Structural models have been produced for the agonist binding and transmembr
ane domains of two NMDA ionotropic glutamate receptors: homomeric NMDA-R2C
and heteromeric NMDA-R1/R2C. These models-produced using homology modelling
techniques in conjunction with distance restraints derived from the access
ibility of substituted cysteines-have aided our understanding of (1) ligand
selectivity and (2) channel activity. The model of the agonist binding dom
ain of NMDA-R2C indicates that T691 forms an essential hydrogen bond with g
lutamate ligand. This interaction is absent in the NMDA-R1 model-where a va
line replaces the threonine-explaining why NMDA-R1 binds glycine rather tha
n glutamate. For the transmembrane region, the models suggest that a number
of positive residues, located in the cytoplasmic loop between the M1 and M
2 segments, create a large electrostatic energy barrier that could explain
why homomeric NMDA-R2C channels are non-functional. Introducing NMDA-R1 to
form heteromeric NMDA-R1/R2C channels is predicted to rescue channel activi
ty because the corresponding region in NMDA-R1 contains negative residues t
hat more than compensate for the electrostatic energy barrier in NMDA-R2C.
These studies suggest that replacing the positively charged region in the M
1-M2 loop of NMDA-R2C with the corresponding negatively charged region of N
MDA-R1 could transform NMDA-R2C into a functional homomeric channel.