Font Size: a A A

A stochastic model of mechanisms underlying plasticity in depressing excitatory cortical synapses

Posted on:2003-06-13Degree:Ph.DType:Dissertation
University:Boston UniversityCandidate:Okatan, MuratFull Text:PDF
GTID:1464390011978185Subject:Biology
Abstract/Summary:PDF Full Text Request
This dissertation explores the mechanisms that mediate short-term depression (STD) and long-term potentiation (LTP) in synapses between cortical pyramidal neurons (CPNs). It is proposed that the major processes that contribute to STD in these synapses are the depletion of the releasable vesicle pool, priming of vesicles for release, and postsynaptic receptor desensitization.; A stochastic model of these processes is proposed. By estimating the optimal model parameters that fit average excitatory postsynaptic potential (EPSP) traces recorded from the same synapse before and after LTP, the model's prediction of what synaptic parameters are changed is obtained. This analysis uses the data of Markram and Tsodyks (Nature, 382, pp. 807, 1996). The model predicts that, in synapses between CPNs, LTP is expressed in terms of a slight increase in the postsynaptic sensitivity to transmitter, and a significant increase in the release probability, possibly through an increased sensitivity of release machinery to calcium influx, or the potentiation of the latter.; The model is shown to exhibit STD, redistribution of synaptic efficacy (RSE), frequency-dependent synaptic potentiation, and 1/f dependence of the average EPSP on the firing rate f near steady-state transmission. It predicts that increasing the concentration of extracellular calcium ions results in more potentiation at firing rates of 5--15 Hz than what would have been obtained after paired pre- and postsynaptic activity (Hebbian pairing). After modification to accommodate the morphological and physiological properties of the neuromuscular junction, the model also exhibits the RSE associated with post-tetanic potentiation of that junction.; Novel methods are used to estimate the model parameters and to obtain 95% confidence intervals for the latter. These methods include the derivation of a deterministic model that approximates the average behavior of the stochastic model, which is an inhomogeneous Markov process.; It is noted that STD and LTP interact to give rise to RSE in cortical synapses but not in synapses made onto hippocampal CA1 pyramidal neurons. As a result of RSE, only the onset of a burst of action potentials experiences potentiated transmission, while the rest of the burst is transiently depressed. The dissertation discusses a possible functional role of this property.
Keywords/Search Tags:Synapses, Model, Cortical, STD, LTP, Potentiation, RSE
PDF Full Text Request
Related items