AMPA receptor diffusion trap following LTP induction
There are two ways that AMPA receptors can move in and out of the postsynaptic density found immediately opposite to the presynaptic terminal button. One is by lateral diffusion within the membrane, and the other is by endo / exocytosis from intracellular vescicles.
One of the phenomena that any model of receptor trafficking has to account for is the AMPA-receptor-mediated induction of LTP, which can take place in just 10 seconds. Some protein (it is not known which) is believed to confine the AMPA receptor's diffusion to the post-synaptic density during LTP.
In order to account for LTP, Tolle and Novere model the AMPA receptors as remaining in the postsynaptic density due to their affinity to some scaffold binding molecule. They speculate that these scaffold molecules are already present in the postsynaptic density but must be activated by the Ca2+ influx through NMDA receptors upon glutamate release. This speculation is consistent with the finding that Ca2+ uncaging prevents AMPA receptors from diffusing (see here). Here is are two examples of AMPA receptor diffusion pathways, where maroon is free diffusion and green is confined to an area such as the postsynaptic density:
And here is a schematic diagram of their model:
This activation-dependent "diffusion trap" is a useful way to think about receptor diffusion and a good way of showing how rapidly postsynaptic effects can be realistically induced.
Reference
Brownian diffusion of AMPA receptors is sufficient to explain fast onset of LTP . Dominic P Tolle and Nicolas Le Novère. BMC Systems Biology 2010, 4:25doi:10.1186/1752-0509-4-25. Link here.