Arrangement of contact sites from single excitatory fibers on medial superior olive dendrites

Neurons in the medial superior olive (MSO) detect interaural time differences (ITD) in the microsecond time range. The temporal precision of the underlying cellular integration process for this coincidence detection is based on pre- and postsynaptic and morphological specializations. Thick bipolar dendrites with large potassium conductances accelerate the EPSP time course while still introducing a propagation time for distal EPSPs towards the soma, which is at least in the same range as physiologically relevant ITDs. Therefore, the arrangement of synaptic contacts of an input fiber may also be crucial to the size of the binaural coincidence time window. To quantify the morphological arrangement of excitatory inputs on dendrites of MSO neurons and its impact on coincidence detection we combined axonal tracing, quantitative morphometry and computational modelling. Labelled axons terminating on MSO neurons followed closely the dendritic structure from distal sites to the terminal bouton with very little branching. Contact sites of labelled axons were identified as swellings adjacent to MAP2 labelled dendrites, a morphological feature that was VGlut positive. Single axons carried usually more than one swelling. Axonal swellings tended to cluster along the dendrite and the majority of terminal boutons were close to the soma. A single contact site harbors more than one active zone. Thus, a single fiber exerts strong excitation to the MSO dendrite along its whole extent. The influence of the distributions of axonal contact sites and travel times were examined in multi-compartmental models of an MSO neuron with conduction velocities known from either myelinated or un-myelinated axons. Distributed contact sites along the dendrite produced shorter EPSP peak latencies, normalized the EPSP time course and lead to larger summed EPSPs when compared to the same synaptic drive applied at a single site. Distributed synaptic sites also improved ITD sensitivity by sharpening the coincidence detection window, but sharpening only occurred when active low threshold potassium channels were present in the model. In a passive membrane model, distributed synapses actually widened the coincidence detection window, as predicted by cable theory. Thus, the arrangement of excitatory inputs of single fibers generally strongly affects the neuron’s ability to transfer information about ITDs. Simulations also showed that, as expected, myelination further shortened EPSP peak latencies, but, interestingly, had no effect on coincidence window and EPSP amplitude, and hence did not facilitate ITD encoding.