Effects of variability in anatomical reconstruction techniques on models of synaptic integration by dendrites: a comparison of three Internet archives

Abstract

The first step in building a realistic computational neuron model is to produce a passive electrical skeleton on to which active conductances can be grafted. For this, anatomically accurate morphological reconstructions of the desired cell type are required. In this study compartmental models were used to compare from a functional perspective three on-line archives of rat hippocampal CA1 pyramidal cell morphologies. The topological organization of cells was found to be similar for all archives, but several morphometric differences were observed. The three-dimensional size of the cells, the diameter and tortuosity of dendrites, and the electrotonic length of the main apical dendrite and of the branches in stratum lacunosum moleculare were dissimilar. The experimentally measured kinetics of somatically recorded inhibitory postsynaptic currents evoked in the stratum lacunosum moleculare (data from the literature) could be reproduced only using the archives that contained cells with an electrotonically short main apical dendrite. In the amplitude attenuation of the simulated postsynaptic currents and the voltage escape from the command potential under voltage clamp conditions, a two- to three-fold difference was observed among archives. Upon activation of a single model synapse on distal branches, cells with low dendritic diameter showed a voltage escape larger than 15 mV. The diameter of the dendrites influenced greatly the results, emphasizing the importance of methods that allow an accurate measurement of this parameter. Our results indicate that there are functionally significant differences in the morphometric data available in different archives even if the cell type, brain region and species are the same.