Spontaneously Recycling Synaptic Vesicles Constitute Readily Releasable Vesicles in Intact Neuromuscular Synapses.

Emerging evidence shows that spontaneous synaptic transmission plays crucial roles on neuronal functions through presynaptic molecular mechanisms distinct from that of action potential (AP)-evoked transmission. However, whether the synaptic vesicle (SV) population undergoing the two forms of transmission is segregated remains controversial due in part to the conflicting results observed in cultured neurons. Here we address this issue in intact neuromuscular synapses using transgenic zebrafish larvae expressing two different indicators targeted in the SVs: a pH-sensitive fluorescent protein, pHluorin, and a tag protein, HaloTag. By establishing a quantitative measure of recycled SV fractions, we found that ∼85% of SVs were mobilized by high-frequency AP firings. In contrast, spontaneously recycling SVs were mobilized only from <8% of SVs with a time constant of 45 min at 25°C, although prolonged AP inhibition mobilized an additional population with a delayed onset. The mobilization of the early-onset population was less temperature-sensitive and resistant to tetanus toxin, whereas that of the late-onset population was more sensitive to temperature and was inhibited by tetanus toxin, indicating that prolonged AP inhibition activated a distinct molecular machinery for spontaneous SV fusion. Therefore, the early-onset population limited to <8% was likely the only source of spontaneous release that occurred physiologically. We further showed that this limited population was independent from those reluctant to fuse during AP firing and was used in both the hypertonic stimulation and the immediate phase of AP-evoked releases, thereby matching the characteristics of the readily releasable pool. Synaptic vesicles (SVs) are divided into functionally distinct pools depending on how they respond to action potential (AP) firing. The origin of SVs used for spontaneous fusion remains enigmatic despite intensive studies in cultured preparations. We addressed this question in intact neuromuscular synapses and provided two findings. First, prolonged AP inhibition activated a distinct population of fusion, which needs to be distinguished from genuine spontaneous fusion arising from a highly limited fraction. Second, the limited fraction observed early in the AP inhibition period exhibited the characteristics of readily releasable pool in the subsequent round of stimulation. Our study revealed that the origin of spontaneous SV fusion is restricted to the readily releasable pool among the SV pools involved in AP-evoked fusion.