Shining Light on the Mode and Mechanism of Vesicular Release at Rod Photoreceptor Synapse

Main Text

A key feature of the early stages of sensory processing in the visual and auditory system is the presence of ribbon-type synapses that allow graded transmission of electrical signals (1). Because of the specialized organization of synaptic vesicles around a ribbon, these synapses can support a continuous and high-throughput mode of neurotransmitter release (1). This is particularly well suited for the photoreceptors, rods, and cones, which continuously release glutamate in darkness and need to encode small and graded changes in light intensity over several log units (2).

The classical view of synaptic release is that an action potential causes the fusion of a single synaptic vesicle at a release site (3). This view of univesicular release has been revised by evidence in most central and sensory synapses for the concomitant release of multiple synaptic vesicles from a terminal (4). Multivesicular release (MVR) is a hallmark of ribbon synapses in hair cells, photoreceptors, and retinal bipolar cells (2). Although such a coordinated mode of synaptic release has been well characterized in the retinal ribbon synapses, in particular the rod bipolar cell ribbon synapse (5), several questions remain unresolved: 1) do ribbons play a direct role in organizing vesicles close to the active zone and thus mediating MVR, 2) does MVR arise from concomitant fusion of vesicles from single or separate release sites, and 3) what might be the functional implications of MVR at ribbon synapses for visual signaling?

In this issue of the Biophysical Journal, Hays and colleagues have addressed two of the three above questions in salamander rod photoreceptors using a combination of careful pre- and postsynaptic measurements of synaptic release (6). The authors assay neurotransmitter release from rod photoreceptors by recording a glutamate transporter-activated anion current. Glutamate transporters, also called excitatory amino acid transporters (EAATs), are a family of high-affinity electrogenic antiporters that are localized to glial cells and pre- and postsynaptic neurons. They mediate rapid reuptake of extracellular glutamate from the synaptic cleft to limit postsynaptic receptor-mediated glutamatergic signaling (7). EAATs have been previously shown to gate an uncoupled chloride (Cl⁻) conductance, which is prominent for some isoforms, EAAT4 and EAAT5 (8). In the vertebrate retina, the cellular expression and biophysical properties of EAATs have been well characterized in multiple cell types including the rod photoreceptors (7). In fact, several groups, including the authors, have used the EAAT gated anion current (I_A(glu)) as a reliable presynaptic readout for exocytosis in rod and cone photoreceptors (9,10). By whole-cell voltage-clamp recordings from the rod photoreceptor soma, in a slice preparation, Hays and colleagues monitored spontaneous miniature I_A(glu)—the EAAT current resulting from the exocytosis of glutamate-filled vesicles (6). They observed large events, which presumably reflect simultaneous release from multiple vesicles. The amplitude distribution of the spontaneous events followed the classical multipeak quantal distribution of synaptic release, in which the higher-order peaks are multiples of a fixed quantal size (6). Based on their analysis of this distribution, the authors estimated that ∼30% of the spontaneous events are multiquantal (6). Interestingly, the kinetics of uniquantal and multiquantal events were identical, indicating synchronous release (6).

Synchronous release of multiple vesicles was further verified by paired voltage-clamp recordings simultaneously from the rod photoreceptors and the postsynaptic neuron (horizontal cells), measuring spontaneous I_A(glu) in the rods and the miniature excitatory postsynaptic currents (mEPSCs) in the horizontal cells (6). Presynaptic events were found to occur nearly simultaneously with postsynaptic mEPSCs (6). Such correlated events are not abundant because 1) an individual horizontal cell pools input from several photoreceptors, and 2) only a few of the ribbons per rod photoreceptor makes synaptic contacts with a horizontal cell. Similar to the presynaptic events, scaled small and large mEPSCs displayed identical rise times indicative of simultaneous vesicle fusion (6).

The authors further probed the spatial scale of MVR from the rod ribbon relative to the calcium channels located beneath the ribbon. It has been shown in hair cells that Ca²⁺ can spread beyond a single release site and cause coordinated fusion of vesicles at distinct release sites (11). By introducing slow or fast calcium buffers, EGTA or BAPTA, respectively, in the intracellular pipette solution, the authors were able to separate Ca²⁺-triggered events closer or farther from the Ca²⁺ channel (6). Based on this approach, the authors discovered that multiquantal events in salamander rods tend to occur in close proximity to the Ca²⁺ channel, whereas a fraction of uniquantal events occurred distant from the Ca²⁺ channel (6). The authors discuss that opening of a single Ca²⁺ channel might be insufficient to trigger high rates of MVR as previously estimated in the rod photoreceptors (12). This instead could potentially arise because of stronger Ca²⁺ dependence of the synaptic protein isoforms in play at the rod terminal (e.g., the Ca²⁺ sensor synaptotagmin).

Does synchronous MVR at the rod terminal rely on ribbons, and does it involve fusion of neighboring vesicles as observed in other conventional and ribbon synapses? The authors addressed both of these questions using a couple of acute perturbations. First, the authors inactivated the ribbon using a technique called fluorophore-assisted light inactivation (FALI) (6). Second, they introduced a small inhibitory peptide of a SNARE protein, syntaxin 3B, in the patch pipette to block SNARE complex formation between nearby vesicles on the ribbon (6). Interestingly, both these perturbations selectively attenuated MVR while sparing the smaller uniquantal events (6). These results suggest a model of homotypic fusion of ribbon-associated vesicles that give rise to synchronous MVR at salamander rod terminals. The presence of a few copies of syntaxin 3B on the vesicle membrane might be sufficient for homotypic fusion given the low number of SNARE complexes required for synaptic vesicle fusion (13).

What functional purpose does MVR serve for visual signaling at the salamander rod photoreceptor ribbon synapse? Although all the recordings were performed in light-adapted retina, the authors point out that the relative frequency of multivesicular events is identical under scotopic (dim light) light conditions (6). Rod photoreceptors reliably detect single photons (14). Seminal studies have shown that rod photoreceptors generate small electrical signals of 1–2 mV upon activation of a single rhodopsin molecule (15). Because rod photoreceptors are depolarized in darkness and undergo continued glutamate release, absorption of a photon results in hyperpolarization, which causes a reduction or pause in transmitter release at the rod ribbon synapse (16). Effective transmission of the single photon signal across the rod terminal is limited by the synaptic noise caused by statistical variability in neurotransmitter release. In addition, spontaneous thermal activation of rhodopsin can also lead to spurious single photon-like events and cause reduction/pauses in transmitter release (17). One way to minimize the contribution of synaptic and thermal noise is to have a high rate of spontaneous neurotransmitter release. Capacitance measurements from salamander rods in fact show a high tonic release rate of ∼400 vesicles/s (18). The potential functional role for MVR could be to 1) ensure high release rates at rod terminals and 2) create a presynaptic nonlinearity for contrast sensitivity such that only larger exocytotic events evoked by light decrement are preferentially transmitted. Future measurements of light-evoked changes in transmitter release from rod terminals under scotopic conditions will to further elucidate the functional role of MVR.

Editor: Meyer Jackson.