Pii: S0306-4522(98)00109-2

–Allosteric regulators of á-amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA) receptors include 2,3-benzodiazepines such as GYKI 52466 and GYKI 53655 and the chaotropic anion thiocyanate that inhibit, and benzothiadiazines such as cyclothiazide that potentiate AMPA receptor currents. Here we sought to determine whether the allosteric regulators modulate AMPA receptors at a common or distinct allosteric sites by comparing their actions on AMPAand kainate-evoked currents in cultured rat hippocampal neurons and Xenopus oocytes expressing recombinant AMPA receptor subunits. GYKI 52466 and thiocyanate blocked AMPA-evoked currents in a concentration-dependent manner (IC50 values, 8.2 μM and 1.1 mM, respectively); in contrast, kainate-evoked currents were blocked by GYKI 52466, but were potentiated by high concentrations of thiocyanate (§3 mM). Thiocyanate enhanced the rate of desensitization and slowed recovery from desensitization of AMPA-evoked currents, whereas GYKI 52466 failed to affect desensitization. Among neurons in the hippocampal cultures, there was cell-to-cell variability in the sensitivity to block of AMPA-evoked currents by thiocyanate that was correlated with the degree of potentiation by cyclothiazide. Moreover, cyclothiazide caused a parallel rightward shift in the concentration–response curve for thiocyanate block, and slowed the onset of thiocyanate block to a rate that was similar to that of cyclothiazide dissociation. Together, these observations suggest that thiocyanate and cyclothiazide act at non-distinct allosteric sites. GYKI 52466 blocked AMPA receptor responses to a similar extent, irrespective of the degree of cyclothiazide potentiation. Moreover, the kinetics of GYKI 53655 block in the presence of cyclothiazide were not consistent with a competitive interaction. As is the case for cyclothiazide, SCN exhibited greater affinity for flip than for flop AMPA receptor splice variants. In particular, GluR1flip/GluR2flip was especially sensitive to thiocyanate block. We conclude that thiocyanate, a flip-preferring allosteric modulator like cyclothiazide, appears to act by enhancing desensitization at a site that may overlap the site where cyclothiazide reduces desensitization, whereas 2,3-benzodiazepines act at a distinct site and the block does not involve a modification of desensitization.