Classifying Mouse RPE Morphometric Heterogeneity Using REShAPE: An AI-Based Image Analysis Tool

Abstract

Retinal degenerative diseases caused by retinal pigment epithelium (RPE) dysfunction affect specific areas of the retina. Regions of molecular and phenotypic RPE heterogeneity have been described in the human eye and are thought to underlie geographic differential RPE sensitivity to degeneration. The mouse is one of the most widely used models to study retinal degeneration. A deeper understanding of mouse RPE heterogeneity can facilitate the interpretation of the observed phenotypes in models of retinal degeneration. In this study, we used an AI-based image analysis algorithm (REShAPE) to generate morphometric maps of RPE flatmounts in young adult mice (2 months old). Using cell area, we distinguished three concentric subpopulations of phenotypic heterogeneity. Treatment with sodium iodate, a widely used oxidative agent to model retinal degenerations, induced RPE atrophy of the two centermost subpopulations. RPE morphometric analysis could prove useful when analyzing animal models of retinal degeneration.

Keywords: Cell morphometry; Mouse RPE; Oxidative stress; REShAPE; RPE subpopulations; Regional sensitivity; Retinal degeneration; Sodium iodate.