agonist. Likewise, with appropriate substitutions and replacement, glutamic acid analogs
can also be competitive antagonists. In these antagonists, bioisosteric replacement of
carboxylates with phosphonates is a frequent design strategy. Incorporation of a ketone
functional group into the carbon backbone of 2-amino-5-phosphonovaleric acid (termed
“AP5” in the NMDA literature) affords a potent, orally active NMDA antagonist called
(R)-4-oxo-5-phosphononorvaline (4.213). This latter compound may be conformation-
ally constrained through the creation of a cis-2,3-disubstituted piperidine nucleus to
yield another agent with potent NMDA receptor affinity. Introduction of trans-4-methyl
substituents into this piperidine derivative yields even greater receptor affinity.
A variety of non-competitive NMDA antagonists have also been identified over the
years. Probably the most famous of these molecules is dizocilopine or MK-801 (4.214).
MK-801 has been massively studied as both a potential drug and a molecular probe for
studying and understanding neurodegeneration. Although it never successfully emerged
as a drug, for toxicology considerations, the structure–activity relationships of MK-801
and related iminomethanobenzocycloheptenes have received considerable study.
NMDA antagonists have also emerged from studies targeting the subsites on the
NMDA receptor complex. Work on glycine subsite antagonists is discussed in the
section on glycine (section 4.8). Research on the polyamine site has also been fruitful.
The polyamine antagonist eliprodil (4.215) was shown to protect cultured hippocampal
brain cells from NMDA-mediated toxicity and demonstrated some activity in rat
models of brain trauma and brain ischemia; oxindole variants of eliprodil have also
been prepared and have shown enhanced activity and improved selectivity.
The zinc subsite has proved to be more challenging. Designing drugs to uniquely
interfere with the biochemistry of a metallic anion in the central nervous system is non-
trivial. It can be appreciated, from studies that attempt to design ion-specific chelating
agents, that developing drugs to uniquely target one metal ion (e.g., zinc) over another
(e.g., copper) can be difficult. Nevertheless, tricyclic antidepressants and phenoth-
iazines, including desmethylimipramine and ethopropazine, have been suggested to act
as zinc site ligands. The relationship between the zinc site and the magnesium site has
also been considered. Magnesium itself has been studied as a drug. Models of stroke
and heart attack have shown that magnesium sulphate infusions have neuroprotective
effects equivalent in magnitude to those of noncompetitive NMDA antagonists.
NEUROTRANSMITTERS AND THEIR RECEPTORS 285