Visual System Development and Dysfunction

David R. Hyde   Professor Rev. Howard J. Kenna C.S.C. Memorial Director of the Center for Zebrafish Research Ph.D. 1985, Pennsylvania State University in biochemistry and molecular biology Post-doc 1985-1988: California Institute of Technology in molecular genetics and neurobiology with Seymour Benzer

e-mail | labpage

      My lab studies a variety of processes associated with the zebrafish eye, including development of the retina and lens, retinal cell death (neuronal degeneration), and the role of adult stem cells and glial cells in regeneration of retinal neurons. The retina is an excellent model because it serves as an easily accessible portion of the central nervous system.   Furthermore, the retina is not essential for the viability of the organism, which permits a variety of approaches to be used in studying its development and activity.   The use of zebrafish in these studies permits a detailed analysis of retinal development because a large number of progeny are produced form a single pair mating and that the embryos develop externally, which simplifies their manipulation and analysis.   Because the zebrafish retina grows throughout the life of the individual, a population of stem cells exists in the adult retina.   There presence is partially responsible for the robust neuronal regeneration response that we are studying.   We use genetic, cell biological, biochemical, and molecular approaches to examine the status of the retina and lens.   These studies are directly relevant to an increased understanding of general neuronal processes and a variety of inherited human diseases, such as macular degeneration and retinitis pigmentosa in the retina and cataracts in the lens.

Analysis of retinal development

    The zebrafish retina develops from a neuroepithelial sheet of undifferentiated cells that proceed through a coordinated wave of mitotic activity followed by cell differentiation to produce the mature laminated retina (Figure 1A).   We are studying how a group of proteins, which are involved in generating the tight junction that separates the apical membrane from the basal-lateral membrane in neuroepithelial cells, are involved in the development of the eyes, the patterning of the different layers in the retina, and the differentiation of the various neuronal types in the retina.   We are pursuing this work using a combination of classical and molecular genetic techniques.   For example, we cloned the zebrafish pard3 gene (Wei et al ., submitted), which is alternatively spliced to encode two different proteins that localize to the apical side of the tight junction.   Using molecular genetic techniques that block the translation of the Pard3 protein produced two phenotypes.   The first phenotype is cyclopia, which is the fusion of the eyes.   We found that the absence of Pard3 expression in the developing ventral diencephalon in the brain resulted in cell death and the inability of the developing eyefields to separate.   Later, the retina exhibits a lack of lamination (Figure 1B), even though all the different neuronal classes are present in the disorganized retina.   We are continuing to identify and characterize additional proteins that are expressed in the tight junction.   While these proteins are all expressed at the tight junction of the neuroepithelial cells, we found that loss of these proteins results in a variety of different developmental phenotypes.   Further characterization of these phenotypes and protein interactions will reveal the role of these proteins in retinal and brain development.

Figure 1.   A wild-type retina (A) and a retina that develops in the absence of the Pard3 protein (B).   The wild-type retina is laminated and contains the outer nuclear layer (ONL), which is where the rod and cone cell nuclei are located, the inner nuclear layer (INL), where the stem cell population is located, and the ganglion cell layer (GCL).   These three nuclear layers are separated by two synaptic layers, the outer plexiform layer (IPL) and inner plexiform layer (IPL).   The retina that develops in the absence of the Pard3 protein is not laminated and lacks any neuronal organization.

Retinal degeneration and regeneration

      We demonstrated that treating albino zebrafish with intense light causes apoptosis (cell death) of the rod and cone cells.   However, cells in the underlying inner nuclear layer begin to proliferate and migrate along Muller (glial) cells to the outer nuclear layer where they differentiate into both rods and cones (Vihtelic and Hyde, 2000).   The ability of these inner nuclear layer cells to differentiate into rod and cone cells after light damage suggests that they are pluripotent and may represent stem cells. We recently found that the light-induced damage varies in different regions of the retina.   Surprisingly, retinal regions that contain primarily rod cell death exhibit low levels of proliferating inner nuclear layer cells, while regions that contain excessive rod and cone cell death exhibit large numbers of proliferating inner nuclear layer cells.   This suggests that the cell proliferation and subsequent neuronal differentiation is dependent on the type and amount of neuronal cell death in the overlying retinal layer. Recent experiments also demonstrate that the Muller glia also begin to express genes and proteins that are expressed in early neuronal differentiation, which suggests that the Muller glia begin to take on a neuronal identity in response to the light-induced retinal degeneration.   To further examine the neuronal regeneration response, we recently began injecting ouabain into the zebrafish eye, which results in cell death of all the neuronal classes throughout the retina (Figure 2).   Within seven days of the massive cell death, there is a large increase in cell proliferation that results in the regeneration of all the neuronal classes in the retina (Figure 2).   This further supports the idea that the proliferating inner nuclear layer cells are stem cells and possess the capacity to differentiate into any retinal cell type.   We are very interested in determining the breadth of neuronal classes that can be regenerated from the inner nuclear layer stem cells, the mechanism that permits these stem cells to recognize which neuronal classes are lost, and the processes that control the differentiation of one neuronal class over another class.   We are employing gene microarray experiments and genetic screens to identify the genes that are involved in these processes.

Figure 2.   Ouabain-induced retinal degeneration is followed by regeneration.   A control retina exhibits the standard lamination pattern.   Four days after ouabain injection, the ganglion cell layer (GCL) is dramatically reduced and the outer and inner nuclear layers (ONL and INL, respectively) are disorganized.   Seven days after ouabain injection, the three nuclear layers can no longer be differentiated.   By 14 days, the ONL is reorganizing.   By 90 days after ouabain injection, the retina is again laminated.    A few misplaced nuclei are present in the synaptic layer (IPL), which may be due to their migration between nuclear layers or a stable displacement.

      Our identification that the Muller glial cells begin to express neuronal cell proteins after the light-induced damage suggests that they may dedifferentiate or trans-differentiate into neurons.   A transgenic zebrafish line that expresses Green Fluorescent Protein (GFP) in neuronal precursor cells, also exhibits GFP expression in a subset of Muller glial cells after light-induced damage.   A subset of radial glial cells in the mammalian brain has also been suggested to differentiate into neurons during specific periods of brain development.   Thus, we are interested in confirming if the Muller glia can differentiate into retinal neurons and elucidating the neuronal classes can be derived from these glial cells.   To perform these experiments, we are developing novel transgenic techniques that will express GFP from a Muller glial-specific promoter and then switch the GFP expression to a ubiquitous promoter to continue expressing the GFP marker in any cell that is differentiated from the Muller glial cell.   This novel technology will be useful in addressing a number of interesting questions in both neuronal regeneration and in retinal and brain development.

Transgenic techniques to characterize retinal degeneration and regeneration

     We cloned several different genes that are expressed in different neuronal classes or the Muller glia in the zebrafish retina (for example, Vihtelic et al., 1999).   We isolated the upstream regions of these genes to identify the essential elements of the corresponding promoters, with the intent to generate transgenic zebrafish.   This involves cloning the GFP gene downstream of different promoter sequences and then cloning these constructs into a transposable element and microinjecting them into zebrafish embryos.   The GFP expression then reveals the expression pattern that is associated with a specific promoter fragment.   Examples of various promoters are shown in Figure 3.   In Panel A, a rhodopsin promoter directs GFP expression in only the rod photoreceptor cells (Kennedy et al., 2001).   In Panel B, an ubiquitous promoter drives GFP expression throughout the zebrafish embryo (a non-transgenic embryo shows the autofluorescence in the residual yolk sac).   We are also cloning promoters that are active in only Muller glial cells, lens epithelial cells, and neuronal precursor cells.   In addition to expressing GFP from these promoters to mark different cell types to characterize development and regeneration, we intend to express toxins from these promoters to ablate specific cell types to examine their role in development, regeneration, and retinal function (such as visual behaviors).

Figure 3.   Transgenic zebrafish developed at Notre Dame.   A transgenic zebrafish line that expresses Green Fluorescent Protein (GFP) from the rhodopsin promoter in only the rod photoreceptor cells (A).   The GFP is expressed from the ef1 ubiquitous promoter and is found throughout the zebrafish embryo (B, top).   A control zebrafish embryo shows autofluorescence in the residual yolk sac (B, bottom).

Gene microarray analysis during retinal regeneration

      We recently initiated gene microarray studies of mRNA expressed in the retina during light-induced retinal degeneration and regeneration.   These experiments reveal that a large number of genes alter their expression in response to the light treatment.   The patterns suggest an increased number of cells reentering the cell cycle, an increased number of neuronal precursor cells and their movement through the differentiation pathway, and an increased number of migrating cells.   These results are consistent with our immunohistochemical analysis of the regenerating retina.   We are now utilizing real-time PCR to confirm these changes in gene expression levels.