Within the retina, neurons are organized vertically into functionally distinct layers, and horizontally into mosaic patterns, in which cells of the same type (homotypic cells) are evenly spaced from one another. This organization is the anatomical basis for the circuits that form within the retina, and is a general principle governing the cellular anatomy of other parts of the brain. The molecular signals that allow cells of the same type to recognize one another and space themselves accordingly are not well understood. It is our goal to understand the molecular mechanisms that direct these neurodevelopmental processes, focusing on matrix and adhesion molecules as the molecular labels that will define homotypic cells. We have identified a mutation in mice that eliminates self-recognition in a subset of retinal neurons. This is the first mutation identified to alter mosaic formation in the retina. Positional cloning identified the mutation as an internal deletion in Down Syndrome Cell Adhesion Molecule (Dscam), which creates a loss of function allele. The subtypes of amacrine cells that would normally express Dscam fail to arborize their processes, which instead fasciculate and pull the cell bodies out of their mosaic pattern. This phenotype is consistent with a failure of these cells to recognize their homotypic partners and maintain their mosaic spacing. Ganglion cells are also affected by the Dscam mutation, but their phenotypes may, in part, be secondary to the disruption of upstream amacrine cell inputs. In this proposal, we will examine how homophilic adhesion of cells expressing Dscam, or the related Dscam-like1, serves to identify homotypic populations of amacrine and ganglion cells during the formation of retinal circuits. We propose three specific aims. In the first, we will examine the morphology, spacing, and the maintenance of axonal connectivity of retinal ganglion cells using a newly generated Dscam conditional allele to delete the gene inducibly or specifically in ganglion cells. We will determine whether the defects observed are primary effects from the loss of Dscam function in the ganglion cells, or are secondary to perturbations in amacrine/ganglion cell circuits. In the second, we will study intracellular effectors of DSCAM function. We have identified the PDZ-domain containing proteins MAGI-2 and -3 as binding the C-terminus of DSCAM. We will examine the significance and specificity of these interactions in vitro and in vivo. Finally, the Dscam-like1 gene is expressed in a similar but non-overlapping population of retinal neurons, leading us to hypothesize that other populations of neurons will use different DSCAM-related proteins for self-recognition. We will test this in our third aim by deleting Dscam-like1 in the mouse. We anticipate a similar phenotype in a different population of retinal cells, as suggested by our preliminary studies of these knockout mice. In total, these studies address the molecular basis of homotypic recognition and the formation of neuronal mosaics, fundamental processes in neurodevelopment, and have implications for neurodevelopmental pathologies seen in human congenital diseases such as Down Syndrome.
We are proposing to study the involvement of the mouse ortholog of Down Syndrome Cell Adhesion Molecule (DSCAM) in the cellular recognition events that direct retinal development. The DSCAM gene in humans is in the region of Chromosome 21 associated with Down Syndrome trisomies. Our preliminary results make it clear that Dscam in mice plays an important role in neurodevelopment. Understanding the basic molecular mechanisms of Dscam's function will help to define its possible role in the neurodevelopmental phenotypes associated with Down Syndrome, and possibly human congenital retinopathies as well.
Showing the most recent 10 out of 12 publications
