Re-Evaluate Fusion Genes in Prostate Cancer

Area of Science:

• Genomics
• Cancer Biology
• Bioinformatics

Background:

• Thousands of gene fusions implicated in prostate cancer lack thorough validation of their authenticity, incidence, and tumor specificity.
• Genomic characteristics of these gene fusions remain underexplored.

Purpose of the Study:

• To validate reported gene fusions in prostate cancer using RNA-seq data.
• To explore the genomic characteristics of validated fusion genes.
• To investigate the transcriptomic and DNA methylomic consequences of E26 transformation-specific (ETS) fusions.

Main Methods:

• Development of FusionVet tool for validating gene fusions via RNA-seq alignments.
• Re-assessment of 2727 gene fusions from 36 studies using The Cancer Genome Atlas (TCGA) RNA-seq data.
• Analysis of genomic characteristics and transcriptomic/DNA methylomic consequences of ETS fusions.

Main Results:

• Over 90% of reported gene fusions could not be validated using TCGA RNA-seq data.
• Only 4% of validated fusions were tumor-specific; many occurred in both tumor and normal samples.
• ETS family and androgen response genes were significantly enriched in prostate cancer-specific fusions.
• ERG fusions were associated with repressed transcription and increased DNA methylation, and showed specific co-occurrence/mutual exclusivity patterns with other genomic alterations.

Conclusions:

• The majority of previously reported prostate cancer gene fusions are not supported by TCGA data.
• Prostate cancer-specific fusion genes are significantly enriched for ETS family and androgen response genes.
• ERG fusion is linked to repressed transcription and increased DNA methylation, highlighting its functional impact.