Among the non-NMDA (non-N-methyl-D-aspartic acid) glutamate receptors, the AMPA (alpha-amino-2,3-dihydro-5-methyl-3-oxo-4-isoxazolepropanoic acid) selective receptors are characterized by a fast occurring desensitization. We and others have searched for specific modifiers of the rapid desensitization of AMPA responses in hippocampal slices using the patch-clamp technique. Aniracetam (1-(4-methoxybenzoyl)-2-pyrrolidinone) and diazoxide (7-chloro-3-methyl-2H-1,2,4-benzo-thiadiazine 1,1-dioxide) (1 mM) increased glutamate-activated currents recorded from voltage-clamped CA1 pyramidal neurons in presence of 5 microM MK-801 (dizocilpine; 10,11-dihydro-5-methyl-5H-dibenzo[a,d]cyclohepten-5,10-imine) by 2.5 fold. Cyclothiazide (3-bicyclo[2.2.1]hept-5-en-2-yl-6-chloro-3,4-dihydro-2H-1,2,4-benzoth ia diazine-7-sulfonamide 1,1-dioxide) (100 microM), a chemical congener of diazoxide, completely removed the desensitization of the AMPA response measured with fast application in excised outside-out patches. At this concentration cyclothiazide produced an 18 fold enhancement of the glutamate current. Eighteen diazoxide analogues (2H-1,2,4-benzothiadizines: IDRA 2-19) were then tested but none of them was as effective as diazoxide. Three analogues of cyclothiazide (3,4-dihydro-2H-1,2,4-benzothiadiazines: IDRA 20-22) were also tested and none of them were as potent as the parent compound. However, IDRA 21 produced a response 3 times larger than diazoxide. Moreover, while cyclothiazide and diazoxide potentiated kainate responses for all the doses that decreased AMPA receptor desensitization, IDRA 21, similarly to aniracetam, inhibited AMPA receptor desensitization preferentially. These results suggest that similarly to NMDA receptors the structure of AMPA receptors may include a center that regulates desensitization.