An integrative model of AMPA receptor trafficking reveals the central contribution of local translation in subtype-specific kinetics

AMPA-type glutamate receptors (AMPARs) underlie most of the excitatory synaptic transmission in the brain and are crucial for implementing long-term synaptic plasticity. AMPARs are multi-protein complexes composed of two types of subunits: pore-forming subunits GluA1-4 that assemble in the endoplasmic reticulum and form the glutamate-gated ion channel, and auxiliary subunits that modulate receptor bio-physical properties and mediate their forward trafficking to the plasma membrane. Here, using a combination of theoretical and experimental approaches, we elucidate the kinetics of essential trafficking steps and the protein sources necessary to explain the experimentally observed distribution of AMPARs and the response of different AMPAR subtypes to LTP induction. Our data indicate that the mRNA coding for one of the most abundant AMPAR auxiliary subunits, CNIH-2, is abundant in dendrites. Consistent with this mRNA distribution, CNIH-2 is locally synthesized. In contrast, the pore-forming subunits GluA1 and GluA2 are mostly synthesized in the cell body. CNIH-2 synthesis increases after the (chemical) induction of long-term potentiation. Strikingly, the translation of CNIH-2 is required for the plasma membrane insertion of GluA2-containing receptors and not GluA1-homomeric AMPARs. Using the selective trafficking of GluA2-containing AMPARs by CNIH-2, our computational model can explain the distinct temporal profiles in response to plasticity induction of two major subtypes of AMPARs, the slow-response of the calciumimpermeable (GluA2-containing) and fast kinetics of the calcium-permeable (GluA2-lacking) AMPARs.