Characterization of AMPA receptors (AMPARs) purified from wild-type mice and from a mouse strain carrying an epitope-tagged version of the GluR2 subunit shows that AMPARs interact with only a subset of proteins previously thought to associate with the receptors, report Fukata et al. (page 399).

Changes in the number of AMPARs alter how well synapses conduct currents in the brain and are thought to control memory storage. A large number of proteins have been found to associate with AMPARs by in vitro assays, such as two-hybrid screens or coexpression in cell lines, but evidence for functional interaction is limited.

When Fukata et al. purified AMPAR components from brain extracts of transgenic mice expressing the epitope-tagged GluR2, only BiP, a common ER chaperone, and transmembrane AMPAR regulatory proteins (TARPs) came down in a quantitative manner. Because BiP and TARPs were in distinct AMPAR complexes, the researchers hypothesized that BiP–GluR2 may be an immature transitory complex.

Why didn't the other known AMPAR proteins come down? The authors think that only a few of the previously identified proteins associate with the majority of AMPARs in brain cells. The other proteins might interact with the receptors in a small subset of cells or under specific conditions. And the new work is consistent with emerging genetic data showing that, when these other proteins are removed, there are limited effects as compared with the phenotype of the stargazer mouse, a TARP mutant that lacks functional AMPARs in a common neural cell type in the brain.