- Email: [email protected]
Long-Term Potentiation of Neuronal Glutamate Transporters
Neuron, Vol. 46, 715–722, June 2, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.neuron.2005.04.033
Long-Term Potentiation of Neuronal Glutamate Transporters Ying Shen and David J. Linden* Department of Neuroscience The Johns Hopkins University School of Medicine 725 North Wolfe Street Baltimore, Maryland 21205
Summary Persistent, use-dependent modulation of synaptic strength has been demonstrated for fast synaptic transmission mediated by glutamate and has been hypothesized to underlie persistent behavioral changes ranging from memory to addiction. Glutamate released at synapses is sequestered by the action of excitatory amino acid transporters (EAATs) in glia and postsynaptic neurons. So, the efficacy of glutamate transporter function is crucial for regulating glutamate spillover to adjacent presynaptic and postsynaptic receptors and the consequent induction of plastic or excitotoxic processes. Here, we report that tetanic stimulation of cerebellar climbing fiber-Purkinje cell synapses results in long-term potentiation (LTP) of a climbing fiber-evoked glutamate transporter current recorded in Purkinje cells. This LTP is postsynaptically expressed and requires activation of an mGluR1/PKC cascade. Together with a simultaneously induced long-term depression (LTD) of postsynaptic AMPA receptors, this might reflect an integrated antiexcitotoxic cellular response to strong climbing fiber synaptic activation, as occurs following an ischemic episode.
Report
both presynaptic and postsynaptic compartments (Asztely et al., 1997; Carter and Regehr, 2000; Brasnjo and Otis, 2001; Oliet et al., 2001; Arnth-Jensen et al., 2002; DiGregorio et al., 2002; Dzubay and Otis, 2002; Reichelt and Knopfel, 2002; Huang and Bordey, 2004; Huang et al., 2004a; Takayasu et al., 2004; Stoffel et al., 2004). The sequelae of transporter-regulated spillover signaling are many and include triggering of both (presumably) adaptive plastic processes such as LTP and LTD (Brasnjo and Otis, 2001; Tsvetkov et al., 2004) as well as pathological processes such as excitotoxicity and epileptiform discharges (Tanaka et al., 1997b; Maragakis and Rothstein, 2004). The cerebellar climbing fiber-Purkinje cell synapse is an advantageous model system for the study of neuronal transporter currents (Otis et al., 1997; Auger and Attwell, 2000). Most Purkinje cells are innervated by a single climbing fiber axon, which ramifies to form w1500 release sites. Each of these sites appears to release multiple vesicles of glutamate with each action potential invasion (Wadiche and Jahr, 2001). A single shock delivered to the climbing fiber axon results in the activation of AMPA and kainate receptors (Huang et al., 2004b), and when these are blocked, a synaptic glutamate transporter current is revealed. Recently, this has been shown to be mediated by the transporter EAAT4 (Huang et al., 2004b), which is strongly expressed in the perisynaptic membranes of climbing fiber-Purkinje cell synapses (Tanaka et al., 1997a). Here, we have recorded climbing fiber-evoked synaptic transporter currents in Purkinje cells to test the hypothesis that neuronal glutamate transport can undergo persistent use-dependent changes in efficacy.
Introduction Results Glutamate released from nerve terminals into the synaptic cleft is subject to diffusion and dilution in the extracellular space as well as uptake into neurons and glia by excitatory amino acid transporters (EAATs). The efficacy of glutamate transport is crucial not only for maintaining the background concentration of glutamate at a low level, but also for determining the amplitude and kinetics of glutamate transients that result from vesicular release. At most synapses, glutamate transporters have a relatively small effect on the rise and peak of the glutamate transient as experienced by receptors in the postsynaptic density directly across the synaptic cleft from the fusion event. However, transporters localized to the perisynaptic membrane of the postsynaptic neuron as well as those on ensheathing glial cells have a major role in controlling the “spillover” of glutamate to adjacent structures (see Huang and Bergles, 2004, for review). Glutamate spillover is enhanced with high-frequency burst activation of afferent fibers, and the control of this spillover by transporters has been shown to regulate the activation of ionotropic and metabotropic glutamate receptors in *Correspondence: [email protected]
Whole-cell voltage-clamp recordings were made from Purkinje cells in sagittal slices of juvenile rat cerebellum. In order to isolate the stoichiometric portion of synaptic glutamate transporter current, recordings were made using a gluconate-based internal saline, as this species does not permeate during the uncoupled anion channel mode of the transporter (Wadiche et al., 1995; Wadiche and Kavanaugh, 1998). Stimulation of the climbing fiber evokes a large EPSC (w1–2 nA, recorded with Vhold = −15 mV), which is strongly inhibited by NBQX (100 M), an antagonist of AMPA and kainate receptors (Figure 1A). Functional NMDA receptors are not expressed on Purkinje cells after the first postnatal week (Häusser and Roth, 1997). The current recorded in NBQX is small (40.6 ± 2.1 pA, recorded with Vhold = −70 mV; n = 16) and has similar onset kinetics to the control EPSC (10%–90% rise time; control EPSC: 1.3 ± 0.1; NBQX-resistant current: 1.6 ± 0.1 ms) but somewhat slower offset kinetics (50% decay time, 8.0 ± 0.6 versus 11.4 ± 1.1 ms; n = 25). Addition of the drug DLthreo-β-benzoylaspartic acid (TBOA, 300 M), a selective inhibitor of glutamate transporters (Shimamoto et al., 1998; Shigeri et al., 2001), strongly attenuated the
Neuron 716
Figure 1. The Climbing Fiber-Evoked Synaptic Current Recorded in a Purkinje Cell Bathed in a High Concentration of NBQX Is Predominantly Mediated by Glutamate Transport (A) Single shocks delivered to the granule cell layer evoke all-or-none climbing fiber responses recorded in Purkinje cells. A control EPSC trace, in which AMPA receptor-mediated current predominates, was recorded with Vhold = −15 mV and is superimposed upon a scaled recording from the same cell in which 100 M NBQX has been added and Vhold changed to −70 mV. (B) Vhold = −70 mV. NBQX (100 M), TBOA (300 M), and UBP-301 (70 M) were added serially. Each trace is the average of three evoked responses. Bar graphs depict the mean ± SEM of 16 cells. (C) A trace recorded in the presence of the specific AMPA receptor antagonist GYKI 53655 (100 M) in which kainate receptormediated current predominates (Huang et al., 2004b) is superimposed upon a scaled recording in the presence of NBQX (100 M) plus TBOA (300 M), which shows the NBQX-resistant fraction of kainate current.
residual current, leaving 9.4 ± 0.5 pA unblocked by NBQX plus TBOA (Figure 1B), similar to previous findings (Brasnjo and Otis, 2004; Huang et al., 2004b). A recent report has shown that this residual component is mediated by NBQX-resistant GluR5-containing kainate receptors (Huang et al., 2004b). Consistent with this finding, we observed that the current remaining in NBQX plus TBOA was completely obliterated (1.0 ± 0.4 pA) by the kainate receptor antagonist UBP-301 (70 M; More et al., 2003). Furthermore, the kinetics of the current recorded in NBQX + TBOA (10%–90% rise time = 3.1 ± 0.6 ms; 50% decay time = 19.4 ± 2.4 ms; n = 9) were similar to that of the much larger total kainate current recorded in the presence of the AMPA receptor-specific antagonist GYKI 53566 (100 M) plus TBOA (10%–90% rise time = 3.0 ± 0.5 ms; 50% decay time = 19.3 ± 1.3 ms; n = 9; Figure 1C). Previous work has shown that tetanic activation of the climbing fiber-Purkinje cell synapse (5 Hz, 30 s) can give rise to long-term synaptic depression (LTD), which requires activation of an mGluR1/PKC cascade (Hansel and Linden, 2000) and which is expressed postsynaptically as a downregulation of AMPA receptors (Shen et al., 2002). We used this same stimulation protocol to see if glutamate transporter function would be persistently changed (Figure 2). In this experiment, external saline supplemented with NBQX (100 M) + UBP-301 (70 M) was applied locally to the slice surface with a flow pipe to block AMPA and kainate receptors, and the remaining transporter currents were measured using test pulses delivered every 20 s for 3 min with Vhold = −70 mV. Following this, Vhold was switched to −15 mV (this improves the quality of the voltage clamp for AMPA-EPSCs by reducing driving force and inactivating voltage-gated channels), and NBQX and UBP301 were washed out until a stable baseline was achieved. Then, tetanic stimulation (5 Hz, 30 s) was delivered and test pulses were resumed. This gave rise to
a persistent depression of the AMPA receptor-mediated EPSC (peak amplitude: 72% ± 3% of baseline at t = 19 min, n = 7), which was not associated with a significant change in paired-pulse ratio (pre, t = −1 min: 0.62 ± 0.02; post, t = 20 min: 0.64 ± 0.02; interpulse interval = 80 ms), similar to that previously reported (Hansel and Linden, 2000; Shen et al., 2002). When NBQX and UBP301 were reintroduced (and Vhold was returned to −70 mV), it was revealed that the transporter current had undergone LTP (peak amplitude: 123% ± 8% of baseline; charge transfer: 146% ± 14% of baseline; t = 25 min). The potentiated transporter currents were completely abolished by subsequent addition of the antagonist TBOA, arguing against the recruitment of a novel conductance by 5 Hz stimulation. When this experiment was repeated at elevated temperature (32°C; n = 5), very similar results were seen (see Figure S1 in the Supplemental Data available online). Interestingly, the degree of LTD of AMPA-EPSCs was positively correlated with the degree of transporter current LTP (p = 0.0004), suggesting that these two phenomena might result from a common signaling cascade. Neither changes in series resistance (Rs; p = 0.48) nor input resistance (Ri; p = 0.58) were significantly correlated with transporter current LTP (Figure S2). It is important to note the caveat that while we presume that transporter current was rapidly potentiated and that this potentiation was then revealed by addition by NBQX and UBP-301 w20 min later, we do not have measurements of NBQX-resistant current during the period t = 0–20 min. Experiments were also performed with NBQX present throughout, but these failed to show potentiation of NBQX-resistant current at any time point (peak amplitude: 96% ± 4% of baseline; charge transfer: 93% ± 3% of baseline; t = 15 min; Figure S3). This is consistent with the finding that LTD of AMPA receptor-mediated currents in cultured Purkinje cells also requires AMPA receptor activation in addition to mGluR1 activation (Linden et al., 1993).
LTP of Neuronal Glutamate Transporters 717
Figure 2. Brief Tetanic Stimulation of Climbing Fibers Gives Rise to LTP of Transporter Current and Simultaneous LTD of AMPA Receptor-Mediated EPSCs in Purkinje Cells (A) These time-course data show the mean ± SEM of each group. Test pulses were delivered to activate climbing fibers, and responses were recorded with Vhold = −70 mV in the presence of NBQX (100 M) plus UBP301 (70 M), delivered locally with a flow pipe (transporter current, IEAAT, open squares). To record AMPA receptor-mediated EPSCs (filled circles), NBQX and UBP-301 were washed out, and Vhold was set to −15 mV. Tetanic stimulation (5 Hz, 30 s) was delivered at t = 0 min, as indicated by the upward arrows, and test pulses were resumed until t = 20 min, when NBQX and UBP-301 were reintroduced, and later Vhold was returned to −70 mV to allow for posttetanus measurements of transporter current. Control group, n = 7. Drugs added prior to patch formation: chelerythrine (10 M), n = 5; CPCCOEt (100 M), n = 5; phorbol-12,13-diacetate (PDA, 8 M), n = 5. Error bars show mean ± SEM. (B) Representative traces recorded from a control cell at the time points indicated in (A). (C) Population measures of currents recorded in the experiments shown in (A). Percent of baseline EPSC measures are derived from the ratio of measurements made at t = 15–20 min to that at t = −1–0 min. *p < 0.05 compared to baseline, ***p < 0.005. Error bars show mean ± SEM.
Pretreatment in the bath using either the PKC inhibitor chelerythrine (n = 5) or the group I mGluR antagonist CPCCOEt (n = 5) produced a complete blockade of LTP of transporter current as well as a complete blockade of LTD of AMPA receptor-mediated EPSCs. Similar effects were produced by pretreatment with the PKC activator phorbol-12,13-diacetate (PDA, 8 M; n = 5). The effects of PDA pretreatment could result from either occlusion of subsequent AMPA receptor LTD and transporter current LTP or blockade of the induction these phenomena resulting from downregulation of mGluR1. If an mGluR1/PKC cascade is required for LTP of NBQX-resistant current, then perhaps exogenous PKC activators would mimic the effects of tetanic climbing fiber stimulation. Bath application of PDA produced a potentiation of NBQX-resistant synaptic current (Figure 3A, upper right; n = 6, measured at t = 18 min after PDA application) and, in a separate population of Purkinje cells, a depression of AMPA receptor mediated EPSCs (Figure 3A, upper left; n = 7). A similar effect was produced by application of the PKC activator (−)-indolactam-V in the patch pipette and comparing time points soon after break in (t = 3 min) and somewhat later, to allow for dendritic perfusion (t = 24 min; Figure 4). Perfusion of the inactive enantiomer, (+)-indolactam-V, had no effect on isolated transporter currents recorded in NBQX + UBP-301 (n = 6).
Are the effects of these exogenous PKC activators mediated by activation of PKC in the Purkinje cell or in some other compartment? Neither bath PDA nor postsynaptic (−)-indolactam-V produced a significant change in Pr as indexed by EPSC paired-pulse ratios, arguing against an effect in climbing fiber terminals (PDA: pre, t = 10 min, 0.64 ± 0.04; post, t = 30 min, 0.64 ± 0.03; (−)-indolactam-V, pre, t = 3 min, 0.60 ± 0.03; post, t = 24 min, 0.60 ± 0.03). Furthermore, perfusion of the Purkinje cell with the cell-impermeant PKC inhibitor peptide PKC 19-36 (10 M) completely blocked the potentiation of transporter current produced by subsequent bath application of PDA (Figure 3A, lower left, n = 5). This last experiment argues that the locus of PKC activation to evoke these effects is the Purkinje cell dendrite. The NBQX-resistant current is predominantly (w78%) carried by EAAT4 glutamate transporters, with the remainder carried by NBQX-resistant kainate receptors. Could kainate current contribute to the potentiation produced by PKC activation (Figure 3) as recorded in NBQX? Previous work in a perirhinal slice preparation showed that bath application of a group I mGluR1 agonist could transiently potentiate synaptic kainate currents in a manner that required PKC activation (Cho et al., 2003). To address this, PDA was applied in two different recording configurations. First, NBQX plus TBOA
Neuron 718
Figure 3. PDA, an Exogenous PKC Activator, Mimics the Effects of Synaptic Tetani on AMPA Receptor-Mediated EPSCs and Transporter Currents, but Not Kainate Current in Purkinje Cells (A) Sample traces recorded before and 18 min after bath application of the exogenous PKC activator PDA (8 M). Each trace is the average of three consecutive records. Recordings were made in separate populations of cells in control conditions (to record EPSCs; n = 7), NBQX (100 M; n = 6), GYKI 53655 (100 M) + TBOA (300 M; n = 7), or NBQX (100 M) plus TBOA (300 M; n = 5). An additional group was recorded in NBQX using cells in which the membrane-impermeant PKC inhibitor peptide PKC 19-36 was included in the pipette at a concentration of 10 M and allowed to perfuse for >17 min prior to PDA application (n = 5). (B) Population measures of currents recorded in the experiments shown in (A) and (B). *p < 0.05, **p < 0.01, ***p < 0.005. Error bars show mean ± SEM.
was used to isolate the small NBQX-resistant kainate current. In this recording configuration, NBQX-resistant kainate current was found to be unaltered by bath application of PDA (Figure 3A, middle right, n = 5, t = 18 min). Second, GYKI 53655 plus TBOA was used to isolate the (much larger) total kainate current (Huang et al.,
2004a), which was also found to be unaltered by PDA (Figure 3A, middle left, n = 7, t = 18 min). We conclude that PKC activation produced by tetanic stimulation of climbing fibers or bath application of a PKC-activating phorbol ester produces LTP of neuronal glutamate transporter current and LTD of AMPA receptor-mediFigure 4. (−)-indolactam-V, a PKC Activator Added to the Internal Saline, Mimics the Effects of Synaptic Tetani on AMPA ReceptorMediated EPSCs and Transporter Currents in Purkinje Cells (A) Sample traces recorded with the PKC activator (−)-indolactam-V (300 M, n = 6) in the pipette. Recordings were made 3 min after achieving a stable whole-cell recording configuration (pre) or after 24 min of internal perfusion [(−)-indolactam]. This method may underestimate the amplitude of the indolactam effect, as it is not certain that there was no action of indolactam at the “pre” time point. (B) A similar experiment performed using the inactive control compound (+)-indolactam-V (300 M, n = 6). (C) Population measures of currents recorded in the experiments shown in (A) and (B). *p < 0.05. Error bars show mean ± SEM.
LTP of Neuronal Glutamate Transporters 719
ated EPSCs, but no change in kainate receptor-mediated EPSCs in the Purkinje cell. The present findings suggest a model in which postsynaptic receptors and transporters are differentially regulated by climbing fiber tetanization. However, it is formally possible that the increases in neuronal glutamate transport current reflect an increase in glutamate release. This increase would then have to be coupled with a large reduction in AMPA receptor function sufficient to yield LTD of the EPSC and a reduction in kainate receptor function that exactly offsets the increase in glutamate release. However, several lines of evidence argue against increases in transmitter release in our experiments. First, the LTD of AMPA-EPSCs produced by climbing fiber tetanization, bath application of PDA, or postsynaptic application of (−)-indolactam-V is not associated with changes in paired-pulse ratio, consistent with previous reports (Hansel and Linden, 2000; Shen et al., 2002). Second, LTD of AMPAEPSCs is not associated with a change in the degree of block by the low-affinity competitive AMPA receptor antagonist γ-D-glutamylglycine (Shen et al., 2002). Third, potentiation of neuronal transporter current evoked by PDA could be blocked by postsynaptic application of a membrane-impermeant PKC inhibitor. This indicates that the relevant locus of PKC activation is the Purkinje cell dendrite, thereby necessitating a retrograde messenger to increase glutamate release. To directly test the idea that neuronal glutamate transport is postsynaptically regulated in Purkinje cells, test pulses consisting of photolytic release of caged glutamate with a UV laser were delivered to Purkinje cells bathed in NBQX (100 M) and UBP-301 (70 M). The diameter of the uncaging spot was set to cover the entire Purkinje cell dendritic arbor, and the duration and intensity of the flash were adjusted to have the responses in NBQX + UBP-301 roughly approximate the amplitude of synaptically evoked transporter current and to be subthreshold for activation of mGluR1 (Figures 5A and 5B). It should be noted that this uncaging flash will activate a broad spatial range of glutamate transporters on Purkinje cells, not just those triggered by climbing fiber stimulation, but also those that might be triggered by parallel fiber synapses, which outnumber climbing fiber synapses by approximately 100:1. UV light flashes produced inward currents of 57.7 ± 5.8 pA (recorded with Vhold = −70 mV; n = 12), which had w2fold slower kinetics compared with synaptically evoked transporter current (10%–90% rise: 4.7 ± 0.5 ms; 50% decay: 16.9 ± 1.6 ms). These responses were stable upon repeated measurement at a frequency of 0.05 Hz (103% ± 4% of baseline at t = 27 min; n = 5). Application of PDA in the bath produced a significant potentiation of test pulse responses (147% ± 9% of baseline at t = 27 min; n = 7). Potentiation was not associated with changes in the kinetics of the evoked current (data not shown). The potentiated responses in PDA were completely abolished by addition of the transporter antagonist TBOA (at t > 28 min), arguing against the recruitment of a novel conductance by PDA treatment (Figure 5C). Application of the inactive compound 4α-PDA left the transporter currents unchanged (n = 5). Previous reports have shown that cultures of cerebellar glia derived from embryonic chicks show downregu-
lation of the transporter GLAST following long-term phorbol ester treatment. This takes the form of reduced GLAST protein and mRNA as well as reduced uptake of [3H]-D-aspartate, a GLAST substrate (González and Ortega, 1997; Espinoza-Rojo et al., 2000). Thus, it is possible that the effects of bath-applied PDA herein are mediated in part through the downregulation of GLAST in cerebellar Bergmann glia and a consequent increase in the availability of synaptically or photolytically released glutamate to bind Purkinje cell transporters (or AMPA receptors). To test this idea, we combined photolytic uncaging of glutamate with patch-clamp recording from identified Bergmann glia (Figure S4). UV light flashes produced inward currents of peak amplitude 77.3 ± 6.6pA (recorded with Vhold = −70 mV; n = 5), which had much slower kinetics compared to similar currents evoked in Purkinje cells (10%–90% rise: 26.9 ± 2.8 ms; 50% decay: 135.6 ± 10.2 ms). Application of PDA in the bath produced no significant alteration of test pulse response amplitudes (96% ± 2% of baseline at t = 19 min; n = 5) or kinetics (data not shown). The photolytically evoked responses in Bergmann glia were completely abolished by addition of the transporter antagonist TBOA (at t > 19 min). These data indicate that, in our preparation, PDA is not exerting its effects through downregulation of Bergmann glial glutamate transport. Discussion The main finding of this study is that neuronal transporter currents in Purkinje cells, which are mediated by the EAAT4 glutamate transporter (Huang et al., 2004b), undergo LTP in response to brief tetanic activation of climbing fibers. This LTP appears to be postsynaptically expressed, it requires activation of a postsynaptic mGluR1/PKC cascade, and it persists for the duration of typical whole-cell recordings, qualities that are shared by a simultaneously induced LTD of AMPA receptor-mediated responses. We suggest that, together, LTP of neuronal glutamate transport and LTD of AMPA receptors may constitute a coherent antiexcitotoxic response of the Purkinje cell to strong synaptic activation by glutamate. While climbing fiber activation of the strength and duration used herein (5 Hz, 30 s) may not occur during typical behavior, similar patterns of activity are likely to be evoked following an ischemic episode (Welsh et al., 2002). Glutamate spillover from the climbing fiber-Purkinje cell synapse is regulated by both neuronal and glial glutamate transport. Blockade of both with TBOA facilitates glutamate spillover, resulting in diverse sequelae including a gradual increase in ambient glutamate concentration (Takayasu et al., 2004), activation of mGluR1 in Purkinje cell perisynaptic membrane (Brasnjo and Otis, 2001; Dzubay and Otis, 2002; Reichelt and Knopfel, 2002), and activation of NMDA receptors on the presynaptic terminals of molecular layer interneurons (Huang and Bordey, 2004). LTP of neuronal glutamate transport described herein could modulate these events and the related sequelae of strong climbing fiber activation, including excitotoxic Purkinje cell death. This model is consistent with the tight correspondence be-
Neuron 720
Figure 5. An Exogenous PKC Activator Potentiates Purkinje Cell Transporter Currents Evoked by Photolytic Glutamate Uncaging Recordings were made in the presence of 100 M NBQX and 70 M UBP-301. UV laser pulses (355 nm; 8 ms duration) were used to uncage MNI-glutamate. (A) A population time course graph showing responses to photolytic test pulses in control conditions (n = 5) and in which the PKC activator PDA (n = 7) or the inactive control compound 4αPDA (n = 5) were bath applied starting at t = 7.5 min, as indicated by the horizontal bar. (B) Representative current traces recorded at the time points indicated on the graph in (A). At the end of the experiments (t > 28 min), the transporter antagonist TBOA (300 M) was added. Each trace is the average of three consecutive responses. (C) Group means show the percent of baseline transporter current (IEAAT) recorded at t = 27 min and again after the addition of TBOA. ***p < 0.001. Statistical comparisons were not performed for TBOA addition. Error bars show mean ± SEM.
tween those parasagittal zones of the cerebellum that show low levels of EAAT4 immunoreactivity and those that are particularly vulnerable to postischemic Purkinje cell death (Welsh et al., 2002). It should be noted, however, that the EAAT4-low parasagittal zones also have low immunoreactivity for aldolase C (also called zebrin I) and possibly other markers that may be relevant for excitotoxicity. In addition, the glutamate transport capacity of Purkinje cells in EAAT4-high versus EAAT4low zones has yet to be quantified. In addition to EAAT4, Purkinje cells express the glutamate transporter EAAC1 (Rothstein et al., 1994). EAAC1 undergoes translocation to the plasma membrane and a consequent increase in transport activity upon stimulation with PKC-activating phorbol esters in a glioma cell line or synaptoneurosomes (Davis et al., 1998; Fournier et al., 2004) as well as in hippocampal slices (Levenson et al., 2002). Furthermore, biochemical analysis of hippocampal slices in which LTP of Schaffer collateral EPSPs had been induced showed an increase in [14C]-glutamate uptake together with a translocation of EAAC1 to the membrane fraction. Interestingly, these changes could also be detected following contextual fear conditioning, but not unpaired control training (Levenson et al., 2002). It is intriguing that phorbol ester treatment results in the binding of both the α isoform of PKC and the syn-
aptic scaffolding molecule PICK1 to EAAC1 (González et al., 2003; M. Gonzalez and M. Robinson, 2003, Soc. Neurosci., abstract). Both PKCα (Leitges et al., 2004) and PICK1 (Xia et al., 2000) are required for LTD of AMPA receptors in cultured Purkinje cells. In the future, it will be useful to determine whether EAAT4 translocation to the membrane is also regulated by these interactions in Purkinje cells. It will also be informative for a more general understanding of excitotoxicity and synaptic plasticity to determine if synaptic activation of an mGluR/PKC cascade causes increased EAAC1-mediated glutamate transport in other cell types in the brain. Experimental Procedures Sagittal slices of the cerebellar vermis (250 m thick) were prepared from P17–P20 Sprague-Dawley rats using a Vibratome 3000 and ice-cold standard ACSF containing 124 mM NaCl, 2.5 mM KCl, 1 mM Na2HPO4, 1.3 mM MgSO4, 2.5 mM CaCl2, 26.2 mM NaHCO3, and 20 mM D-Glucose bubbled with 95% O2 and 5% CO2. Following a recovery period of at least 1 hr, the slices were placed in a submerged chamber that was perfused at a flow rate of 2 ml/min with room temperature ACSF supplemented with 20 M GABAzine to block GABAA receptors. Recordings were performed using the visualized whole-cell patch-clamp technique using a Zeiss Axioskop 2FS microscope and Axopatch 200A or 200B amplifiers. The recording electrodes (resistance 2–3 M⍀) were typically filled with a solution containing 135 mM CsOH, 135 mM D-gluconic acid, 10 mM CsCl, 10 mM HEPES, 4 mM Na2ATP, 0.4 mM Na3GTP, and 0.2
LTP of Neuronal Glutamate Transporters 721
mM EGTA (pH 7.25). Currents were filtered at 1 kHz, digitized at 10 kHz, and acquired using either Axodata 4.5 or Clampex 9.0 software (Axon Instruments). For climbing fiber stimulation, theta glass pipettes were used, which were filled with external saline and placed in the granule cell layer. Stimulating electrode placement was optimized to reduce the stimulus artifact and thereby reveal the rising phase of the evoked currents without using a subtraction scheme. Test responses were evoked at a frequency of 0.05 Hz using 8–20 A pulses that were applied for 80–100 s. Only cells that were innervated by a single climbing fiber input, as indicated by a single step in the climbing fiber input-output relation and paired-pulse depression, were used. These criteria will exclude contaminating parallel fiber and ascending granule cell axons inputs as well, as these are recruited in a graded fashion and show paired-pulse facilitation (Hansel and Linden, 2000; Nishiyama and Linden, 2004). Blockade of AMPA receptors was produced in some experiments by local application of external saline supplemented with 100 M NBQX using a gravity-driven large bore (w60 m O.D.) flow pipe placed w200 m from the Purkinje neuron soma. Pilot studies showed that local flow pipe application produced effects on climbing fiber-evoked currents that were indistinguishable from bath application of 100 M NBQX. For photolytic uncaging of glutamate, external saline supplemented with 4-methoxy-7-nitroindolinyl glutamate (MNI-glu, 400 M), and in some cases, other test compounds, was applied locally from a flow pipe positioned at w200 m away from the recorded cell soma in the molecular layer. When test compounds (TBOA, UBP-301, PDA, etc.) were applied, these were simultaneously delivered to the flow pipe and the bath. Uncaging was produced by directing the 355 nm output of a rapidly pulsed diode laser (DPSS Inc., power at the laser head = 1.1 W) into the epifluorescence train of a Zeiss Axioskop 2FSmot microscope with a 50 m multimode fused silica fiber and coupling optics. Light was directed to the objective by reflection from a 400 nm dichroic mirror. Uncaging test pulses (8 ms duration, 0.05 Hz) were originated by Clampex 9.0 software (Axon Instruments) and controlled by a combination of switching the laser and activation of a mechanical shutter (Uniblitz, Vincent Associates). Cells were excluded from the study if series resistance or input resistance varied by >15% over the course of the experiment. All group data are shown as mean ± SEM. Comparisons were made using Student’s t test. Drugs were purchased from Sigma, with the exception of GABAzine, (−)-indolactam-V, (+)-indolactam-V, DLTBOA, PDA, 4α-PDA, UBP-301, and NBQX, from Tocris; and MNIglutamate, from Molecular Probes. GYKI 53655 was a gift from Dr. Antal Simay, IVAX Drug Research Institute.
Supplemental Data The Supplemental Data for this article can be found at http:// www.neuron.org/cgi/content/full/46/5/715/DC1/.
Acknowledgments Y.S. would like to dedicate this work to his wife, Xueqin Xia, with thanks for her support. Thanks to R. Bock and D. Gurfel for skillful technical assistance and to members of the Linden and Bergles labs for helpful comments. C.-M. Tang provided invaluable assistance in designing the photolysis rig. Dr. A. Simay made a generous gift of GYKI 53655. This work was supported by USPHS MH51106 and MH01590 and the Develbiss Fund (D.J.L.). Received: October 20, 2004 Revised: February 2, 2005 Accepted: April 28, 2005 Published: June 1, 2005 References Arnth-Jensen, N., Jabaudon, D., and Scanziani, M. (2002). Cooperation between independent hippocampal synapses is controlled by glutamate uptake. Nat. Neurosci. 5, 325–331.
Asztely, F., Erdemli, G., and Kullmann, D.M. (1997). Extrasynaptic glutamate spillover in the hippocampus: dependence on temperature and the role of active glutamate uptake. Neuron 18, 281–293. Auger, C., and Attwell, D. (2000). Fast removal of synaptic glutamate by postsynaptic transporters. Neuron 28, 547–558. Brasnjo, G., and Otis, T.S. (2001). Neuronal glutamate transporters control activation of postsynaptic metabotropic glutamate receptors and influence cerebellar long-term depression. Neuron 31, 607–616. Brasnjo, G., and Otis, T.S. (2004). Isolation of glutamate transportcoupled charge flux and estimation of glutamate uptake at the climbing fiber-Purkinje cell synapse. Proc. Natl. Acad. Sci. USA 101, 6273–6278. Carter, A.G., and Regehr, W.G. (2000). Prolonged synaptic currents and glutamate spillover at the parallel fiber to stellate cell synapse. J. Neurosci. 20, 4423–4434. Cho, K., Francis, J.C., Hirbec, H., Dev, K., Brown, M.W., Henley, J.M., and Bashir, Z.I. (2003). Regulation of kainate receptors by protein kinase C and metabotropic glutamate receptors. J. Physiol. 548, 723–730. Davis, K.E., Straff, D.J., Weinstein, E.A., Bannerman, P.G., Correale, D.M., Rothstein, J.D., and Robinson, M.B. (1998). Multiple signaling pathways regulate cell surface expression and activity of the excitatory amino acid carrier 1 subtype of Glu transporter in C6 glioma. J. Neurosci. 18, 2475–2485. DiGregorio, D.A., Nusser, Z., and Silver, R.A. (2002). Spillover of glutamate onto synaptic AMPA receptors enhances fast transmission at a cerebellar synapse. Neuron 35, 521–533. Dzubay, J.A., and Otis, T.S. (2002). Climbing fiber activation of metabotropic glutamate receptors on cerebellar Purkinje neurons. Neuron 36, 1159–1167. Espinoza-Rojo, M., López-Bayghen, E., and Ortega, A. (2000). GLAST: gene expression regulation by phorbol esters. Neuroreport 11, 2827–2832. Fournier, K.M., González, M.I., and Robinson, M.B. (2004). Rapid trafficking of the neuronal glutamate transporter, EAAC1: evidence for distinct trafficking pathways differentially regulated by protein kinase C and platelet-derived growth factor. J. Biol. Chem. 279, 34505–34513. González, M.I., and Ortega, A. (1997). Regulation of the Na+-dependent high affinity glutamate/aspartate transporter in cultured Bergmann glia by phorbol esters. J. Neurosci. Res. 50, 585–590. González, M.I., Bannerman, P.G., and Robinson, M.B. (2003). Phorbol myristate acetate-dependent interaction of protein kinase Cα and the neuronal glutamate transporter EAAC1. J. Neurosci. 23, 5589–5593. Hansel, C., and Linden, D.J. (2000). Long-term depression of the cerebellar climbing fiber-Purkinje neuron synapse. Neuron 26, 473–482. Häusser, M., and Roth, A. (1997). Dendritic and somatic glutamate receptor channels in rat cerebellar Purkinje cells. J. Physiol. 501, 77–95. Huang, Y.H., and Bergles, D.E. (2004). Glutamate transporters bring competition to the synapse. Curr. Opin. Neurobiol. 14, 346–352. Huang, H., and Bordey, A. (2004). Glial glutamate transporters limit spillover activation of presynaptic NMDA receptors and influence synaptic inhibition of Purkinje neurons. J. Neurosci. 24, 5659–5669. Huang, Y.H., Dykes-Hoberg, M., Tanaka, K., Rothstein, J.D., and Bergles, D.E. (2004a). Climbing fiber activation of EAAT4 transporters and kainate receptors in cerebellar Purkinje cells. J. Neurosci. 24, 103–111. Huang, Y.H., Sinha, S.R., Tanaka, K., Rothstein, J.D., and Bergles, D.E. (2004b). Astrocyte glutamate transporters regulate metabotropic glutamate receptor-mediated excitation of hippocampal interneurons. J. Neurosci. 24, 4551–4559. Leitges, M., Kovac, J., Plomann, M., and Linden, D.J. (2004). A unique PDZ-ligand in PKCα confers induction of cerebellar longterm synaptic depression. Neuron 18, 585–594. Levenson, J., Weeber, E., Selcher, J.C., Kategaya, L.S., Sweatt,
Neuron 722
J.D., and Eskin, A. (2002). Long-term potentiation and contextual fear conditioning increase neuronal glutamate uptake. Nat. Neurosci. 5, 155–161. Linden, D.J., Smeyne, M., and Connor, J.A. (1993). Induction of cerebellar long-term depression in culture requires postsynaptic action of sodium ions. Neuron 11, 1093–1100. Maragakis, N.J., and Rothstein, J.D. (2004). Glutamate transporters: animal models to neurologic disease. Neurobiol. Dis. 15, 461–473. More, J.C., Troop, H.M., Dolman, N.P., and Jane, D.E. (2003). Structural requirements for novel willardiine derivatives acting as AMPA and kainate receptor antagonists. Br. J. Pharmacol. 138, 1093– 1100. Nishiyama, H., and Linden, D.J. (2004). Differential maturation of climbing fiber innervation in cerebellar vermis. J. Neurosci. 24, 3926–3932. Oliet, S.H., Piet, R., and Poulain, D.A. (2001). Control of glutamate clearance and synaptic efficacy by glial coverage of neurons. Science 292, 923–926. Otis, T.S., Kavanaugh, M.P., and Jahr, C.E. (1997). Postsynaptic glutamate transport at the climbing fiber-Purkinje cell synapse. Science 277, 1515–1518. Reichelt, W., and Knopfel, T. (2002). Glutamate uptake controls expression of a slow postsynaptic current mediated by mGluRs in cerebellar Purkinje cells. J. Neurophysiol. 87, 1974–1980. Rothstein, J.D., Martin, L., Levey, A.I., Dykes-Hoberg, M., Jin, L., Wu, D., Nash, N., and Kuncl, R.W. (1994). Localization of neuronal and glial glutamate transporters. Neuron 13, 713–725. Shen, Y., Hansel, C., and Linden, D.J. (2002). Glutamate release during LTD at cerebellar climbing fiber-Purkinje cell synapses. Nat. Neurosci. 5, 725–726. Shigeri, Y., Shimamoto, K., Yasuda-Kamatani, Y., Seal, R.P., Yumoto, N., Nakajima, T., and Amara, S.G. (2001). Effects of threo-βhydroxyaspartate derivatives on excitatory amino acid transporters (EAAT4 and EAAT5). J. Neurochem. 79, 297–302. Shimamoto, K., Lebrun, B., Yasuda-Kamatani, Y., Sakaitani, M., Shigeri, Y., Yumoto, N., and Nakajima, T. (1998). DL-threo-β-benzyloxyaspartate, a potent blocker of excitatory amino acid transporters. Mol. Pharmacol. 53, 195–201. Stoffel, W., Körner, R., Wachtmann, D., and Keller, B.U. (2004). Functional analysis of glutamate transporters in excitatory synaptic transmission of GLAST1 and GLAST1/EAAC1 deficient mice. Brain Res. Mol. Brain Res. 128, 170–181. Takayasu, Y., Iino, M., and Ozawa, S. (2004). Roles of glutamate transporters in shaping excitatory synaptic currents in cerebellar Purkinje cells. Eur. J. Neurosci. 19, 1285–1295. Tanaka, J., Ichikawa, R., Watanabe, M., Tanaka, K., and Inoue, Y. (1997a). Extra-junctional localization of glutamate transporter EAAT4 at excitatory Purkinje cell synapses. Neuroreport 8, 2461– 2464. Tanaka, K., Watase, K., Manabe, T., Yamada, K., Watanabe, M., Takahashi, K., Iwama, H., Nishikawa, T., Ichihara, N., Kikuchi, T., et al. (1997b). Epilepsy and exacerbation of brain injury in mice lacking the glutamate transporter GLT-1. Science 276, 1699–1702. Tsvetkov, E., Shin, R.M., and Bolshakov, V.Y. (2004). Glutamate uptake determines pathway specificity of long-term potentiation in the neural circuitry of fear conditioning. Neuron 41, 139–151. Wadiche, J.I., and Jahr, C.E. (2001). Multivesicular release at climbing fiber-Purkinje cell synapses. Neuron 32, 301–313. Wadiche, J.I., and Kavanaugh, M.P. (1998). Macroscopic and microscopic properties of a cloned glutamate transporter/chloride channel. J. Neurosci. 18, 7650–7661. Wadiche, J.I., Amara, S.G., and Kavanaugh, M.P. (1995). Ion fluxes associated with excitatory amino acid transport. Neuron 15, 721– 728. Welsh, J.P., Yuen, G., Placantonakis, D.G., Vu, T.Q., Haiss, F., O’Hearn, E., Molliver, M.E., and Aicher, S.A. (2002). Why do Purkinje cells die so easily after global brain ischemia? Aldolase C, EAAT4,
and the cerebellar contribution to posthypoxic myoclonus. Adv. Neurol. 89, 331–359. Xia, J., Chung, H.J., Wihler, C., Huganir, R.L., and Linden, D.J. (2000). Cerebellar long-term depression requires PKC-regulated interactions between GluR2/3 and PDZ domain-containing proteins. Neuron 28, 499–510.
Long-Term Potentiation of Neuronal Glutamate Transporters Ying Shen and David J. Linden* Department of Neuroscience The Johns Hopkins University School of Medicine 725 North Wolfe Street Baltimore, Maryland 21205
Summary Persistent, use-dependent modulation of synaptic strength has been demonstrated for fast synaptic transmission mediated by glutamate and has been hypothesized to underlie persistent behavioral changes ranging from memory to addiction. Glutamate released at synapses is sequestered by the action of excitatory amino acid transporters (EAATs) in glia and postsynaptic neurons. So, the efficacy of glutamate transporter function is crucial for regulating glutamate spillover to adjacent presynaptic and postsynaptic receptors and the consequent induction of plastic or excitotoxic processes. Here, we report that tetanic stimulation of cerebellar climbing fiber-Purkinje cell synapses results in long-term potentiation (LTP) of a climbing fiber-evoked glutamate transporter current recorded in Purkinje cells. This LTP is postsynaptically expressed and requires activation of an mGluR1/PKC cascade. Together with a simultaneously induced long-term depression (LTD) of postsynaptic AMPA receptors, this might reflect an integrated antiexcitotoxic cellular response to strong climbing fiber synaptic activation, as occurs following an ischemic episode.
Report
both presynaptic and postsynaptic compartments (Asztely et al., 1997; Carter and Regehr, 2000; Brasnjo and Otis, 2001; Oliet et al., 2001; Arnth-Jensen et al., 2002; DiGregorio et al., 2002; Dzubay and Otis, 2002; Reichelt and Knopfel, 2002; Huang and Bordey, 2004; Huang et al., 2004a; Takayasu et al., 2004; Stoffel et al., 2004). The sequelae of transporter-regulated spillover signaling are many and include triggering of both (presumably) adaptive plastic processes such as LTP and LTD (Brasnjo and Otis, 2001; Tsvetkov et al., 2004) as well as pathological processes such as excitotoxicity and epileptiform discharges (Tanaka et al., 1997b; Maragakis and Rothstein, 2004). The cerebellar climbing fiber-Purkinje cell synapse is an advantageous model system for the study of neuronal transporter currents (Otis et al., 1997; Auger and Attwell, 2000). Most Purkinje cells are innervated by a single climbing fiber axon, which ramifies to form w1500 release sites. Each of these sites appears to release multiple vesicles of glutamate with each action potential invasion (Wadiche and Jahr, 2001). A single shock delivered to the climbing fiber axon results in the activation of AMPA and kainate receptors (Huang et al., 2004b), and when these are blocked, a synaptic glutamate transporter current is revealed. Recently, this has been shown to be mediated by the transporter EAAT4 (Huang et al., 2004b), which is strongly expressed in the perisynaptic membranes of climbing fiber-Purkinje cell synapses (Tanaka et al., 1997a). Here, we have recorded climbing fiber-evoked synaptic transporter currents in Purkinje cells to test the hypothesis that neuronal glutamate transport can undergo persistent use-dependent changes in efficacy.
Introduction Results Glutamate released from nerve terminals into the synaptic cleft is subject to diffusion and dilution in the extracellular space as well as uptake into neurons and glia by excitatory amino acid transporters (EAATs). The efficacy of glutamate transport is crucial not only for maintaining the background concentration of glutamate at a low level, but also for determining the amplitude and kinetics of glutamate transients that result from vesicular release. At most synapses, glutamate transporters have a relatively small effect on the rise and peak of the glutamate transient as experienced by receptors in the postsynaptic density directly across the synaptic cleft from the fusion event. However, transporters localized to the perisynaptic membrane of the postsynaptic neuron as well as those on ensheathing glial cells have a major role in controlling the “spillover” of glutamate to adjacent structures (see Huang and Bergles, 2004, for review). Glutamate spillover is enhanced with high-frequency burst activation of afferent fibers, and the control of this spillover by transporters has been shown to regulate the activation of ionotropic and metabotropic glutamate receptors in *Correspondence: [email protected]
Whole-cell voltage-clamp recordings were made from Purkinje cells in sagittal slices of juvenile rat cerebellum. In order to isolate the stoichiometric portion of synaptic glutamate transporter current, recordings were made using a gluconate-based internal saline, as this species does not permeate during the uncoupled anion channel mode of the transporter (Wadiche et al., 1995; Wadiche and Kavanaugh, 1998). Stimulation of the climbing fiber evokes a large EPSC (w1–2 nA, recorded with Vhold = −15 mV), which is strongly inhibited by NBQX (100 M), an antagonist of AMPA and kainate receptors (Figure 1A). Functional NMDA receptors are not expressed on Purkinje cells after the first postnatal week (Häusser and Roth, 1997). The current recorded in NBQX is small (40.6 ± 2.1 pA, recorded with Vhold = −70 mV; n = 16) and has similar onset kinetics to the control EPSC (10%–90% rise time; control EPSC: 1.3 ± 0.1; NBQX-resistant current: 1.6 ± 0.1 ms) but somewhat slower offset kinetics (50% decay time, 8.0 ± 0.6 versus 11.4 ± 1.1 ms; n = 25). Addition of the drug DLthreo-β-benzoylaspartic acid (TBOA, 300 M), a selective inhibitor of glutamate transporters (Shimamoto et al., 1998; Shigeri et al., 2001), strongly attenuated the
Neuron 716
Figure 1. The Climbing Fiber-Evoked Synaptic Current Recorded in a Purkinje Cell Bathed in a High Concentration of NBQX Is Predominantly Mediated by Glutamate Transport (A) Single shocks delivered to the granule cell layer evoke all-or-none climbing fiber responses recorded in Purkinje cells. A control EPSC trace, in which AMPA receptor-mediated current predominates, was recorded with Vhold = −15 mV and is superimposed upon a scaled recording from the same cell in which 100 M NBQX has been added and Vhold changed to −70 mV. (B) Vhold = −70 mV. NBQX (100 M), TBOA (300 M), and UBP-301 (70 M) were added serially. Each trace is the average of three evoked responses. Bar graphs depict the mean ± SEM of 16 cells. (C) A trace recorded in the presence of the specific AMPA receptor antagonist GYKI 53655 (100 M) in which kainate receptormediated current predominates (Huang et al., 2004b) is superimposed upon a scaled recording in the presence of NBQX (100 M) plus TBOA (300 M), which shows the NBQX-resistant fraction of kainate current.
residual current, leaving 9.4 ± 0.5 pA unblocked by NBQX plus TBOA (Figure 1B), similar to previous findings (Brasnjo and Otis, 2004; Huang et al., 2004b). A recent report has shown that this residual component is mediated by NBQX-resistant GluR5-containing kainate receptors (Huang et al., 2004b). Consistent with this finding, we observed that the current remaining in NBQX plus TBOA was completely obliterated (1.0 ± 0.4 pA) by the kainate receptor antagonist UBP-301 (70 M; More et al., 2003). Furthermore, the kinetics of the current recorded in NBQX + TBOA (10%–90% rise time = 3.1 ± 0.6 ms; 50% decay time = 19.4 ± 2.4 ms; n = 9) were similar to that of the much larger total kainate current recorded in the presence of the AMPA receptor-specific antagonist GYKI 53566 (100 M) plus TBOA (10%–90% rise time = 3.0 ± 0.5 ms; 50% decay time = 19.3 ± 1.3 ms; n = 9; Figure 1C). Previous work has shown that tetanic activation of the climbing fiber-Purkinje cell synapse (5 Hz, 30 s) can give rise to long-term synaptic depression (LTD), which requires activation of an mGluR1/PKC cascade (Hansel and Linden, 2000) and which is expressed postsynaptically as a downregulation of AMPA receptors (Shen et al., 2002). We used this same stimulation protocol to see if glutamate transporter function would be persistently changed (Figure 2). In this experiment, external saline supplemented with NBQX (100 M) + UBP-301 (70 M) was applied locally to the slice surface with a flow pipe to block AMPA and kainate receptors, and the remaining transporter currents were measured using test pulses delivered every 20 s for 3 min with Vhold = −70 mV. Following this, Vhold was switched to −15 mV (this improves the quality of the voltage clamp for AMPA-EPSCs by reducing driving force and inactivating voltage-gated channels), and NBQX and UBP301 were washed out until a stable baseline was achieved. Then, tetanic stimulation (5 Hz, 30 s) was delivered and test pulses were resumed. This gave rise to
a persistent depression of the AMPA receptor-mediated EPSC (peak amplitude: 72% ± 3% of baseline at t = 19 min, n = 7), which was not associated with a significant change in paired-pulse ratio (pre, t = −1 min: 0.62 ± 0.02; post, t = 20 min: 0.64 ± 0.02; interpulse interval = 80 ms), similar to that previously reported (Hansel and Linden, 2000; Shen et al., 2002). When NBQX and UBP301 were reintroduced (and Vhold was returned to −70 mV), it was revealed that the transporter current had undergone LTP (peak amplitude: 123% ± 8% of baseline; charge transfer: 146% ± 14% of baseline; t = 25 min). The potentiated transporter currents were completely abolished by subsequent addition of the antagonist TBOA, arguing against the recruitment of a novel conductance by 5 Hz stimulation. When this experiment was repeated at elevated temperature (32°C; n = 5), very similar results were seen (see Figure S1 in the Supplemental Data available online). Interestingly, the degree of LTD of AMPA-EPSCs was positively correlated with the degree of transporter current LTP (p = 0.0004), suggesting that these two phenomena might result from a common signaling cascade. Neither changes in series resistance (Rs; p = 0.48) nor input resistance (Ri; p = 0.58) were significantly correlated with transporter current LTP (Figure S2). It is important to note the caveat that while we presume that transporter current was rapidly potentiated and that this potentiation was then revealed by addition by NBQX and UBP-301 w20 min later, we do not have measurements of NBQX-resistant current during the period t = 0–20 min. Experiments were also performed with NBQX present throughout, but these failed to show potentiation of NBQX-resistant current at any time point (peak amplitude: 96% ± 4% of baseline; charge transfer: 93% ± 3% of baseline; t = 15 min; Figure S3). This is consistent with the finding that LTD of AMPA receptor-mediated currents in cultured Purkinje cells also requires AMPA receptor activation in addition to mGluR1 activation (Linden et al., 1993).
LTP of Neuronal Glutamate Transporters 717
Figure 2. Brief Tetanic Stimulation of Climbing Fibers Gives Rise to LTP of Transporter Current and Simultaneous LTD of AMPA Receptor-Mediated EPSCs in Purkinje Cells (A) These time-course data show the mean ± SEM of each group. Test pulses were delivered to activate climbing fibers, and responses were recorded with Vhold = −70 mV in the presence of NBQX (100 M) plus UBP301 (70 M), delivered locally with a flow pipe (transporter current, IEAAT, open squares). To record AMPA receptor-mediated EPSCs (filled circles), NBQX and UBP-301 were washed out, and Vhold was set to −15 mV. Tetanic stimulation (5 Hz, 30 s) was delivered at t = 0 min, as indicated by the upward arrows, and test pulses were resumed until t = 20 min, when NBQX and UBP-301 were reintroduced, and later Vhold was returned to −70 mV to allow for posttetanus measurements of transporter current. Control group, n = 7. Drugs added prior to patch formation: chelerythrine (10 M), n = 5; CPCCOEt (100 M), n = 5; phorbol-12,13-diacetate (PDA, 8 M), n = 5. Error bars show mean ± SEM. (B) Representative traces recorded from a control cell at the time points indicated in (A). (C) Population measures of currents recorded in the experiments shown in (A). Percent of baseline EPSC measures are derived from the ratio of measurements made at t = 15–20 min to that at t = −1–0 min. *p < 0.05 compared to baseline, ***p < 0.005. Error bars show mean ± SEM.
Pretreatment in the bath using either the PKC inhibitor chelerythrine (n = 5) or the group I mGluR antagonist CPCCOEt (n = 5) produced a complete blockade of LTP of transporter current as well as a complete blockade of LTD of AMPA receptor-mediated EPSCs. Similar effects were produced by pretreatment with the PKC activator phorbol-12,13-diacetate (PDA, 8 M; n = 5). The effects of PDA pretreatment could result from either occlusion of subsequent AMPA receptor LTD and transporter current LTP or blockade of the induction these phenomena resulting from downregulation of mGluR1. If an mGluR1/PKC cascade is required for LTP of NBQX-resistant current, then perhaps exogenous PKC activators would mimic the effects of tetanic climbing fiber stimulation. Bath application of PDA produced a potentiation of NBQX-resistant synaptic current (Figure 3A, upper right; n = 6, measured at t = 18 min after PDA application) and, in a separate population of Purkinje cells, a depression of AMPA receptor mediated EPSCs (Figure 3A, upper left; n = 7). A similar effect was produced by application of the PKC activator (−)-indolactam-V in the patch pipette and comparing time points soon after break in (t = 3 min) and somewhat later, to allow for dendritic perfusion (t = 24 min; Figure 4). Perfusion of the inactive enantiomer, (+)-indolactam-V, had no effect on isolated transporter currents recorded in NBQX + UBP-301 (n = 6).
Are the effects of these exogenous PKC activators mediated by activation of PKC in the Purkinje cell or in some other compartment? Neither bath PDA nor postsynaptic (−)-indolactam-V produced a significant change in Pr as indexed by EPSC paired-pulse ratios, arguing against an effect in climbing fiber terminals (PDA: pre, t = 10 min, 0.64 ± 0.04; post, t = 30 min, 0.64 ± 0.03; (−)-indolactam-V, pre, t = 3 min, 0.60 ± 0.03; post, t = 24 min, 0.60 ± 0.03). Furthermore, perfusion of the Purkinje cell with the cell-impermeant PKC inhibitor peptide PKC 19-36 (10 M) completely blocked the potentiation of transporter current produced by subsequent bath application of PDA (Figure 3A, lower left, n = 5). This last experiment argues that the locus of PKC activation to evoke these effects is the Purkinje cell dendrite. The NBQX-resistant current is predominantly (w78%) carried by EAAT4 glutamate transporters, with the remainder carried by NBQX-resistant kainate receptors. Could kainate current contribute to the potentiation produced by PKC activation (Figure 3) as recorded in NBQX? Previous work in a perirhinal slice preparation showed that bath application of a group I mGluR1 agonist could transiently potentiate synaptic kainate currents in a manner that required PKC activation (Cho et al., 2003). To address this, PDA was applied in two different recording configurations. First, NBQX plus TBOA
Neuron 718
Figure 3. PDA, an Exogenous PKC Activator, Mimics the Effects of Synaptic Tetani on AMPA Receptor-Mediated EPSCs and Transporter Currents, but Not Kainate Current in Purkinje Cells (A) Sample traces recorded before and 18 min after bath application of the exogenous PKC activator PDA (8 M). Each trace is the average of three consecutive records. Recordings were made in separate populations of cells in control conditions (to record EPSCs; n = 7), NBQX (100 M; n = 6), GYKI 53655 (100 M) + TBOA (300 M; n = 7), or NBQX (100 M) plus TBOA (300 M; n = 5). An additional group was recorded in NBQX using cells in which the membrane-impermeant PKC inhibitor peptide PKC 19-36 was included in the pipette at a concentration of 10 M and allowed to perfuse for >17 min prior to PDA application (n = 5). (B) Population measures of currents recorded in the experiments shown in (A) and (B). *p < 0.05, **p < 0.01, ***p < 0.005. Error bars show mean ± SEM.
was used to isolate the small NBQX-resistant kainate current. In this recording configuration, NBQX-resistant kainate current was found to be unaltered by bath application of PDA (Figure 3A, middle right, n = 5, t = 18 min). Second, GYKI 53655 plus TBOA was used to isolate the (much larger) total kainate current (Huang et al.,
2004a), which was also found to be unaltered by PDA (Figure 3A, middle left, n = 7, t = 18 min). We conclude that PKC activation produced by tetanic stimulation of climbing fibers or bath application of a PKC-activating phorbol ester produces LTP of neuronal glutamate transporter current and LTD of AMPA receptor-mediFigure 4. (−)-indolactam-V, a PKC Activator Added to the Internal Saline, Mimics the Effects of Synaptic Tetani on AMPA ReceptorMediated EPSCs and Transporter Currents in Purkinje Cells (A) Sample traces recorded with the PKC activator (−)-indolactam-V (300 M, n = 6) in the pipette. Recordings were made 3 min after achieving a stable whole-cell recording configuration (pre) or after 24 min of internal perfusion [(−)-indolactam]. This method may underestimate the amplitude of the indolactam effect, as it is not certain that there was no action of indolactam at the “pre” time point. (B) A similar experiment performed using the inactive control compound (+)-indolactam-V (300 M, n = 6). (C) Population measures of currents recorded in the experiments shown in (A) and (B). *p < 0.05. Error bars show mean ± SEM.
LTP of Neuronal Glutamate Transporters 719
ated EPSCs, but no change in kainate receptor-mediated EPSCs in the Purkinje cell. The present findings suggest a model in which postsynaptic receptors and transporters are differentially regulated by climbing fiber tetanization. However, it is formally possible that the increases in neuronal glutamate transport current reflect an increase in glutamate release. This increase would then have to be coupled with a large reduction in AMPA receptor function sufficient to yield LTD of the EPSC and a reduction in kainate receptor function that exactly offsets the increase in glutamate release. However, several lines of evidence argue against increases in transmitter release in our experiments. First, the LTD of AMPA-EPSCs produced by climbing fiber tetanization, bath application of PDA, or postsynaptic application of (−)-indolactam-V is not associated with changes in paired-pulse ratio, consistent with previous reports (Hansel and Linden, 2000; Shen et al., 2002). Second, LTD of AMPAEPSCs is not associated with a change in the degree of block by the low-affinity competitive AMPA receptor antagonist γ-D-glutamylglycine (Shen et al., 2002). Third, potentiation of neuronal transporter current evoked by PDA could be blocked by postsynaptic application of a membrane-impermeant PKC inhibitor. This indicates that the relevant locus of PKC activation is the Purkinje cell dendrite, thereby necessitating a retrograde messenger to increase glutamate release. To directly test the idea that neuronal glutamate transport is postsynaptically regulated in Purkinje cells, test pulses consisting of photolytic release of caged glutamate with a UV laser were delivered to Purkinje cells bathed in NBQX (100 M) and UBP-301 (70 M). The diameter of the uncaging spot was set to cover the entire Purkinje cell dendritic arbor, and the duration and intensity of the flash were adjusted to have the responses in NBQX + UBP-301 roughly approximate the amplitude of synaptically evoked transporter current and to be subthreshold for activation of mGluR1 (Figures 5A and 5B). It should be noted that this uncaging flash will activate a broad spatial range of glutamate transporters on Purkinje cells, not just those triggered by climbing fiber stimulation, but also those that might be triggered by parallel fiber synapses, which outnumber climbing fiber synapses by approximately 100:1. UV light flashes produced inward currents of 57.7 ± 5.8 pA (recorded with Vhold = −70 mV; n = 12), which had w2fold slower kinetics compared with synaptically evoked transporter current (10%–90% rise: 4.7 ± 0.5 ms; 50% decay: 16.9 ± 1.6 ms). These responses were stable upon repeated measurement at a frequency of 0.05 Hz (103% ± 4% of baseline at t = 27 min; n = 5). Application of PDA in the bath produced a significant potentiation of test pulse responses (147% ± 9% of baseline at t = 27 min; n = 7). Potentiation was not associated with changes in the kinetics of the evoked current (data not shown). The potentiated responses in PDA were completely abolished by addition of the transporter antagonist TBOA (at t > 28 min), arguing against the recruitment of a novel conductance by PDA treatment (Figure 5C). Application of the inactive compound 4α-PDA left the transporter currents unchanged (n = 5). Previous reports have shown that cultures of cerebellar glia derived from embryonic chicks show downregu-
lation of the transporter GLAST following long-term phorbol ester treatment. This takes the form of reduced GLAST protein and mRNA as well as reduced uptake of [3H]-D-aspartate, a GLAST substrate (González and Ortega, 1997; Espinoza-Rojo et al., 2000). Thus, it is possible that the effects of bath-applied PDA herein are mediated in part through the downregulation of GLAST in cerebellar Bergmann glia and a consequent increase in the availability of synaptically or photolytically released glutamate to bind Purkinje cell transporters (or AMPA receptors). To test this idea, we combined photolytic uncaging of glutamate with patch-clamp recording from identified Bergmann glia (Figure S4). UV light flashes produced inward currents of peak amplitude 77.3 ± 6.6pA (recorded with Vhold = −70 mV; n = 5), which had much slower kinetics compared to similar currents evoked in Purkinje cells (10%–90% rise: 26.9 ± 2.8 ms; 50% decay: 135.6 ± 10.2 ms). Application of PDA in the bath produced no significant alteration of test pulse response amplitudes (96% ± 2% of baseline at t = 19 min; n = 5) or kinetics (data not shown). The photolytically evoked responses in Bergmann glia were completely abolished by addition of the transporter antagonist TBOA (at t > 19 min). These data indicate that, in our preparation, PDA is not exerting its effects through downregulation of Bergmann glial glutamate transport. Discussion The main finding of this study is that neuronal transporter currents in Purkinje cells, which are mediated by the EAAT4 glutamate transporter (Huang et al., 2004b), undergo LTP in response to brief tetanic activation of climbing fibers. This LTP appears to be postsynaptically expressed, it requires activation of a postsynaptic mGluR1/PKC cascade, and it persists for the duration of typical whole-cell recordings, qualities that are shared by a simultaneously induced LTD of AMPA receptor-mediated responses. We suggest that, together, LTP of neuronal glutamate transport and LTD of AMPA receptors may constitute a coherent antiexcitotoxic response of the Purkinje cell to strong synaptic activation by glutamate. While climbing fiber activation of the strength and duration used herein (5 Hz, 30 s) may not occur during typical behavior, similar patterns of activity are likely to be evoked following an ischemic episode (Welsh et al., 2002). Glutamate spillover from the climbing fiber-Purkinje cell synapse is regulated by both neuronal and glial glutamate transport. Blockade of both with TBOA facilitates glutamate spillover, resulting in diverse sequelae including a gradual increase in ambient glutamate concentration (Takayasu et al., 2004), activation of mGluR1 in Purkinje cell perisynaptic membrane (Brasnjo and Otis, 2001; Dzubay and Otis, 2002; Reichelt and Knopfel, 2002), and activation of NMDA receptors on the presynaptic terminals of molecular layer interneurons (Huang and Bordey, 2004). LTP of neuronal glutamate transport described herein could modulate these events and the related sequelae of strong climbing fiber activation, including excitotoxic Purkinje cell death. This model is consistent with the tight correspondence be-
Neuron 720
Figure 5. An Exogenous PKC Activator Potentiates Purkinje Cell Transporter Currents Evoked by Photolytic Glutamate Uncaging Recordings were made in the presence of 100 M NBQX and 70 M UBP-301. UV laser pulses (355 nm; 8 ms duration) were used to uncage MNI-glutamate. (A) A population time course graph showing responses to photolytic test pulses in control conditions (n = 5) and in which the PKC activator PDA (n = 7) or the inactive control compound 4αPDA (n = 5) were bath applied starting at t = 7.5 min, as indicated by the horizontal bar. (B) Representative current traces recorded at the time points indicated on the graph in (A). At the end of the experiments (t > 28 min), the transporter antagonist TBOA (300 M) was added. Each trace is the average of three consecutive responses. (C) Group means show the percent of baseline transporter current (IEAAT) recorded at t = 27 min and again after the addition of TBOA. ***p < 0.001. Statistical comparisons were not performed for TBOA addition. Error bars show mean ± SEM.
tween those parasagittal zones of the cerebellum that show low levels of EAAT4 immunoreactivity and those that are particularly vulnerable to postischemic Purkinje cell death (Welsh et al., 2002). It should be noted, however, that the EAAT4-low parasagittal zones also have low immunoreactivity for aldolase C (also called zebrin I) and possibly other markers that may be relevant for excitotoxicity. In addition, the glutamate transport capacity of Purkinje cells in EAAT4-high versus EAAT4low zones has yet to be quantified. In addition to EAAT4, Purkinje cells express the glutamate transporter EAAC1 (Rothstein et al., 1994). EAAC1 undergoes translocation to the plasma membrane and a consequent increase in transport activity upon stimulation with PKC-activating phorbol esters in a glioma cell line or synaptoneurosomes (Davis et al., 1998; Fournier et al., 2004) as well as in hippocampal slices (Levenson et al., 2002). Furthermore, biochemical analysis of hippocampal slices in which LTP of Schaffer collateral EPSPs had been induced showed an increase in [14C]-glutamate uptake together with a translocation of EAAC1 to the membrane fraction. Interestingly, these changes could also be detected following contextual fear conditioning, but not unpaired control training (Levenson et al., 2002). It is intriguing that phorbol ester treatment results in the binding of both the α isoform of PKC and the syn-
aptic scaffolding molecule PICK1 to EAAC1 (González et al., 2003; M. Gonzalez and M. Robinson, 2003, Soc. Neurosci., abstract). Both PKCα (Leitges et al., 2004) and PICK1 (Xia et al., 2000) are required for LTD of AMPA receptors in cultured Purkinje cells. In the future, it will be useful to determine whether EAAT4 translocation to the membrane is also regulated by these interactions in Purkinje cells. It will also be informative for a more general understanding of excitotoxicity and synaptic plasticity to determine if synaptic activation of an mGluR/PKC cascade causes increased EAAC1-mediated glutamate transport in other cell types in the brain. Experimental Procedures Sagittal slices of the cerebellar vermis (250 m thick) were prepared from P17–P20 Sprague-Dawley rats using a Vibratome 3000 and ice-cold standard ACSF containing 124 mM NaCl, 2.5 mM KCl, 1 mM Na2HPO4, 1.3 mM MgSO4, 2.5 mM CaCl2, 26.2 mM NaHCO3, and 20 mM D-Glucose bubbled with 95% O2 and 5% CO2. Following a recovery period of at least 1 hr, the slices were placed in a submerged chamber that was perfused at a flow rate of 2 ml/min with room temperature ACSF supplemented with 20 M GABAzine to block GABAA receptors. Recordings were performed using the visualized whole-cell patch-clamp technique using a Zeiss Axioskop 2FS microscope and Axopatch 200A or 200B amplifiers. The recording electrodes (resistance 2–3 M⍀) were typically filled with a solution containing 135 mM CsOH, 135 mM D-gluconic acid, 10 mM CsCl, 10 mM HEPES, 4 mM Na2ATP, 0.4 mM Na3GTP, and 0.2
LTP of Neuronal Glutamate Transporters 721
mM EGTA (pH 7.25). Currents were filtered at 1 kHz, digitized at 10 kHz, and acquired using either Axodata 4.5 or Clampex 9.0 software (Axon Instruments). For climbing fiber stimulation, theta glass pipettes were used, which were filled with external saline and placed in the granule cell layer. Stimulating electrode placement was optimized to reduce the stimulus artifact and thereby reveal the rising phase of the evoked currents without using a subtraction scheme. Test responses were evoked at a frequency of 0.05 Hz using 8–20 A pulses that were applied for 80–100 s. Only cells that were innervated by a single climbing fiber input, as indicated by a single step in the climbing fiber input-output relation and paired-pulse depression, were used. These criteria will exclude contaminating parallel fiber and ascending granule cell axons inputs as well, as these are recruited in a graded fashion and show paired-pulse facilitation (Hansel and Linden, 2000; Nishiyama and Linden, 2004). Blockade of AMPA receptors was produced in some experiments by local application of external saline supplemented with 100 M NBQX using a gravity-driven large bore (w60 m O.D.) flow pipe placed w200 m from the Purkinje neuron soma. Pilot studies showed that local flow pipe application produced effects on climbing fiber-evoked currents that were indistinguishable from bath application of 100 M NBQX. For photolytic uncaging of glutamate, external saline supplemented with 4-methoxy-7-nitroindolinyl glutamate (MNI-glu, 400 M), and in some cases, other test compounds, was applied locally from a flow pipe positioned at w200 m away from the recorded cell soma in the molecular layer. When test compounds (TBOA, UBP-301, PDA, etc.) were applied, these were simultaneously delivered to the flow pipe and the bath. Uncaging was produced by directing the 355 nm output of a rapidly pulsed diode laser (DPSS Inc., power at the laser head = 1.1 W) into the epifluorescence train of a Zeiss Axioskop 2FSmot microscope with a 50 m multimode fused silica fiber and coupling optics. Light was directed to the objective by reflection from a 400 nm dichroic mirror. Uncaging test pulses (8 ms duration, 0.05 Hz) were originated by Clampex 9.0 software (Axon Instruments) and controlled by a combination of switching the laser and activation of a mechanical shutter (Uniblitz, Vincent Associates). Cells were excluded from the study if series resistance or input resistance varied by >15% over the course of the experiment. All group data are shown as mean ± SEM. Comparisons were made using Student’s t test. Drugs were purchased from Sigma, with the exception of GABAzine, (−)-indolactam-V, (+)-indolactam-V, DLTBOA, PDA, 4α-PDA, UBP-301, and NBQX, from Tocris; and MNIglutamate, from Molecular Probes. GYKI 53655 was a gift from Dr. Antal Simay, IVAX Drug Research Institute.
Supplemental Data The Supplemental Data for this article can be found at http:// www.neuron.org/cgi/content/full/46/5/715/DC1/.
Acknowledgments Y.S. would like to dedicate this work to his wife, Xueqin Xia, with thanks for her support. Thanks to R. Bock and D. Gurfel for skillful technical assistance and to members of the Linden and Bergles labs for helpful comments. C.-M. Tang provided invaluable assistance in designing the photolysis rig. Dr. A. Simay made a generous gift of GYKI 53655. This work was supported by USPHS MH51106 and MH01590 and the Develbiss Fund (D.J.L.). Received: October 20, 2004 Revised: February 2, 2005 Accepted: April 28, 2005 Published: June 1, 2005 References Arnth-Jensen, N., Jabaudon, D., and Scanziani, M. (2002). Cooperation between independent hippocampal synapses is controlled by glutamate uptake. Nat. Neurosci. 5, 325–331.
Asztely, F., Erdemli, G., and Kullmann, D.M. (1997). Extrasynaptic glutamate spillover in the hippocampus: dependence on temperature and the role of active glutamate uptake. Neuron 18, 281–293. Auger, C., and Attwell, D. (2000). Fast removal of synaptic glutamate by postsynaptic transporters. Neuron 28, 547–558. Brasnjo, G., and Otis, T.S. (2001). Neuronal glutamate transporters control activation of postsynaptic metabotropic glutamate receptors and influence cerebellar long-term depression. Neuron 31, 607–616. Brasnjo, G., and Otis, T.S. (2004). Isolation of glutamate transportcoupled charge flux and estimation of glutamate uptake at the climbing fiber-Purkinje cell synapse. Proc. Natl. Acad. Sci. USA 101, 6273–6278. Carter, A.G., and Regehr, W.G. (2000). Prolonged synaptic currents and glutamate spillover at the parallel fiber to stellate cell synapse. J. Neurosci. 20, 4423–4434. Cho, K., Francis, J.C., Hirbec, H., Dev, K., Brown, M.W., Henley, J.M., and Bashir, Z.I. (2003). Regulation of kainate receptors by protein kinase C and metabotropic glutamate receptors. J. Physiol. 548, 723–730. Davis, K.E., Straff, D.J., Weinstein, E.A., Bannerman, P.G., Correale, D.M., Rothstein, J.D., and Robinson, M.B. (1998). Multiple signaling pathways regulate cell surface expression and activity of the excitatory amino acid carrier 1 subtype of Glu transporter in C6 glioma. J. Neurosci. 18, 2475–2485. DiGregorio, D.A., Nusser, Z., and Silver, R.A. (2002). Spillover of glutamate onto synaptic AMPA receptors enhances fast transmission at a cerebellar synapse. Neuron 35, 521–533. Dzubay, J.A., and Otis, T.S. (2002). Climbing fiber activation of metabotropic glutamate receptors on cerebellar Purkinje neurons. Neuron 36, 1159–1167. Espinoza-Rojo, M., López-Bayghen, E., and Ortega, A. (2000). GLAST: gene expression regulation by phorbol esters. Neuroreport 11, 2827–2832. Fournier, K.M., González, M.I., and Robinson, M.B. (2004). Rapid trafficking of the neuronal glutamate transporter, EAAC1: evidence for distinct trafficking pathways differentially regulated by protein kinase C and platelet-derived growth factor. J. Biol. Chem. 279, 34505–34513. González, M.I., and Ortega, A. (1997). Regulation of the Na+-dependent high affinity glutamate/aspartate transporter in cultured Bergmann glia by phorbol esters. J. Neurosci. Res. 50, 585–590. González, M.I., Bannerman, P.G., and Robinson, M.B. (2003). Phorbol myristate acetate-dependent interaction of protein kinase Cα and the neuronal glutamate transporter EAAC1. J. Neurosci. 23, 5589–5593. Hansel, C., and Linden, D.J. (2000). Long-term depression of the cerebellar climbing fiber-Purkinje neuron synapse. Neuron 26, 473–482. Häusser, M., and Roth, A. (1997). Dendritic and somatic glutamate receptor channels in rat cerebellar Purkinje cells. J. Physiol. 501, 77–95. Huang, Y.H., and Bergles, D.E. (2004). Glutamate transporters bring competition to the synapse. Curr. Opin. Neurobiol. 14, 346–352. Huang, H., and Bordey, A. (2004). Glial glutamate transporters limit spillover activation of presynaptic NMDA receptors and influence synaptic inhibition of Purkinje neurons. J. Neurosci. 24, 5659–5669. Huang, Y.H., Dykes-Hoberg, M., Tanaka, K., Rothstein, J.D., and Bergles, D.E. (2004a). Climbing fiber activation of EAAT4 transporters and kainate receptors in cerebellar Purkinje cells. J. Neurosci. 24, 103–111. Huang, Y.H., Sinha, S.R., Tanaka, K., Rothstein, J.D., and Bergles, D.E. (2004b). Astrocyte glutamate transporters regulate metabotropic glutamate receptor-mediated excitation of hippocampal interneurons. J. Neurosci. 24, 4551–4559. Leitges, M., Kovac, J., Plomann, M., and Linden, D.J. (2004). A unique PDZ-ligand in PKCα confers induction of cerebellar longterm synaptic depression. Neuron 18, 585–594. Levenson, J., Weeber, E., Selcher, J.C., Kategaya, L.S., Sweatt,
Neuron 722
J.D., and Eskin, A. (2002). Long-term potentiation and contextual fear conditioning increase neuronal glutamate uptake. Nat. Neurosci. 5, 155–161. Linden, D.J., Smeyne, M., and Connor, J.A. (1993). Induction of cerebellar long-term depression in culture requires postsynaptic action of sodium ions. Neuron 11, 1093–1100. Maragakis, N.J., and Rothstein, J.D. (2004). Glutamate transporters: animal models to neurologic disease. Neurobiol. Dis. 15, 461–473. More, J.C., Troop, H.M., Dolman, N.P., and Jane, D.E. (2003). Structural requirements for novel willardiine derivatives acting as AMPA and kainate receptor antagonists. Br. J. Pharmacol. 138, 1093– 1100. Nishiyama, H., and Linden, D.J. (2004). Differential maturation of climbing fiber innervation in cerebellar vermis. J. Neurosci. 24, 3926–3932. Oliet, S.H., Piet, R., and Poulain, D.A. (2001). Control of glutamate clearance and synaptic efficacy by glial coverage of neurons. Science 292, 923–926. Otis, T.S., Kavanaugh, M.P., and Jahr, C.E. (1997). Postsynaptic glutamate transport at the climbing fiber-Purkinje cell synapse. Science 277, 1515–1518. Reichelt, W., and Knopfel, T. (2002). Glutamate uptake controls expression of a slow postsynaptic current mediated by mGluRs in cerebellar Purkinje cells. J. Neurophysiol. 87, 1974–1980. Rothstein, J.D., Martin, L., Levey, A.I., Dykes-Hoberg, M., Jin, L., Wu, D., Nash, N., and Kuncl, R.W. (1994). Localization of neuronal and glial glutamate transporters. Neuron 13, 713–725. Shen, Y., Hansel, C., and Linden, D.J. (2002). Glutamate release during LTD at cerebellar climbing fiber-Purkinje cell synapses. Nat. Neurosci. 5, 725–726. Shigeri, Y., Shimamoto, K., Yasuda-Kamatani, Y., Seal, R.P., Yumoto, N., Nakajima, T., and Amara, S.G. (2001). Effects of threo-βhydroxyaspartate derivatives on excitatory amino acid transporters (EAAT4 and EAAT5). J. Neurochem. 79, 297–302. Shimamoto, K., Lebrun, B., Yasuda-Kamatani, Y., Sakaitani, M., Shigeri, Y., Yumoto, N., and Nakajima, T. (1998). DL-threo-β-benzyloxyaspartate, a potent blocker of excitatory amino acid transporters. Mol. Pharmacol. 53, 195–201. Stoffel, W., Körner, R., Wachtmann, D., and Keller, B.U. (2004). Functional analysis of glutamate transporters in excitatory synaptic transmission of GLAST1 and GLAST1/EAAC1 deficient mice. Brain Res. Mol. Brain Res. 128, 170–181. Takayasu, Y., Iino, M., and Ozawa, S. (2004). Roles of glutamate transporters in shaping excitatory synaptic currents in cerebellar Purkinje cells. Eur. J. Neurosci. 19, 1285–1295. Tanaka, J., Ichikawa, R., Watanabe, M., Tanaka, K., and Inoue, Y. (1997a). Extra-junctional localization of glutamate transporter EAAT4 at excitatory Purkinje cell synapses. Neuroreport 8, 2461– 2464. Tanaka, K., Watase, K., Manabe, T., Yamada, K., Watanabe, M., Takahashi, K., Iwama, H., Nishikawa, T., Ichihara, N., Kikuchi, T., et al. (1997b). Epilepsy and exacerbation of brain injury in mice lacking the glutamate transporter GLT-1. Science 276, 1699–1702. Tsvetkov, E., Shin, R.M., and Bolshakov, V.Y. (2004). Glutamate uptake determines pathway specificity of long-term potentiation in the neural circuitry of fear conditioning. Neuron 41, 139–151. Wadiche, J.I., and Jahr, C.E. (2001). Multivesicular release at climbing fiber-Purkinje cell synapses. Neuron 32, 301–313. Wadiche, J.I., and Kavanaugh, M.P. (1998). Macroscopic and microscopic properties of a cloned glutamate transporter/chloride channel. J. Neurosci. 18, 7650–7661. Wadiche, J.I., Amara, S.G., and Kavanaugh, M.P. (1995). Ion fluxes associated with excitatory amino acid transport. Neuron 15, 721– 728. Welsh, J.P., Yuen, G., Placantonakis, D.G., Vu, T.Q., Haiss, F., O’Hearn, E., Molliver, M.E., and Aicher, S.A. (2002). Why do Purkinje cells die so easily after global brain ischemia? Aldolase C, EAAT4,
and the cerebellar contribution to posthypoxic myoclonus. Adv. Neurol. 89, 331–359. Xia, J., Chung, H.J., Wihler, C., Huganir, R.L., and Linden, D.J. (2000). Cerebellar long-term depression requires PKC-regulated interactions between GluR2/3 and PDZ domain-containing proteins. Neuron 28, 499–510.