, 2007, Milstein et al., 2007, Soto et al., 3-MA mouse 2009 and Tomita et al., 2003), the cornichon homologs (CNIH-2, CNIH-3; Schwenk et al., 2009), and the CKAMP44 protein ( von Engelhardt et al., 2010). Alone or in combination, these auxiliary subunits

control the gating and pharmacology of the AMPARs and profoundly impact their biogenesis and protein processing ( Bats et al., 2007, Chen et al., 2000, Gill et al., 2011, Harmel et al., 2012, Kato et al., 2010, Schober et al., 2011, Schwenk et al., 2009, Soto et al., 2007, Tomita et al., 2005, Vandenberghe et al., 2005 and von Engelhardt et al., 2010). It is not clear, however, whether these auxiliary proteins represent the whole set of building blocks for native AMPARs or whether they contain additional yet unknown protein constituents. Likewise, quantitative data on the subunit composition of native AMPAR complexes are not yet available. This information may be obtained from comprehensive

and quantitative proteomic analyses as have recently been presented for the Cav2 family of voltage-gated calcium channels (Müller et al., 2010). Here we used two orthogonal biochemical strategies, multiepitope and target knockout-controlled affinity purifications (Bildl et al., 2012 and Müller et al., 2010) and newly developed high-resolution quantitative analyses of protein complexes separated on native gels (BN-MS), for investigation of the subunit composition of AMPARs PLX3397 cell line from total brain. These analyses unravel native AMPARs as macromolecular complexes of

unanticipated complexity and identify 21 novel protein constituents, mostly transmembrane or secreted proteins of low molecular mass and with distinct functions. Subsequent studies using antibody shift assays, binding studies, and electrophysiological recordings reveal the architecture of native AMPARs and demonstrate that properties and function of the receptor complexes may be quite distinct strongly depending on the particular subunit composition. For Resminostat comprehensive proteomic analysis of native AMPARs, we performed multiepitope affinity purifications (ME-APs) (Müller et al., 2010 and Schwenk et al., 2010) with ten different antibodies (ABs) specific for the GluA1-4 proteins on membrane fractions prepared from total brains of adult rats, wild-type (WT) mice, and AB-target knockout mice (see Experimental Procedures). For ME-APs the membrane fractions were treated with detergent buffers of either mild (CL-47) or intermediate (CL-91) stringency (Müller et al., 2010 and Schwenk et al., 2010) solubilizing ∼40% and 100% of the total pool of AMPARs, respectively (Figures S1A and S1B). These buffers were selected as the two extremes in a test series probing the solubilization efficiency of various CL-buffers as well as of RIPA and Triton X-100, the buffers most widely used with AMPARs (Kim et al.