Rational drug design and the discovery of the Δ2-1,2,3-triazolines, a unique class of anticonvulsant and antiischemic agents

The Delta(2)- 1,2,3- triazoline anticonvulsants (TRs) may be considered as representing a unique class of "built-in" heterocyclic prodrugs where the active ''structure element" is an integral part of the ring system and can be identified only by a knowledge of their chemical reactivity and metabolism. Investigations on the metabolism and pharmacology of a lead triazoline, ADD17014 suggest that the triazolines function as "prodrugs" and exert their anticonvulsant activity by impairing excitatory amino acid (EAA) L-Glutamate (L-Glu) neurotransmission via a unique "dual-action" mechanism. While an active primary P-amino alcohol metabolite from the parent prodrug acts as an N-methyl-D-aspartate (NMDA)/MK -801 receptor antagonist, the parent triazoline impairs the presynaptic release of L-Glu. Various pieces of theoretical reasoning and experimental evidence have led to the elucidation of the dual-action mechanism. Based on the unique chemistry of the triazolines, and their metabolic pathways, biotransformation products of TRs were predicted to be the beta-amino alcohols V and VA, the alpha-amino acid VI, the triazole VII, the aziridine VIII and the ketimine IX. In vivo and in vitro pharmacological studies of the TR and potential metabolites, along with a full quantitative urinary metabolic profiling of TR indicated the primary P-amino alcohol V as the active species. It was the only compound that inhibited the specific binding of [3H]MK-801 to the MK-801 site, 56% at 10 muM drug concentration, but itself had no anticonvulsant activity, suggesting TR acted as a prodrug. Three metabolites were identified; V was the most predominant (45.7 +/- 7.6)% of administered drug, with lesser amounts of VA, (17.3 +/- 5.1)% and very minor amounts of aziridine VIII (4.0 +/- 0.02)%. Since only VIII can yield VA, its formation indicated that the biotransformation of TR occurred, at least in part, through aziridine. No amino acid metabolite was detected, which implied that no in vivo oxidation of V or oxidative biotransformation of TR or aziridine by hydroxylation at the methylene group occurred. While triazoline significantly decreased Ca2+ -dependent, k(+)-evoked L-Glu release (83% at 100 muM drug concentration), some triazolines showed an augmentation of 50-63%, in the Cl- channel activity, a useful membrane action that reduces the excessive L-Glu release that occurs during epileptic seizures. The high anticonvulsant activity of TRs in a variety of seizure models including their effectiveness in the kindling model of complex partial seizures may be due to their unique dual-action mechanism whereby the TR and V together effectively impair both pre- and postsynaptic aspects of EAA neurotransmission; thus the TRs have clinical potential in the treatment of complex partial epilepsy which is refractory to currently available drugs. Since there is strong evidence that L-Glu plays an important role in human epilepsy as well as in brain ischemia/stroke, and since the TRs act by inhibiting EAA neurotransmission, it was logical to expect that the anticonvulsant TRs may evince beneficial therapeutic potential in cerebral ischemia resulting from stroke as well. And indeed, several TRs, when tested in the standard gerbil model of global ischemia did evince remarkable ability to prevent neuronal death.