We measured concentration-response curves in the A665C mutant at the beginning of the application of oxidizing conditions (peak, when the receptors were still reduced) and in steady-state oxidizing conditions, when we assumed trapping was complete. We obtained the EC50 from fits to the Hill equation: IImax=[Glu]n[Glu]n+EC50n,where n is the Hill coefficient, and [Glu] is glutamate concentration. We measured trapping after the application of CuPhen in different concentrations of glutamate, plotting

the immediate active fraction of the current against different concentrations of glutamate. This was normalized against the current following oxidizing conditions plus 10 mM glutamate. A log normal function was fitted to the data. The GluA2 receptor contains 11 cysteines, 4 of them involved in disulfide bonds (C63 with C315 and C718 with C773C). In order to obtain a construct running monomerically under denaturing conditions, CFTR activator we serially removed free cysteines, eventually constructing the 7 × Cys(−) mutant by introducing the following mutations into the GluA2 WT receptor: C89S, C190A, C436S, C425S, check details C528S, C589S, and C815S (Figure S3A). This channel remained functional and had similar properties to WT (Figure S3B). All cysteine mutants studied by western blotting were made on this background. All mutations were introduced by overlap PCR and confirmed by double-stranded

sequencing. HEK293T cells were plated in 10 cm dishes and transfected with different plasmids (5 μg) using polyethylenimine (PEI; 1 mg/μl). After Astemizole 72 hr, cells were collected in PBS, centrifuged 5 min at 1,000 × g, and pellets were resuspended in a buffer containing 300 mM NaCl, 50 mM Tris (pH 8), 1% DDM (Anatrace), and a protease inhibitor mixture (Roche). For treatments under reducing and oxidizing

conditions, dishes were rinsed with PBS followed by incubation with 100 mM DTT or 100 μM CuPhen in serum-free medium for 30 min before lysis. After sonication, the lysates were rotated (10 rpm) for 1 hr at 4°C and subsequently centrifuged at 20,000 × g to obtain cleared lysates. Protein extracts (50 μg) were then separated by 4%–12% Bis-Tris Glycine SDS/PAGE and transferred to nitrocellulose membranes. Blots were immunostained overnight at 4°C, or for 5 hr at RT, with anti-GluA2 N terminus (1:1,000; Millipore) or anti-β-actin (1:2,000; Cell Signaling) primary antibodies. Following exposure to appropriate peroxidase-conjugated secondary antibodies (Biozol), blots were visualized with chemiluminescence reagent (SuperSignal West Pico; Thermo Scientific). Densitometric analysis was performed using ImageJ ( Schneider et al., 2012). The signal from β-actin was used as a loading control, and the results were normalized as the ratio of dimer band intensity versus the total intensity of dimer and monomer bands.