A bioinformatics approach to identify molecular pathways that characterize the trabecular meshwork

Abstract

Purpose : The pathogenesis of primary open angle glaucoma (POAG) has not been clarified and its treatment is still not based on insights in the underlying molecular mechanisms. We performed in silico pathway analysis based on gene expression datasets of healthy human trabecular meshwork (TM) cells. Identification of the active pathways in this tissue contributes to a better insight in the molecular processes and may provide candidate target genes for new treatment options of POAG.

Methods : A search for datasets was conducted in Gene Expression Omnibus and ArrayExpress. We used five microarray datasets (GSE27276, GSE65240, GSE6298, GSE37474, and GSE16643; n=23 individuals), which contained gene expression profiles of healthy trabecular meshwork cells. All data were subjected to a thorough quality control based on ArrayAnalysis.org and R scripts. Divergent samples were removed from the datasets. Pathway analysis and visualisation was conducted with PathVisio using a curated pathway collection from WikiPathways. Ubiquitous pathways (pathways that are common pathways in almost all body tissues) were excluded based on data from TissueAnalyzer. From each dataset, the top 25% highest expressed genes were identified and used for pathway overrepresentation analysis. The top 20 Z-score ranked pathways from the pathway analysis of each dataset were selected as these are likely the most relevant TM active pathways.

Results : Seven ubiquitous pathways were found in the pathway analysis and excluded. One dataset (GE6298) expressed none of these common pathways and did not show any correlation with the results of the other four datasets. Therefore, this dataset was removed from further analysis. Pathways that were expressed in at least three out of four datasets were considered to be active, non-ubiquitous pathways of the healthy TM. Four pathways were significantly expressed in all the datasets and four pathways in three out of four datasets. Several of these pathways have been associated with processes in TM, for example the pathways TNF alpha signalling (WP231), trans-sulfuration (WP2333), and miRNA targets in extracellular matrix and membrane receptor (WP2911).

Conclusions : We identified eight non-ubiquitous pathways that appear to be active in healthy human trabecular meshwork tissue.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.