In the CSP-α KO, dynasore also induced a reduction in ΣQC in comparison to control conditions (59,135 ± 7,207 in control and 39,961 ± 5,525 in dynasore) (Figure 6I). We interpreted that under such stimulation conditions, 139.1 ± 5.2% of the vesicles in WT junctions were reused by a dynasore-sensitive mechanism (Maeno-Hikichi et al., 2011), in contrast to only 46.0 ± 7.4% of vesicles recycled in mutant synapses (Figure 6J). A recent study at the frog NMJ (Douthitt et al., 2011) has reported that dynasore treatment increases find more release probability at low (1 Hz) but not at high (50 Hz) stimulation frequency, so we cannot rule

out that the fluorescence increase in dynasore attributed to endocytosis inhibition might have, in the worst of the cases, a minor component due to increased exocytosis. We have evaluated the increase in cumulative quantal content (QC) in dynasore compared to control conditions to find that such a ratio is the same for WT and mutant junctions (Figure S4G). In any case, we do not consider that such an effect of dynasore on neurotransmitter release interferes significantly with its major blocking effect of endocytosis that we have used in our study. In summary,

in the absence of CSP-α, dynasore-sensitive recycling of synaptic vesicles was impaired and that could contribute to the strong synaptic depression under repetitive stimulation at the terminals from CSP-α KO mice. We analyzed the uptake and release of the stiryl dye FM2-10 at the NMJ in CSP-α KO VX-770 cost mice that were not spH transgenic. We depolarized motor nerve terminals (600 s at 30 Hz) in the presence of FM2-10 to label the entire recycling pool (Perissinotti et al., 2008), washed out the noninternalized dye and induced dye release (600 s at 30 Hz). The WT junctions loaded the dye efficiently and, upon stimulation, underwent almost total destaining (Figures 7A, arrowheads, and 7B) as expected for normal synaptic vesicle endo- and exocytosis. The mutant terminals internalized the dye very Sitaxentan efficiently too (Figure 7A, arrows). However, upon stimulation, the dye released from the mutant terminals

was dramatically low (Figure 7B). FM2-10 loading at mutant terminals was even higher than at the controls (Figure 7C) (51%, p = 0.04 Student’s t test). Nevertheless, in mutant nerve terminals stimulated to release, most of the dye (66.9 ± 2.7% of the total loaded) became trapped inside, whereas in the control terminals the residual dye was very little (19.9 ± 4.4%, p < 0.001, Student’s t test) (Figure 7D). To go deeper in our study, we carried out ultrastructural analysis of synaptic terminals with electron microscopy. In general, mutant junctions fixed in resting conditions presented normal postsynaptic foldings and similar nerve terminal size and vesicle density to WT terminals (Figure 7E, panels a–c, and S5A).