The concept that sensory information was encoded as topographic maps in the brain existed since the time of Descartes (1644 and 1647). Recent studies using modern histological and tracing techniques have demonstrated that topographic mapping is a basic principle of brain architecture. Neuronal inputs to the central nervous system (CNS) are topographic in many sensory systems including the retinofugal, the auditory, and the somatosensory projections. Topographic maps are not restricted to the sensory systems. They also exist in the limbic circuits, which mediate learning, memory, and emotions. One of the major limbic components, the hippocampus for example, is connected to the subcortical target, the septum, in a highly organized manner.

It has been proposed by Roger Sperry in the 1940's that axons are guided to their proper target by matching cytochemical tags on the presynaptic and postsynaptic neurons. To address the problematic requirement of astronomically large number of tags for the wiring of the nervous system. Sperry further proposed that the tags might be distributed in the projecting and target fields in gradients, and only matching concentrations of the complementary tags allowed formation of synapses. Such tag pairs might be oriented along perpendicular axis to define multi-dimensional maps.

The molecular nature of topographic guidance cues has only been recently defined. We have shown that the Eph family tyrosine kinase receptors and their ligands are expressed in opposing gradients in multiple neural circuits including the hippocamposeptal, thalamocortica, and midbrain dopaminergic projections, fulfilling the long-standing prediction of Sperry's chemoaffinity theory. We have further defined in vitro using a novel co-culture assays developed in the laboratory that the Eph ligands can differentiate neurons from different spatial origins and negatively regulate the growth of topographically inappropriate axons. Our studies demonstrated for the first time that these ligands establish inhibitory domains in many different regions of the brain and function as general regulators of topographic map formation in multiple neural pathways. Currently we are using a number of approaches. including transgenic and knockout technology. to study the function of both negative and positive candidate axonal guidance molecules in vivo. Our long-term objective is to elucidate how the concerted actions of different guidance cues establish neuronal networks.