| Pyramidal neurons have extensively branched dendritic trees with non-uniform morphological and membrane properties. They receive thousands of inputs throughout their dendritic arbors, which are integrated into a decision of whether or not to fire an action potential. One major avenue to examining pyramidal neuron function is through multi-compartmental simulations of reconstructed, biophysically realistic neuronal models. Since neuronal voltage signals are created by ion channels in the membrane, simulations of pyramidal neurons require models and distributions of ion channels (both voltage-gated and ligand-gated) that agree with experimental observations. This thesis focuses on two specific topics: the generation of sodium channel models exhibiting complex kinetics over various timescales, and the determination of synapse distribution in the basal dendrites of pyramidal neurons.;To generate a voltage-gated sodium channel model that reproduces experimentally observed behavior on vastly different timescales, we develop an algorithm to generate state-dependent ion channel models by optimizing the topology and rate constants, given experimental data. Our algorithm's flexibility in searching over both the space of topologies and parameters generates a novel sodium channel model exhibiting both fast and prolonged inactivation. We then incorporate our channel model into a cell-wide model to show that differential distribution of this type of sodium channel in pyramidal cell dendrites is sufficient to explain activity-dependent attenuation of backpropagating action potentials, a phenomenon that has been linked to synaptic plasticity and associative learning.;Regarding ligand-gated channels, we used a combined experimental and computational approach to determine the distribution and strength of synapses in the basal dendrites of pyramidal neurons. We show that the strength and density of synapses on basal dendrites of pyramidal neurons are regulated in a region-specific, bidirectional manner. This distribution optimizes the contribution of synapses to both somatic voltage and dendritic spikes, suggesting that these cells employ a coherent and efficient strategy to normalize their inputs regardless of location. |
