Trapping was again barely detectable when the receptors were saturated with 10 mM glutamate and 100 μM CTZ (Figure 5A), indicating that the crosslink also does not trap the receptor in a fully activated state. In contrast to the observations above, we detected substantial trapping at concentrations of glutamate over

the range 100 μM to 2 mM. The exponential decay of the current following the jump into oxidizing conditions at 500 μM glutamate (τ = 500 ± 100 ms; n = 5) was indistinguishable from the extent of trapping plotted against different time intervals (τ = 400 ± 100 ms; Figure 5B), suggesting that this relaxation reflects the inhibition Enzalutamide clinical trial due to formation of the disulfide bond. WT receptors showed no inhibition in oxidizing conditions when activated by 500 μM glutamate (Figure 5C). The relationship between the active (untrapped) fraction and log concentration followed an inverted bell shape and thus was well described by an inverted log normal function with a minimum at 248 μM (Figure 5D). Kinetic simulations of channel activation demonstrated that this relationship mandates trapping in partially bound states (Figure S5). Crosslinking also lowered apparent glutamate potency, consistent with a reduced occupancy GSK1210151A solubility dmso of the receptor by glutamate (Figure 5D). Consistent with strong trapping being associated with incomplete activation, saturating

the receptor with the partial Lepirudin agonist kainate also promoted trapping (data not shown; see Discussion). We obtained similar results by oxidizing with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) in the presence of heavy metal chelators (Figure S6), demonstrating that the absence of trapping at high glutamate concentrations is not due to copper chelation and consequent loss of oxidizing activity. The inhibitory effect of DTNB was less strong than that of CuPhen, leading to a smaller rightward shift in the glutamate concentration response curve. The weak action

of DTNB is unsurprising given that it is approximately twice as bulky as phenanthroline and that the interdimer region is closely packed. Poor access of DTNB to the cysteines at position 665 probably results in a mixed population of receptors with reduced and intact crosslinks (Gielen et al., 2008). Zinc bridging between subunits inhibited the HHH mutant with an apparent affinity of 95 ± 30 nM (Figure 6A). We exploited multibarrel fast perfusion to assess the state dependence of this zinc bridging in the HHH mutant. Consistent with the structure of the full-length receptor, we could not detect inhibition due to zinc bridging at rest, and zinc alone did not activate a current in the presence of CTZ (n = 6 patches). No inhibition was seen following desensitizing exposures (100 μM glutamate without CTZ), at rest or after full activation (10 mM glutamate with CTZ).