Abstract
Germline variants in high-penetrance breast cancer susceptibility genes BRCA1 and BRCA2 are often identified through routine diagnostic gene screening, typically performed for individuals from high-risk breast-ovarian families. Many BRCA1 and BRCA2 variants are known to increase breast cancer risk by disrupting mRNA splicing and compromising the tumour suppressor function of these genes, which work to repair single and double stranded breaks in DNA. A significant number of BRCA1/2 sequence variants are located in or near splice sites and splicing regulatory regions and may disrupt mRNA splicing; however, the relative level to which mRNA splicing is modulated by these common or rare DNA sequence variants has not been ascertained.
Splicing assays undertaken to assess the clinical relevance of rare sequence variants in BRCA1 and BRCA2 typically utilise a PCR-based approach and are able to detect the presence of aberrant isoforms and/or the absence of naturally occurring isoforms. The isoforms expressed are useful for determining the disruptive potential of each variant using the classification guidelines recommended by the ENIGMA (Evidence-based Network for the Interpretation of Germline Mutant Alleles) consortium. However, to date PCR-based assays used to evaluate potential BRCA1 and BRCA2 spliceogenic variants have typically only provided qualitative expression profiles. Thus, detection of quantitative and allele-specific expression changes, which may be associated with variants conferring disease risk, has been limited. Advances in sequencing platforms over the last decade have produced technologies that generate quantitative and high throughput expression information, simultaneously overcoming many of the limitations previously reported for PCR-based approaches in splicing assays.
The work presented here exploits the capabilities of a targeted RNA-seq platform to generate the first comprehensive expression profile of normally expressed BRCA1/BRCA2 mRNA isoforms from lymphoblastoid cell lines (LCLs). Although a high degree of mRNA expression variability was identified across samples, the calculated expression levels made it possible to highlight changes present in rare variant samples outside the expected natural range. Additionally, results from this work identified instances where PCR-based assay primer design prevented isoform detection in variant samples.
Targeted RNA-seq was coupled with allele-specific expression (ASE) analysis to further explore the potential spliceogenic impact of genetic variation in BRCA1 and BRCA2, highlighting ASE changes to natural mRNA isoforms for carriers of a rare variant. While the common variants included in this work were not found to have an obvious impact on splicing, observed allelic imbalances indicate that additional factors are likely to be influencing the mRNA expression variation seen. Exploration of this hypothesis found that common culturing practices, including liquid N2 storage and treatment with a nonsense mediated decay inhibitor, did not impact the mRNA isoforms expressed in a single LCL over time. However, the technology used for mRNA detection was found to play a significant role, with a direct relationship between the number of alternative events detected and the read depth in each sample. Further work into the extent to which cellular heterogeneity contributes to the observed mRNA variability was undertaken with a novel in situ hybridisation platform (RNAscope) to establish the level of variability in mRNA expression between individual cells. The results from this work highlighted how the inter-cell variation in BRCA1 and BRCA2 expression patterns is considerable, potentially explaining why variability is commonly observed when studying mRNA expression at a cell population level. Candidate gene analysis was completed for four patients who have a history of breast cancer but do not carry any disease-associated variants in BRCA1/2. This work did not identify any variants in other known susceptibility genes that are likely to have contributed towards their disease and further investigation into unexplored regions of the genome would be required to identify an underpinning genetic cause.
Many women predicted to be high risk for breast cancer have yet to have their genetic basis successfully identified through genetic testing. The work undertaken here has established a technique to quantitatively assess BRCA1 and BRCA2 mRNA isoforms, while identifying technical and biological factors that influence the observed variability. This comprehensive analysis would benefit patient management as it provides a more informative base for variant classification from a better understanding of how disruptive any given genetic variant is likely to be. Clinicians and genetic counsellors will have the capacity to council patients more effectively as more variants of unknown clinical significance will be able to be given a classification that more informatively highlights their associated genetic risk. These more conclusive genetic test results will mean that patients are also less likely to be subjected to the stress and uncertainty that would otherwise be present with reported variants that remain unclassified.
This work provides the basis for further studies to extend this work to other known breast cancer susceptibility genes, providing a more comprehensive assessment to identify variants that are likely to be influencing disease risk in high risk women. Such data will be critical for the future interpretation of splicing analyses in a diagnostic setting.