Retrograde modulation of transmitter release by postsynaptic subtype 1 metabotropic glutamate receptors in the rat cerebellum

Abstract

1. The aim of the study was to elucidate the mechanisms underlying the depressant effect of the group I/II metabotropic glutamate receptor (mGluR) agonist 1S,3R-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD) on parallel fibre (PF) to Purkinje cell (PC) synaptic transmission. Experiments were performed in rat cerebellar slices using the whole-cell patch-clamp technique and fluorometric measurements of presynaptic calcium variation 2. Analysis of short-term plasticity, fluctuation of EPSC amplitude and responses of PCs to exogenous glutamate showed that depression caused by 1S,3R-ACPD is presynaptic. 3. The effects of 1S,3R-ACPD were blocked and reproduced by group I mGluR antagonists and agonists, respectively. 4. These effects remained unchanged in mGluR5 knock-out mice and disappeared in mGluR1 knock-out mice. 5. 1S,3R-ACPD increased calcium concentration in PFs. This effect was abolished by AMPA/kainate (but not NMDA) receptor antagonists and mimicked by focally applied agonists of these receptors. Thus, it is not directly due to mGluRs but to presynaptic AMPA/kainate receptors indirectly activated by 1S,3R-ACPD. 6. Frequencies of spontaneous and evoked unitary EPSCs recorded in PCs were respectively increased and decreased by mGluR1 agonists. Similar results were obtained when mGluR1s were activated by tetanic stimulation of PFs. 7. Injecting 30 mM BAPTA into PCs blocked the effects of 1S,3R-ACPD on unitary EPSCs. 8. In conclusion, 1S,3R-ACPD reduces evoked release of glutamate from PFs. This effect is triggered by postsynaptic mGluR1s and thus implies that a retrograde messenger, probably glutamate, opens presynaptic AMPA/kainate receptors and consequently increases spontaneous release of glutamate from PF terminals and decreases evoked synaptic transmission.

Figure 1. Effects of 1S,3R-ACPD PF EPSCs

A, left trace, superimposed sweeps of PF EPSCs elicited by 2 successive PF stimulations with an interstimulus interval of 30 ms in control (1), in 50 μm 1S,3R-ACPD (2) and after washout (3); right trace, scaling of sweep 2 to sweep 1 shows that the slopes of their rising phase are not changed by 1S,3R-ACPD. Note that 1, 2 and 3 correspond to numbers indicated in C. B, bar graph of mean CV (+ s.e.m., n = 16) calculated with CV values obtained during control and in the presence of 50 μm 1S,3R-ACPD. C, plot of the mean ( ± s.e.m.) normalized amplitude (a) or paired-pulse facilitation (PPF) (b) of PF EPSCs recorded from 9 PCs. c, plot of the mean ( ± s.e.m.) amplitude of the current induced in these 9 PCs by 400 ms glutamate pulses. Inset shows an example of these currents recorded from one PC. Y-axis labels are the percentage of the mean amplitude in control (calculated over the 5 min preceding ACPD application).

Figure 2. The depressant effect of 1S,3R-ACPD is due to group I mGluRs

Plot of mean ( ± s.e.m.) PF EPSC amplitude over time. A, effect 1S,3R-ACPD in the presence of the group I antagonist AIDA (300 μm, n = 5 cells), plotted on the same axis as the effect of 1S,3R-ACPD alone. B, effect of 1S,3R-ACPD plotted on the same axis as the effect of the specific group I agonist S-DHPG (50 μm, n = 7 cells).

Figure 6. Effects of 1S,3R-ACPD in the presence of mGluR4 or GABA_B or CB1 receptor antagonists

A, plot of normalized amplitudes (mean ± s.e.m., upper graph) and PPF (mean ± s.e.m., lower graph) of PF EPSCs against time before, during and after bath application of 50 μm 1S,3R-ACPD, in control (⋄ or ▴) or in the presence of 200 μm of the group III mGluR antagonist MSOP (♦ or ▴, n = 5). B, plot of normalized amplitudes (mean ± s.e.m.) of PF EPSCs against time before, during and after bath application of 50 μm 1S,3R-ACPD (filled horizontal bar). Top panel, control conditions (⋄) superimposed on experiments in the presence of 300 nm GABA_B receptor antagonist CGP 55845-A (♦). Lower panel, control conditions (○) superimposed on experiments in the presence of 1 μm CB1 receptor antagonist SR141716-A (•).

Figure 3. Effects of 1S,3R-ACPD on calcium fluorometric measurements from PFs and granule cells

Each panel represents a separate experiment. Data are means ± s.e.m. of 3 consecutive points obtained during acquisition except for C, which shows raw sweeps. A, example of fluorescence measurements of calcium changes (ΔF/F) recorded from the soma of fluo-3-loaded granule cells revealed the lack of calcium changes during bath application of 50 μm 1S,3R-ACPD (left), in contrast to the calcium transient increase evoked in the same cells by bath application of 50 μm glutamate (right). B, example of fluorescence measurements of calcium changes (ΔF/F) in a population of fluo-3-labelled PFs during the application of 50 μm 1S,3R-ACPD (filled bar) in the presence of CNQX (10 μm) and d-APV (100 μm) (hatched bar) (left), or after wash-out of these antagonists (right). C, left panel, example of calcium increases (in the presence of CNQX and d-APV) evoked in a set of PFs loaded with the low affinity dye fluo-4-FF AM by their stimulation, in control, in the presence of 1S,3R-ACPD and during wash-out of 1S,3R-ACPD. Right panel, same as left, in control, in the presence of the group III agonist and during wash-out of l-AP4. D, example of fluorescence measurements of calcium changes (ΔF/F) in a population of fluo-3-labelled PFs during the local co-application through a theta-tube of NMDA + kainate (100 μm each, 1 min, horizontal filled bar) on the set of loaded PFs. Recordings made in the presence of the group I, II and III mGluR antagonists (200 μm each) and of TTX (1 μm) (hatched bar). Vertical scale bars are ΔF/F multiplied by 100 to give the percentage change.

Figure 5. Effects of 1S,3R-ACPD on miniature events in PCs loaded with BAPTA and in the presence of the NMDA receptor antagonist d-APV

All experiments from this figure were made in strontium-containing medium. A, left, superimposed cumulative distributions of normalized amplitude of miniature EPSCs recorded from 5 PCs in TTX alone (control, ⋄) or in TTX + 1S,3R-ACPD (ACPD, ♦). Right, bar graph of mean detection rates (+s.e.m.) of miniature EPSCs presented on the left. The histograms are normalized relative to the mean detection rate of miniature EPSCs in control solution (TTX). B, data pooled from 7 PCs loaded with 30 mm BAPTA. Left, superimposed cumulative distributions of normalized amplitude of spontaneous (circles) and evoked (diamonds) events recorded in control (open symbols) or in the presence of 1S,3R-ACPD (filled symbols). Right, histograms of mean detection rates (+s.e.m.) of events described on the left. The histograms are normalized relative to the mean detection rate of spontaneous EPSCs recorded under control conditions (strontium-containing medium, BAPTA-loaded PCs, no 1S,3R-ACPD). C, d-APV does not block the effects of 1S,3R-ACPD. Left, a raw example of fluorescence (ΔF/F) recorded from a population of fluo-3-labelled PFs during 5 min bath application of 50 μm 1S,3R-ACPD, in the presence of d-APV alone and in the presence of d-APV + CNQX, as indicated by horizontal bars. Right, bar graph of mean detection rates (+s.e.m.) of spontaneous and evoked events recorded from 7 PCs in control (hatched bars) and in the presence of 50 μm 1S,3R-ACPD (filled bars). The bars are normalized relative to the mean detection rate of spontaneous EPSCs recorded in control.

Figure 4. Effects of 1S,3R-ACPD on spontaneous and evoked excitatory events

A–C, experiments in external calcium medium. A, example of 30 s sweeps recorded from a PC in control (top) or in the presence of 25 μm 1S,3R-ACPD (bottom). B, bar graph of mean detection rates (+s.e.m., n = 5) of spontaneous EPSCs recorded in calcium standard bathing medium, in control (hatched bar), and in the presence of 50 μm 1S,3R-ACPD (filled bar). The histograms are normalized relative to the mean detection rate of spontaneous unitary EPSCs in control solution. Cells analysed here are the same as in C. C, superimposed distributions of cumulative normalized amplitude (mean ± s.e.m., n = 5) of spontaneous events, in control calcium medium (□), and in calcium medium containing 50 μm 1S,3R-ACPD (▪). D-F, experiments in external strontium medium. D, examples of desynchronized PF EPSCs recorded in a PC when external calcium was replaced by strontium, in control (top) and in the presence of 50 μm 1S,3R-ACPD (bottom). Total acquisition sweeps were of 1.4 s duration; spontaneous and evoked EPSCs were detected in two time windows as described in Methods. E, bar graph of mean detection rates ( ± s.e.m., n = 6) of spontaneous (left) and evoked EPSCs (right), in control (hatched bars), and in the presence of 1S,3R-ACPD (filled bars). The histograms are normalized relative to the mean detection rate of spontaneous events in control solution. Cells analysed here are the same as in F. F, left graph, superimposed distributions of cumulative normalized amplitudes of spontaneous EPSCs (mean ± s.e.m., n = 6 cells), in control strontium medium (⋄), and in the presence of 50 μm 1S,3R-ACPD (♦). Right graph, same as left graph but for evoked events.

Figure 8. Scheme summarizing the retrograde control of PF activity by PCs

1, activation of mGluR1 in the PC, either pharmacologically or by tetanic stimulation of PFs. 2, subsequent increase in calcium (Ca²⁺) concentration in the PC. 3, calcium-dependent release of a retrograde messenger, most probably glutamate (Glu). 4, activation of ionotropic glutamate receptors (iGluRs) located in PFs and subsequent depolarization of PFs which propagates to their terminals through action potential (AP)-dependent process. 5, PF terminals undergo vesicular release of glutamate, observed at the level of PCs as an increase in frequency of spontaneous EPSCs and a decrease in frequency of evoked ones. This retrograde depolarization of PFs caused by the activation of postsynaptic mGluR1s explains the decrease in the amplitude of PF EPSCs during mGluR1 activation.

Figure 7. A tetanic stimulation of PFs increases the frequency of spontaneous EPSCs

A, example of a sweep showing the increase in the frequency of spontaneous EPSCs elicited by a 100 Hz train of 7 PF stimulations (arrows). B, example of a sweep recorded from the same PC as in A, showing that 1 PF stimulation does not change the frequency of spontaneous EPSCs. C, bar graph summarizing the effects of a train of 7 PF stimulations on the frequency of spontaneous EPSCs (the three filled bars) normalized relative to their frequency during the pre-tetanus control period (hatched bar) in control, for PCs loaded with 30 mm BAPTA (n = 4) or in the presence of the group I mGluR antagonist AIDA (200 μm, n = 3).