Abstract
Genomic rearrangements are a hallmark of cancer biology and progression, allowing cells to rapidly transform through alterations in regulatory structures, changes in expression patterns, reprogramming of signaling pathways, and creation of novel transcripts via gene fusion events. Though functional gene fusions encoding oncogenic proteins are the most dramatic outcomes of genomic rearrangements, we investigated the relationship between rearrangements evidenced by fusion transcripts and local expression changes in cancer using transcriptome data alone. 9,953 gene fusion predictions from 418 primary serious ovarian cancer tumors were analyzed, identifying depletions of gene fusion breakpoints within coding regions of fused genes as well as an N-terminal enrichment of breakpoints within fused genes. We identified 48 genes with significant fusion-associated upregulation and furthermore demonstrate that significant regional overexpression of intact genes in patient transcriptomes occurs within 1 megabase of 78 novel gene fusions that function as central markers of these regions. We reveal that cancer transcriptomes select for gene fusions that preserve protein and protein domain coding potential. The association of gene fusion transcripts with neighboring gene overexpression supports rearrangements as mechanism through which cancer cells remodel their transcriptomes and identifies a new way to utilize gene fusions as indicators of regional expression changes in diseased cells with only transcriptomic data.
Highlights
High grade serous ovarian cancer (HGSOC) is the second most common gynecological cancer and the most aggressive form of ovarian cancer[1]
We further identify a novel phenomenon in which expressed fusion events are frequently associated with perturbation in gene expression in the neighborhood of the fusion event, an effect that can extend up to a megabase from the site of rearrangement, consistent with other studies linking more comprehensive genomic data such as whole exome sequencing[13], Single Nucleotide Polymorphism (SNP) array data[14], or whole genome sequencing[15] with altered expression of nearby genes
A MUC16 fusion (MUC16-OR7G2) was detected in only one cell line transcriptome (OVCAR3) (Supplementary Table S2). These results indicate that the presence of fusion transcripts for certain genes is associated with expression changes of the genes directly involved in the fusion events
Summary
High grade serous ovarian cancer (HGSOC) is the second most common gynecological cancer and the most aggressive form of ovarian cancer[1]. HGSOC is a disease of the damaged genome, characterized by widespread copy number alterations, a high frequency of TP53 mutations, but a paucity of dominant acting oncogenes[3]. HGSOC is characterized with gains and losses of chromosome arms or entire chromosomes, and is amongst the most genomically unstable tumors of those that have been studied to date with a median 33% of the genome undergoing copy number alterations[6]. The high level of genomic instability in HGSOC creates a fertile environment for the acquisition of fusion genes which can create neomorphic cancer drivers through the creation of novel domain combinations[10]. Our study utilizes transcriptome data alone to demonstrate remodeling of the transcriptome in HGSOC through gene fusion events that may contribute to the overall fitness of the evolving tumor
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
