Regulating the number of synapse-localized glutamate receptors, including those of the NMDAR variety, is one major strategy for controlling signaling from that synapse. In the short term, decreased synaptic activity promotes the loss of NMDARs and vice versa. But over the long haul, the number of NMDARs is up-regulated at quiet synapses.
Ehlers' group now demonstrates that this enduring modification is a result of selective mRNA splicing to produce a fast-moving NMDAR variety. Preventing neuronal firing (with sodium channel blockers) increased synaptic NMDARs by increasing the proportion of newly synthesized NMDAR mRNA (and protein) with a C2′—rather than C2—tail. Inclusion of the C2′ tail increased the number of these receptors at synapses because they quickly exited the ER, the bottleneck for most secreted proteins, thanks to a novel COPII-recruiting signal in C2′. In contrast, continuous neuronal firing favored production of the slow-moving C2 variant and thus decreased the number of synaptic NMDARs.
NMDAR activity is needed for this splicing adjustment. As NMDARs are calcium permeable, Ehlers postulates that calcium-regulated splicing factors might process certain neuronal mRNAs. The feedback regulation probably prevents a synapse from reaching an absolute maximum or minimum activity, so that later changes in activity can be detected. “It's a countering activity,” says Ehlers. “If a synapse keeps getting potentiated, it gets to a maximum rate, and further strengthening would do no good.” ▪