Identification of gene fusion transcripts by transcriptome sequencing in BRCA1-mutated breast cancers and cell lines
1 Department of Human Genetics, McGill University, Room N5-13, Stewart Biology Building, 1205 Dr. Penfield Ave, Montreal, Quebec, H3A 1B1, Canada
2 McGill University and Genome Quebec Innovation Centre, 740 Dr. Penfield Ave, Montreal, Quebec, H3A 1A4, Canada
3 Program in Cancer Genetics, McGill University, 546 Pine Ave, Montreal, Quebec, H2W 1S6, Canada
4 Segal Cancer Centre, Lady Davis Institute, Jewish General Hospital, 3755 Côte-Ste-Catherine Road, Montreal, Quebec, H3T 1E2, Canada
5 The Breakthrough Breast Cancer Research Centre, The Institute of Cancer Research, 237 Fulham Road, London, SW3 6JB, UK
6 Institut Curie, 26 Rue d'Ulm, 75248, Paris, France
7 Department of Oncology, Lady Davis Institute, Jewish General Hospital, McGill University, 3755 Côte-Ste-Catherine Road, Montreal, H3T 1E2, Canada
8 Department of Surgery, Lady Davis Institute, Jewish General Hospital, McGill University, 3755 Côte-Ste-Catherine Road, Montreal, Quebec, H3T 1E2, Canada
BMC Medical Genomics 2011, 4:75 doi:10.1186/1755-8794-4-75Published: 27 October 2011
Gene fusions arising from chromosomal translocations have been implicated in cancer. However, the role of gene fusions in BRCA1-related breast cancers is not well understood. Mutations in BRCA1 are associated with an increased risk for breast cancer (up to 80% lifetime risk) and ovarian cancer (up to 50%). We sought to identify putative gene fusions in the transcriptomes of these cancers using high-throughput RNA sequencing (RNA-Seq).
We used Illumina sequencing technology to sequence the transcriptomes of five BRCA1-mutated breast cancer cell lines, three BRCA1-mutated primary tumors, two secretory breast cancer primary tumors and one non-tumorigenic breast epithelial cell line. Using a bioinformatics approach, our initial attempt at discovering putative gene fusions relied on analyzing single-end reads and identifying reads that aligned across exons of two different genes. Subsequently, latter samples were sequenced with paired-end reads and at longer cycles (producing longer reads). We then refined our approach by identifying misaligned paired reads, which may flank a putative gene fusion junction.
As a proof of concept, we were able to identify two previously characterized gene fusions in our samples using both single-end and paired-end approaches. In addition, we identified three novel in-frame fusions, but none were recurrent. Two of the candidates, WWC1-ADRBK2 in HCC3153 cell line and ADNP-C20orf132 in a primary tumor, were confirmed by Sanger sequencing and RT-PCR. RNA-Seq expression profiling of these two fusions showed a distinct overexpression of the 3' partner genes, suggesting that its expression may be under the control of the 5' partner gene's regulatory elements.
In this study, we used both single-end and paired-end sequencing strategies to discover gene fusions in breast cancer transcriptomes with BRCA1 mutations. We found that the use of paired-end reads is an effective tool for transcriptome profiling of gene fusions. Our findings suggest that while gene fusions are present in some BRCA1-mutated breast cancers, they are infrequent and not recurrent. However, private fusions may still be valuable as potential patient-specific biomarkers for diagnosis and treatment.