Properties of a double gradient model for retinotectal specificity.

The properties of a double gradient model for retinotectal specificity are discussed. The model utilizes only two complementary molecules, each located on both retina and tectum, to determine position along the dorsoventral axis. Two possible modes of interaction between these molecules are assumed. One of these allows all possible bonds between a retinal cell and tectal loci to be formed. This results in a rigid retinotectal projection in which the adhesion of each retinal cell to its normal tectal locus is maximal. The other assumes stochastic interactions between the molecules, and results in retinotectal specificity only if additional constraints are imposed on the system. Either of these modes of interaction predicts adhesive preferences for retinal cells to tectal halves similar to those observed experimentally.