Abstract
Purpose: :
To further annotate the retina transcriptome and thereby facilitate cell–specific gene discovery, expression, and mapping using the next generation of Serial Analysis of Gene Expression (SAGE), LongSAGE. Microarray– and short SAGE–based large scale profiling of the retina has been extremely useful for identifying rod photoreceptor–specific genes but has not yielded the same wealth of gene discovery in the cones or RPE. We are particularly interested in expression profiling in these cells as this may have an impact on our understanding of AMD.
Methods: :
The first longSAGE library of human retina, 4cRet, was generated from 41–66–year–old normal donors to improve the annotation of existing retinal short SAGE data. The 21 bp long tags were mapped directly to the genomic assembly and annotated as ‘sense’ or ‘antisense’ using a novel approach. Retina longSAGE data was compared to existing public longSAGE libraries from other tissues to assay for retina specificity. The longSAGE data was linked to the 8 existing retina and RPE short SAGE libraries and compared to expression in other tissues in the EyeSAGE Database, a tool for analysis of large scale expression data in the eye. This tool was used to infer cell type–associated expression profiles, particularly cone–enriched transcripts. Novel retina–specific transcripts, antisense transcripts of known and novel retina–specific genes, and transcript variants of known retina–specific genes identified by longSAGE were further tested by standard, direction–specific, and real–time quantitative RT–PCR.
Results: :
Annotation of SAGE data was significantly improved with the added length of the longSAGE tags. SAGE data was more accurately mapped to the appropriate inherited dystrophy and disease susceptibility loci. Widespread natural antisense transcription in the retina was detected and further tested in known retinal genes such as PDE6G and rhodopsin. Transcript variants with higher expression than the reference sequences were identified for several known retinal genes such as PDE6G and PDE6A. New cone–enriched candidate genes that were not detected by the existing SAGE profiles were identified, including a candidate whose sense transcript is ubiquitously expressed while the antisense transcript has a retina–specific and cone–enriched expression profile.
Conclusions: :
LongSAGE analysis of retina has enabled detection of transcriptional events not easily detected through other methods and is especially useful in conjunction with existing short SAGE data.
Keywords: retina • gene/expression • photoreceptors