Abstract
Purpose :
Uveal melanoma (UM) is the most common primary intraocular cancer found in adults, with a propensity to metastasise to the liver in ∼50% of patients. Due to the lack of effective therapies, this is generally fatal within 2 years of diagnosis. Factors contributing to metastatic spread include loss of chromosome 3, and other specific gene alterations. Patients can be stratified into low (LR) or high (HR) metastatic risk groups and biological differences between these groups may determine potential treatment avenues. In this study, we used transcriptomics from in-house and public domain datasets (The Cancer Genome Atlas (TCGA)) to identify differentially expressed genes/miRNAs and dysregulated signalling pathways that can be targeted using drug repurposing.
Methods :
University of Liverpool (UoL) gene datasets were generated from RNA extracted from 20 UM, 10 LR and 10 HR, hybridised onto Affymetrix GeneChip microarrays. Bioinformatic pipelines were used to identify differentially expressed genes/miRNAs, which were then compared with TCGA datasets from 80 UM samples: 49 LR and 31 HR, using RNA/miRNA-Seq. For this analysis, results were reported with a false discovery rate (FDR) of ≤5% and P<0.05. The library of integrated network-based cellular signatures (LINCS) L1000 database was used to determine relevant drugs.
Results :
UoL and TCGA expression profiling showed a high degree of overlap of differentially expressed genes with 416 genes upregulated and 885 downregulated in the HR c.f LR groups for both datasets. The most differentially expressed upregulated genes included: CYSLTR2, HTR2B, PLN; whilst downregulated genes included: PALMD and MEGF10. miRNA analysis revealed 13 downregulated miRs in both UoL and TCGA datasets including: hsa-miR-509-3p and hsa-1296-5p; with no overlapping upregulated miRNAs. Validation of a subset of genes and miRNA by qPCR with UM patient samples demonstrated concordance with microarray data. UM cell lines,examined with the same panel, revealed variation in expression levels. The top 10 predicted drugs targeting pathways dysregulated in HR UM included mTOR inhibitors and DNA-dependent protein kinase inhibitors.
Conclusions :
Transcriptomic analysis of microarray data, strengthened with TCGA data, identified genes and miRNAs that are dysregulated in UM. Drug repurposing has identified several candidates for future testing in both UM patient samples and cell lines.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.