Hereditary breast cancer accounts for a small proportion of total breast cancer cases – with estimates varying up to ten percent – however this minority of cases can provide clues to the far more common non-hereditary form, as illustrated in a recent study in Genome Medicine. Mutations in the BRCA1 and BRCA2 genes are well known to greatly increase the risk of developing breast cancer. Martin Widschwendter from University College London, UK, and colleagues reveal that genome wide epigenetic changes identified in women carrying BRCA1 mutations can be used to predict the chances of developing breast cancer in those lacking these mutations. Here Widschwendter explains how these findings came about, what their implications are, and the next steps to further refine their predictions.
What is the relative prevalence of hereditary versus non-hereditary breast cancer? To what extent do we understand the genetics of breast cancer?
Breast cancer is the most common cancer worldwide in women, with nearly 1.7 million new cases diagnosed each year. Only 5-10 percent of these cases are associated with a strong inherited genetic component, of which BRCA mutations comprise a subset. Other moderate penetrance genes have been identified including RAD51, ATM, CHEK2 and PALB2 for example, but their contribution to the risk of developing breast cancer is far less than that of BRCA mutations. Genome wide association studies have added further to the genetic picture by highlighting an association between specific single nucleotide polymorphisms (SNPs) and breast cancer. However, the gain in power of risk prediction afforded when adding these SNPs to conventionally used clinical epidemiological models has so far proven to be very small. Studies of identical and non-identical twins tell us that approximately 73 percent of the risk of developing breast cancer is due to non-heritable factors (e.g. lifestyle, environment, reproductive factors, etc.).
Your study provides a DNA methylation signature to predict non-hereditary breast cancer. Can you explain how BRCA mutation carriers were employed to achieve this?
As it is known that women with BRCA1/2 genetic mutations have a high lifetime risk for breast cancer development (up to 85 percent), we were interested in looking at ‘epigenetic’ changes in the blood of these women and to see whether these same changes are also seen in non-BRCA1/2 carriers who develop breast cancer (and who have therefore acquired the same ‘epigenetic’ changes via other routes).
Epigenetics describes a system of processes that can alter gene function via the modification of DNA and not as a consequence of underlying sequence mutations. Whilst the genome is referred to as the hardware of our cells, the epigenome can be thought of as the software that can potentially be misprogrammed. We studied the most well described epigenetic event known as DNA methylation which can cause genes to be switched ‘on’ or ‘off’. We used a platform allowing us to simultaneously assess 27,000 of these methylated sites throughout the genome per sample. We derived a methylation signature from BRCA mutation carriers consisting of about 1800 methylated sites that was not present in the blood of non-carriers. This signature was then applied to blood sample sets from women included in the MRC National Survey of Health and Development (NSHD) 1946 birth cohort, and the UKCTOCS ovarian cancer screening trial. In both collections, blood samples were available from women predating their diagnosis of breast cancer. We discovered that this methylation signature is also present in the blood of non-BRCA1/2 carriers and is detectable in advance of disease development.
What were the key findings from your study?
The key findings of the study were the discovery of an epigenetic (methylation) signature in white blood cells of BRCA1/2 mutation carriers that was also present in women without BRCA mutations and who developed breast cancer several years later. Furthermore, the signature was able to identify those women more likely to die from the disease.
In your study you note the tissue-specificity of your DNA methylation signature for the prediction of breast cancer. Do you think certain tissues types may provide more accurate signatures for prediction?
In our study we demonstrated that the epigenetic risk prediction signature was present in the blood but not in buccal cells from the same women, reiterating that DNA methylation changes are tissue specific. This is unsurprising since DNA methylation is crucial for early development and cellular differentiation – all cells with the same genetic code are able to develop into different tissue types owing to the expression of different gene sets, which in turn is made possible by epigenetic modulation. As the cellular source of breast cancer is epithelial, we hypothesise that more powerful risk predicting signatures could be derived from the comparison of epithelial cells between BRCA mutation carriers and non-carriers. Cervical cells for instance could be potentially more suited to this approach as they are both hormone sensitive (hormone exposure is one of the biggest risk factors for breast cancer development) and easily accessible.
Do you think DNA methylation changes in breast cancer is causative or consequential?
In our study we describe a BRCA mutation associated epigenome wide methylation signature in white blood cells affecting numerous other genes. Notably the signature was enriched for genes important in cellular differentiation suggesting that DNA methylation mediated silencing of these genes could result in suppression of immune cell differentiation in women harbouring the signature. The fact that our signature was also associated with the development of other cancers supports the view that the DNA methylation signature might be epigenetically compromising immune-defence against tumour cells in affected women. Further studies are required to confirm whether this is the case.
Do you think this DNA methylation signature could one day be used to inform decisions around preventative mastectomies, in the same way BRCA mutation tests are?
It is far too early to address this question. Further research is required to produce an epigenetic risk predicting test with sufficient power to justify implementation in the clinic. The hope would be that such a test would afford a large enough window of opportunity to allow for preventative measures (both lifestyle and clinical interventions) to prevent breast cancer development in all women.
What’s next for your research?
As well as looking further into the specifics of the blood based epigenetic signature we have discovered, we plan to study the epigenome further with the perspective to predict cancer development – this time employing a more advanced analysis platform allowing for assessment of half a million DNA methylation sites as opposed to 27,000. We plan to specifically look into epithelial cells to determine whether this tissue type offers significantly high enough levels of breast cancer risk prediction to justify the initiation of a clinical trial.