The Use of Stereologic Counting Methods to Assess Retinal Ganglion Cell (RGC) Loss

Abstract

Abstract: : Purpose:To determine an unbiased estimate of RGC death in a rat glaucoma model. Methods:RGC were backlabeled with Fluorogold (FG). One week later, elevated intraocular pressure (IOP) was induced in one eye of each rat using the Morrison model. Following 10 days of elevated IOP, eyes were removed, sectioned (16 microns) and immunostained with rabbit–anti–FG antibody. A random number generator was used to determine the first sampling section. Additional sections were then selected at preset intervals systematically distributed throughout the retina with each section having an equal chance of being counted using the Olympus CAST stereology system. The optical fractionator method was applied to step through known volumes of retinal tissue for interactive RGC identification. Using this sampling scheme, total numbers of RGC were estimated from six rats with unilaterally elevated IOP. Results:The variance of RGC number estimated was small and essentially identical when counting 3% or 5% of every 10th or 20th retinal section from a rat with normal IOP. Counting every 30th, 40th or 50th retinal section yielded values that were highly variable. Thus, a sampling scheme of counting 5% of every 20th section was used. Using this systematic random sampling scheme, as few as 250 RGC/retina needed to be counted to accurately estimate total RGC number. The estimated total number of RGC of the eyes with normal IOP was 8.83x10⁴ ± 5.99x10³ (n=6), whereas the estimated total number of RGC for the eyes with increased IOP was 5.58x10⁴ ± 4.59x10³ (n=6). Thus, a 33 ± 2% (n=6, p<0.05) RGC loss was seen in rats with IOP elevated for 10 days (peak IOP = 43.2 ± 1.53 mmHg). Conclusions:Methods for manually counting RGC within a selected sample are prone to subjective and systematic errors. Stereological counting yielded consistent RGC number estimates. Systematic random sampling using stereology may lead to more accurate and unbiased results, and therefore may be preferred for a more thorough investigation of the effects of elevated intraocular pressure on rat RGC death. This method illustrates the consistent and marked loss of RGC in the Morrison model of experimental glaucoma and will provide a robust technique to assess neuroprotective strategies.

Keywords: ganglion cells • microscopy: light/fluorescence/immunohistochemistry • cell death/apoptosis