Figure 4.

Glutamate dialysis increases larger mEPSCs to a smaller extent. A, Left, The mean mEPSC, averaged over every 60 s, was increased after a presynaptic pipette containing 100 mm glutamate broke into a calyx (arrow). Right, Sample mEPSCs (top) and the mean mEPSC (bottom) collected before (Ctrl) (black) and after glutamate (Glu) (red) dialysis. The mean EPSC amplitude was 26.5 ± 0.7 pA (n = 421 mEPSCs) (black) for control and 39.3 ± 0.8 pA (n = 452 mEPSCs) (red) after glutamate dialysis. Data in A–D were from the same synapse. B, The mEPSC amplitude distribution before (Ctrl) (black) and after glutamate dialysis (Glu) (red). Background noise distribution (peaks centered at ∼0 pA) was obtained from ∼10–20 ms of recordings in which no mEPSCs were evident (also applies to Figs. 5B, 6A,D). C, The cumulative probability plotted versus the mEPSC amplitude before (Ctrl) (black) and 2 min after presynaptic break-in with a pipette containing 100 mm glutamate (Glu) (red). The dotted lines indicate that, at a given cumulative probability level, Y, the corresponding mEPSC amplitude in control (A_Ctrl) and that after glutamate dialysis (A_Glu) can be found. The ratio between A_Glu and A_Ctrl is R_Glu/Ctrl. D, R_Glu/Ctrl decreased as A_Ctrl increased. R_Glu/Ctrl was calculated for every 5% increase in the cumulative probability level and every 1% increase between 95 and 99% of the cumulative probability level. Thus, for every paired recording such as the one shown in C, 23 R_Glu/Ctrl values were obtained with corresponding A_Ctrl values. E, Pooled data from 16 recorded synapses showing that R_Glu/Ctrl decreased as A_Ctrl increased. R_Glu/Ctrl (23 values) and the corresponding A_Ctrl from each of 16 paired recordings were pooled and then sorted such that A_Ctrl ascended monotonically and binned for every 16 values. The binned R_Glu/Ctrl values were plotted as a function of the corresponding binned A_Ctrl values. Error bars indicate SEM.