The development of neural connections from retina to thalamus: Maturation and plasticity of receptive fields

Neural connections in the mammalian nervous system can be extremely precise. This is best demonstrated in the adult central visual system. Single relay neurons in the thalamus, for instance, receive strong synaptic connections from only one or two retinal ganglion cells. This adult state emerges from an immature one where neurons synapse onto many inappropriate targets. The program by which neurons select appropriate targets consists of activity-independent (molecular) and activity-dependent phases. During the activity-dependent phase of neural development, circuits are susceptible to pathologic reorganization from sensory deprivation or abnormal sensory experience. The two studies presented in this thesis represent an attempt to understand the activity-dependent mechanisms by which single neurons refine their synaptic connections onto their targets. First, we used white-noise stimuli to make high-resolution receptive-field maps of cells in the developing ferret retinothalamic pathway. This allowed us to test a fundamental assumption in present models of thalamocortical development: that thalamic inputs to the cortex are a homogeneous population, each with concentric on- or off-center receptive fields. Receptive fields in the lateral geniculate nucleus (LGN) during development (from P31 to ∼P50) had a variety of shapes. Some were concentric, others had elongated centers, while some had isolated ‘hot spots’ of sensitivity. We find that the convergence of multiple retinal afferents onto immature LGN neurons causes inhomogeneous receptive fields. Building on these results, we propose a Hebbian-based model in which imprecise connections between retina and LGN help refine connections between the LGN and cortex.; We next asked whether visual experience alone could instructively pattern connections from retina to thalamus. In developing ferrets, when we synchronized the activity of retinal neurons with a horizontal stimulus for 25 minutes, the receptive fields of their thalamic targets became more horizontal. Subsequent stimulation with a vertical stimulus showed that this effect was dynamic and stimulus-specific. Our findings suggest that cells with synchronous activity—such as spontaneously firing retinal ganglion cells in the immature mammalian retina—have their synapses strengthened onto common targets. Our finding of specific, experience-induced plasticity bears on a lengthy debate on the instructive versus permissive role of visual experience in brain development.