The endocytic blocker Dynasore or Pitstop-2 inhibits slow, fast-accelerating and fast endocytosis at the calyx of Held

(A) Average traces of slow endocytic membrane capacitance changes (ΔCm) in response to a 5-ms depolarizing pulse (stepping from -70 mV to +10 mV) in the absence (control, black trace) or presence of Dynasore (100 µM, brown trace) or Pitstop-2 (25 µM, green trace), recorded from the calyx of Held presynaptic terminal in slices from P13-15 mice at physiological temperature (PT, 37°C) and in 2.0 mM Ca2+ aCSF. The 4th panel from the left shows the superimposed average ΔCm traces under control, Dynasore and Pitstop-2. The rightmost bar graph shows the endocytic decay rate (calculated from the slope 0.45-5.45 s after stimulation) that was slower in the presence of Dynasore (8.5 ± 2.8 fF/s; n = 5; p = 0.004, Student’s t-test) or Pitstop-2 (11.9 ± 3.4 fF/s; n = 5; p = 0.015) than control (28.8 ± 3.7 fF/s; n = 5). Significance level was set at p < 0.05, denoted with asterisks (*p < 0.05, **p < 0.01, ***p < 0.001)

(B) Fast-accelerating endocytosis induced by a train of 20-ms depolarizing pulses (repeated 15 times at 1 Hz) in the absence (control) or presence of Dynasore or Pitstop-2. Averaged traces are shown as in (A). The 4th panel shows the endocytic rates (fF/s) calculated from the slope of Cm decay, 0.45-0.95 ms after each stimulation pulse under control, Dynasore, or Pitstop-2 (superimposed). The rightmost bar graph shows the endocytic rate averaged from stimulations #12-15 (bar in 4th panel) that was slower in the presence of Dynasore (121 ± 19 fF/s, n = 6; p = 0.0045, Student’s t-test) or Pitstop-2 (133 ± 16 fF/s, n = 5; p = 0.007) than control (251 ± 28 fF/s; n = 6), indicating significant inhibition of the fast-accelerating endocytosis by Dynasore or Pitstop-2 (also see Table S1).

(C) Fast-endocytosis (average traces) evoked by a train of 20-ms pulses (repeated 10 times at 10 Hz) in the absence (control) or presence of Dynasore or Pitstop-2. The 4th panel shows cumulative ΔCm traces (superimposed) in an expanded timescale during and immediately after the 10-Hz train. The rightmost bar graph indicates endocytic decay rates (measured 0.45-1.45 s after the 10th stimulation) in Dynasore (204 ± 25.3 fF; n = 4; p = 0.02; Student’s t-test) or in Pitstop-2 (131 ± 24.6 fF; n = 4; p = 0.002) both of which were significantly slower than control (346 ± 40.1 fF; n = 6).

Endocytic blockers rapidly enhance synaptic depression in a stimulation frequency-dependent manner, but do not prolong the recovery from depression at the calyx.

(A1, B1) A train of 30 EPSCs were evoked at the calyx of Held by afferent fiber stimulation at 10 Hz (A1) or 100 Hz (B1) in the absence (control, black traces) or presence of Dynasore (brown traces) or Pitstop-2 (green traces) at PT (37°C) in 1.3 mM Ca2+ aCSF. Panels from left to right: left-top: sample EPSC traces; left-bottom: normalized average EPSC amplitudes at each stimulus number; right-top bar graph: steady-state depression of EPSC amplitudes (mean of EPSCs #26-30, bar in the second panel); right-bottom: 1st-3rd EPSC amplitudes in expanded timescale.

(A2, B2) The recovery of EPSCs from STD in control, or in the presence of Dynasore or Pitstop-2 at the calyx of Held measured using a stimulation protocol (shown on top in A2); a train of 30 stimulations at 10 Hz (A2) or 100 Hz (B2) followed by test pulses after different time intervals (Δt: 0.02, 0.1, 0.3, 1, 3, 8, 12, and 20s). The EPSC amplitude after Δt (IΔt) relative to the first EPSC in the stimulus train (I1st) was normalized by subtracting the steady state EPSCs (Iss) to measure the recovery rates.

(A1) During 10-Hz stimulation, the steady-state depression under control (0.55 ± 0.02; n = 11) was slightly enhanced in the presence of Dynasore (0.48 ± 0.02; n = 12; p = 0.03, Student’s t-test) but not in Pitstop-2 (0.53 ± 0.02; n = 7; p = 0.4, no significant difference).

(A2) After 10-Hz stimulation, the time constant of EPSCs recovery in control (2.3 ± 0.4 s; n = 9) was unchanged in the presence of either Dynasore (1.7 ± 0.2 s; n = 8; p = 0.2, Student’s t-test) or Pitstop-2 (1.9 ± 0.3 s; n = 7; p = 0.4).

(B1) During 100-Hz stimulation, EPSCs underwent a significant depression starting at the 2nd stimulation (10 ms) in Dynasore (0.6 ± 0.03; n =10; p < 0.001, t-test) or Pitstop-2 (0.65 ± 0.06; n = 7; p = 0.005) than control (0.91 ± 0.05; n = 11). Bar graph indicates STD magnitudes; control: (0.42 ± 0.025), Dynasore: (0.25 ± 0.02; p < 0.001), and Pitstop-2: (0.26 ± 0.03; p < 0.001).

(B2) After 100-Hz stimulation, both fast and slow recovery time constants were significantly faster in the presence of Dynasore or Pitstop-2 than control (bar graphs, Table S2).

Scaffold machinery inhibitors have no effect on endocytic membrane retrievals at the calyx of Held

The CDC42 inhibitor ML141 (10 µM, cyan) or actin depolymerizer Latrunculin B (Lat-B, 15 µM, red) had no effect on slow endocytosis in response to a 5-ms depolarizing pulse (A) or on fast accelerating endocytosis (B; induced by 1-Hz train of 20 ms x 15 pulses) or fast endocytosis (C; evoked by a 10-Hz train of 20 ms x 10 pulses) at PT and in 2.0 mM Ca2+ aCSF.

(A) Averaged and superimposed ΔCm traces in response to a 5-ms pulse. The rightmost bar graph indicates the endocytic decay rate in control (28.8 ± 3.7 fF; n = 5), unchanged by ML141 (21.9 ± 5.8 fF; n = 6; p = 0.42, Student’s t-test) or Lat-B (28.5 ± 5.2 fF; n = 6; p = 0.97, Table S1).

(B) The average endocytic rate (#12-15) in control (251 ± 28 fF/s; n = 6) unaltered by ML141 (214 ± 19 fF/s; n = 5; p = 0.3, Student’s t-test) or Latrunculin-B (233 ± 48 fF/s; n = 4; p = 0.7; Table S1).

(C) Fast endocytic decay rate in the presence of ML141 (242 ± 21.5 fF; n = 4; p = 0.054, t-test) or Lat-B (350 ± 33.3 fF; n = 5; p = 0.95) was not different from control (346 ± 40.1 fF; n = 6; Table S1).

Scaffold machinery inhibitors strongly enhance synaptic depression in a stimulation frequency-independent manner, without prolonging the recovery from depression at the calyx

(A1, B1) EPSCs (30x) evoked at the calyx of Held by afferent fiber stimulation at 10 Hz (A1) or 100 Hz (B1) in the absence (control, black) or presence of ML141 (cyan) or Lat-B (red) at PT and in 1.3 mM Ca2+ aCSF. Panels from left to right: like Figure 2.

(A2, B2) The recovery of EPSCs from STD in control, or in the presence of ML141 or Latrunculin-B at the calyx of Held measured at 10 Hz (A2) or 100 Hz (B2) – like figure 2.

(A1) At 10-Hz stimulation, enhancement of depression became significant from the 3rd stimulation (200 ms) in the presence of ML141 (0.61 ± 0.03; n = 9; p = 0.013, Student’s t-test) or Lat-B (0.55 ± 0.04; n = 7; p = 0.006) compared to control (0.74 ± 0.04, n = 11). Bar graph indicates the steady-state depression (STD) strongly enhanced from control (0.55 ± 0.02) by ML141 (0.4 ± 0.02; p < 0.001) or Lat-B: (0.36 ± 0.02; p < 0.001).

(A2) After a train of 30 stimulations at 10 Hz, the time constant of EPSCs recovery under control was unchanged by ML141 or Latrunculin-B (Table S2).

(B1) At 100-Hz stimulation, EPSCs showed significant enhancement of depression at the 2nd stimulation (10 ms) in the presence of ML141 (0.65 ± 0.05; n = 10; p < 0.001) or Lat-B (0.58 ± 0.05; n = 8; p < 0.001) from control (0.91 ± 0.05; n = 11). Bar graph indicates strong STD produced by ML141 (0.22 ± 0.02; p < 0.001) or Lat-B (0.18 ± 0.013; p < 0.001) compared to control (0.42 ± 0.025).

(B2) The time course of EPSC recovery from STD induced by a train of 30 stimulations at 100 Hz, indicating no significant change caused by ML141 or Lat-B in the recovery time constants (Table S2). The stimulation and recovery protocols were the same as those used in Figure 2.

Endocytic and scaffold machineries co-operate for rapid vesicle replenishment during high-frequency transmission at the calyx of Held.

(A) Exponential curve fits to the time-course of synaptic depression during 100-Hz stimulation under control and in the presence of either endocytic blockers or scaffold protein inhibitors. The control time-course was best fit to a single exponential, whereas the time-course in the presence of endocytic blockers or scaffold blockers was fit best to double exponential function with fast and slow time constants.

(B) Parameters for the curve-fit, including fast and slow time constants (τfast, τslow), weighted mean time constant (τmean) and relative ratio of fast and slow components (Af/As). Similar fast time constant (τfast) in the presence of endocytic inhibitors or scaffold-inhibitors suggests simultaneous operation of endocytic and scaffold mechanisms for countering the synaptic depression. Similar slow decay time constant irrespective of the presence or absence of blockers suggests that the endocytosis and scaffold mechanisms for vesicle replenishment operates predominantly at the beginning of the high frequency stimulations.

(C) Enhancement of synaptic depression by co-application of Dynasore and Latrunculin-B was like Latrunculin-B alone and stronger than Dynasore alone. Sample EPSC traces and normalized depression time courses are shown on the upper and lower left panels, respectively. The STD magnitudes are compared in bar graph; control: (0.42 ± 0.025; n = 11), Dynasore: (0.25 ± 0.02; n = 10, p = 0.009), Latrunculin-B: (0.18 ± 0.013; n = 8, p = 0.005 vs Dynasore), and Dyn + Lat-B together: (0.17 ± 0.02; n = 11; p = 0.005 vs Dynasore).

Endocytic blockers attenuate synaptic facilitation, but scaffold inhibitors have no effect at hippocampal CA1 synapses

(A-D) A train of 30 EPSCs evoked in hippocampal CA1 pyramidal cells by Schaffer collateral stimulation at 10 Hz (A and C) or 25 Hz (B and D) in the absence (control, black) or presence of endocytic blocker Dynasore (100 μM, brown) or Pitstop-2 (25 μM, green) (A, B), or scaffold protein inhibitor ML141 (10 µM, cyan) or Latrunculin-B (15 µM, red) (C, D) at PT (37°C) and in 1.3 mM Ca2+ aCSF. Top panels show sample EPSC traces. Lower panels show average EPSC amplitudes normalized and plotted against stimulation numbers. Bar graphs show EPSCs amplitudes averaged from #26-30 events.

(A) At 10 Hz stimulation, EPSCs in control showed facilitation reaching a peak at the 7th stimulation (2.34 ± 0.3; n = 17). At this point, in the presence of Dynasore (1.44 ± 0.15; n = 10; p = 0.012, Student’s t-test) facilitation was significantly attenuated. The effect of Pitstop-2 (1.85 ± 0.14; n = 14; p = 0.06) was not different. Towards the end of stimulus train (#26-30), synaptic facilitation in control (1.85 ± 0.17) was significantly attenuated by Dynasore (0.83 ± 0.07; p < 0.001) or Pitstop-2 (1.21 ± 0.08; p = 0.002).

(B) At 100 Hz stimulation, synaptic facilitation peaked at the 12th stimulation in control (3.5 ± 0.4; n = 16), at which the facilitation was significantly attenuated by Dynasore (1.6 ± 0.11; n = 11; p < 0.001, t-test) or Pitstop-2 (2.11 ± 0.22; n = 14; p = 0.004). Also, at #26-30, synaptic facilitation in control (2.65 ± 0.3) was strongly attenuated by Dynasore (0.94 ± 0.1; p < 0.001) or Pitstop-2 (1.62 ± 0.2; p = 0.006).

(C) At 10 Hz stimulation, the peak facilitation at the 7th stimulation in control (2.34 ± 0.3; n = 17) was not significantly changed by ML141 (2.5 ± .22, n = 10; n = 0.75, Student’s t-test) or Lat-B (2.3 ± 0.23; n = 12; p = 0.93). Likewise, the facilitation at #26-30 in control (1.85 ± 0.17) was not altered by ML141 (2.03 ± 0.16; p = 0.44) or Lat-B (1.84 ± 0.1; p = 0.96).

(D) At 100 Hz stimulation, peak facilitation at 12th stimulation in control (3.5 ± 0.4; n = 16) was unchanged by ML141 (3.3 ± 0.34; n = 10; p = 0.72, t-test) or Lat-B (3.1 ± 0.4; n = 11; p = 0.42). Facilitation at #26-30 events in control (2.65 ± 0.3) was also unaltered by ML141 (2.53 ± 0.2; p = 0.74) or Lat-B (2.5 ± 0.25; p = 0.7).

Hypothetical scheme on the roles of endocytosis and scaffold machineries in vesicle replenishment during high-frequency transmission at brainstem relay synapses and hippocampal plastic synapses.

During high-frequency transmission, endocytosis plays site-clearance role allowing new vesicle to be replenished to release site. This endocytic function counteracts synaptic depression caused by vesicle depletion, thereby maintaining excitatory synaptic strength above firing threshold, enabling high-fidelity fast neurotransmission at sensory relay synapses. This synaptic strengthening function of endocytosis augments synaptic facilitation at synapses with long-term plasticity, thereby boosting its induction capability for memory formation. Presynaptic scaffold machinery plays a powerful vesicle replenishment role rapidly translocating new vesicles to open release sites or those just opened by endocytic site-clearance. This scaffold function is specifically devoted to fast synapses with high release probability but not to plastic synapses, where vesicle depletion is minimal due to low release probability.