The protein was concentrated to 12 mg/ml using an Amicon Ultracentrifugal filter device (Millipore) with a 10 kDa MWCO. The final concentration of DNQX (Tocris Bioscience) added to the protein solution was 3 mM. Crystals were grown at 4°C in hanging drops containing a 1:1 (v/v) ratio of protein solution to reservoir solution. The crystallization buffer contained 19% PEG 1450, 300 mM lithium sulfate, and 100 mM sodium cacodylate at pH 6.5. The crystals were soaked in crystallization buffer supplemented with 10% glycerol

and 3 mM DNQX prior to flash freezing in liquid nitrogen. X-ray diffraction data were collected at the Advanced Photon Source (Argonne National Laboratory) www.selleckchem.com/products/Y-27632.html at beamline 23-ID-D using a MARMOSAIC 300 CCD detector. A wavelength of Akt inhibitor 1.0332 Å was used. The temperature of the crystal was ∼100 K. The data were indexed, integrated, and scaled

using Mosflm and Scala from the CCP4 suite of programs (Potterton et al., 2003). The structure was solved by molecular replacement using Phaser (McCoy et al., 2007). The LBD-L483Y-DNQX dimer structure (PDB ID 1LB9) (Sun et al., 2002) was used as the search probe. Model building and crystallographic refinement were performed using Coot (Emsley et al., 2010) and PHENIX (Adams et al., 2010) until the R/Rfree values converged. The amount of domain closure in each LBD is determined according to two center-of mass distances ξ1 (between residues 479 and 481 in lobe 1 and residues 654 and 655 in lobe 2) and

ξ2 (between residues 401 and 403 in lobe 1 and residues 686 and 687 in lobe 2) (Lau and Roux, 2007 and Lau and Roux, 2011). A one-dimensional projection of these distances is ξ12 = (ξ1 + ξ2)/2. Rigid body rotation of the LBD dimers was performed using CHARMM (Brooks et al., first 2009). The model of how the LBD tetramer 2 conformation (see Table S1) might influence the ion channel gate was also generated using CHARMM. NMA was performed using the ANM server (Eyal et al., 2006). For modeling of the zinc-binding sites, mutant histidine residues were built using SCWRL4, and the zinc ion was roughly centered between the side chains. The loop segments containing the histidines and the zinc were subjected to energy minimization with CHARMM, with the histidines and zinc restrained to be in close proximity. The rest of the model was held fixed. The flip splice variant of GluA2 was used for electrophysiology studies. Mutants were generated using overlap PCR. The presence of the mutant codons was corroborated by double-stranded DNA sequencing of the amplified region. WT and mutant AMPA receptors were expressed by transient expression in HEK293 cells for outside-out patch recording. The external solution in all the experiments contained 150 mM NaCl, 0.1 mM MgCl2, 0.1 mM CaCl2, 5 mM HEPES, titrated to pH 7.3 with NaOH, to which we added different drugs either in oxidizing or reducing conditions.