Abstract
Abstract: :
Purpose: To estimate the relative spatial location of NMDA receptors postsynaptic to ribbon synapses of bipolar cells and test the hypothesis that extrasynaptic NMDA receptor location contributes to their differential role in miniature versus light–evoked EPSCs. Methods: We modeled transmitter release from a single synaptic vesicle into a 2 µm3 volume of "inner plexiform layer" (IPL) made up of an array of cubes (125 to 729 elements) arranged to yield differing amounts of tortuosity and volume fraction in the 3D simulation space. The program MCell was used to track the position of molecules in 3D space and determine the kinetic states of glutamate receptors and transporters. Glutamate concentration was measured in these simulations directly below the mouth of the pore and at various distances from the release site. Simulations were run for diffusion rates of 2.0 to 0.5 x 10–6 cm2 per sec. AMPA and NMDA receptors were placed at different locations relative to the release site. Glutamate transporters were placed on the surfaces of all "cell processes" in the simulations. 200 stochastic MCell runs were carried out for every set of parameters. The mean number of open NMDA receptors was measured at distances from 0 to 900 nm from the release site in order to find a distance where channels do not open, as predicted by experimental data (Taylor et al, J. Physiol. 1995). Kinetic models for rapidly (NR1/NR2A) and slowly (NR1/NR2B) desensitizing NMDA receptors (Chen and Murphy, Molec. Pharmacol. 2001) were tested in these simulations. Results: For a 25 nm diameter vesicle filled with 2000 molecules of glutamate, the peak concentration at all distances declined as the diffusion rate of glutamate increased. The maximum value at the release site was 1.1 mM. The highest concentration reached 0.36 and 0.08 mM at 150 and 250 nm away from the release site, respectively. Only when NMDA receptors (all kinetic models) were stimulated with a concentration waveform consistent with their being 250 to 500 nm from the release site was there no NMDA response as, physiological data predict. Conclusions: If the lack of an NMDA component to mini–EPSCs in ganglion cells results from a spatial localization of receptors, NMDA receptors must be between 250 to 500 nm equidistant from all ribbon synapses. Changes in vesicle glutamate concentration, tonic levels of glutamate in the extracellular space, or possible modulation of co–agonist D–serine will affect NMDA contributions at these locations.
Keywords: computational modeling • excitatory amino acid receptors • ganglion cells