Abstract
Purpose::
Photoreceptor cell fate determination in the developing retina is a well-studied area, due not only to its potential therapeutic application to treat retinal degeneration, but also because the developing retina has long been considered a good model of CNS development. There are a number of published studies that have profiled changes in gene expression during normal retinal development. However, the large number of genes profiled at comparatively few time points or conditions has made extraction of gene networks from this data extremely difficult.
Methods::
To begin to address this issue we have begun using a ‘seed-network’ to query multiple large-scale expression data sets. The seed network used was composed of genes with a published relationship to rod-photoreceptor differentiation. The genes included Rb, Cyclin D1, Cdk 4 and 6, Nrl, Nr2e3, Rhodopsin, NeuroD1, Crx, Chx10, and Otx2. The seed-network was constructed based on phosphorylation, binding, transcriptional activation and expression studies published over the last 10 years. However, many of the ‘links’ in the seed-network were confirmed by querying high throughput expression data. To identify additional network components (individual genes or signaling network modules) we first identified genes whose expression is highly correlated, in multiple data sets, to genes already present in the network. By analyzing these gene lists using GO annotations, KEGG pathway data, literature mining and biological intuition we have identified and prioritized multiple candidate genes and signaling pathways for further investigation.
Results::
Several genes and signaling pathways identified by this approach have already been shown to be prevalent in the developing retina. Pathways identified include wnt/frizzled, bmp/smad and notch signaling. Several other signaling pathways have been identified as good candidates for further investigation including IGF1/IGF2, VEGF/HIF and N-Myc downstream regulated genes. Our preliminary experimental wet-lab results have identified an effect of IGF-1 on retinal development that acts via MAPK signaling.
Conclusions::
We have demonstrated that the use of an experimental based seed-network is an effective strategy for querying large-scale expression analysis and may be useful for generating hypothesis-based experiments using ‘omics’ data.
Keywords: photoreceptors • gene/expression • proteomics