The role of experience in the development and maturation of neural circuits

At the onset of visual experience, which corresponds to eye opening in some mammals and birth in others, visual cortex has already undergone significant development. The neurons of the cortex have been generated and have migrated to their mature positions, and molecular cues have guided afferent axons from the lateral geniculate nucleus (LGN, the thalamic relay of retinal signals) to their approximate final locations within the map of visual space. In addition, activity-dependent plasticity mechanisms that rely on spontaneous neuronal firing (independent of experience) have helped to refine the retinotopic organization of the LGN afferents.

Yet we also know that experience has a major impact on the maturation of cortical circuits. The density of synaptic connections in visual cortex doubles in the first year of human life. If vision in one of the two eyes is obscured for a period of time in the first year after birth, then signals from that eye become functionally disconnected from visual cortex (amblyopia). If both eyes are obscured, then motion processing suffers dramatically. While there has been significant research addressing how visual cortical inputs can be rewired during a “critical” period after receptive field properties have been initially established, we are interested in understanding how experience influences the initial assembly of neuronal circuits and the initial emergence of receptive field properties.

My colleagues and I have recently identified a useful model system for exploring the impact of experience on the development of neural circuits using ferret visual cortex, where motion selectivity develops in an experience-dependent manner.  At the time of eye opening, cortical neurons exhibit orientation selectivity but they respond equally to stimulation in either of two opposite directions of motion. In the weeks after eye opening, most neurons develop a strong preference for motion in a single direction (top right figure). We employ a 2-photon calcium imaging system that can follow the activity of dozens of the same single neurons over time (left figure). Using this system, my colleagues and I discovered that a motion training protocol consisting of visual stimulation with gratings that moved back and forth in opposite directions led to the rapid (3-6 hours) emergence of direction selective responses in cortical layer 2/3 of anesthetized, visually naïve ferrets that had just opened their eyes (bottom right figure).

In current experiments, members of the Van Hooser lab are focusing on identifying the changes in visual cortical circuitry that underlie the emergence and maturation of motion selectivity.

Watching the brain as it learns to see

Using 2-photon microscopy and fluorescent calcium dyes, it is possible to study how response profiles of dozens of the same single neurons change over time. Here, we assessed the motion selectivity of a group of cells in a naive animal that had just opened its eyes (left), and examined the motion selectivity of the same cells after 6 hours of exposure to moving bars (right). The two highlighted cells were indifferent to two directions of stimulus motion before training (left), but, after training, strongly preferred motion that was upward and rightward. This result suggests that visual experience is sufficient for the emergence of motion selectivity. From Li/Van Hooser et al., 2008.
Motion selectivity emerges after the onset of 
visual experience



At eye opening (left example), most neurons are selective for oriented bars but respond equally to motion in one of two opposite directions (green symbols; symbol orientation indicates cell orientation preference). After several days of natural visual experience (right example), a majority of neurons exhibit a strong preference for one direction of motion (red arrows). From Li/Van Hooser et al., 2008.


Emergence of motion selectivity depends on
experience with a moving stimulus
  

At eye opening (left), most neurons are unselective for motion, but after several hours of training with a motion stimulus (moving in the directions in white), many of the same neurons exhibit a strong preference for one direction of motion (red arrows, top 3 images on right). If flashing bars are used in the training period instead of moving bars, then there is no increase in motion selectivity (bottom). From Li/Van Hooser et al., 2008.