & Sah, 2005). In the present study we examined the distribution of Ca²⁺ release evoked by metabotropic glutamate and muscarinic acetylcholine receptors and determined how these two systems interact to raise cytosolic Ca²⁺. We show that Ca²⁺ waves can result from either muscarinic or metabotropic glutamate receptor activation and that coactivation of the two receptor types can lead to a superlinear amplification of the Ca²⁺ response. We also show that synaptically evoked Ca²⁺ waves can invade the nucleus. Furthermore, IP₃ generated during synaptic stimulation can interact synergistically with action potentials (APs) to amplify the nuclear Ca²⁺ response. This IP₃-mediated store release is restricted to the soma, nucleus and proximal dendrites, and is not observed in the distal spine-dense dendrites where the majority of excitatory synaptic contacts are made. This compartmentalization of IP₃ sensitive Ca²⁺ stores suggests a primary role of metabotropic-evoked Ca²⁺ store release in nuclear signalling.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Coronal brain slices (350–400 µm) were prepared using standard techniques (Power & Sah, 2002). Wistar rats (21–28 days old) were anaesthetized with halothane and decapitated. These procedures were in accordance with the guidelines of the Institutional Animal Ethics Committee of the University of Queensland. Slices were incubated at 33°C for 30 min and then maintained at room temperature in an artificial cerebral spinal fluid (ACSF) solution containing (mM) 119 NaCl, 2.5 KCl, 1.3 MgCl₂, 2.5 CaCl₂, 1.0 Na₂PO₄, 26.2 NaHCO₃, 11 glucose, equilibrated with 95% CO₂, 5% O₂. Slices were perfused with ACSF heated to 33°C, and whole-cell recordings were made from the soma of BLA neurons using infrared differential interference contrast videomicroscopy. Patch pipettes (2–5 M) were filled with an internal solution containing (mM) 135 KMeSO₄, 8 NaCl, 10 Hepes, 2 Mg₂ATP, 0.3 Na₃GTP, 0.1 spermine (pH 7.3 with KOH, osmolarity 280–290 mosmol l^–1) and one of the following Ca²⁺ indicators: 50–100 µM Oregon Green BAPTA-1, or 300 µM Fluo-5F (Molecular Probes). In some experiments, Alexa 594 (30 µM) was also added to the internal solution to allow imaging of the cell structure. Electrophysiological signals were amplified with either an Axopatch 1D or a Multiclamp 700A amplifier (Molecular Devices), filtered at 5 kHz and digitized at 20 kHz with an ITC-16 board (Instrutech), and controlled using Axograph (Axon Instruments). Synaptic stimuli were generated using either a DS2A (Digitimer) or an Isolator-11 (Axon Instruments) isolated stimulator. Electrophysiological data were analysed using Axograph. Whole-cell recordings were obtained from projection neurons in the basal nucleus of the BLA. Only cells that had resting potentials more negative than –55 mV, AP amplitudes >100 mV, and membrane resistances (R_m) >60 M, were included in the data set.

Whole-field fluorescence measurements were made using a monochromator based imaging system, Polychrome II (TILL Photonics GmbH). Neurons were visualized using a BX50 microscope (Olympus) equipped with a 60x water immersion objective (NA 0.9; Olympus) and illuminated with 488 nm light. IP₃ was uncaged using a 2 ms stimulation from a pulsed xenon arc lamp (TILL Photonics GmbH) that illuminated the entire field of view. Light from both the monochromator and the flash were delivered to the BX50 microscope via a quartz light guide and a custom epifluorescence attachment provided by TILL Photonics, and then to the cells through the objective. Images were acquired with an interline transfer, cooled CCD camera (TILL Photonics GmbH) in which the scan lines were binned by two in both horizontal and vertical directions, giving a spatial resolution of 0.33 µm per pixel. Frames were collected at 25–33 Hz with a 10 ms exposure time. Images were analysed offline using Vision (TILL Photonics GmbH).

Two-photon and single-photon confocal fluorescence images were obtained using a Zeiss Axioskop 2FS with a 510 laser scanning head. The Axioskop was equipped with an argon laser for single photon illumination and either a Chameleon laser (Coherent Scientific) or a Verdi solid-state pump laser with a Mira 900-F femptosecond Ti:S pulsed laser (Coherent Scientific) for two-photon excitation. The excitation wavelengths were 488 nm for single-photon excitation and 810 nm for two-photon excitation. Fluorescence images were acquired in line-scan mode (50–200 Hz) at a resolution of 10–20 pixels µm^–1. When acquiring single-photon confocal data, the detector pinhole aperture was set to give an axial resolution of <1.5 µm. Small segments (3–5 µm) of the line were selected over each subcellular region and the fluorescence over this segment was averaged. Measurements of somatic Ca²⁺ were taken from the extranuclear soma, closest to the dendrite where the wave initiated. Nuclear measurements were taken from the centre of the nucleus.

Kinetic sequences were constructed over time for each of the selected regions. Kinetic sequences were calculated as the relative change in fluorescence over baseline fluorescence (F/F). F/F_t = (F_t – F₀)/(F₀ – B) where F_t is the fluorescence at time t, F₀ is the average baseline fluorescence prior to the stimulus, and B is the background fluorescence measured in an adjacent extracellular region. In some two-photon experiments, Ca²⁺ indicators were coloaded with the marker dye Alexa 594 (30 µM) which is also excited at 810 nM but fluoresces red. The emitted light was split with a dichroic (DT560), bandpass filtered (green channel, 500–560 nm; red channel, 575–640 nM), and detected with separate non-descanned detectors. In these instances, kinetic sequences were calculated as the change in green fluorescence normalized to the red fluorescence, G/R_t.

Fluorescence kinetic sequences were analysed offline using custom software. Measurements for each region of interest included peak amplitude, peak latency, area of the Ca²⁺ response integrated over time, half-width, response onset latency, and rise-time. The onset of the response was calculated as the time point at which the response exceeded 3 standard deviations above the baseline period. The rise-time was calculated as the time taken to rise from 10% to 90% of the peak response. The wave propagation speed was measured as the time for propagation of the wave-front a distance of 10 µm along the dendrite. To measure Ca²⁺-induced Ca²⁺ release at IP₃ receptors (IP₃-CICR), the peak and integrated area of the AP-evoked Ca²⁺ response in the presence of agonist were normalized to the AP-evoked responses without agonist for each region of interest. All data are presented as means ± S.E.M. unless otherwise stated. Statistical comparisons made using a paired Student's t test or ANOVA and Fisher's PLSD post hoc tests as appropriate.

In some cases the Ca²⁺ concentration was estimated using methods published by Maravall et al. (2000). To estimate Ca²⁺ changes ([Ca²⁺]_i)_, and resting Ca²⁺ ([Ca²⁺]_rest)_, we used the high-affinity indicator Oregon Green BAPTA-1 and the following formula:

(1)

where R_f and K_D are the dynamic range and dissociation constant of the indicator, with R_f = 5.7 and K_D = 210 nM for Oregon Green BAPTA-1 (Maravall et al. 2000). Estimates of F_max and F_max were obtained by measuring the amplitude of a saturating fluorescence plateau evoked by a prolonged high-frequency train of APs (100 Hz).

In some instances we also estimated the large Ca²⁺ amplitudes with the moderate affinity fluo-5F (K_D 2 µM) using the following formula (Yasuda et al. 2004):

(2)

On rare occasions the peak fluorescence signal was well beyond the linear range of the indicator and quite unreliable. To minimize the effects of these error prone values, estimates of peak Ca²⁺ were capped at 10 µM.

Low molecular weight heparin (500 µg ml^–1; Sigma) and myo-inositol 1,4,5-trisphosphate caged IP₃ (25–100 µM; Molecular Probes) were added to the internal solution as required. Acetylcholine, muscarine, carbachol, glutamate, and (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (t-ACPD) were applied either by focal pressure application (2–100 µM in ACSF) through a patch pipette using a picospritzer (10–30 p.s.i. (69–207 kPa), 50–500 ms; Parker Hannifin Fairfield, NJ, USA) or via iontophoresis. For iontophoretic application, current was delivered using the Multiclamp 700A amplifier and we adjusted the pH of acetylcholine (ACh) and t-ACPD to 4 and 8, respectively. ACh (10–50 mM) and muscarine (5 mM) were applied with a 10–50 nA ejection current with a –10 nA retention current. t-ACPD (2 mM) was applied with a –50 nA ejection current and a 10 nA retention current. All other drugs were bath applied. Cyclopiazonic acid (CPA), 1,2,3,4-tetrahydro-6-nitro-2,3-dioxo-benzo[f]quinoxaline-7-sulphonamide (NBQX), and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) were prepared as stock solutions in DMSO, and diluted in ACSF when required. NBQX and 2-amino-5-phosphonovaleric acid (APV) were purchased from RBI. t-ACPD was purchased from Tocris. Tetrodotoxin (TTX) was purchased from Alomone Laboratories. All other drugs were obtained from Sigma.

The dendritic arborization of BLA projection neurons has no readily discernable orientation, so there was an inherent bias in the selection of dendrites. In whole-field fluorescence imaging experiments there was a strong bias toward large apical-like dendrites. Confocal images were restricted to the first 150 µm of the dendritic tree. Imaged dendrites were located near the surface of the slice (<70 µm). z-Stacks were constructed to measure morphological features such as dendrite diameter, distance from soma, distance from the iontophoretic electrode and spine density. Spine density was estimated by making two dimensional projections from z-stacks and counting the number of spine protrusions visible over a 10 µm dendritic segment. While these measurements are likely to underestimate the actual spine density, there was a clear difference between spine-sparse and spine-dense dendritic segments. We defined branch order solely based on the branch points, with primary branches being between the soma and first branch point, secondary branches between the second and third branch point, and higher order branches being 2–3 branch points removed from the soma.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Synaptic stimulation raises nuclear Ca²⁺

Whole-cell patch-clamp recordings and fluorescence images were made from projection neurons within the basal division of the BLA. These neurons had a resting membrane potential of –66.0 ± 4.7 mV (n = 195; mean ± S.D.). Neurons in the basal nucleus receive glutamatergic input primarily from the cortex, hippocampus and lateral amygdala (Sah et al. 2003). To evoke Ca²⁺ waves, bipolar stimulating electrodes were placed either in the external capsule (EC), to preferentially activate cortical and hippocampal afferents, in the lateral amygdala (LA) to preferentially stimulate afferents from the LA, or locally with in the basal nucleus (B) to activate all afferent pathways. Neurons in the basal nucleus also receive extensive cholinergic innervation from the basal forebrain (Nagai et al. 1982; Mesulam et al. 1983; Carlsen et al. 1985) which traverses through the EC (Ichikawa & Hirata, 1986). EC stimulation has been shown to activate cholingeric afferents on BLA projection neurons (Washburn & Moises, 1992b; Womble & Moises, 1993). Unlike the hippocampus and cortex, the basolateral amygdala is not a layered structure and the dendritic trees of its neurons do not have any clear organization. Thus, neither the arborization patterns of these inputs nor the subcellular distribution of mAChRs or mGluRs on these neurons is known. Results from the lateral amygdala suggest that glutamatergic inputs to projection neurons may be dispersed throughout the dendritic tree (Humeau et al. 2005).

Using whole-field imaging and the Ca²⁺ indicator Oregon Green BAPTA-1 (50 µM) we found that Ca²⁺ waves could be evoked by tetanic stimulation (33–100 Hz, 0.1–2 s) at all stimulator locations. Ca²⁺ waves evoked by different stimulating electrode positions were indistinguishable. In each instance, Ca²⁺ waves began as a focal increase in the proximal dendrite (15–50 µm from the soma; EC, 21 ± 7 µm; LA, 33 ± 8 µm; B, 30 ± 5 µm) 300–1200 ms after the first stimulus and propagated toward the soma (Fig. 1B and C). A limited back propagation of the Ca²⁺ wave was also observed (see also Power & Sah, 2005). Similar to observations in CA1 pyramidal neurons (Nakamura et al. 2002), the wave initiation site tended to be near the first branch point. In 9 of 11 neurons, where both the wave initiation site and cell morphology were determined, Ca²⁺ waves initiated within 2 µm of the first or second branch point. On some occasions, synaptic stimulation evoked waves that propagated simultaneously from multiple dendrites (Fig. 1D). The propagation speed of the wave front along the dendrite varied considerably between neurons (median 95 µm s^–1; range 32–333 µm s^–1; n = 14).

View larger version (65K): [in this window] [in a new window]   Figure 1.  Synaptic stimulation evokes Ca2+ waves A, fluorescence image of a cell loaded with 50 µM Oregon Green BAPTA-1. B, selected pseudo-colour frames (F/F) are shown at the indicated times after stimulation (50 Hz, 1 s) of the external capsule. C, rises in Ca2+, plotted as F/F measured at the extranuclear soma, nuclear soma and different distances from the soma (µm). The voltage response is shown beneath the F/F plots. Resting membrane potential was –63 mV and no receptor or channel blockers are present in this example. Note that summating synaptic potentials remain subthreshold during the train. D, fluorescence image of a neuron loaded with Oregon Green BAPTA-1 with two primary dendrites visible. Pseudo-colour frames (F/F), taken at the indicated times after stimulation (100 Hz, 1 s) of the basal nucleus in the presence of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 20 µM) and 2-amino-5-phosphonovaleric acid (D-APV; 60 µM) under voltage-clamp conditions (Vm –70 mV). Ca2+ waves are observed emanating from two different dendrites, each propagating toward the soma. Arrows indicate the initiation point of the dendritic waves. E, Ca2+ waves evoked by tetanic stimulation (100 Hz, 1 s) of the external capsule in the presence of kynurenic acid (2 mM) under voltage clamp (Vm –70 mV). With low-intensity stimulation (40 V, 100 µs; low stim), Ca2+ rises are restricted to the dendrite (red and green). With higher stimulus intensity (45 V, 100 µs; high stim), waves propagate into the soma (blue). A fluorescence image of a neuron loaded with Oregon Green BAPTA-1 together with the selected regions of interest is shown on the left. Scale bar, 10 µm.

Wave propagation was dependent on stimulus intensity such that near the threshold for wave generation, the Ca²⁺ rise was restricted to a local dendritic region (Fig. 1E, low stim); higher stimulus intensities resulted in more robust waves which propagated further (Fig. 1E, high stim). Increasing either the stimulus intensity (40–45 V DS2A stimulator, 500–700 µA Isolator-11 stimulator) or the number of stimuli also augmented the peak Ca²⁺ rise in the soma (0.37 ± 0.20 versus 1.16 ± 0.28 F/F; n = 6; P < 0.05) as well as the integrated area of the Ca²⁺ transient in the soma (0.32 ± 0.13 versus 2.32 ± 0.88 F/F ; n = 5; P < 0.05) and the proximal dendrite (0.67 ± 0.20 versus 1.46 ± 0.42 F/F ; n = 6; P < 0.05). While Ca²⁺ waves clearly invaded the soma, the Ca²⁺ rise was smaller in amplitude, and had a slower time course in the soma as compared to the dendrite. Thus, with Oregon Green BAPTA-1 (50 µM), peak amplitude was 1.35 ± 0.25 F/F in the proximal dendrite as compared with 0.87 ± 0.16 F/F in the extranuclear soma (P < 0.05). The time-to-peak and half-width of the Ca²⁺ transient were 275 ± 34 and 515 ± 60 ms in the dendrite, significantly (P < 0.01) faster than in the extranuclear soma (471 ± 34 and 1059 ± 158 ms, respectively; n = 22).

Whole-field imaging cannot separate signals arising from Ca²⁺ rises in extranuclear cytosolic regions from those arising within the nucleus. We therefore examined whether Ca²⁺ waves invade the nucleus using confocal imaging using the moderate affinity Ca²⁺ indicator Fluo-5F (K_D 2 µM) along with the Ca²⁺-insensitive fluorophore Alexa 594. With confocal imaging, the nucleus appears as an area of brighter fluorescence of the red marker dye Alexa 594 (Fig. 2A). This difference in fluorescence intensity is most likely due to differences in free intracellular space, as there is a relative lack of organelles in the nucleus, resulting in larger free intracellular volume per sampled volume and thus a brighter fluorescence signal in the nucleus for both the Ca²⁺ indicator and the marker dye (Connor, 1993). The two- to threefold difference in Alexa 594 fluorescence between the nucleus and cytoplasm that we observed corresponds well to published differences in freely diffusible space between the two compartments (Horowitz & Moore, 1974). Confocal imaging confirms (Fig. 2B and C) synaptically evoked Ca²⁺ waves can invade the nucleus. Furthermore, as can be seen in Fig. 2C, under voltage-clamp conditions synaptic stimulation (50 Hz, 1 s) generated an inward current due to summating EPSCs that do not reach threshold. Similar to results obtained using whole-field imaging, the amplitude of the Ca²⁺ wave varied between neuronal compartments (Fig. 2D) with the dendrite showing the greatest amplitude (P < 0.05, dend versus soma; P < 0.001, dend versus nucleus; n = 5). Compared with the extranuclear soma, the rise in nuclear Ca²⁺ was smaller in amplitude (0.40 ± 0.14 versus 0.61 ± 0.15 G/R, P < 0.001) and had a slower time course with the rise time and half-width in the nucleus being 1059 ± 110 and 1770 ± 322 ms as compared with 659 ± 115 ms (P < 0.05) and 860 ± 322 ms (P < 0.001), respectively, in the extranuclear soma. Ca²⁺ transients evoked by APs that open voltage-dependent Ca²⁺ channels in the plasmalemma showed similar slow kinetics in the nucleus (Fig. 2E). It was notable that the rise in nuclear Ca²⁺ following influx during single (Fig. 2E) or brief trains of APs was markedly smaller than that seen during invasion by a Ca²⁺ wave. We estimate (see Methods) that a single AP raises free Ca²⁺ by 70 ± 20 nM in the extranuclear soma and 14 ± 2 nM in the nucleus (estimated by evoking a single AP and the indicator Oregon Green BAPTA-1; n = 6). In contrast, synaptically evoked Ca²⁺ waves were able to raise Ca²⁺ concentration in the soma and nucleus by several hundred nanomoles (Fig. 2D, Fluo-5F; similar estimates were obtained in neurons loaded with Oregon Green BAPTA-1). The smaller amplitude and slower kinetics of the nuclear Ca²⁺ signal are consistent with diffusion of Ca²⁺ from the cytosol into the nucleus due to the slight diffusional barrier presented by the nuclear pores in the nuclear membrane (O'Malley, 1994).

View larger version (94K): [in this window] [in a new window]   Figure 2.  Synaptically evoked Ca2+ waves invade the nucleus A, red fluorescence two-photon image for a basolateral amygdala (BLA) projection neuron loaded with the Fluo-5F (300 µM; KD 2 µM) and Alexa 594 (30 µM). The straight red line indicates the location of the scan line. Nuclear (N) somatic (S) and dendritic (D) segments are indicated by dashed lines. B, local synaptic stimulation (50 Hz, 1 s) causes a rise in Ca2+ in the dendrite, soma and nucleus. The fluorescence response along the scan line over time, plotted as G/R on a pseudocoloured scale, is shown in a and on an expanded time scale in b. Horizontal scale bars, 250 ms; vertical scale bar, 10 µm. C, the Ca2+ response, plotted as G/R, is shown for the nucleus, soma, and dendrite in response to synaptic stimulation (50 Hz, 1 s). The whole-cell current (holding potential –55 mV) is shown below the Ca2+ response. The inset to the right of the current trace shows the voltage response to a single synaptic stimulation. D, summary data plotting the peak amplitude of synaptically activated Ca2+ waves, in the nucleus (N), soma (S) and proximal dendrite (D), 10–15 µm from the soma. Lines connect data from the same neuron. Data in A–D were acquired using moderate-affinity Fluo-5F. Waves that did not propagate to the somatic border are not included in D. E, fluorescence response, plotted as F/F, to a single AP in the nucleus and soma for a neuron loaded with Oregon Green BAPTA-1 (50 µM); note the slow kinetics of the nuclear Ca2+ signal.

Cooperative action of mAChRs and mGluRs in wave generation

In pyramidal neurons, activation of either metabotropic glutamate receptors (mGluRs; Nakamura et al. 1999; Larkum et al. 2003) or muscarinic acetylcholine receptors (mAChRs; Power & Sah, 2002) can generate Ca²⁺ waves. Both mGluR5 metabotropic glutamate receptors (Rodrigues et al. 2002) and mAChRs (Mash & Potter, 1986) are present at high concentrations in the BLA. As the location of our stimulating electrodes will activate both glutamatergic and cholinergic afferents to BLA neurons, we tested for the involvement of these systems in wave generation. Waves were evoked at 3–5 min intervals under voltage-clamp conditions (holding potential –50 to –55 mV) in the presence of the ionotropic receptor blockers APV and CNQX. After obtaining a stable baseline, we bath-applied saturating concentrations of either the mAChR antagonist atropine or the mGluR5 antagonist 6-Methyl-2-(phenylethynyl) pyridine (MPEP) (Gasparini et al. 1999). Atropine (1 µM) reduced the peak Ca²⁺ response by 81 ± 10% in the soma (P < 0.01) and 80 ± 16% in the proximal dendrite (Fig. 3B; n = 5; P < 0.05). MPEP (10 µM) reduced the Ca²⁺ response by 65 ± 17% in the soma (n = 6; P < 0.05) and 40 ± 20% in the proximal dendrite (Fig. 3C). Following application of either atropine or MPEP, there was no change in the initiation point of the Ca²⁺ wave. However, the onset of the Ca²⁺ wave was delayed by 34 ± 7 ms (n = 5; P < 0.01; atropine 1, MPEP 4). There was no detectable rise in Ca²⁺ in the presence of both antagonists (n = 3). Consistent with a role of both mAChRs and mGluRs in the synaptic generation of Ca²⁺ waves, robust Ca²⁺ waves were reliably evoked with focal pressure application of either muscarine (5–20 µM) or t-ACPD (5–20 µM), or with iontophoretic application of ACh (10–50 mM), muscarine (5 mM) or t-ACPD (2 mM) (Fig. 3C). Similar to synaptically evoked waves, Ca²⁺ transients evoked by exogenously applied agonists propagated along the dendrite with little reduction in amplitude, and thus did not result from passive diffusion of Ca²⁺ from the initiation point. Agonist-evoked waves also readily invaded the soma and nucleus. There was no obvious difference between the kinetics of waves generated by either mGluR or mAChR stimulation, with either class of agonist being able to evoke a robust Ca²⁺ wave (Fig. 3C). As with synaptic stimulation (Fig. 3A and B), coapplication of the two agonists revealed the cooperative nature of mGluR and mAChR activation on wave generation. Thus, when agonists were applied at levels that evoked little or no Ca²⁺ rise individually, coapplication of the two agonists evoked a large Ca²⁺ wave (Fig. 3D). This cooperative action was particularly evident in the large rise in nuclear Ca²⁺ when the two agonists were applied together (Fig. 3E). This cooperative activation is most likely to be due to summation of the relatively low levels of IP3 generated by each transmitter on its own and the regenerative nature of Ca²⁺ waves.

View larger version (37K): [in this window] [in a new window]   Figure 3.  Ca2+ waves are mediated by metabotropic receptor activation A, Ca2+ waves repetitively evoked once every 5 min by synaptic stimulation (100 Hz, 1 s) in the presence of kynurenic acid (2 mM), before and after sequential application of atropine (1 µM) and MPEP (10 µM). Rises in Ca2+, plotted as F/F, were measured in soma and dendrite (15 µm from soma) as indicated on the inset image. B, summary data showing the percentage blockade of the Ca2+ wave amplitude (mean ± S.E.M.) following bath application of atropine or MPEP, or coapplication of atropine and MPEP. C, robust Ca2+ waves can be evoked by application of either ACh or (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (t-ACPD) . Ca2+ rises are shown in the same neuron to iontophoretic application of ACh and t-ACPD. D, cholinergic and glutamatergic receptor activation act cooperatively to evoke Ca2+ waves. The response to somatic iontophoretic application of subthreshold levels of t-ACPD and ACh is shown along with the response to coapplication of t-ACPD and ACh. E, summary data showing that the nuclear Ca2+ rise evoked by coapplication of low levels of t-ACPD and ACh is greater than the arithmetic sum of the Ca2+ rise evoked by separate applications of ACPD and ACh. A and B, data were obtained using the Ca2+ indicator Oregon Green BAPTA-1 (50 µM). C, D and E, data were obtained using Fluo-5F (300 µM).

View larger version (57K): [in this window] [in a new window]   Figure 4.  Propagation of agonist-evoked Ca2+ waves A, iontophoretic application of t-ACPD onto the proximal dendrite (top) results in a Ca2+ wave that initiates in the dendrite and propagates toward the soma. Application of t-ACPD onto the soma (bottom) produces a wave that propagates from the soma. A projection of the neuron is shown on the left. The yellow arrow indicates the part of the neuron that is closest to the iontophoretic electrode. Scale bar, 5 µm. The middle plot shows the green fluorescence response (indicator, Fluo-5F) for a scan line along the somato-dendritic axis shown on the left over time. Coloured traces in the right panel show rises in Ca2+, plotted as G/R, measured in the soma and along the dendrite as indicated by the coloured brackets in the left panel. B, picospritzer application of muscarine (10 µM, 20 p.s.i. (138 kPa), 300 ms) onto the proximal dendrites results in a Ca2+ wave that initiates at the first dendritic branch point and propagates toward the soma. The muscarine-evoked Ca2+ rise did not propagate along the secondary dendrites (red and green traces). A fluorescence image of the neuron loaded with Oregon Green BAPTA-1 shown on the left together with coloured boxes showing selected regions of interest. Scale bar, 10 µm. Rises in Ca2+, measured along the dendrite as indicated by the coloured boxes are shown in response to muscarine application (middle). The response to a train of four APs (50 Hz) is shown on the right. C, increasing agonist stimulation increases wave propagation. The Ca2+ rise (coloured traces) evoked by iontophoresis of ACh (50 mM, 1500 ms) onto the proximal dendrite is shown for various intensities of ejection current (indicator, Fluo-5F). Cellular regions of interest along with the scan-line are indicated on the confocal section (left). Graphs of the iontophoretic intensity versus onset and peak for each region of interest are shown on the right. Scale bar, 10 µm.

Ca²⁺ signals evoked by activation of metabotropic receptors in neurons result from generation of IP₃ and subsequent release of Ca²⁺ from IP₃-sensitive intracellular Ca²⁺ stores (Finch & Augustine, 1998; Takechi et al. 1998; Nakamura et al. 1999; Power & Sah, 2002; Larkum et al. 2003). In BLA neurons, waves evoked either synaptically (n = 3; Fig. 5A) or by exogenous agonist application (t-ACPD, n = 4; Ach, n = 4; Fig. 5B) were completely and reversibly blocked by cyclopiazonic acid (CPA) which blocks the endoplasmic reticulum (ER) Ca²⁺ATPase and empties intracellular Ca²⁺ stores. We have previously shown (Power & Sah, 2005) that, in BLA neurons, ryanodine receptors (RyRs), are not required for agonist-evoked Ca²⁺ rises, even though they share a common intracellular Ca²⁺ pool with IP₃ receptors. In agreement with this, agonist-evoked waves could not be evoked when the IP₃ receptor antagonist heparin (500 µg ml^–1) was included in the patch pipette (n = 5; t-ACPD; data not shown). Synaptically evoked TILL Photonics waves were also blocked by application of 2-aminoethoxydiphenyl borate (2APB; 500 µM; n = 3; Fig. 5C), a membrane-permeable blocker of IP₃ receptors and store-operated Ca²⁺ entry (Maruyama et al. 1997; Bootman et al. 2002b; Peppiatt et al. 2003) indicating that wave generation requires the production of IP₃ and release of intracellular Ca²⁺ stores. Consistent with this, agonist-evoked Ca²⁺ rises could be mimicked by photolytic uncaging of IP₃, and blocked by emptying stores with CPA (Fig. 5D). However, unlike synaptic and agonist-evoked Ca²⁺ rises that had a clear initiation point, photolytic release of IP₃ evoked a near simultaneous Ca²⁺ rise in both the soma and the proximal dendrites, showing that IP₃-sensitive Ca²⁺ stores are present in both locations. Ca²⁺ rises were generally not associated with any change in membrane potential or whole-cell current, consistent with Ca²⁺ release from intracellular Ca²⁺ stores. However occasionally, Ca²⁺ rises were accompanied by a membrane hyperpolarization (Fig. 5B and D, lower traces) or an outward current in voltage clamp. Both the outward current and the resultant hyperpolarization were also blocked by disruption of IP₃-evoked store release (Fig. 5B and D), suggesting that the current results from activation of a Ca²⁺-dependent conductance secondary to release of Ca²⁺ from intracellular stores. This current appears to be mediated by activation of SK Ca²⁺-activated K⁺ channels (Power & Sah, 2006), similar to observations in midbrain dopamine neurons (Fiorillo & Williams, 1998; Morikawa et al. 2003) and cortical neurons (Yamada et al. 2004; Gulledge & Stuart, 2005).

View larger version (25K): [in this window] [in a new window]   Figure 5.  Ca2+ waves require release of Ca2+ from inositol 1,4,5-trisphosphate (IP3)-sensitive intracellular Ca2+ stores A, cyclopiazonic acid (CPA, 30 µM) reversibly blocks synaptically evoked Ca2+ waves. Waves were evoked by external capsule stimulation (100 Hz, 250 ms) at 5 min intervals under voltage clamp (HP –70 mV; indicator, Oregon Green BAPTA-1). Neurons were depolarized to –50 mV between tetani to maintain Ca2+ stores. Changes in dendritic and somatic Ca2+ (F/F) are plotted before application of CPA, in the presence of CPA (5 min and 10 min), and during washout of CPA. B, CPA blocks agonist-evoked Ca2+ waves. Ca2+ rises, plotted as G/R (indicator, 300 µM Fluo-5F), and the voltage response evoked by iontophoretic application of t-ACPD (bottom traces) are shown before (thin traces) and 15 min after (bold traces) application of CPA (30 µM). C, rises in Oregon Green BAPTA-1 fluorescence evoked by external capsule stimulation (100 Hz, 1 s) before (left and middle) and 5–10 min after application of 2-aminoethoxydiphenyl borate (2APB; 500 µM; right). The bar beneath the fluorescence traces indicates the timing of synaptic stimulation. D, Ca2+ rises evoked by the photolytic uncaging of IP3 (50 µM caged IP3 and 50 µM Oregon Green BAPTA-1 in pipette), and the voltage change in response to the Ca2+ change (bottom traces), are blocked by bath application of CPA (10 min, 30 µM).

Distribution of Ca²⁺ release in BLA projection neurons

In hippocampal and cortical pyramidal neurons, metabotropic-receptor evoked Ca²⁺ rises are largely restricted to the soma and proximal apical dendrite (Nakamura et al. 2000; Larkum et al. 2003). Although the amygdala is not a layered structure, projection neurons in the BLA are often described as ‘pyramidal-like’ (McDonald, 1992; Washburn & Moises, 1992a; Sah et al. 2003). Indeed, some projection neurons in the basal nucleus have a stereotypical pyramidal morphology, with a single long, broad (>1.5 µm diameter) ‘apical-like’ dendrite emanating from the soma, and a number of thin (<1.5 µm diameter) rapidly branching ‘basal-like’ dendrites radiating from the soma. However, similar to previous observations (McDonald, 1982; Rainnie & Shinnick-Gallagher, 1992; Washburn & Moises, 1992a), projection neurons were often found to be multipolar, with multiple broad dendrites radiating from the soma (e.g. Fig. 7A). Morphological examination of dye-filled BLA projection neurons revealed that, as in many other pyramidal neurons (Elston & DeFelipe, 2002), the primary branch of the ‘apical-like’ dendrites contains few if any spines. Secondary branches are thinner and generally spiny while tertiary and higher branch orders are thin and spine dense (>0.5 spines µm^–1). The spine density of the primary and secondary branches of ‘basal-like’ dendrites is heterogeneous, ranging from few if any spines to spine dense. Tertiary and higher order branches of these basal processes are spine dense. As shown above, the soma and proximal dendrite of BLA projection neurons contain intracellular Ca²⁺ stores that can be released by generation of IP₃. We next determined the distribution of these Ca²⁺ stores through the dendritic tree.

View larger version (50K): [in this window] [in a new window]   Figure 7.  mGluR-mediated amplification of AP-evoked Ca2+ rises are localized to the soma and proximal dendrites A, two-dimensional projection of a confocal image stack (Alexa 594). Note that two broad ‘apical-like’ dendrites can be seen projecting from the soma. B, amplification of AP-evoked rises by t-ACPD is differentially distributed throughout the dendritic tree. Application of t-ACPD to the soma or primary ‘apical-like’ dendrite augments the AP-evoked Ca2+ rise in the nucleus, soma, and primary ‘apical-like’ dendrite. Application of t-ACPD onto the tertiary dendrite (3° dend) has little or no effect on the AP Ca2+ response in the tertiary dendrite. The Ca2+ rise evoked by two AP trains (four APs at 100 Hz) is shown with (bold traces) and without (thin traces) iontophoretic application of t-ACPD. Regions where measurements were made on the primary and tertiary dendrites are indicated by the arrows in a. Ca, summary data showing the amplitude of the AP-evoked Ca2+ rise with and without application of t-ACPD to the nucleus (nuc), soma (soma), primary (1), secondary (2) and higher-order (3+) dendritic branches. Cb and c, summary data illustrating the peak amplitude (b) and area (c) of the AP-evoked Ca2+ rise in the presence of t-ACPD normalized to the AP-only response. D, the mGluR-mediated augmentation of the AP-evoked Ca2+ rise was greater in spine-sparse (n = 11) than in spine-dense (n = 18) dendritic segments (unpaired t test P < 0.05). Bars show the integrated area of the AP-evoked Ca2+ response in the presence of t-ACPD, normalized to the AP-only response, for secondary dendritic branches which are spine sparse (<0.5 spines µm–1) and spine-dense (>0.5 spines µm–1). E, augmentation of AP-evoked Ca2+ response in the presence of t-ACPD (bold traces) is blocked by CPA (thin traces). F, summary data showing the integrated area of the AP-evoked Ca2+ rise in the presence of t-ACPD, normalized to the AP-evoked response in the absence of t-ACPD, before and after bath application of CPA (30 µM). G, when heparin was included in the patch pipette, coapplication of t-ACPD (bold traces) had no effect on the AP-evoked Ca2+ response (thin traces). Bars indicate means ± S.E.M. *P < 0.05.

View larger version (23K): [in this window] [in a new window]   Figure 6.  Amplification of AP-evoked Ca2+ rises by metabotropic receptor stimulation Ca2+ rises in the soma and proximal dendrite (indicator, Oregon Green BAPTA-1) are shown in response to picospritzer application of muscarine (10 µM; grey traces), a train of four APs (50 Hz; thin traces), and an AP train delivered immediately following application of muscarine (bold traces). The voltage responses are shown below the Ca2+ responses. The arrow indicates the timing of the muscarine application (100 ms, 20 p.s.i. (138 kPa)).

View larger version (37K): [in this window] [in a new window]   Figure 8.  Distribution of AChR-mediated Ca2+ rises A, amplification of AP-evoked Ca2+ rises evoked by mAChR activation is differentially distributed throughout the dendritic tree. The Ca2+ rise evoked by AP trains is shown with (bold traces) and without (thin traces) iontophoretic application of ACh (20 mM, 1 s) to various cellular regions (indicator, Fluo-5F). B, summary data showing the amplification of the peak amplitude (a) and integrated area (b) of the AP-evoked Ca2+ rise with and without mAChR stimulation in the nucleus (n), soma (s), primary (1), secondary (2) and higher order (3+) dendritic branches. C, bars show the mAChR-evoked amplification of the AP-evoked Ca2+ transient for secondary branches that are classified as either spine-sparse (n = 5) or spine-dense (n = 7). D, augmentation of the AP-evoked Ca2+ response by ACh (bold traces) is blocked by CPA (15 µM). E, summary data showing ACh-mediated amplification of the amplitude (a) and integrated area (b) of the AP-evoked Ca2+ rise before and after bath-application of CPA (30 µM; n = 5). *P < 0.05. F, bath-application of muscarine (5 µM) augments the AP-evoked Ca2+ response in the soma and proximal dendrites, but is without effect on more distal dendrites. The AP-evoked Ca2+ rise, plotted as G/R (indicator, Fluo-5F), is shown before (thin traces) and after bath-application of muscarine (bold traces) measured at the indicated cellular compartments. Muscarinic augmentation of the AP Ca2+ signal was blocked by atropine (1 µM; grey traces). B, C and E, the AP-evoked Ca2+ response in the presence of ACh is normalized to the AP-evoked Ca2+ response in the absence of ACh for each region of interest.

View larger version (30K): [in this window] [in a new window]   Figure 9.  Synaptic stimulation leads to Ca2+-induced Ca2+ release at IP3 receptors (IP3-CICR) in the nucleus A, Ca2+ rises in the soma and dendrite (indicator, Oregon Green BAPTA-1) are shown in response to local synaptic stimulation (100 Hz, 1 s; grey traces), in the presence of 2-amino-5-phosphonovaleric acid (APV; 60 µM) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 20 µM), to a train of four APs (50 Hz; thin black traces), and to an AP train delivered during synaptic stimulation (bold black traces). Voltage responses are plotted below the fluorescence responses. The inset voltage traces show an expanded view of the slow afterhyperpolarization that follows the AP train. Note the suppression of the slow afterhyperpolarization following tetanic stimulation. B, confocal red fluorescence image of a BLA projection neuron loaded with Fluo-5F (300 µM) and Alexa 594 (30 µM) showing the location of the scan line along with the nuclear (N) and somatic (S) and dendritic (D) regions of interest. C, pairing local synaptic stimulation (50 Hz, 2 s) with two AP trains (four APs 100 Hz, 200 ms intertrain interval) evoked a nuclear Ca2+ response that was greater than the sum of the AP and synaptically evoked responses. The Ca2+ rise evoked by AP trains (thin traces), synaptic stimulation (grey traces), and AP trains paired with synaptic stimulation (bold traces) are shown for the regions of interest indicated in B. Voltage responses are shown below the fluorescence responses. APV (30 µM) and NBQX (10 µM) were present in the bath. Note that the wave evoked by unpaired synaptic stimulation failed to invade the nucleus. The responses to paired stimulation showing the component due to amplified Ca2+ release obtained by subtracting the AP and synaptic responses are shown in D. Note the supra-linear response in each compartment. Dashed lines indicate the timing of the AP trains. E, the nuclear Ca2+ response to paired presentations of APs and synaptic stimulation (50 Hz, 1–2 s) was greater than the arithmetic sum of the responses to unpaired presentations. The line plot shows the integrated area of the nuclear Ca2+ responses to paired synaptic and AP stimulation, along with the arithmetic sum of unpaired presentations for individual neurons. Individual values are an average of 2–4 repetitions. Dashes indicate the group mean (n = 4).

View larger version (22K): [in this window] [in a new window]   Figure 10.  Ca2+ rise evoked by IP3 uncaging is restricted to the soma and proximal dendrites A, the Ca2+ response, plotted as F/F (Oregon Green BAPTA-1, 50 µM), to photolytic uncaging of IP3 (75 µM) is shown for the soma, and segments of the primary, secondary and tertiary dendrites. The dashed line indicates the timing of the UV flash. B, summary data showing the integrated area (mean ± S.E.M.) of the Ca2+ rise evoked by uncaging of IP3 and AP trains for the soma, primary dendrite (1°), and tertiary dendrite (3°). The number of neurons in each group is indicated in parentheses. **P < 0.01 (Fisher's PLSD post hoc test). C, the AP-evoked Ca2+ response in the distal dendrites is unaffected by photolytic uncaging of IP3. The Ca2+ rise evoked by an AP train (four APs at 50 Hz) is shown with (bold traces) and without (thin traces) photolytic uncaging of IP3, in the primary (1°) and tertiary (3°) dendrites. D, summary data illustrating the area of the AP-evoked Ca2+ rise when APs were paired with photolytic uncaging of IP3, normalized to the responses to the APs alone for the soma, primary (1°) and tertiary (3°) dendrites. The number of neurons in each group is indicated in parentheses. **P < 0.01 (Fisher's PLSD post hoc test).

	Discussion

Top Abstract Introduction Methods Results Discussion References

Neurons in the BLA receive glutamatergic as well as dense cholinergic innervation (Ben-Ari et al. 1977). We have found that these two transmitter systems act cooperatively to release intracellular Ca²⁺ (Fig. 3). Synaptic stimulation of local afferents to BLA neurons activates both types of afferents. Thus, Ca²⁺ waves are attenuated by >70% when muscarinic receptors are blocked by atropine, and by 40–65% when mGluR5 receptors are blocked by MPEP (Fig. 3). This cooperative action is most likely to be due to summation of the relatively low levels of IP₃ generated by each transmitter on its own. As Ca²⁺ release from IP₃-sensitive stores is a regenerative function, coapplication evokes a superlinear Ca²⁺ response (Fig. 3C and D).

In BLA neurons, IP₃-evoked Ca²⁺ responses are most prominent in the soma and primary dendrites. For the dendritic tree, there is an inverse correlation between spine density and the capacity of metabotropic receptor activity to evoke IP₃-mediated Ca²⁺ release in secondary dendrites. Thus, store release is present in dendrites that have a low density of spines (primary and secondary dendrites) but is greatly reduced or absent in spine-dense regions. A similar distribution has been suggested in hippocampal and cortical pyramidal cells where regenerative release of Ca²⁺ from IP₃-sensitive stores in response to metabotropic receptor activation has been found almost exclusively in the soma and spine-sparse proximal apical dendritic shaft (Kapur et al. 2001; Nakamura et al. 2002; Power & Sah, 2002; Larkum et al. 2003), but not in the spine-dense oblique branches (Nakamura et al. 2002). Why does IP₃-CICR show such a restricted distribution? The possibilities are that it may reflect a differential distribution of either metabotropic receptors, IP₃ receptors or, as in cultured neurons, some of the subcellular components of the IP₃ signalling pathway (Jacob et al. 2005). In hippocampus (Marino et al. 1998) and cortex (Mrzljak et al. 1993), M1 mAChRs, which are coupled to phosphoinositide hydrolysis (Caulfield, 1993), are found on the soma, proximal and distal dendrites, as well as on the spine head. Group I mGluR receptors are localized on the soma and dendrites, and they are particularly prevalent the spines (Lujan et al. 1997). In the basal nucleus, cholinergic terminals have been shown to make contact with soma, dendrites and spines (Carlsen & Heimer, 1986; Li et al. 2001); however, the subcellular distribution of mAChR subtypes in is unknown. While the distribution of mGluR subtypes in the basal nucleus is not known, mGluR-5 receptors in the lateral amygdala are located primarily on spines and dendrites, and to a lesser extent on the soma (Rodrigues et al. 2002). Our data show that the Ca²⁺ response to IP₃ photolysis is also restricted to the soma and proximal dendrites. Thus, the restricted distribution of Ca²⁺ waves and IP₃-CICR reflects in part differences in the distribution of IP₃ receptors: either a difference in IP₃ receptor density, or a difference in the sensitivity of the receptor to IP₃ such that IP₃ receptors located proximally are more readily activated.

As in other neuronal types (