Aging human retinal pigment epithelial (RPE) cells accumulate lipofuscin and live on: histological basis of quantitative autofluorescence (AF)

Abstract

Purpose: Fundus AF imaging is established in clinical diagnosis and treatment monitoring. The histological basis of fundus AF as well as the validity of the lipofuscin toxicity theory of cell death in aging (Dorey; PMID 2759786) would be furthered by comprehensive data on histological AF and RPE cell density (cells/sq mm) at locations relative to the fovea in young and aged adult eyes.

Methods: RPE-Bruch membrane flat-mounts were dissected from donor eyes without gross chorioretinal pathology (<50 yr and >80 yr; n=10 each group). RPE cytoskeleton and lipofuscin AF were monitored microscopically at up to 90 locations chosen in a systematic and unbiased manner. Maps were assembled from 83,336 cells counted in 1,470 locations. Each cell was represented as a Voronoi region for determining the number of each cell’s neighbors, area, and total AF intensity (a surrogate for lipofuscin mass). Histological AF values were normalized to a quantitative AF standard (Delori; PMID 22016060).

Results: A ring of high AF signal centered on the fovea is visible at the perifovea in all tissues, corresponding to areas of highest rod photoreceptor densities (Curcio; PMID 2324310) and fundus AF (Delori, PMID 11431454, 22016060). AF maps show large variability that is only partly explained by age. RPE cell density peaks at the fovea, independent of age. While AF difference maps (<50 vs >80) highlight increasing perifoveal AF with age, no net RPE cell loss is detectable in the corresponding cell density difference maps. Larger cells had more total AF. A small percentage of cells have ≥8 neighbors consistent with remodeling at all ages.

Conclusions: Conclusion: 1) Lipofuscin accumulation follows rod topography. Cones also participate, as AF of foveal RPE, visualized without screening by macular pigment, is non-zero. 2) The lack of correlation between an age-related increase in lipofuscin AF and RPE numerical density may impact our understanding of lipofuscin’s role in AMD pathogenesis. 3) Stable RPE cell number in aging contrasts with 30% loss of central rods (Curcio PMID 8225863), which are further away from the choroid than the RPE. A vascular insufficiency theory will be offered as an explanation for these results.

Keywords: 413 aging • 701 retinal pigment epithelium • 582 ipofuscin