The syntheses of several novel N-(hydroxydioxocyclobutenyl)-containing analogues of gamma-amino-butyric acid and L-glutamate were undertaken to test the hypothesis that derivatives of 3,4-dihydroxy-3-cyclobutene-1,2-dione (squaric acid), such as 3-amino-4-hydroxy-3-cyclobutene-1,2-dione, could serve as a replacement for the carboxylate moiety in neurochemically interesting molecules. The syntheses were successfully accomplished by preparation of a suitably protected diamine or diamino acid followed by reaction with diethyl squarate. Subsequent deprotection resulted in the isolation of the corresponding N-(hydroxydioxocyclobutenyl)-containing analogues 13, 14, and 18. These analogues were screened as displacers in various neurochemical binding site assays. The L-glutamate analogue 18, which showed high affinity as a displacer for kainate and AMPA binding, was also examined for agonist potency for CA1 pyramidal neurons of the rat hippocampal slice preparation. It rivaled AMPA as one of the most potent agonists for depolarizing pyramidal neurons in medium containing 2.4 mM Mg+2 ions in which kainate/AMPA receptors are active but NMDA receptors are inhibited (IC50 = 1.1 microM). It was 1 order of magnitude less potent for depolarizing pyramidal neurons under conditions in which kainate/AMPA receptors were inhibited by 10 microM CNQX but NMDA receptors were active in 0.1 mM Mg(+2)-containing medium (IC50 = 10 microM). Compound 18 did not induce sensitization of CA1 pyramidal cells to depolarization by phosphonate analogues of glutamate (the QUIS-effect).