Abstract
Background
Genomic instability is a critical feature of breast cancers that manifests in genome-wide copy number aberrations (CNA), often causing “gene breakage” and the generation of fusion genes. We aimed to identify aborted transcripts with underlying CNAs and to investigate the molecular landscape of breast cancers harbouring such events.
Methods
A walking student’s t-test algorithm was applied to Affymetrix Exon 1.0ST array data of 123 breast cancers to identify regions of aborted transcription and overlaid with DNA breakpoints derived from matched Affymetrix SNP6 ASCAT-segmented copy number. Aborted transcripts were investigated as potential fusion gene partners through RNA-seq analysis of 151 breast cancer samples (TCGA) and 51 breast cancer cell lines using ChimeraScan. Clinical correlates were established for clinicopathological features, measures of genomic instability, and gene expression-based molecular classifiers including PAM50, TNBCtype, IntClust subtypes and immune signatures.
Results
One hundred and six genes with recurrent CNA-induced aborted transcription were identified. Aborted transcription showed hormone receptor subtype-specificity for 7 genes (nTNBC=1, nNon-TNBC=6) and were less prevalent in samples of IntClust 2 and IntClust 4 subtypes (p: 0.0043, 0.0011). Aborted transcripts was more frequently observed in samples with greater copy-neutral loss of heterozygosity (p=0.012) while aborted transcription of 54/106 genes significantly affected enrichment of 27 tumor-infiltrating lymphocyte subpopulations. Of the 106, 14 aborted transcripts were found as a fusion gene with one partner in RNAseq of TCGA and breast cancer cell lines, while 19 were involved in multiple fusion events (range=1-6, median=2). Notably, FOXP1, localised to a tumour suppressor locus at 3p14.1, reported the highest number of fusion configurations (n=6) with concurrent aborted transcription across all RNAseq datasets (nPRADA=9, nTCGA=38, nBCCL=6). 9/106 genes displayed gene breakage and fusion events exclusively in samples with an enriched tandom duplication phenotype.
Conclusion
CNA-induced gene breakage may affect the molecular landscape of breast cancers, especially by affecting genes located in tumour suppressor loci.
Genomic instability is a critical feature of breast cancers that manifests in genome-wide copy number aberrations (CNA), often causing “gene breakage” and the generation of fusion genes. We aimed to identify aborted transcripts with underlying CNAs and to investigate the molecular landscape of breast cancers harbouring such events.
Methods
A walking student’s t-test algorithm was applied to Affymetrix Exon 1.0ST array data of 123 breast cancers to identify regions of aborted transcription and overlaid with DNA breakpoints derived from matched Affymetrix SNP6 ASCAT-segmented copy number. Aborted transcripts were investigated as potential fusion gene partners through RNA-seq analysis of 151 breast cancer samples (TCGA) and 51 breast cancer cell lines using ChimeraScan. Clinical correlates were established for clinicopathological features, measures of genomic instability, and gene expression-based molecular classifiers including PAM50, TNBCtype, IntClust subtypes and immune signatures.
Results
One hundred and six genes with recurrent CNA-induced aborted transcription were identified. Aborted transcription showed hormone receptor subtype-specificity for 7 genes (nTNBC=1, nNon-TNBC=6) and were less prevalent in samples of IntClust 2 and IntClust 4 subtypes (p: 0.0043, 0.0011). Aborted transcripts was more frequently observed in samples with greater copy-neutral loss of heterozygosity (p=0.012) while aborted transcription of 54/106 genes significantly affected enrichment of 27 tumor-infiltrating lymphocyte subpopulations. Of the 106, 14 aborted transcripts were found as a fusion gene with one partner in RNAseq of TCGA and breast cancer cell lines, while 19 were involved in multiple fusion events (range=1-6, median=2). Notably, FOXP1, localised to a tumour suppressor locus at 3p14.1, reported the highest number of fusion configurations (n=6) with concurrent aborted transcription across all RNAseq datasets (nPRADA=9, nTCGA=38, nBCCL=6). 9/106 genes displayed gene breakage and fusion events exclusively in samples with an enriched tandom duplication phenotype.
Conclusion
CNA-induced gene breakage may affect the molecular landscape of breast cancers, especially by affecting genes located in tumour suppressor loci.
| Original language | English |
|---|---|
| Journal | NCRI Conference |
| Publication status | Published - 2016 |
