Presynaptic terminals favor intermediate-conductance Ca(V)2.2 (N type) over high-conductance Ca(V)1 (L type) channels for single-channel, Ca(2+) nanodomain-triggered synaptic vesicle fusion. However, the standard Ca(V)1>Ca(V)2>Ca(V)3 conductance hierarchy is based on recordings using nonphysiological divalent ion concentrations. We found that, with physiological Ca(2+) gradients, the hierarchy was Ca(V)2.2>Ca(V)1>Ca(V)3. Mathematical modeling predicts that the Ca(V)2.2 Ca(2+) nanodomain, which is ∼25% more extensive than that generated by Ca(V)1, can activate a calcium-fusion sensor located on the proximal face of the synaptic vesicle.

Somatic sensory neuron somata are located within the dorsal root ganglia (DRG) and are mostly ensheathed by individual satellite glial cell sheets. It has been noted, however, that a subpopulation of these DRG somata are intimately associated, separated only by a single thin satellite glial cell membrane septum. We set out to test whether such neuron-glial cell-neuron trimers (NGlNs) are also linked functionally. The presence of NGlNs in chick DRGs was confirmed by electron microscopy. Selective satellite glial cell immunostains were identified and were used to image the inter-neuron septa in DRG frozen sections. We used a gentle, dispase-based enzymatic method to isolate chick and rat NGlNs in vitro for double patch clamp recordings. In the majority of pairs tested, an action potential-like stimulus train delivered to one soma resulted in a delayed, noisy and long-duration response in its idle partner. The response to a second stimulus train given minutes later was markedly facilitated. Both bidirectional and unidirectional transmission was observed between the paired neurons. Transmission was chemical and block by the general purinergic blocker suramin implicated ATP as a neurotransmitter. We conclude that the two neuronal somata in the NGlN can communicate by chemical transmission, which may involve a transglial, bi-synaptic pathway. This novel soma-to-soma transmission reflects a novel form of processing that may play a role in sensory disorders in the DRG and interneuron communication in the central nervous system.

Evidence that synaptic vesicles (SVs) can be gated by a single voltage sensitive calcium channel (CaV2.2) predict a molecular linking mechanism or “tether” (Stanley, 1993). Recent studies have proposed that the SV binds to the distal C-terminal on the CaV2.2 calcium channel (Kaeser et al., 2011; Wong et al., 2013) while genetic analysis proposed a double tether mechanism via RIM: directly to the C terminus PDZ ligand domain or indirectly via a more proximal proline rich site (Kaeser et al., 2011). Using a novel in vitro SV pull down binding assay, we reported that SVs bind to a fusion protein comprising the C-terminal distal third (C3, aa 2137–2357; Wong et al., 2013). Here we limit the binding site further to the last 58 aa, beyond the proline rich site, by the absence of SV capture by a truncated C3 fusion protein (aa 2137–2299). To test PDZ-dependent binding we generated two C terminus-mutant C3 fusion proteins and a mimetic blocking peptide (H-WC, aa 2349–2357) and validated these by elimination of MINT-1 or RIM binding. Persistence of SV capture with all three fusion proteins or with the full length C3 protein but in the presence of blocking peptide, demonstrated that SVs can bind to the distal C-terminal via a PDZ-independent mechanism. These results were supported in situ by normal SV turnover in H-WC-loaded synaptosomes, as assayed by a novel peptide cryoloading method. Thus, SVs tether to the CaV2.2 C-terminal within a 49 aa region immediately prior to the terminus PDZ ligand domain. Long tethers that could reflect extended C termini were imaged by electron microscopy of synaptosome ghosts. To fully account for SV tethering we propose a model where SVs are initially captured, or “grabbed,” from the cytoplasm by a binding site on the distal region of the channel C-terminal and are then retracted to be “locked” close to the channel by a second attachment mechanism in preparation for single channel domain gating.

The fusion of synaptic vesicles (SVs) at the presynaptic transmitter release face is gated by Ca2+ influx from nearby voltage-gated calcium channels (CaVs). Functional studies favor a direct molecular “tethering” attachment and recent studies have proposed a direct link to the channel C-terminal. To test for direct CaV–SV attachment we developed an in vitro assay, termed SV pull-down (SV-PD), to test for capture of purified, intact SVs. Antibody-immobilized presynaptic or expressed CaV2.2 channels but not plain beads, IgG or pre-blocked antibody successfully captured SVs, as assessed byWestern blot for a variety of protein markers. SV-PD was also observed with terminal fusion proteins of the distal half of the C-terminal, supporting involvement of this CaV region in tethering. Thus our results support a model in which the SV tethers directly to the CaV. Since the tip of the C-terminal could extend as far as 200 nm into the cytoplasm, we hypothesize that this link may serve as the initial SV capture mechanism by the release site. Further studies will be necessary to evaluate the molecular basis of C-terminal tethering and whether the SV binds to the channel by additional, shorter-range attachments.

Biochemical and physiological evidence suggest that pre-synaptic calcium channels are attached to the transmitter release site within the active zone by a molecular tether. A recent study has proposed that 'Rab3a Interacting Molecule' (RIM) serves as the tether for CaV2.1 channels in mouse brain, based in part on biochemical co-immunoprecipitation (co-IP) using a monoclonal antibody, mRIM. We previously argued against this idea for CaV2.2 calcium channel at chick synapses based on experiments using a different anti-RIM antibody, pRIM1,2: while staining for the two proteins co-localized and co-varied at the transmitter release face, consistent with an association, they failed to co-IP from a synaptosome membrane lysate. RIM is, however, a family of proteins and we tested the possibility that the mRIM antibody used in the more recent study identifies a particular channel-tethering variant. We find that co-immunostaining with mRIM and anti-CaV2.2 antibody neither co-localized nor co-varied at the transmitter release face and the two proteins did not co-IP, arguing against a common protein complex and a key CaV2.2 scaffolding role for RIM at the active zone. The differing results might be reconciled, however, in a model where a RIM family member contributes to a protein bridge that anchors the pre-fusion secretory vesicle to the calcium channel protein complex. Keywords: active zone, calcium channel, particle web, presynaptic, RIM1, transmitter release site. RIM proteins are believed to play a scaffolding role in SV docking or priming (Schoch et al. 2002;Koushika et al. 2001;Wang et al. 2000). In vitro binding studies suggest that the C2 domain of RIM fusion protein can bind directly to intracellular domains of the channel as well as other synaptic proteins (Coppola et al. 2001), and it was speculated that such an attachment could contribute to channel scaffolding. We tested this idea using a commercially available anti-RIM antibody, RIM2 [Note: since this antibody cross-reacts with other RIM variants, as shown below, for clarity we now refer to it here as p(oly)RIM1,2]. We found that immunostaining for RIM co-localized and co-varied with staining for CaV2.2 at the pre-synaptic transmitter release face at the chick ciliary ganglion calyx synapse, consistent with at least the idea that the proteins are in the same region, however, a biochemical analysis was negative. Using the pRIM1,2 antibody, we failed to detect RIM and CaV2.2 co-immunoprecipitation (co-IP) from purified synaptosome membrane lysates which are enriched for active zone proteins (Khanna et al. 2006a). A recent study has suggested an alternative mechanism of pre-synaptic calcium channel tethering by RIM. In this model, RIM scaffolds the a1 pore-forming CaV2 family subunit via the intracellular b subunit (Kiyonaka et al. 2007). One observation in favor of such a role in the Kiyonaka paper was the finding that immunoprecipitation (IP) of RIM from mouse brain lysate using a monoclonal anti-RIM antibody, mRIM, co-immunoprecipitates (co-IPs) CaV2.1 ...