Presynaptic NMDARs in the Hippocampus Facilitate Transmitter Release at Theta Frequency
McGuinness et al 2010 (Neuron)
This paper comes from the group of Nigel Emptage in Oxford, who has agreed to take me on for a short placement this summer. So I should probably be careful what I say! However, this does really strike me as a well presented paper with a strong story about the presynaptic role of NMDA receptors.
In rat hippocampus organotypic slices, they patched a CA3 pyramidal neuron and filled it with a calcium-sensitive dye, which enabled them to stimulate action potentials in the cell and observe calcium transients in boutons synapsing onto neurons in CA1. Using fancy stats, they identified less frequent large-amplitude transients which were abolished by the NMDAR-blocker AP5. Perfusing the cell with norketamine, an internal NMDAR blocker also abolished the large transients, confirming that this is definitely a presynaptic phenomenon. Some EM pictures confirm that NMDARs are present at both sides of these CA3-CA1 synapses. They also used glutamate uncaging (my pet technique!) to show that presynaptic NMDARs can produce inward currents visible back at the soma. And in a nice coda, they started to explore the physiological relevance by showing that these large calcium transients are more frequent after induction of LTP by theta-burst stimulation.
So the idea is that a large calcium transient corresponds to the release of a glutamate vesicle at the bouton. This is stochastic - it doesn't happen every time. The kinetics seem to be fast enough to allow for the displacement of magnesium to open up the NMDAR channels within the duration of a single action potential.
Caveat: this work was done in organotypic slices after 7 to 14 days in vitro, so may be more relevant to the physiology of the developing than the adult brain.