Taking a chance on epigenetics

Epigenetics has become the indispensable science for understanding the gene-environment relationships that control risk exposure, therapeutic response and disease progression. Despite the importance of the field, not all scientists agree on how even to define epigenetics (Dupont et al., 2009); some, like me, restrict epigenetics to events that happen at the gene level, such as DNA methylation and histone modifications that constitute epigenetic marks, and the binding of proteins and RNAs involved in transcription control. Others also include miRNAs that interact with messenger RNAs. My purpose here is not to discuss semantics, but rather to consider the enormous expectations that we have placed upon epigenetics for its potential to provide quantifiable markers of risk. The notion of risk itself is a matter of significant debate in public health. It relies on our capacity to determine when a cell or tissue has become altered to the point where a disease might have a chance of onset, recurrence, progression or resistance to treatment. Risk is also linked to statistics; individuals who display a known marker of risk are placed in an “at risk” group from which, usually, only some people will actually develop the disorder. Therefore, to address uncertainty associated with risk assessment, the individual “at risk” should be monitored, and we should prioritize the search for underlying mechanisms in order to develop prevention strategies. Next-generation sequencing is currently the basis for large scale epigenetic analyses. But the balance between the investment of large amounts of funds in genetic sequencing and the paucity of resulting new information for cancer knowledge has raised doubt about the overall efficiency of this approach (Yaffe 2013). With the epigenetic code consisting of more than two-dozen different epigenetic marks, the gathering of data for risk assessment would require an exponential use of such costly technologies. Nonetheless, in my opinion, epigenetics is likely to enable great strides in risk assessment because epigenetic marks are rapidly, and sometimes sustainably, modified in response to risk variations. The highly specific nature of the mechanisms that control epigenetic marks might even provide targets for strategies designed to reduce individual risk. Indeed, our query should not be merely whether epigenetics can contribute to risk assessment, but how the information that we learn can and will be used to provide meaningful measures of risk. However, as developed in the next paragraphs, epigenetic information is unlikely to