Competitive interactions during dendritic growth: a simple stochastic growth algorithm

A simple growth algorithm is presented that deals with one feature of dendritic growth, the distance between branches. The fundamental assumption of our growth algorithm is that the lengths of dendritic segments are determined by the branching characteristics of the growing neurite. Realistic-appearing dendritic trees are produced by computer simulations in which it is assumed that: (1) growth of individual neurons occurs only at the tips of each growing neurite; (2) the growing neurite can either branch (as a bifurcation) or continue to elongate; (3) events at any one growing tip do not affect the events at any other growing tip; and (4) the probability of branching is a function only of the distance grown either from the cell body (if branching has not occurred) or from the previous branch point. An analytic solution of a differential equation based on these same assumptions produces a distribution of dendritic segment lengths that accurately fits an experimentally determined distribution of dendritic segment lengths of reconstructed neurons, accounting for about 89% of the sample variance. Our analysis indicates that, immediately following branching, the temporary suppression of further branching during dendritic growth may be an important mechanism for regulating the distance between branches.