Initiation of depolarization-evoked release depends on the timing of the GCs. Experiments performed at 10°C (n = 4–5 for each histogram). Protocols shown on the left of each row correspond to the entire row. Test = test pulse (−0.5 μA, 0.5 ms); Pre = prepulse (−1.0 μA, 0.1 ms); Wide pre = wide prepulse (−0.2 μA, 0.9 ms). (A) In WT mice, application of the prepulse 1 ms before the test pulse increased the amount of release (A₁) and caused release to initiate earlier (A₂, Kolmogorov-Smirnov [KS] test, P < 0.03). Gal (insets) abolished the effect of the prepulse on release. (B) In M₂KO mice the strong and brief prepulse did not affect the amount (B₁) or the initiation (B₂) of release (P > 0.9). (C) In WT mice, application of a wide prepulse, which causes release by itself (C₁, short dashed line), increased the amount of test pulse release (C₁) but did not affect initiation of release (C₂, P > 0.9). Inset shows representative (n = 3) Ca²⁺ currents derived from ENTCs without (solid line) and with (dashed line) the wide prepulse preceding the ENTC. (D) In M₂KO mice, the wide prepulse increased the amount of release (D₁) and caused release to initiate earlier (D₂, P = 0.008). (E) Initiation of test pulse release in WT mice without (solid line) and with (dashed line) the brief and strong prepulse and in M₂KO mice without a prepulse (short dashed line) shown on the same axes.