This Week in The Journal

doi:10.1523/JNEUROSCI.twij.42.17.2022

Spontaneous Release Draws from Readily Releasable Pool

Yoshihiro Egashira, Ayane Kumade, Akio Ojida, and Fumihito Ono

Axon terminals contain numerous synaptic vesicles. Vesicles docked at the presynaptic active zone are the first to fuse when the neuron begins to spike and are therefore called the readily releasable pool. During prolonged spiking, the readily releasable pool is depleted and previously undocked vesicles from the so-called reserve pool begin to fuse. After vesicles fuse, their membrane and proteins are retrieved to construct new vesicles, forming the recycling pool. Yet some vesicles in the synaptic terminal apparently never fuse with the plasma membrane; these form the resting pool. Which, if any, of these pools includes vesicles that fuse spontaneously in the absence of action potentials continues to be debated, but Egashira et al. provide evidence that these vesicles are members of the readily releasable pool.

To visualize synaptic vesicle release and recycling at intact zebrafish neuromuscular junctions, the authors targeted two proteins to the vesicle lumen. One protein, pHluorin, fluoresces when exposed to the extracellular environment and is quenched when it is recycled into an acidified vesicle. The other protein, HaloTag, covalently binds a fluorescent ligand (applied extracellularly) upon vesicle fusion and retains this fluorescence when recycled into new vesicles. These indicators suggested that vesicles released spontaneously when action potentials were blocked were drawn predominantly from a restricted population containing ∼8% of all vesicles. To determine whether this population overlapped with the readily releasable or reserve pools, the authors compared HaloTag labeling after strong depolarization to that after the same strong depolarization plus an hour-long period of spontaneous release. There was no difference in the amount of labeling in the two conditions. suggesting vesicles came from the same pool. Next, the authors compared HaloTag labeling induced by hypertonic stimulation, which empties the readily releasable pool, to labeling induced by hypertonic stimulation plus an hour of spontaneous release. Again, no difference in labeling was seen. Finally, they labeled vesicles during spontaneous release and recycling and found these vesicles were depleted rapidly after subsequent depolarization-induced spiking.

Together, these results suggest that spontaneously released vesicles are drawn from the readily releasable pool of vesicles. Future experiments will need to reconcile this finding with previous results, replicated here, that different vesicular SNARE proteins mediate evoked and spontaneous release.

After a dark-adapted retina is exposed to light, arrestin-1 (green) translocates from rod inner segments to the outer segments to help terminate the light response by binding to rhodopsin (red). See Hsieh et al. for details.

Loss of Arrestin-1 Slows Rhodopsin Dephosphorylation

Chia-Ling Hsieh, Yun Yao, Vsevolod V. Gurevich, and Jeannie Chen

(see pages 3537–3545)

Rod phototransduction begins when absorption of a photon by 11-cis retinal causes a conformational change in rhodopsin, enabling rhodopsin to activate its coupled G-protein, transducin. Transducin activates phosphodiesterase, which hydrolyzes cGMP, leading to closure of cGMP-gated channels and, consequently, to photoreceptor hyperpolarization. To terminate the light response, rhodopsin is deactivated by phosphorylation at several sites and subsequent binding of arrestin-1. Rhodopsin must then be returned to the basal state—dephosphorylated, dissociated from arrestin-1, and incorporating 11-cis retinal—before it can contribute to further phototransduction.

Studies of bovine rod outer segments in vitro suggested that arrestin-1 binding to phosphorylated rhodopsin prevents dephosphorylation by blocking phosphatase access. Surprisingly, however, Hsieh et al. provide evidence that this is not the case in vivo. Dark-adapted arrestin-1-knock-out and wild-type mice were exposed to bright light then returned to darkness for different times before retinae were isolated. Immediately after light exposure, rhodopsin was similarly phosphorylated at six sites in wild-type and arrestin-1-deficient retinae. Whereas most rhodopsin molecules were dephosphorylated in wild-type retinae after 1 h of darkness, however, most rhodopsin in arrestin-1-deficient retinae remained at least partially phosphorylated for 2 h, and some molecules remained phosphorylated even after 5 h of darkness. This delay in dephosphorylation also occurred in mice lacking both arrestin-1 and transducin, indicating that it did not result from persistent activation of the phototransduction cascade. Furthermore, arrestin-1 knockout did not affect expression levels of protein phosphatase 2 or the reincorporation of 11-cis retinal into rhodopsin molecules.

These results suggest that arrestin-1 promotes, rather than prevents, rhodopsin dephosphorylation in vivo. How it does so, given evidence that it binds to phosphates on rhodopsin and blocks access by phosphatases, is unclear. Noting that β-arrestin is present in a complex with phosphatase 2A and D₂ dopamine receptors in neurons, the authors propose that arrestin-1 has a role in recruiting phosphatase to rhodopsin in rods.