Abstract

The medial superior olive (MSO) is an auditory brainstem nucleus within the superior olivary complex. Its functional role for sound source localization has been thoroughly investigated (for review see Grothe et al., 2010). However, few quantita tive data about the morphology of these neuronal coincidence detectors are available and computational models incorporating detailed reconstructions do not exist. This leaves open questions about metric characteristics of the morphology of MSO neurons as well as about electrophysiological properties that can be discovered using detailed multicompartmental models: what are the passive parameters of the membrane? What is the axial resistivity? How do dendrites integrate synaptic events? Is the medial dendrite symmetric to the lateral dendrite with respect to integration of synaptic events? This thesis has two main aspects: on the one hand, I examined the shape of a MSO neuron by developing and applying various morphological quantifications. On the other hand, I looked at the impact of morphology on basic electrophysiological properties and on characteristics of coincidence detection. As animal model I used Mongolian gerbils (Meriones unguiculatus) during the late phase of development between postnatal day 9 (P9) and 37 (P37). This period of time is of special interest, as it spans from just before hearing onset at P12 – P13 (Finck et al., 1972; Ryan et al., 1982; Smith and Kraus, 1987) to adulthood. I used single cell electroporation, microscopic reconstruction, and compartmentalization to extract anatomical parameters of MSO neurons, to quantitatively describe their morphology and development, and to establish multi-compartmental models. I found that maturation of the morphology is completed around P27, when the MSO neurons are morphologically compact and cylinder-like. Dendritic arbors become less complex between P9 and P21 as the number of branch points, the total cell length, and the amount of cell membrane decrease. Dendritic radius increases until P27 and is likely to be the main source of the increase in cell volume. In addition, I showed that in more than 85% of all MSO neurons, the axonal origin is located at the soma. I estimated the axial resistivity (80 Ω·cm) and the development of the resting conductance (total conductance during the state of resting potential) which reaches 3 mS/cm2 in adult gerbils. Applying these parameters, multi-compartmental models showed that medial versus lateral dendritic trees do not equally integrate comparable synaptic inputs. On average, latencies to peak and rise times of lateral stimulation are longer (12 μs and 5 μs, respectively) compared to medial stimulation. This is reflected in the fact that volume, surface area, and total cell length of the lateral dendritic trees are significantly more larger in comparison to the medial ones. Simplified models of MSO neurons showed that dendrites improve coincidence detection (Agmon-Snir et al., 1998; Grau-Serrat et al., 2003; Dasika et al., 2007). Here, I confirmed these findings also for multi-compartmental models with biological realistic morphologies. However, the improvement of coincidence detection by dendrites decreases during early postnatal development.