One of the major problems of modern neurobiology is to understand how genetic information gives rise to complex organization of organs such as the brain and spinal cord. As a highly accessible extension of the brain, the retina can serve as a particularly good model for the development of brain organization. We have chosen to study how the retina self-organizes during development. To understand this process, it is important to realize that genes cannot directly specify all retinal cells. While genes specify tens of thousands of proteins, there are hundreds of millions of retinal cells, each with a different receptive field. Hence, genes do not specify cell by cell, but rather generate mechanisms through which cells influence each other, giving rise to complex organization. These are called epigenetic mechanisms. Our goal is to examine the epigenetic mechanisms controlling the development of ganglion-cells.
Here, we provide the following two examples of studies in our laboratory of how epigenetic mechanisms contribute to retinal development: First, the size of a ganglion cell’s receptive field determines the amount of spatial integration that the cell can perform. Thus, it directly indicates the sensitivity of the cell (and by extension, the visual system as a whole) to spatial scale. We discovered that spontaneous waves of neural activity in the retina during development are a critical component in determining ganglion-cell receptive-field size. If one eliminates waves, growth ceases. Alternatively, if one causes waves to continue beyond their normal cessation, receptive fields overgrow. (One way to do this is to rear animals in the dark, since it is exposure to light that normally ends the spontaneous waves of activity.) However, growth is not entirely regulated by the waves. We now know that receptive-field growth will eventually stop even in the continued presence of waves.
Second, if one dark rears an animal during development, the density of particular kinds of inter-neurons may change. For instance, we found that the density of acetylcholine-containing cells increases in turtle retinas. This result can be proven both by immunocytochemical studies and by measuring with western blotting levels of acetylcholine-related enzymes. This increase in acetylcholine is important, since cholinergic cells contribute to the retinal response to motion. Furthermore, they contribute to spontaneous waves, which are so important for receptive-field growth.
This line of developmental research has several possible practical applications. First, any understanding of normal retinal development may prove useful in deriving treatments for developmental pathologies, such as retinopathy of prematurity, retinal degeneration, and Leber hereditary optic neuropathy. Moreover, one can now transplant healthy retinal tissue into an eye with a damaged or defective retina. However, the transplanted tissue rarely forms neatly organized layers, as in normal retina. And the transplant does not integrate with the extant, surviving retinal tissue. By mimicking the conditions present during development, it might be possible to induce successful incorporation of the transplant. A final possible application of our developmental work is technological. The rules discovered for self-organization in neural tissue may prove useful in the further development of artificially intelligent neural networks.