| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Neurobiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261
Submitted 13 March 2003; accepted in final form 2 July 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Caird and Klinke 1983; for review, see Oertel 1999). Neurons in this nucleus are excited by sound at the ipsilateral ear and are inhibited by sound at the contralateral ear. Excitation reaches the LSO via a glutamatergic projection from spherical bushy cells in the ipsilateral anteroventral cochlear nucleus (AVCN, reviewed in Thompson and Schofield 2000). Inhibition reaches the LSO via a glycinergic projection from the medial nucleus of the trapezoid body (MNTB). Both excitatory and inhibitory inputs are tonotopically organized and converge on single LSO neurons in a precise frequency-specific manner (Oertel 1999; Sanes and Rubel 1988).
Normal development of the LSO depends on the presence of afferent inputs and synaptic activity (reviewed in Sanes and Friauf 2000). Abolishing the ipsilateral excitatory inputs by cochlea ablation decreases LSO volume and cell number (Moore 1992) and reduces excitatory synaptic transmission onto surviving neurons (Kotak and Sanes 1997). Decreasing glycinergic synaptic activity alters normal dendritic growth of LSO neurons and changes the axonal termination pattern of MNTB axons. These anatomical changes are accompanied by a number of physiological changes including abnormal inhibitory, as well as excitatory, synaptic transmission (review by Sanes and Friauf 2000).
Despite the importance of synaptic activity for normal LSO development, the mechanisms by which synaptic activity exerts its effect on immature LSO neurons are still poorly understood. A number of studies in other neuronal systems have demonstrated the central role of calcium-dependent mechanisms in mediating the effects of activity on neuronal survival and development (Berridge 1998; Franklin and Johnson 1992; Spitzer et al. 2000). In the developing auditory system, the role of calcium-dependent mechanisms in activity-dependent neuronal survival is best understood in the cochlear nucleus of chickens. In this nucleus, glutamate released from 8th nerve terminals activates postsynaptic metabotropic glutamate receptors (mGluRs) that mobilize intracellular calcium. This activates protein kinases A and C, which in turn inhibit voltage-gated calcium channels, thereby preventing accumulation of toxic levels of [Ca2+]i (Lachica et al. 1995; Zirpel and Rubel 1996; Zirpel et al. 1998).
Recent evidence has indicated that calcium-dependent mechanisms are also crucial for the survival and maturation of neurons in the LSO (Lohmann et al. 1998), but the mechanisms by which synaptic activity increases [Ca2+]i in the LSO are poorly understood. We have previously shown that glycinergic/GABAergic synapses in the MNTB-LSO pathway increase [Ca2+]i during the period when these synapses are depolarizing and are functionally eliminated (Kandler and Friauf 1995; Kim and Kandler 2003; Kullmann et al. 2002). However, the mechanisms by which glutamatergic afferents from the cochlear nucleus regulate postsynaptic [Ca2+]i in the LSO are unknown. Developing LSO neurons express all major classes of ionotropic and metabotropic glutamate receptors (Caicedo and Eybalin 1999; Kandler and Friauf 1995; Kotak and Sanes 1996), indicating that glutamatergic synapses could elicit calcium responses by a variety of mechanisms such as calcium influx through Ca2+-permeable AMPA, kainate, or N-methyl-D-aspartate (NMDA) receptors, depolarization-induced activation of voltage-gated calcium channels (VGCCs), or stimulation of metabotropic glutamate receptors (mGluRs) and mobilization of intracellular calcium.
In the present study, we have examined the contribution of GluRs to calcium responses in the developing LSO using Fura-2 calcium imaging in brain stem slices from neonatal mice. We found that all major types of GluRs can contribute to synaptically elicited calcium responses and that activation of specific types of GluRs and their associated calcium entry pathways is dependent on the frequency of presynaptic stimulation. These results indicate that developing LSO neurons can rate-code spontaneous synaptic activity patterns from the cochlear nucleus by activating distinct calcium entry routes that may be linked to distinct second-messenger pathways and gene expression (West et al. 2002). Some of the results have previously been presented in abstract form (Negoita and Kandler 2002).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Experiments were performed in brain stem slices of C57Bl/6J mice (Jackson Laboratory, Bar Harbor, ME) of both sexes aged between postnatal day 0 (P0: the date of birth) and P7. Experimental procedures were in accordance with National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh. Animals were anesthetized by hypothermia and decapitated, and the brains were removed and placed into cold (4–8°C) artificial cerebrospinal fluid (ACSF, composition in mM: 124 NaCl, 26 NaHCO3, 10 glucose, 5 KCl, 1.25 KH2PO4, 1.3 MgSO4, 2 CaCl2, and 1 kynurenic acid, pH 7.4 when aerated with 95% O2-5% CO2). Transverse slices (200–300 µm thickness) were cut with a vibrotome (DTK-1500E, Ted Pella, Redding, CA), and slices in which the LSO and MNTB could be identified were selected for labeling.
Fura-2 labeling
Slices were first incubated in a concentrated Fura-2AM solution (100 µM; Molecular Probes, Eugene, OR) for 2–5 min followed by submersion in aerated ACSF containing 10 µM Fura-2AM, for 2–5 h at 30–32°C (Peterlin et al. 2000). All Fura-2AM staining solutions were prepared from 1 mM stock solution in DMSO. Previously, we showed that in the immature LSO, Fura-2AM preferentially labels neurons (Kullmann et al. 2002).
Calcium imaging
Individual slices were transferred into a recording chamber and were continuously superfused with oxygenated ACSF (30–32°C, perfusion rate 2–3 ml/min) that contained no kynurenic acid but 100 µM Trolox (Scheenen et al. 1996). Slices were imaged with an inverted epifluorescence microscope (Nikon Eclipse TE200) equipped with x10 and x20 air objectives (NA: 0.5 and 0.75 respectively). Ratiometric imaging was performed as previously described (Kullmann et al. 2002).
Fluorescence intensity was converted to intracellular calcium concentration ([Ca2+]i) using the equation [Ca2+]i = Kd
(R – Rmin)/(Rmax – R), where Rmin is the fluorescence ratio of Ca2+-free Fura-2, Rmax is the ratio of Ca2+-bound Fura-2, and is the ratio of the fluorescence intensity of Ca2+-free Fura-2 at 380 nm to the fluorescence intensity of Ca2+-bound Fura-2 at 380 nm (Grynkiewicz et al. 1985). A Kd of 224 mM was used. Rmin was determined in slices incubated in Ca2+-free ACSF (Ca2+ was replaced by Mg2+) with 2 mM EGTA and 4 µM Ca2+ ionophore Br-A23187 (Alomone Labs, Israel). Rmax was measured in slices incubated in 10 mM CaCl2 (Kao 1994). Over the course of this study, the system was calibrated repeatedly, and the values for Rmin ranged from 0.25 to 0.31, for Rmax from 1.22 to 1.8, and from 2.3 to 3.7.
Electrical stimulation
Afferent fibers from the AVCN to the LSO were stimulated using bipolar stainless steel electrodes (tip distance: 100–200 µm; FHC, Bowdoingham, ME). Electrodes were placed in the ipsilateral ventral acoustic stria (VAS), lateral to the LSO (Fig. 1A). Electrical stimuli consisted of constant current pulses (duration: 100 µs) with amplitudes between 50 and 300 µA (Master-8 and ISO-Flex; AMPI) delivered in single pulses or stimulus trains (10 pulses at 5, 10, 20, 50, and 100 Hz).
|
Drug application
Drugs were dissolved in ACSF from concentrated stock solutions and delivered via bath application or pressure application using patch-electrodes (10 psi, 100- to 1,000-ms pulse duration; PV 820 picopump, WPI, Sarasota, FL).
Drugs used in this study were as follows.
AGONISTS OF GLURS. N-methyl-D-aspartate (NMDA), 
-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA), kainic acid (KA), (S)-3,5-dihydroxyphenylglycine (DHPG), 4-carboxy-3-hydroxyphenylglycine (4C3HPG), L-(+)-2-amino-4-posphonobutyric acid (L-AP4), and 1S, 3R,-aminocyclopentane-trans-dicarboxylate (ACPD).
ANTAGONISTS OF GLURS. 6-Cyano-7-nitroquinoxaline-2,3-dione disodium (CNQX), D-2-amino-5-phosphonopentanoic acid (D-APV), -methyl-4-carboxyphenylglycine (-MCPG), and 1-(4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine hydrochloride (GYKI 52466). With the exception of GYKI 52466 (gift from E. Aizenman), all drugs were from Tocris (Ballwin, MO).
VGCCS BLOCKERS. NiCl2 (nonspecific blocker of VGCCs at 2–5 mM) (Gu et al. 1994), nifedipine (L-type channel inhibitor), 
-conotoxin GVIA (N-type channel inhibitor), and -agatoxin TK (P/Q-type channel inhibitor, Alomone Labs, Jerusalem, Israel).
Other drugs used were thapsigargin and tetrodotoxin (TTX, Alomone Labs).
Electrical stimulation was performed in the presence of bicuculline (10 µM) and strychnine (10 µM), and GluR agonists were applied in TTX (1 µM). For calcium-free ACSF, 2 mM EGTA was added and calcium was exchanged for equimolar concentrations of magnesium. If not stated otherwise, chemicals used were from Sigma (St. Louis, MO).
Data analysis
Image analysis was performed with the program Tillvision (T.I.L.L. Photonics). Images were low-pass filtered (Gaussian 3 x 3 kernel) and background subtraction was performed as described previously (Betz and Bewick 1993). Briefly, for each 380 image sequence, the lowest pixel value found in the sequence was subtracted from all pixels. For 340 images series, the histogram of pixel values for each image in a series was determined. The average of the mode values (peak of histogram) of images during baseline was calculated, and 80% of this value was subtracted from all pixels in the sequence.
Calcium concentrations were measured from the soma of cells in the focal plane. Cells with resting calcium levels >250 nM, with responses >1500 nM, and cells whose calcium concentration failed to return to baseline were excluded from analysis. Only changes in [Ca2+]i that exceeded 2 standard deviations of the baseline were regarded as responses.
Electrical recordings
For AMPA ionophoresis, visualized whole cell patch-clamp recordings were performed in voltage clamp (Vhold: –65 mV) at room temperature. Recording electrodes had resistances of 2–3 M
and were filled with a solution containing (in mM): 110 d-gluconic acid, 110 CsOH-H2O, 11 EGTA, 10 CsCl, 1 MgCl2, 1 CaCl2, 10 HEPES, 0.3 Na-GTP, and 2 Mg-ATP-3.5H20. To isolate AMPAR-mediated currents, D-APV (50 µM), MCPG (1 mM), bicuculline (10 µM), strychnine (5 µM), and TTX (1 µM) were added to the ACSF. For electrical recordings only, desensitization of AMPARs was prevented with CTZ (50 µM). AMPA (10 mM in 0.15 M NaCl, pH 7.5) was ionophoretically applied close to the soma using pipettes with resistances 
70 M. Retaining current was 20–50 nA (BVC-700, Dagan, Minneapolis, MN), and ejection current was 30–50 nA for 2s. Membrane currents were filtered at 1 kHz (Axopatch 1D, Axon Instruments), digitized (AD board, National Instrument, Austin, TX), and stored for off-line analysis. Data acquisition and analysis was performed using custom-written software running under the LabVIEW environment (National Instruments). For recordings of postsynaptic responses, d-gluconic acid and CsOH in the the pipette solution were substituted by 110 mM K-gluconate, and recordings were performed in current clamp, adjusting current injection to yield a Vm of approximately –60 mV. The ipsilateral pathway was stimulated using low- and high-intensity single stimuli and low frequency stimulus trains. The minimal current that elicited reliable EPSPs (failure rate <10%) was used for low intensity stimuli. An amplitude three times threshold was used for high-intensity stimulation. For low-frequency stimulus trains, 10 stimuli were applied at 10 Hz with low stimulus intensities.
For analysis, 5–10 excitatory postsynaptic potential (EPSP) traces were averaged, and peak amplitude and area of response were measured. The latter was determined by fitting a monoexponential curve to the decaying phase of responses, using the intersection of the fit with the Vm baseline as the endpoint of the response.
Statistical analysis
Statistical significance was analyzed using paired t-test, ANOVA followed by Student Newman-Keuls post hoc test, Fisher's exact test and linear (Pearson) correlation test. If data were not normally distributed, the Mann-Whitney nonparametric statistical test was used. Throughout the text, values are expressed as means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Ollo and Schwartz 1979; Rietzel and Friauf 1998). Marginal cells and very small round cells (diameter <5 µm) were excluded from further analysis. Results presented here include data from 801 neurons from 72 slices from animals aged between P0 and P7. The average resting Ca2+ concentration of neurons included in this study was 92 ± 2 nM (n = 647). Synaptically elicited calcium responses
Electrical stimulation of the ipsilateral VAS, the major excitatory input from the AVCN to the LSO, consistently evoked Ca2+-responses in neurons throughout the LSO (Fig. 1). In accordance with previous electrophysiological studies (Caspary and Faingold 1989; Kandler and Friauf 1995; Kotak and Sanes 1995; Wu and Kelly 1992), VAS-elicited responses were glutamatergic because they were abolished by a mixture of the GluR antagonists CNQX (20 µM), APV (50–100 µM), and MCPG (1 mM; n = 45 cells, data not shown). The amplitudes of these Ca2+ responses increased with stimulation frequency (Fig. 1, B and C). Single electrical shocks elicited only small increases in [Ca2+]i (51 ± 7 nM, n = 54 cells), and responses returned to baseline within 2–5 s. Low-frequency stimulation (10 pulses at 5 and 10 Hz) elicited larger responses (138 ± 13 nM, n = 66 cells), and these responses were longer, returning to baseline within 5–15 s. High-frequency stimulation (10 pulses at 20, 50, and 100 Hz) elicited the largest responses (232 ± 19 nM, n = 66 cells) with the longest durations, returning slowly to the baseline within 3.5 to >15 s (limited by the duration of recordings).
The effects of stimulus frequency on response amplitude and duration could be due to frequency-dependent recruitment of the number of GluRs and/or the type of ionotropic and metabotropic GluRs. To address these possibilities, we investigated the recruitment of different GluR types as a function of stimulation frequency.
Contribution of AMPA and kainate receptors
To examine the contribution of AMPARs to the observed Ca2+ signals, the VAS was stimulated in the presence of the AMPA/kainate receptor antagonist CNQX (20 µM). Under these conditions, responses elicited by a single stimulus were nearly abolished (reduction: 92 ± 2%; peak amplitudes in control: 55 ± 8 nM, in CNQX: 4 ± 1 nM, n = 40 cells; paired t-test, P < 0.01; Fig. 2) while responses to low-frequency trains were reduced by 73 ± 3% (control: 170 ± 17 nM, CNQX: 54 ± 11 nM, n = 43 cells; paired t-test, P < 0.01). Responses elicited by high-frequency stimulation were the least affected by CNQX and were reduced by only 48 ± 3% (control: 277 ± 27 nM, CNQX: 148 ± 27 nM, n = 43 cells; paired t-test, P < 0.01). These data indicate that the relative contribution of AMPA/kainate receptors to Ca2+-responses decreases with increasing stimulus frequency (Fig. 2B).
|
Several nuclei in the mature and developing auditory brain stem express kainate receptors (Lachica et al. 1998; Lohrke and Friauf 2002; Petralia et al. 1994, 1996; review Petralia et al. 2000). To test whether these receptors contribute to an increase in [Ca2+], in LSO neurons, KA was applied in the presence of 25–50 µM GYKI 52466, an AMPAR antagonist (Lachica et al. 1998; Vizi et al. 1997). Under this condition, KA (5–25 µM, 30 s) increased [Ca2+]i (157 ± 23 nM, n = 37) and CNQX (20 µM) reduced these responses to 28 ± 5% of control (n = 37 cells, P0 - P3; P < 0.01; paired t-test; Fig. 2, C and D).
AMPA/kainate receptors can increase [Ca2+]i by two mechanisms: depolarization-induced activation of VGCCs or calcium influx through Ca2+-permeable AMPA/kainate receptors. We thus compared AMPA/kainate receptor-mediated responses before and after blockade of VGCCs with Ni2+ (2–5 mM, Gu et al. 1994) (Fig. 3). Ni2+ blocked responses elicited by AMPA (1 mM, 50 ms pressure application, n = 31 cells) or kainate in the presence of GYKI 52466 (n = 8 cells). Nickel (2 mM) had no effect on AMPA-elicited whole cell currents (P = 0.39, Student's t-test; n = 4 cells; Fig. 3C). This indicates that AMPA and kainate receptors in neonatal LSO neurons are not Ca2+ permeable, and thus that AMPA and kainate receptor-elicited Ca2+ responses are mediated via activation of VGCCs.
|
Contribution of NMDA receptors
The contribution of NMDA receptors (NMDARs) to synaptic Ca2+ responses was examined using the NMDAR antagonist APV (50–100 µM; Fig. 4, A and B). APV had no effect on very small responses, indicating that for small responses, NMDARs do not contribute to somatic calcium changes (control: 8 ± 2 nM, APV: 10 ± 4 nM; n = 6 cells; paired t-test, P > 0.1). For larger responses, however, APV reduced Ca2+ changes by 50%, and this reduction was independent of the stimulus train frequency (Fig. 4, A and B). The absence of NMDAR-mediated calcium influx at low response amplitudes suggests that NMDARs in LSO neurons are blocked by magnesium (Mayer et al. 1984; Nowak et al. 1984). In support of this, application of NMDA (50 µM, 60 s) failed to elicit Ca2+ responses at physiological Mg2+ concentrations (1.3 mM; n = 83 cells, P1–P4; Fig. 4C) but elicited strong Ca2+ responses in Mg2+-free ACSF (467 ± 87 nM, n = 83 cells) that were completely blocked by APV (50 µM, n = 83).
|
The 50% contribution of NMDARs to calcium responses was higher than expected based on previous electrophysiological studies indicating that ipsilateral excitation to LSO neurons is mediated primarily or exclusively by non-NMDARs (Caspary and Faingold 1989; Kandler and Friauf 1995; Kotak and Sanes 1996; Wu and Kelly 1992). We thus investigated the contribution of NMDARs to ipsilaterally elicited EPSPs in our preparation. These experiments revealed that in neonatal mice (P3-P5), NMDARs contribute to EPSP amplitudes between 35 and 64% and to response areas between 56 and 67%, depending on stimulus conditions (Table 1, Fig. 4D). For high-intensity stimuli and stimulus trains, these values agree well with those we observed on calcium level, but for low-intensity responses, the obvious NMDAR contribution was higher for EPSPs. Taken together, these results demonstrate that in neonatal mice, NMDARs contribute considerably to ipsilaterally elicited calcium and voltage responses.
|
Metabotropic glutamate receptors
To examine whether and to what extent mGluRs contribute to synaptic Ca2+ responses, the VAS was stimulated while blocking ionotropic GluR receptors with APV (50–100 µM) and CNQX (20 µM). These antagonists abolished Ca2+ responses elicited by single stimuli in all cells and abolished responses to low-frequency trains in about half the cells (51%, 23 of 45 cells). Responses to high-frequency trains were abolished in only 13% of cells (6 cells of the same 45 cells; Fisher's exact test, P = 0.0002). Application of the mGluR antagonist MCPG (1 mM) significantly reduced responses to low frequency stimuli that remained after APV/CNQX treatment (57% reduction, n = 39 cells; paired t-test, P < 0.01; Fig. 5A). In contrast, MCPG only slightly reduced APV/CNQX -resistant responses to low-frequency trains (from 16 ± 3 to 11 ± 2 nM, n = 22 cells; Fig. 5B; paired t-test, P > 0.05). These results indicate that mGluRs in LSO neurons are recruited by high-frequency stimulus trains. For individual cells, there was no correlation between the response amplitude and the contribution of mGluRs (Pearson correlation coefficient r = 0.18), suggesting that stimulation frequency and not response amplitude determines mGluR activation.
|
Pharmacological analysis of mGluR-mediated responses
Previous electrophysiological studies demonstrated that stimulation of mGluRs can induce long-lasting membrane depolarizations of neonatal LSO neurons (Kotak and Sanes 1995). We thus performed an additional series of pharmacological experiments to investigate the mechanisms underlying mGluR-mediated calcium responses in more detail. Metabotropic glutamate receptors comprise a large family of receptors that are categorized into three groups (group I–III) based on their pharmacology, structure, and activation of intracellular pathways (Pin and Duvoisin 1995). We first determined which of these mGluRs increase [Ca2+]i in LSO neurons using specific agonists.
GROUP I mGluRs. Activation of group I mGluRs by the specific agonist DHPG (10–20 µM) (Ito et al. 1992) or the group I/II agonist ACPD consistently increased [Ca2+]i (Fig. 6A). These responses had a distinctive profile consisting of a rapid, initial increase in [Ca2+]i (peak response) followed by a long-lasting increase (plateau response). The amplitude of the plateau was 35% of the amplitude of the peak response (average peak amplitude: 463 ± 26 nM, plateau amplitude: 162 ± 9 nM, n = 83 cells).
|
GROUP II mGluRs. Activation of group II mGluRs by LCCG-I (5–10 µM) (Keele et al. 1999; Maiese et al. 1999) or by the group II-agonist/group I-antagonist 4C3HPG (200 µM) (Hayashi et al. 1994; Keele et al. 1999), elicited Ca2+ responses with distinct peak and plateau components (Fig. 6B). Compared with group-I-elicited responses, peak responses of group II responses were significantly smaller (group II: 265 ± 28 nM, n = 45 cells; group I: 463 ± 26 nM, n = 83 cells, P < 0.01; t-test) while plateau responses had similar amplitudes (group II: 183 ± 17 nM, n = 45 cells; group I: 162 ± 9 nM, n = 83 cells; P > 0.1; t-test).
GROUP III mGluRs. The group III mGluR agonist L-AP4 (100–500 µM for 30–90 s) (Tanabe et al. 1993) never elicited Ca2+ responses (n = 42 cells; Fig. 6C). In the same slices, however, LSO neurons responded vigorously to KCl depolarizations (60 mM, 30 s), indicating that the absence of a response to L-AP4 was not due to decreased viability of these cells.
These results indicate that immature LSO neurons express group I and group II mGluRs, whose activation increases [Ca2+]i. Group III mGluRs, however, are either not expressed in LSO neurons or are not coupled to intracellular calcium mobilization.
Mechanisms of mGluR-elicited Ca2+ responses
The best-characterized pathway by which mGluRs increase [Ca2+]i is through activation of PLC and the production of IP3, which releases calcium from internal stores (Conn and Pin 1997). To test the contribution of internal calcium stores to mGluR-elicited calcium responses in LSO neurons, calcium was depleted from the endoplasmatic reticulum with thapsigargin (10 µM, 1- to 2-h incubation). This treatment completely blocked calcium responses after application of ACPD (100 µM; n = 33 cells, 2 slices; Fig. 6D), indicating that Ca2+ release from intracellular stores is necessary for mGluR-mediated responses.
To test the possible contribution of extracellular calcium, mGluRs were activated in Ca2+-free ACSF. Under this condition, ACPD still increased [Ca2+]i, but the initial peak was greatly reduced, and the plateau phase was completely abolished (peak reduced to 31 ± 8%; area reduced to 18 ± 5%, n = 9 cells, Fig. 6, E and F). Similar results were also observed with the group-I-specific agonists DHPG (peak reduced to 32 ± 2%, area reduced to 12 ± 1%, n = 82 cells) and the group-II-specific agonist 4C3HPG (peak reduced to 16 ± 2%, area reduced to 23 ± 3%, n = 19 cells). These results indicate that Ca2+ released from internal stores contributes only 30% to the peak response, whereas extracellular Ca2+ contributes 70% to the peak response and is the sole source of the plateau phase.
Because activation of mGluRs can induce membrane depolarizations (Crepel et al. 1994; Kotak and Sanes 1995), we also examined whether mGluRs activate VGCCs. To this end, we applied ACPD (20 µM) in the presence of specific antagonists to L-type (10 µM nifedipine), N-type (50–1,000 nM -conotoxin GVIA), and P/Q-type calcium channels (50 nM -agatoxin TK). None of these VGCC blockers significantly decreased the amplitudes of the peak or the plateau (Fig. 7) indicating that mGluR-activated influx of extracellular calcium does not involve VGCCs.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
Technical considerations
In this study, calcium measurements were made exclusively from cell bodies due to the relatively high background staining in bulk-labeled slices and the thin diameter of LSO dendrites, both of which prevented the reliable detection of small fluorescent changes in dendritic segments.
Because in mature LSO neurons most glutamatergic inputs are located on dendrites, caution must be used in applying calcium measurements obtained from the soma to the dendrites. For example, due to the smaller intracellular volume of dendrites, the same amount of calcium influx will produce a larger increase in [Ca2+]i in dendrites than in cell bodies. As seen in other neuronal types (Berridge 1998; Magee et al. 1998), VGCCs and intracellular calcium mobilization mechanism might be unequally distributed along single LSO neurons, resulting in locally distinct calcium responses. Thus while our approach allows us to draw basic conclusions about specific GluR activation, detailed dendritic calcium imaging data are necessary to understand the properties of synapse-specific calcium responses as well as synaptic interaction and integration from the standpoint of local calcium dynamics.
AMPA/kainate receptor-mediated Ca2+ responses
Single stimuli to the VAS elicited synaptic calcium responses that were primarily mediated by AMPARs. This agrees with previous electrophysiological studies in developing and adult animals that demonstrated that ipsilateral excitation to the LSO is primarily conveyed by non-NMDARs (Caspary and Faingold 1989; Kotak and Sanes 1996; Wu and Kelly 1992). Our results further demonstrate that these AMPAR-mediated Ca2+ responses result from activation of VGCCs rather than from Ca2+ influx though AMPARs. Thus AMPARs in neonatal mouse LSO neurons are Ca2+ impermeable and most likely contain GluR2 subunits (Hollmann et al. 1991; Verdoorn et al. 1991). This is consistent with previous studies that demonstrated the expression of GluR2 subunits during circuit maturation in the LSO of neonatal rats (Caicedo and Eybalin 1999; Caicedo et al. 1998) and the nucleus magnocellularis of embryonic chickens (Lawrence and Trussell 2000; Sugden et al. 2002). The calcium impermeability of GluR2-containing AMPARs reduces calcium entry during low synaptic activity. This could restrict activation of calcium-dependent intracellular pathways to periods of high synaptic activity, when, in addition to VGCCs, NMDARs, and mGluRs are also activated. Because GluR2 subunits play an important role in regulating activity-dependent synaptic plasticity (Jia et al. 1996) and internalization and stabilization of synaptic AMPARs (Borgdorff and Choquet 2002; Kim et al. 2001; Man et al. 2000), expression of GluR2 early in auditory brain stem development could play a role in synaptic maturation and organization.
Calcium responses mediated by kainate receptors
Kainate receptors are expressed in several nuclei of the mature and developing auditory brain stem (Lachica et al. 1998; Lohrke and Friauf 2002; Petralia et al. 1994, 1996; reviewed in Petralia et al. 2000), but their role in auditory processing or neuronal development is poorly understood. Preliminary results indicated that kainate receptors can elicit low-amplitude inward currents in developing LSO neurons (Vitten et al. 1999). Our results confirm and expand these previous studies by demonstrating that kainate receptors in the LSO are calcium impermeable and are thus most likely composed of the Q/R-edited forms of GluR5 and GluR6 subunits (Burnashev et al. 1995; Lee et al. 2001). It remains to be shown to what extent these kainate receptors participate in glutamatergic synaptic transmission (Frerking et al. 1998; Kidd and Isaac 1999) and, if they do, whether such participation is frequency-dependent or perhaps restricted to specific glutamatergic afferent pathways (DeVries and Schwartz 1999; Frerking and Nicoll 2000).
NMDAR-mediated Ca2+ responses
Previous studies indicated that LSO neurons express mRNA and protein of NMDARs (Caicedo and Eybalin 1999; Sato et al. 1999). Our results demonstrate that these NMDARs can be activated by synaptic inputs from cochlear nucleus fibers and that they contribute significantly to postsynaptic Ca2+ responses. Consistent with the expression of NR2A subunits (Caicedo and Eybalin 1999; Sato et al. 1999), we found that NMDARs in developing LSO neurons are magnesium sensitive, providing an explanation as to why NMDARs contributed to calcium responses only in larger responses, which most likely were accompanied by strong membrane depolarizations.
Once activated, NMDARs accounted for 50% of the calcium influx and 60% of the EPSP area. This contribution is higher than that reported in other studies, which indicated that ipsilateral excitation to LSO neurons is mediated primarily or exclusively by non-NMDARs (Caspary and Faingold 1989; Kandler and Friauf 1995; Kotak and Sanes 1996; Wu and Kelly 1992). However, because our animals were 
1 week younger than those used in previous studies, this difference most likely indicates that, as in many other brain regions, NMDAR-mediated responses are downregulated during maturation of the LSO. Because NMDAR-mediated calcium influx plays a central role in neuronal development, plasticity, and cell death in several systems (West et al. 2002; Wong and Ghosh 2002), the large contribution of NMDARs to calcium responses here suggests that NMDARs play an important part in the maturation of LSO neurons as well.
Metabotropic glutamate receptor-mediated Ca2+ responses
Based on subunit composition, pharmacology, and coupling to signal transduction pathways, mGluRs are classified into three categories, group I, II, and III (reviewed in Conn and Pin 1997). Group I mGluRs (mGluR1 and mGluR5) are coupled to Gq-type G proteins, which activate the PLC-IP3 pathway and mobilize intracellular calcium. Group II mGluRs (mGluR2 and mGluR3) and group III mGluRs (mGluR4 and mGluR6–8) are associated with Gi/o-type proteins and generally are negatively coupled to adenylate cyclase activity.
In the auditory brain stem, group I and II mGluRs are expressed in several auditory nuclei (Bilak and Morest 1998; Elezgarai et al. 1999, 2001; Jaarsma et al. 1998; Molitor and Manis 1997; Schwartz and Eager 1999; Takahashi et al. 1996; Zirpel et al. 2000; reviewed in Petralia et al. 2000). In the LSO, electrophysiological recordings have indicated expression of group II and/or III mGluRs on presynaptic glutamatergic terminals (Wu and Fu 1998) and group I mGluRs on postsynaptic LSO neurons (Kotak and Sanes 1995). Our results extend these findings by demonstrating that synaptic activation of postsynaptic group I and II mGluRs contributes to calcium responses in immature LSO neurons. These calcium responses were only observed with high-frequency stimulation, which is consistent with a perisynaptic location (Baude et al. 1993; Petralia et al. 1997; Takumi et al. 1999) and activation by glutamate spillover (Huang 1998; Isaacson 2000).
Stimulation of group I, as well as group II, mGluRs increased [Ca2+]i. The responses elicited by group II mGluR agonists were unexpected because group II mGluRs usually modulate adenylate cyclase activity rather than increase IP3. It is unlikely that these responses were due to unspecific activation of group I mGluRs because [Ca2+]i also increased with 4C3HPG, which is not only an agonist for group II mGluRs but is also a potent antagonist for group I mGluRs (Hayashi et al. 1994). Although the exact intracellular pathways that underlie group II-elicited calcium responses remain unclear, the similarity of group II to group I responses suggests the involvement of IP3-dependent calcium mobilization. Very recently, group II mGluR-mediated activation of the PLC-IP3 pathway has been demonstrated in prefrontal cortex neurons (Otani et al. 2002), where group II mGluRs are critically involved in mediating activity-dependent synaptic plasticity.
Calcium responses elicited by mGluRs were characterized by a rapid, initial increase in [Ca2+]i that was followed by a lower-amplitude, prolonged calcium plateau. These responses were completely abolished if calcium stores were depleted by thapsigargin, indicating that calcium mobilization from internal stores is necessary for these responses. Recordings in zero extracellular calcium, however, indicated that mobilization of intracellular calcium contributed only a minor fraction (30%) to the initial peak response. This is consistent with the idea that mobilization of intracellular calcium activates membrane calcium channels and causes an influx of extracellular calcium. VGCCs do not contribute to this calcium influx because responses were unaffected by specific antagonists of VGCCs. Potential candidates for this channel are members of the transient receptor potential channels, some of which are activated by mobilization of intracellular calcium (Clapham et al. 2001). Further experiments to characterize this calcium channel in LSO neurons are important in light of its large contribution to mGluR-mediated calcium responses during the period when the survival of LSO neurons depends on calcium (Lohmann et al. 1998) and synaptic maturation depends on glutamatergic inputs (Kotak and Sanes 1997).
Stimulus-dependent recruitment of different classes of glutamate receptors
Prior to hearing onset, spontaneous activity is present in the avian and mammalian central auditory pathway (Glowatzki and Fuchs 2000; Gummer and Mark 1994; Jones and Perez 2001; Kotak and Sanes 1995; Kros et al. 1998; Lippe 1994; Romand and Ehret 1990). Accumulating evidence indicates that this spontaneous activity plays a critical role in neuronal survival (Lachica et al. 1995; Zirpel and Rubel 1996) and in the normal physiological and anatomical development of central auditory connections (Gabriele et al. 2000; Kitzes et al. 1995; Russell and Moore 1995; reviewed in Sanes and Friauf 2000). Spontaneous activity in the immature auditory system is rhythmic and oscillatory, consisting of bursts of activity that are separated by periods of low-frequency events (Jones and Perez 2001; Kotak and Sanes 1995; Lippe 1994), a pattern that is reminiscent of the activity patterns in other developing systems (reviewed in Zhang et al. 2000). With the exception of the visual system, in which burst-like synaptic activity in the form of retinal waves is thought to play an instructive role in the formation of eye-specific layers in the lateral geniculate nucleus (Stellwagen and Shatz 2002), the exact function of the burst-like pattern of spontaneous activity has remained largely elusive. Our results demonstrate that distinct temporal patterns of synaptic activity in the LSO activate distinct calcium-entry pathways (Fig. 8) and, most likely, distinct cellular pathways and gene expression (West et al. 2002). Although the exact nature of these intracellular events remains to be shown, our results are consistent with the idea that different patterns of afferent activity serve distinct developmental roles, highlighting the importance of the temporal pattern of spontaneous activity in the immature auditory system.
|
DISCLOSURES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: K. Kandler, Dept. of Neurobiology, University of Pittsburgh School of Medicine, W1447 Biomedical Science Tower, 3500 Terrace St., Pittsburgh, PA 15261 (E-mail: kkarl{at}pitt.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Berridge MJ. Neuronal calcium signaling. Neuron 21: 13–26, 1998.[ISI][Medline]
Betz WJ and Bewick GS. Optical monitoring of transmitter release and synaptic vesicle recycling at the frog neuromuscular junction. J Physiol 460: 287–309, 1993.
Bilak SR and Morest DK. Differential expression of the metabotropic glutamate receptor mGluR1alpha by neurons and axons in the cochlear nucleus: in situ hybridization and immunohistochemistry. Synapse 28: 251–270, 1998.[ISI][Medline]
Borgdorff AJ and Choquet D. Regulation of AMPA receptor lateral movements. Nature 417: 649–653, 2002.[Medline]
Boudreau JC and Tsuchitani C. Cat superior olive S-segment cell discharge to tonal stimulation. Contrib Sens Physiol 4: 143–213, 1970.[Medline]
Burnashev N, Zhou Z, Neher E, and Sakmann B. Fractional calcium currents through recombinant GluR channels of the NMDA, AMPA, and kainate receptor subtypes. J Physiol 485: 403–418, 1995.[ISI][Medline]
Caicedo A and Eybalin M. Glutamate receptor phenotypes in the auditory brainstem and mid-brain of the developing rat. Eur J Neurosci 11: 51–74, 1999.[ISI][Medline]
Caicedo A, Kungel M, Pujol R, and Friauf E. Glutamate-induced Co2+ uptake in rat auditory brain stem neurons reveals developmental changes in Ca2+ permeability of glutamate receptors. Eur J Neurosci 10: 941–954, 1998.[ISI][Medline]
Caird D and Klinke R. Processing of binaural stimuli by cat superior olivary complex neurons. Exp Brain Res 52: 385–399, 1983.[ISI][Medline]
Caspary DM and Faingold CL. Non-N-methyl-D-aspartate receptors may mediate ipsilateral excitation at lateral superior olivary synapses. Brain Res 503: 83–90, 1989.[ISI][Medline]
Clapham DE, Runnels LW, and Strubing C. The TRP ion channel family. Nat Rev Neurosci 2: 387–396, 2001.[ISI][Medline]
Conn PJ and Pin JP. Pharmacology and functions of metabotropic glutamate receptors. Annu Rev Pharmacol Toxicol 37: 205–237, 1997.[ISI][Medline]
Crepel V, Aniksztejn L, Ben Ari Y, and Hammond C. Glutamate metabotropic receptors increase a Ca(2+)-activated nonspecific cationic current in CA1 hippocampal neurons. J Neurophysiol 72: 1561–1569, 1994.
DeVries SH and Schwartz EA. Kainate receptors mediate synaptic transmission between cones and "off" bipolar cells in a mammalian retina. Nature 397: 157–160, 1999.[Medline]
Elezgarai I, Benitez R, Mateos JM, Lazaro E, Osorio A, Azkue JJ, Bilbao A, Lingenhoehl K, Van Der PH, Hampson DR, Kuhn R, Knopfel T, and Grandes P. Developmental expression of the group III metabotropic glutamate receptor mGluR4a in the medial nucleus of the trapezoid body of the rat. J Comp Neurol 411: 431–440, 1999.[ISI][Medline]
Elezgarai I, Bilbao A, Mateos JM, Azkue JJ, Benitez R, Osorio A, Diez J, Puente N, Donate-Oliver F, and Grandes P. Group II metabotropic glutamate receptors are differentially expressed in the medial nucleus of the trapezoid body in the developing and adult rat. Neuroscience 104: 487–498, 2001.[ISI][Medline]
Franklin JL and Johnson EM. Suppression of programmed neuronal death by sustained elevation of cytoplasmic calcium. Trends Neurosci 15: 501–508, 1992.[ISI][Medline]
Frerking M, Malenka RC, and Nicoll RA. Synaptic activation of kainate receptors on hippocampal interneurons. Nat Neurosci 1: 479–486, 1998.[ISI][Medline]
Frerking M and Nicoll RA. Synaptic kainate receptors. Curr Opin Neurobiol 10: 342–351, 2000.[ISI][Medline]
Gabriele ML, Brunso-Bechtold JK, and Henkel CK. Plasticity in the development of afferent patterns in the inferior colliculus of the rat after unilateral cochlear ablation. J Neurosci 20: 6939–6949, 2000.
Glowatzki E and Fuchs PA. Cholinergic synaptic inhibition of inner hair cells in the neonatal mammalian cochlea. Science 288: 2366–2368, 2000.
Grynkiewicz G, Poenie M, and Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 260: 3440–3450, 1985.
Gu X, Olson EC, and Spitzer NC. Spontaneous neuronal calcium spikes and waves during early differentiation. J Neurosci 14: 6325–6335, 1994.[Abstract]
Gummer AW and Mark RF. Patterned neural activity in brain stem auditory areas of a prehearing mammal, the tammar wallaby (Macropus eugenii). Neuroreport 5: 685–688, 1994.[ISI][Medline]
Hayashi Y, Sekiyama N, Nakanishi S, Jane DE, Sunter DC, Birse EF, Udvarhelyi PM, and Watkins JC. Analysis of agonist and antagonist activities of phenylglycine derivatives for different cloned metabotropic glutamate receptor subtypes. J Neurosci 14: 3370–3377, 1994.[Abstract]
Helfert RH and Schwartz IR. Morphological features of five neuronal classes in the gerbil lateral superior olive. Am J Anat 179: 55–69, 1987.[ISI][Medline]
Hollmann M, Hartley M, and Heinemann S. Ca2+ permeability of KA-AMPA–gated glutamate receptor channels depends on subunit composition. Science 252: 851–853, 1991.
Huang EP. Synaptic transmission: spillover at central synapses. Curr Biol 8: R613–R615, 1998.[ISI][Medline]
Isaacson JS. Spillover in the spotlight. Curr Biol 10: R475–R477, 2000.[ISI][Medline]
Ito I, Kohda A, Tanabe S, Hirose E, Hayashi M, Mitsunaga S, and Sugiyama H. 3, 5-Dihydroxyphenyl-glycine: a potent agonist of metabotropic glutamate receptors. Neuroreport 3: 1013–1016, 1992.[ISI][Medline]
Jaarsma D, Dino MR, Ohishi H, Shigemoto R, and Mugnaini E. Metabotropic glutamate receptors are associated with non-synaptic appendages of unipolar brush cells in rat cerebellar cortex and cochlear nuclear complex. J Neurocytol 27: 303–327, 1998.[ISI][Medline]
Jia Z, Agopyan N, Miu P, Xiong Z, Henderson J, Gerlai R, Taverna FA, Velumian A, MacDonald J, Carlen P, Abramow-Newerly W, and Roder J. Enhanced LTP in mice deficient in the AMPA receptor GluR2. Neuron 17: 945–956, 1996.[ISI][Medline]
Jones SJ and Perez N. The auditory "C-process": analyzing the spectral envelope of complex sounds. Clin Neurophysiol 112: 965–975, 2001.[ISI]
; *, P < 0.01; paired t-test). B: reduction of Ca2+-responses by CNQX. Responses elicited by single stimuli are almost abolished by CNQX. Responses elicited by low- and high-frequency stimulation are reduced 
) and after (
