Glutamate unbinding reveals new insights into NMDA receptor activation



Understanding the relationships between the structure and function of neurotransmitter receptors has long been a goal of physiological research. Over the past two decades, developments in molecular genetic techniques have led to rapid advances in this area that are of major significance to our understanding of receptor function and also to drug development in the pharmaceutical industry. A new paper by Wyllie et al. (2006) in this issue of The Journal of Physiology investigating the glutamate binding site of the NMDA subtype of glutamate receptors registers another significant insight in this area.

In contrast to the very fast synaptic transmission mediated by AMPA-type receptors, NMDA receptors mediate a slow synaptic current that peaks in about 10 ms and decays over a time course of 100 ms or more (reviewed by Cull-Candy & Leszkiewicz, 2004; Popescu & Auerbach, 2004; Erreger et al. 2004). Compared to other ion channel receptors, the NMDA receptor is definitely built for stamina rather than speed! This slow activation time course is determined by the kinetic properties of the receptor and is crucial for the role of NMDA receptors as coincidence detectors of pre- and postsynaptic activity. Wyllie et al. (2006) have investigated the effect of mutation of a conserved serine residue (serine 670 to glycine) in the glutamate binding (S2) domain of the NR2A subunit. Even small neurotransmitters like glutamate interact with their binding site at multiple points of contact (Mayer, 2006). This makes understanding transmitter–receptor interaction complex, although in one sense is very convenient: changing a single point of contact doesn't necessarily mean complete loss of the interaction, but allows exploration of the effect of a change in binding on receptor function.

Wyllie et al. use a combination of macroscopic recordings (responses of a large population of receptors) and single channel recordings (studying activation of single receptor molecules) to show that the effect of this mutation can be explained solely by assuming a 124-fold increase in the dissociation rate of glutamate from the binding site. The authors demonstrate this by comparing their data with the predictions of a recently published NMDA receptor mechanism that fits wild-type NR2A receptor single channel data (Schorge et al. 2005). Changing only a single rate constant of that model – the glutamate unbinding rate – is sufficient to describe nearly all the data from the S670G mutant receptor. Receptor properties such as the single channel conductance and duration of channel open times and closed times within each receptor activation were unaffected by this mutation. However the data also show some surprisingly long channel closed times, not predicted by the model, which are unexpected given the agonist concentrations employed and may represent periods of receptor ‘desensitization’ not included in this model. The duration of the receptor activation and the time course of the response to a brief 1.0 ms pulse of glutamate (hence mimicking synaptic transmission) are reduced, as expected if the receptor activation is terminated by dissociation of glutamate from the receptor. Thus the data are consistent with the hypothesis that the S670G mutation reduces the average time that glutamate spends docked in the binding site.

Although this result has a satisfying simplicity – changing the binding site affects binding, but not other aspects of receptor function – it is surprising that it is so. In principle, changing even a single amino acid could have knock-on effects throughout the protein that may result in alteration of the rates of protein conformational changes at sites far removed from the position of the original amino acid (Colquhoun, 1998). The fact that it does not produce marked changes in this case is rather interesting because it suggests that some domains of the protein function at least to some extent independently of remote domains. Yet there obviously is communication between protein domains because binding of glutamate in the extracellular domain results in opening of the ion channel formed by the transmembrane domains of the receptor.

Until recently, further progress would have been limited by lack of knowledge of the three-dimensional structure of the receptor. However, this gap in our knowledge is partly being filled by structural information from protein crystals of the extracellular domains of the glutamate receptor subunits (Mayer, 2006). In the future molecular dynamic simulations, which currently can be used to predict only the initial structural rearrangements of receptors that occur on agonist binding (Kaye et al. 2006), will begin to predict those that lead to ion channel opening. When this happens, the physical reasons for the prolonged nature of the NMDA receptor activation may begin to become clear.