A candidate protein for the regulation of AMPAR transport
A candidate protein for the regulation of AMPAR transport is CaMKII, a Ca2+-dependent kinase with diverse cellular functions. With respect to synaptic transmission, CaMKII is recognized as a key synaptic protein that is required to modify the number of synaptic AMPARs in response to LTP and LTD stimulation protocols (Hell, 2014, Nicoll and Roche, 2013, Shonesy et al., 2014). In C. elegans, the unc-43 gene encodes the sole CaMKII homolog (Reiner et al., 1999, Rongo and Kaplan, 1999). In unc-43 loss-of-function (lf) mutants, GLR-1 accumulates in neuronal cell bodies suggesting that UNC-43/CaMKII might have a role in GLR-1 transport (Rongo and Kaplan, 1999). We now demonstrate that voltage-gated Ca2+ Remoxipride hydrochloride synthesis (VGCCs) and CaMKII regulate activity-dependent transport of synaptic AMPARs. Additionally, using an optogenetics strategy, we demonstrate the dependence of synapse-specific plasticity on CaMKII. Remarkably, we find that AMPAR transport, AMPAR-mediated currents, and the regulation of synaptic strength are all disrupted in unc-43/CaMKII mutants. Our findings indicate that the number of functional postsynaptic AMPARs is tightly regulated by CaMKII and suggest that synaptic plasticity is controlled in part by regulated motor-driven transport.
Discussion Our experiments have revealed that UNC-43/CaMKII has unexpected and fundamentally important roles in the maintenance of synaptic strength. In unc-43 mutants, transport of AMPARs out of cell bodies is dramatically reduced. This defect was also induced by selective spatiotemporal inactivation of UNC-43/CaMKII in the cell bodies of adult worms and was rescued by heat-shock expression of UNC-43/CaMKII at the adult stage. AMPAR transport was rescued cell autonomously when either UNC-43/CaMKII or mouse alpha-CaMKII was specifically expressed in the AVA neurons. Our FRAP experiments confirmed that synaptic delivery of GLR-1 homomeric and GLR-1/GLR-2 heteromeric AMPARs was greatly reduced in unc-43 mutants, and the photoconversion experiments demonstrated that receptor removal was also dramatically reduced. These defects are similar to those observed in unc-116/KIF5 mutants (Hoerndli et al., 2013) and are consistent with UNC-43/CaMKII having either a direct or indirect role in the loading of AMPAR cargo onto UNC-116/KIF5 kinesin motors, which is critical for the synaptic delivery and removal of AMPARs (Figure 8). We previously reported a significant increase in synaptic accumulation of GLR-1 homomeric receptors in unc-116 mutants. Conversely, GLR-1/GLR-2 heteromeric receptors were diminished in these mutants because they were selectively degraded in the absence of motor-driven transport. We found similar changes in the distribution and abundance of AMPARs in unc-43 mutants. Furthermore, glutamate-gated current was greatly diminished in both unc-116 and unc-43 mutants, which is consistent with the current’s dependence on GLR-1/GLR-2 heteromeric receptors (Hoerndli et al., 2013, Mellem et al., 2002). In the absence of kinesin- and CaMKII-mediated synaptic delivery and removal, GLR-1 homomeric receptors reach synapses by diffusing along neural processes where they subsequently become trapped. In support of this model, we found that driving diffusion of heteromeric receptors by greatly overexpressing the GLR-2 subunit in transgenic unc-43 mutants increased synaptic GLR-2 and restored glutamate-gated current to wild-type levels. Interestingly, overexpressing all known components of the AMPAR signaling complex (GLR-1, GLR-2, and the auxiliary proteins) in unc-43 mutants produced currents much larger than those observed when the same components were overexpressed in wild-type controls. Presumably, the increased current was due to heteromeric receptors diffusing to and accumulating at synapses. By selectively inactivating UNC-43/CaMKII in the cell body of adult worms, we could decrease AMPAR export from the cell body without affecting transport of receptors that had already entered the neural process. Conversely, selective inactivation in the neural process decreased synaptic removal. Together, these data support a model in which UNC-43/CaMKII has a critical role in the loading of AMPAR cargo onto UNC-116/KIF5 motors in the cell body for motor-driven export, and in loading at synapses for their subsequent removal (Figure 8). In vertebrates, there is some evidence that AMPARs bind to scaffolding proteins that serve as adaptors for cargo transport (Setou et al., 2002). However, the identity of these adaptors has not been resolved in vertebrates (Mao et al., 2010) or in C. elegans. Of particular interest is whether adaptors that function at the cell body for the initial loading of AMPARs are different from those at synapses required for the removal of AMPARs.