MCDB Home > Faculty > Steven Fisher > Current Research

Retinal Injury

The retina shares with the rest of the central nervous system limited regenerative capabilities. As in the brain and spinal cord, this has profound consequences for the return of function after injury. There are very limited options when it comes to reducing the damaging results of injury or promoting recovery in the central nervous system. The major goal of this part of my research is to gain a better understanding of the cellular events in a common sight-threatening injury to the retina called retinal detachment.

Retinal detachment initiates more than just photoreceptor cell degeneration. The degeneration of rod and cone photoreceptors accounts for the vast majority of blinding conditions in humans, whether genetic based (i.e., retinitis pigmentosa), disease related, or due to physical trauma (retinal detachment, penetrating injuries, etc). Injury initiates the degeneration of the light-sensitive portion of the photoreceptor, eventually leading to cell death; in some diseases this approaches 100% of the cells. It is well known that the treatment for detachment, physical reattachment surgery takes advantage of the ability of surviving cells to regenerate their outer segment. However recovery is not perfect. Although surgical success rates approach 95%, only 20-40% of eyes in which the macula (central focal point for high-acuity vision) is affected recover corrected visual acuity to 20/50. Many experience worse acuity. Other visual deficits commonly remain, deficits that are not measured by simple acuity tests. These data suggest that events beyond outer segment degeneration and regeneration underlie the degree of visual recovery. Also, in about 10% of successful reattachments sight-threatening complications develop from a process we believe to be analogous to glial scar formation in the brain and spinal cord. In the retina as elsewhere in the CNS there is a poorly understood balance between positive effects of glial cell reactivity and negative consequences.

Thus understanding mechanisms underlying the retina's responses to injury will have direct consequences in terms of the treatment of retinal injuries and will have implications for understanding similar processes elsewhere in the CNS. The reaction of the CNS to injury is an immensely complex reaction that brings into play many different cellular responses from an activation of resident macrophages to the proliferation of non-neural cells that probably produce a population of undifferentiated progenitor cells in the process.

Because photoreceptors have the ability to regenerate their light-sensing portion, we have looked to other changes in the retina that may explain poor functional recovery. The sort of "big picture" of our new discoveries are shown in the drawings below and they range from events such as synaptic and cellular remodeling throughout the retina to the activation of microglia, to the activation of Müller's glia, a critical component of glial scar formation. The differential expression of intermediate filament proteins in different subcellular compartments of Müller's glia appears to play a critical role in determining the directional growth of these cells onto the retinal surfaces as they form glial scars.

There are significant architectural changes in the retina at the tissue, cellular and subcellular levels as a result of retinal detachment. A goal of my research is to correlate these changes with the expression of specific proteins and to understand how these changes result in visual impairment. We are also very much in a "discovery mode" in this project, because there are many aspects of CNS degeneration that remain poorly understood.

Following up on these studies includes ongoing experiments generated by the following specific aims from my NIH grant

In this project we propose to use animal models to better understand the biological mechanisms underlying the responses of the retina and retinal pigmented epithelium to detachment and reattachment. We will also investigate, at the cellular level, the use of experimental procedures to either lessen the impact of detachment, or enhance the successful outcome of surgical reattachment. These experiments also should provide new information on the normal retina and its responses to injury and disease.

1. Cellular responses and remodeling. To test the hypotheses: a) that retinal detachment induces structural remodeling (and correlative molecular changes) in all major classes of retinal neurons and glia; b) that retinal reattachment will induce its own specific structural remodeling (and correlative molecular changes) of retinal neurons and glial cells; and c) that this remodeling results in significant changes in retinal circuitry.
2. Progenitor cells. To test the hypothesis that the intraretinal proliferative response induced by detachment results in the production of a population of progenitor cells, and to determine the fate of cells generated by the proliferative response.
3. Proliferative response. To test the hypotheses that inhibiting the intraretinal proliferative response induced by detachment will reduce the production of subretinal or epiretinal membranes and/or alter the course of degeneration induced by detachment.
4. Hyperoxic rescue. To test the hypothesis that hyperoxia will improve the outcome of retinal reattachment.
5. Mouse model of detachment. To develop and characterize a mouse model of detachment and initiate a series of studies taking advantage of transgenic animals available in this species. We will first test the hypothesis that Müller cell reactivity depends on the intermediate filament proteins GFAP and vimentin.

In addition to these we have a number of other related projects that are involve collaborations with other laboratories

1. Computational Molecular Phenotyping of retinal neurons and glial cells after detachment and reattachment. This involves bringing new technology into my lab from the lab or Robert Marc, University of Utah. CMP = the ability to detect small molecules (free amino acids, small molecule transmitters, e.g.) at high resolution in glutaraldehyde fixed, resin embedded tissue (the type used for EM), thus tissue suitable for very high resolution imaging. The analysis is heavily driven by computational imaging technology and produces quantitative results because only the surface of the ultrathin tissue section is labeled by the 2 nm gold particles. Serial sections can be registered using specific software and projected as a single plane to show the "molecular signatures" of different cell types. We first used this technology in collaboration with the Marc lab in 1998 (see Journal cover below as an example of the data generated), but we are now undertaking a major effort to import it here because of its value to our research goals. This has not been a trivial project because it involves issues of producing consistent serial ultrathin sections (50-100nm thickness), some fairly tedious gold labeling technology, and significant image maniputalion (mosaic production, precise image registration at the cellular level). Twe are working with the University of Utah group and our own bio-image informatics scientists to develop to develop new, more "user-friendly," and less expensive software for use in the technique.

   An example from an earlier study of the results of CMP in retinal tissue. RGB color channel are assigned to signals from immunogold labeling by antibodies to different small molecules and the images superimposed. Marc, Murry, Fisher, Linberg, Lewis & Kalloniatis, IOVS. (1998)

2. Comparative studies or protein expression in human retinal samples. These studies, with retinal surgeons from Santa Barbara, London, Amsterdam, and Berlin have provided us with the opportunity to validate our animal models of detachment by making direct comparisons to results from human tissue samples. There are distinct species differences, hence the need to determine how the cellular reactions in different species used as model systems compares to that in human retina.
3. Nanometer resolution tomography of normal and degenerative photoreceptor synapses. This is a new collaboration with the National Center for Imaging and Microscopy Research at UC San Diego. With their help we are using the Intermediate Voltage EM and appropriate computational methods to obtain accurate 3-dimensional tomographic data from "thick" (0.25 µm) sections of retinal tissue. This is a cross-collaboration with our Bio-Image Informatics project. We know that there changes in structure and protein expression in photoreceptor synapses after detachment. This will study will allow us to make structural comparisons at nanometer levels of resolution, thus comparisons between subcellular events. Examples of such questions are the distribution of synaptic vesicles within synaptic terminals, the relationship of synaptic vesicles to the plasma membrane, and the configuration of other pre- and post-synaptic structures,
4. IV. Physiological response characteristics of retinal ganglion cells after retinal detachment. This is a new collaborative project with Prof. Ming-Liang Pu, Peking Univeristy. Prof. Pu is an expert in electrophysiological recordings from single retinal ganglion cells. Pilot data indicates that there are specific changes in the receptive field characteristics of these retinal output neurons in a detached retina These experiments will basically test the hypothesis that retinal circuitry changes after detachment can effect the physiological output of the retina (the alternative is that redundant pathways salvage the output.

Bio-Image Informatics

Most of the primary data collected in my lab is in the form of biological images. Thus the opportunity to be part of a new research emphasis on campus in bio-image informatics is an exciting new development for me. This field includes the handling, analysis, and processing of large sets of bio-image data. This research is becoming increasingly relevant as digital imaging makes it possible to collect immense amount of data. The means for handling, searching and analyzing this data have not kept up with our ability to generate it.

The Center for Bio-Image Informatics main goal is to develop new information processing technologies appropriate for extracting detailed understanding of biological processes from images depicting the distribution of biological molecules within cells or tissues. This will be accomplished by developing new methods for information processing at the sensor level to enable high speed and super-resolution imaging, by applying pattern recognition and data mining methods to bio-molecular images to fully automate the extraction of information from those images and the construction of statistically-sound models of the processes depicted in them, and by developing distributed database methods for large sets of biological images.

This project is highly-interdisciplinary and under the leadership of Prof. B.S. Manjunath, involves several different laboratories working together. Here are some of the highlights of the research: These summaries are taken from the Center's website (http://www.bioimage.ucsb.edu/) where more details can be found My lab has made contributions in one form or another to each, and will continue to do so.

Bisque (Bio-Image Semantic Query User Environment) was developed for the exchange and exploration of biological images. The Bisque system supports several areas useful for imaging researchers from image capture to image analsysis and querying. Capturing image experimental parameters and meta-data is done through a flexible "Digital Notebook": A simple application which imports images and metadata directly to the Bisque system from the researcher's desktop. The bisque system is centered around a database of images and metadata. Search and comparison of datasets by image data and content is supported. Novel semantic analyses are integrated into the system allowing high level semantic queries and comparison of image content.

Image Based Tool for Counting nuclei (ITCN) is an ImageJ plugin for counting the number cells within an image. The inputs are: (1) an estimation of the diameter of a cell, (2) an estimation of the minimum distance between cells, and (3) either a region of interest (ROI) selected with ImageJ's selection tools or a black and white mask image that is white in regions that are to be counted.

The iWall is a 20 monitor tiled display for viewing high resolution biological images. Advances in microscopy imaging and automatic image mosaicking are providing biologists with images of such resolution that they can't be properly viewed on standard displays. The UCSB iWall provides a 8000x 4800 pixel display (38.4 Megapixels) for viewing every detail of even the largest images . The monitors are arranged in a 5 x 4 grid with one computer driving a column of 4 monitors for a total of 5 computers driving the monitor grid. A sixth computer serves as the main control node, directing the other computers to display portions of the 8000 x 4800 sized window. The entire system was put together with off-the-shelf equipment and open source software. Distributed Multihead X (DMX) is used to create a seamless images across all 20 windows.