T., unpublished data). After glutamate washout, the mEPSC frequency decreased by 20-fold (from 11.4 ± 3.6 Hz to 0.54 ± 0.16 Hz, n = 6) and recovered fully after glutamate uncaging (13.7 ± 3.2 Hz) (Figure 1E). The mEPSC amplitude after glutamate uncaging

(33.9 ± 2.0 pA, n = 6) remained similar to the control before glutamate washout (34.6 ± 1.9 pA, n = 6). When the mEPSC frequency significantly decreases owing to the vesicular glutamate depletion, many mEPSC events will become undetectable, with their amplitudes being merged into a noise level (∼5 pA). In such a condition, the mean amplitude of detectable mEPSCs no longer provides a reliable measure for the quantal size. Therefore, to assess the time course of vesicle refilling, we adopted quantal charge (time integral of mEPSCs in 1 s

windows), which reflects both the amplitude and frequency of mEPSCs. Upon glutamate uncaging, quantal charge increased BMS-754807 supplier with a time constant of 18.3 s (±2.1 s, n = 6) that was similar to the recovery time constant of evoked EPSCs (17.2 s, Figure 1A). Altogether, these results indicate that the recovery of the evoked EPSC amplitude is caused primarily by vesicle refilling with glutamate. Glutamate-binding affinity (Km) and kinetics of VGLUT have been determined in isolated or reconstructed vesicles ( Naito and Ueda, 1985; Carlson et al., 1989; Bellocchio et al., 2000; Gras et al., 2002). The speed of glutamate uptake depends upon the copy number of VGLUT on vesicles and extravesicular glutamate concentrations, whereas the affinity Proteases inhibitor Edoxaban is an intrinsic property of the glutamate transporter. To determine these parameters for vesicles in the calyx of Held terminal, we first examined the relationship between presynaptic cytosolic glutamate concentration ([glu]i) and the magnitude of EPSCs ( Figure 2A) by switching presynaptic whole-cell pipettes containing

different concentrations (0.1–10 mM) of glutamate. Next, we varied [glu]i by photolysis of the MNI-glutamate of different concentrations (2–10 mM) after depleting vesicular glutamate ( Figure 2B). Photolysis of higher concentrations of MNI-glutamate produced faster and larger recoveries of EPSCs, with the recovery time constant (τ) being inversely related to the magnitude of recovery after glutamate uncaging ( Figure 2B). By combining these relationships ( Figures 2A and 2B), we plotted τ against [glu]i ( Figure 2C). In the Lineweaver-Burk plot, Km was estimated as 0.91 mM, which is similar to those estimated for VGLUT1 and VGLUT2 reconstituted in the heterologous expression system (1–2 mM) ( Bellocchio et al., 2000; Gras et al., 2002). These results confirm that the EPSC recovery was caused by vesicle refilling with glutamate via VGLUTs. The maximal refilling time constant 1/Vmax was estimated to be 15.1 s, which is 10–100 times faster than those estimated in isolated vesicles ( Naito and Ueda, 1985; Maycox et al., 1988; Carlson et al., 1989).