Comparing glutamate kinetics between a planar model and reconstructed rod spherules. (A) Schematic representation of an invaginating rod synapse that shows two HC dendrites flanking the synaptic ridge (upper left) and an RBP dendrite terminating beneath the ribbon release site. The image at the lower right shows a simplified planar model of this synaptic arrangement that consists of three slabs intersecting at an angle of 120°. Each slab involves two planes separated by 16 nm to simulate a synaptic cleft. We simulated the release of 2,000 glutamate molecules at the vertex of this narrow cleft. We measured the number of glutamate molecules that entered a small measurement region (gray box, 15 × 100 × 200 nm) with the leading edge 70 nm from the release site. (B) Illustration of a reconstructed synapse (rod 1) with a measurement region (gray box, 15 × 100 × 200 nm) placed just above an RBP dendrite (purple). HC dendrites are shown in yellow and green. The ribbon is in dark blue, and glutamate molecules are in sky blue. (C) Monte Carlo simulations of single-vesicle release showed an abrupt rise and a rapid decline of glutamate in measurement regions in the planar model (blue trace; average of 12 seeds) and four reconstructed synapses (purple trace; average of four rods, 12 seeds apiece). The red trace shows the change in glutamate observed when the rod models included active uptake by 3,000 EAAT5 distributed throughout the synaptic cleft. The measurement regions attained a peak concentration of 0.4 mM in the planar model and an average of 0.775 ± 0.3325 mM in the four rod synapses. Decay kinetics was fit with two exponentials. Planar model: τ_fast = 0.081 ms; τ_slow = 0.558 ms. Passive glutamate decay in the four rod synapses: τ_fast = 0.114 ± 0.0217 ms; τ_slow = 0.589 ± 0.1394 ms. Decay in the rod synapses with active uptake by 3,000 EAAT5: τ_fast = 0.116 ± 0.0238 ms; τ_slow = 0.644 ± 0.1943 ms.