Complexity and Computation at the Synapse: Multilayer Architecture and Role of Diffusion in Shaping Synaptic Activity and Computation - Bioinspired Approaches for Human-Centric Technologies - page 277

the spatial structure of the synapse and the diffusion constrains in its surroundings.
The estimations of neurotransmitter concentration and clearance provided by this in
silico approach confirmed the substantial match of those obtained with
“deconvolution” experiments (Holmes 1995 ; Clements 1996 ; Kleinle et al. 1996 ;
Wahl et al. 1996 ; Glavinov ´ c 1999 ; Franks et al. 2002 ; Ventriglia and Maio 2003 ;
Petrini et al. 2011 ). In addition, these studies highlighted a critical dependence of
these values on several factors including the following (1) number of released
molecules, (2) synaptic geometry, (3) number of binding sites (neurotransmitter
receptors and transporters), (4) neurotransmitter reuptake, and (5) neurotransmitter
diffusion coefficient.
The neurotransmitter concentration time course also depends on the modality of
presynaptic release. The aforementioned model simulations assume the “full-col-
lapse fusion,” the main release mode described at central synapses that involve the
complete fusion of the vesicle membrane into the presynaptic plasma membrane.
However, it has been proposed that synaptic vesicles may fuse transiently and
incompletely with the plasma membrane by forming a reversible “fusion pore”
connecting the vesicle lumen with the synaptic cleft, a releasing mechanism
referred to as “kiss-and-run” or “continuous-release” (Heuser and Reese 1973 ;
Ceccarelli and Hurlbut 1980 ; Harata et al. 2006 ). The impact of this latter mecha-
nism of release has been addressed by Kleinle et al. ( 1996 ) by adding to the Fick's
equation a “release function” that describes the neurotransmitter escape from the
vesicle through a fusion pore formed by the synaptic vesicle and the presynaptic
membrane. The authors found that in conditions of simulated “continuous release,”
the neurotransmitter concentration only peaked at 0.37 mM and decayed in ~2 ms,
more than one order of magnitude slower than in the conventional “full collapse”
mode. Overall, the experimental and modeling studies indicate that, following
typical synaptic release, the neurotransmitter peaks at very high concentration
(mM range) and, due to diffusion, lasts in the synaptic cleft for only few hundreds
of microseconds, similar to an “explosion” at the nanometric scale. It has to be born
in mind, however, that the profile of concentration in the synaptic cleft can be
highly diverse in specific synaptic subtypes, neuronal developmental stages, and
physiological conditions (Barberis et al 2005 ; Karayannis et al. 2010 ).

9.4.2 Functional Implications of Fast Neurotransmitter
Diffusion

Synaptic neurotransmitter exposure in the range of ~100 ms represents a time
considerably briefer than that needed for the full activation of most fast ligand-
gated postsynaptic receptors (AMPA, GABA A , and glycine receptors ~300-
400 ms). As a consequence, unitary synaptic currents evoked by the release of a
single vesicle quantum are elicited under conditions of substantial nonequilibrium .
The most important conceptual consequence of “nonequilibrium activation”