Figure 1. High frequency PFs stimulation induces GABAA receptor dependent LTP at PF to PN synapses.
A schematic representation of the cerebellar microcircuit and experimental setting is shown in panel a. Purkinje neurons (PN) receive excitatory inputs from climbing fibers (CF) and mossy fibers (MF) via granule cells (GC) activation. Parallel fibers excitatory synapses (green+) drive PNs and inhibitory (red −) molecular layer interneurons stellate (SC) and basket (BC) cells. Traces elicited by paired-pulses PFs stimulation (Stim, a) in a voltage-clamped PN (Rec, a) at different time points are showed in the inset of panel b. The Baseline trace was obtained from averaging of all recordings during baseline while the t = 45 min trace is the average of three consecutive PF-mediated responses recorded every 20 seconds at the indicated time point. High frequency PFs stimulation induced a long lasting increase in PNs response (MLIdep-LTP, b: mean ± SEM, N = 5, RM ANOVA P < 0.001). Bath application of the GABAA receptor antagonist SR95531 prevented MLIdep-LTP (c, mean ± SEM, N = 5, RM ANOVA P < 0.001). A summary graph for SR95531-mediated effect on MLIdep-LTP is shown in panel e (bar represents normalized PF-Rsp at t = 45 min, mean ± SEM, data from panel b and c), *indicates a statistically significant difference among values (t-test, P = 0.011). Keeping GABAergic transmission intact only during high frequency PFs stimulation was sufficient to induce MLIdep-LTP (d). For these experiments, a 10 minutes baseline was established with SR95531 and the antagonist was washed out for at least 15 minutes before the induction protocol was applied; SR95531 was added back to the recording chamber immediately or 15 minutes after high frequency stimulation (d, mean ± SEM, N = 7, RM ANOVA P < 0.001; e, bar represents normalized PF-Rsp at t = 65 min, mean ± SEM, data from panel d). PPRs value (mean ± SEM) for the baseline (t = 10) and the post induction phase (t = 45) with or without bath application of SR95531 are shown in panel f.