Microarray Analysis of Laser Dissected Photoreceptor and RPE Cells as a Tool to Select Novel Positional Candidate Retinal Disease Genes

Abstract

Purpose:: To identify new positional candidate genes for human retinal diseases using celltype-specific gene expression of photoreceptors and retinal pigment epithelium (RPE) cells from healthy human eyes.

Methods:: Using laser microdissection, we specifically isolated photoreceptors and RPE cells from three healthy human donor eyes. Amplified RNA from both cell types was hybridized to three Agilent 44k microarrays and to a custom designed 22k microarray enriched for genes expressed in either RPE, neural retina or brain. Candidate genes were bioinformatically selected based on their expression levels in photoreceptors or RPE, their chromosomal map position in the locus of a mapped but unidentified retinal disease gene, and a functional Gene Ontology (GO) clustering.

Results:: This study presents candidate genes for 49 retinal disease loci on 18 different chromosomes. Using our selection strategy, we reduced the average number of candidate genes per locus from 243 positional candidate genes to 7 genes per locus. Validation of our approach comes from the identification of the causative gene in the analysis of four loci with previously identified genes (AIPL1, CRB1, GUCY2D and NRL).

Conclusions:: We describe a microarray and bioinformatic-based approach for the selection of novel candidate genes for retinal diseases. We demonstrate a more than thirty-fold reduction in the number of candidate genes to be considered in the positional identification of hereditary retinal disease genes. This approach will facilitate cost-effective candidate gene screening and allow the identification of new retinal disease genes.

Keywords: candidate gene analysis • photoreceptors • retinal pigment epithelium