The 8‐fold quadrant dissection method for ex vivo human interventional retinal experimentation

Abstract

PurposeRetinal research relies on animal and in‐vitro models which lack many of the characteristics of human retina. We have instead established a reproducible ex‐vivo model of primary retinal explants derived from human donor eye cups.MethodsA dissection strategy performed independently by two investigators was designed to maximize retinal tissue whilst maintaining experimentally reproducible fragments of retina suitable for experimental purposes. The retina was divided into 4 quadrants through the fovea and equivalent distribution of photoreceptors between quadrants was confirmed with CD73 staining in 7 pairs of donor eyes with flow cytometry. In the 8 quadrants from each of the 7 donors, the standard error in proportion of photoreceptors was 0.8–2.3%.ResultsCellular composition of free floating retinal explants were followed for 2 months in 10 retinas with quadrant dissection and 8 with random dissection. Explants could be maintained for 2 months with live populations of photoreceptors, ganglion cells and Müller cells detected by flow cytometry. In contrast, amacrine and horizontal cells, decreased by 90% at 7 days. Only quadrant dissection from the same patient showed reproducible and reliable proportions of cell populations between dissections at any time point. There was no statistically significant relationship between the proportion of any cell population and donor age, time after death, time to storage or time in storage.ConclusionsThe 8 fold quadrant dissection method from a single donor forms an attractive human experimental model for interventional testing in retinal research. We demonstrate the utility of this model in a simple neuroprotection study, showing that insulin protected against CoCl2 induced hypoxia in human photoreceptors.