Abstract
Purpose: :
A method for automated glial fibrillary acidic protein (GFAP) analysis after immunohistology was developed using eyes of our established Experimental Autoimmune Glaucoma animal model (EAG). This model is non-intraocular pressure dependant. Slow progressive retinal ganglion cell (RGC) loss is a result of immunizing rats with ocular antigens. Intent of this study was to analyze the behavior of Müller glia with software based detection shortly after immunization.
Methods: :
During the EAG-study IOP was measured regularly. After removal of rat’s eyes RGC density was examined. For GFAP-immunofluorescence staining cross-sectioned eyes taken 8, 12 and 22 days after immunization with optic nerve homogenate (ONH) and equal controls (Co) were used (n=5 per group). 2 central areas, near optic nerve head, and 2 peripheral areas of each retina were photographed. GFAP was quantified using 2 computer based methods: Method A was based on area measurements with respective GFAP intensity after threshold setting, to detect Müller cell expansion. Method B was based on different intensities of GFAP immunofluorescence, as initial point for comparison. Integrated density was acquired. Both groups at all time-points were compared using t-test.
Results: :
IOP stayed stable throughout the EAG-study (P=0.9). A significant RGC loss was detectable at 22 days (P=0.001). Examining method A, the mean area for both ONH and control group increased from 8 (% of GFAP positive area: ONH=7.1%±2.3, Co=5%±3.1) to 14 (ONH=7.8%±4.6, Co=5.5%±1.2) to 22 days (ONH=8.5%±3.1, Co=6%±3). However, the immunized group ascended significantly higher each time (P<0.05). Analyzing method B, an increase in integrated density of GFAP could be detected in time. A first significant raise could be ascertained for the 22 day time-point with an integrated density of 218.4±53 for ONH and 173.1±53.2 for Co (P=0.01). Data reveals a reactive gliosis after immunization. At first swelling of Müller glia was observed. Later in process, as RGCs die off, Müller glia produce more GFAP-filament per area.
Conclusions: :
As EAG is a glaucoma model with slow progressive RGC loss, the common glia scoring systems, for example ranging from 0 to 3, were not detailed enough to detect discrete changes. A software based analyzing method, to examine even discrete increase in expression of GFAP, was developed and applied. Analyzing the sequential increase of Müller glia was only possible with the combination of methods A and B (area plus intensity of GFAP). This analytical method could also be useful for other animal models in the future.
Keywords: Muller cells • microscopy: light/fluorescence/immunohistochemistry • glia