Noradrenergic control of associative synaptic plasticity by selective modulation of instructive signals

Abstract

Synapses throughout the brain are modified through associative mechanisms in which one input provides an instructive signal for changes in the strength of a second coactivated input. In cerebellar Purkinje cells, climbing fiber synapses provide an instructive signal for plasticity at parallel fiber synapses. Here, we show that noradrenaline activates alpha2-adrenergic receptors to control short-term and long-term associative plasticity of parallel fiber synapses. This regulation of plasticity does not reflect a conventional direct modulation of the postsynaptic Purkinje cell or presynaptic parallel fibers. Instead, noradrenaline reduces associative plasticity by selectively decreasing the probability of release at the climbing fiber synapse, which in turn decreases climbing fiber-evoked dendritic calcium signals. These findings raise the possibility that targeted presynaptic modulation of instructive synapses could provide a general mechanism for dynamic context-dependent modulation of associative plasticity.

Figure 1. Noradrenaline decreases EPSC amplitude at the climbing fiber to Purkinje cell synapse through activation of α2-adrenergic receptors

(A–D) Voltage-clamp recordings of CF-EPSCs in PCs in response to pairs of CF stimuli separated by 30 ms. (A) EPSCs from a representative experiment are shown in control conditions (black) and in the presence of 5 μM NA (red). (B) Effects of NA and subsequent application of the α2-receptor antagonist yohimbine (15 μM) on the first EPSC (filled circles) and paired-pulse ratio (open circles) are summarized (n=6). (C,D) Experiments similar to those in (A, B) are shown with the exception that the α2-receptor agonist UK14304 (15 μM) is used rather than NA (n=5). (E, F) PFs were activated with a pair of stimuli separated by 30 ms and the resulting EPSCs were recorded from PCs in voltage-clamp with a Cs-based internal solution in control conditions, in the presence of NA, and in the presence of both NA and 15 μM yohimbine (n=5). Responses are normalized to the average amplitude of the first EPSC, before drug application. (G,H) Summary of effects on the first CF-EPSC (G) and paired pulse ratio (H) of: NA; UK14304; the α1- and β-adrenergic receptor agonists phenylephrine and isoproterenol (10 μM each, n=6); and NA in the presence of a cocktail of antagonists for CB1Rs (AM251, 5 μM), adenosine A1Rs (DPCPX, 5 μM), type I/II mGluRs (MCPG, 500 μM), and GABABRs (CGP 55845A, 2 μM) (n=4). Responses are normalized to the amplitude of the first EPSC in control conditions (dashed line).

Figure 2. Noradrenaline decreases release probability at the CF-PC synapse

(A,B) Effects of NA on CF-EPSCs recorded from PCs in the presence of the rapid low-affinity AMPA receptor antagonist DGG (n=7). Data are plotted as in Figure 1. (C–G) Effects of NA on CF-EPSCs recorded with an external recording solution in which calcium was replaced with strontium. (C–F) A representative experiment shows responses in control conditions (black) and in the presence of NA (red) for the average CF-EPSCs (C), individual traces showing asynchronous release events (filled circles) (D), cumulative histograms (E), and average quantal events (F). (G) The effects of NA on the evoked CF-EPSC (eEPSC) amplitude, the amplitude of the quantal events, and the frequency of evoked quantal events in the presence of strontium are summarized (n=6 PCs).

Figure 3. Noradrenaline alters climbing fiber-evoked complex spikes and associated Purkinje cell dendritic calcium signals

(A) A two-photon image of a PC highlighting the region of interest for measurement of dendritic calcium signals. (B,C) A representative experiment is shown in which a CF was stimulated and the resulting complex spike waveforms (B), complex spike spikelets (B, inset) and dendritic calcium transients (C) were simultaneously recorded from a PC in control conditions (black) and in the presence of 5 μM NA (red). (D) Summary (n = 6 PCs) of noradrenergic effects on the number of complex spike spikelets and (E) the duration of the after-hyperpolarization. (F) Effects of NA on dendritic calcium transients in the same PCs. Peak calcium transients in the presence of NA, following subsequent addition of the α2-adrenergic receptor antagonist yohimbine (15 μM), and in the presence of NA with GTPγS (n=6) included in the internal recording solution are each shown normalized to control conditions. NA caused a decrease in CF-evoked calcium elevation that was reversed by coapplication of yohimbine but unaffected by GTPγS.

Figure 4. α2-receptor activation disrupts associative short-term plasticity

PF-EPSPs were measured with test pulses presented at 0.5 Hz before and after conditioning trains consisting of either stimulation of PFs only or PFs and CFs. (A,C,E) Responses of a PC to a PF-only train (left) and PF+CF trains (right) are shown for control conditions (A), in the presence of the α2 agonist UK14304 (15 μM, C, red), and after drug washout (E). Vertical lines beneath traces indicate timing of PF (thin) and CF (thick) stimuli. Insets display the PF-EPSPs measured in response to test pulses before (dark) and after (light) conditioning trains. (B,D,F) Summary (n=7 PCs) of average PF-EPSP amplitudes in response to test pulses before and after conditioning trains that were delivered at time 0. Responses are normalized to the average EPSP amplitude before the conditioning train. PF+CF trains (filled symbols) resulted in greater suppression of PF-EPSP amplitude than PF-only trains (open symbols). The suppression resulting from PF+CF trains was selectively reduced by UK14304.

Figure 5

Parallel fiber plasticity and postsynaptic calcium elevations are not affected by α2 receptor activation. (A,B) A representative experiment is shown in which PFs were activated with test pulses presented at 0.5 Hz and the resulting PF-EPSPs were recorded from PCs in current-clamp with a K-based internal solution. Conditioning trains consisted of 10 PF stimuli at 100 Hz. (A) Responses to the conditioning trains and PF-EPSPs before and after the conditioning trains (insets) were measured in control conditions (top, black) and in the presence of 15 μM UK14304 (bottom, red). (B) Local dendritic calcium signals were measured in response to the conditioning train. Inset, summary (n=5). (C,D) Summaries (n=5) of the time course of PF-EPSP amplitude (C) and the amount of EPSC suppression (D) in control conditions (black) and in the presence of UK14304 (red).

Figure 6

Activation of α2-receptors interferes with the induction of associative LTD at PF-to-PC synapses. (A) A representative experiment is shown in which LTD was induced with a protocol consisting of a train of 10 PF stimuli at 100 Hz followed by 2 CF stimuli at 20 Hz. This conditioning train was repeated 30 times every 10 seconds. Vertical lines beneath traces indicate timing of PF (thin) and CF (thick) stimuli. (A, inset) Average PF-EPSPs measured for the 5 min. before (black) and 15–20 min. after (gray) the induction protocol. (B) A separate representative experiment conducted in the presence of 15 μM UK14304 is shown. (C) Summary of the time course and amplitude of LTD in control conditions (n=13, black) and in the presence of UK14304 (n=12, red). (D) Cumulative histogram of the ratio of EPSP amplitudes 6–30 min. after/0–10 min. before LTD induction.