epi- | ˈɛpi | (also ep-) prefix 1 upon: epigraph. 2 above: epicontinental. 3 in addition: epiphenomenon. ORIGIN from Greek epi ‘upon, near to, in addition’.
Genetics seems to be part of the public consciousness. We are sensitive to discussions about GM crops or animals; we say that our innermost convictions or tendencies are ‘in our DNA’. Genomics and genomes are themselves quite commonly discussed, either in terms of testing or, recently, in terms of editing. DNA is, quite literally, everywhere. The Nuffield Council on Bioethics has recently published several reports in which genetic themes loom large (see, for example, genome editing and non-invasive prenatal testing). So, a recent Council workshop dedicated to the topic of epigenetics was an interesting contribution to the wider discussion of ethics in the context of human biology, health and imagined futures.
As the dictionary definition above indicates, the prefix ‘epi’ refers to something that lies above or upon an existing structure, offering an additional layer of complexity or content. Whilst even scientists might disagree about the best definition of ‘epigenetic’, it is generally used to refer to the way in which genes and genomes can undergo functional changes that are not associated with changes in DNA sequence. Most commonly, we mean changes either to the chemical structure of DNA – such as DNA methylation (DNAm) – or the structure of the histone proteins that associate with DNA in chromatin. Such chemical changes, or ‘marks’ as they are sometimes known, can act as additional layers of information, reflecting the changing activity of genes. It is not always clear whether they are causally responsible for such activity states, or are by-products. Much depends on context. But scientists can now study the complex and dynamic changes to such marks in a genome-wide fashion – characterising what is known as the ‘epigenome’ – using new technologies. And these have yielded interesting insights into gene regulation in basic research and, more recently, associations between epigenomic variation and human health.
The workshop was divided into two main sessions. The first addressed the use of epigenomics tools to examine a number of phenomena, with several scientists giving brief scene-setting talks. There are a number of groups aiming to identify changing epigenetic biomarkers, or signatures, of ageing, characterising so-called ‘epigenetic clocks’. Related studies aim to establish whether epigenomic signatures may be predictive of the risk of subsequent disease and there are convincing data that this may be the case with smoking-induced cancers. These studies are often based on the examination of DNAm in peripheral blood cells (PBC), a tissue not obviously implicated in any causal mechanism leading to pathologies in tissues such as the lungs; but a PBC epigenomic signature can be viewed as a proxy for epigenomic events elsewhere that are causal and mechanisms have been proposed to account for this. As ever, more data are required. We heard claims that epigenomics can be used to infer an epigenetic age that allows the estimation of ‘time-to-cancer’, ‘time-to-coronary heart disease’ and even ‘time-to-death’. Whilst it is possible that such epigenomic tests could indicate the degree of risk faced by an individual, and could be useful for prompting changes in lifestyle, it is also possible that they may be unreliable, due to the relatively small size of any effect, and may cause unnecessary anxiety. These are, of course, themes recognisable in the context of discussions of genetic testing.
We returned several times to a topic that is of great interest to both scientists and ethicists: the possibility of transgenerational epigenetic inheritance. Briefly, this is the phenomenon by which an epigenetic mark might persist over multiple generations, evading erasure by the normal mechanisms that operate in the germ line and early embryo to re-set the epigenome and prepare it for subsequent events. What excites many about this scenario is the possibility that such an epigenetic mark might be established by exposure to some environmental agent i.e. that it might be acquired during the lifespan of an individual. The use of the word ‘acquired’ brings to mind the ‘acquired characteristics’ evolutionary model of Jean-Baptiste de Lamarck, and so has a whiff of paradigm-shifting science about it.
The only problem is that evidence for such a phenomenon is scant indeed, at least in mammals, and is widely contested. It has been reported as a mechanism that explains very rare types of inheritance in the mouse, and other studies in rodents suggest that it may mediate the long-term effects of environmental stressors, including nutritional and behavioural. Scientists disagree about the interpretation of such data. In humans, some have claimed an association between environmental stressors, such as famine, and disease in subsequent generations of the individuals exposed to the stress. But these are only associations and other explanations (including genetic) can be offered for them. As ever in science, causality can be fiendishly difficult to establish in descriptive scenarios in which direct experimental interventions are not possible or simply unethical.
It is important to note that the environment includes both the natural environment and social environment, the latter characterised by possible impacts from smoking, mobile phone masts and education policy etc., and this may account for much of the interest in epigenetics from social scientists. Several contributors at the workshop noted that if a strong link were to be established between the behaviour of parents and impaired health of subsequent generations, mediated perhaps by the transgenerational inheritance of acquired epigenomic signatures, there was the prospect of increased guilt on the part of individuals who could have avoided such behaviour. On the other hand, one potentially attractive feature of epigenomic states is that, unlike the relative stability of DNA sequence, they may not be permanent. If somehow reversible, the possibility opens up of an improvement in the health prospects of any individual with the undesirable epigenomic signature.
The second session was dedicated to ethical questions arising from epigenetic technologies and their application in real-world situations. We heard of epigenomic data being used as evidence in court cases (guilt has psychological and legal connotations); of studies of epigenomics in the education setting, perhaps as a way of assessing the impact of education policy on individual children, acting as a proxy for educational attainment. These uses of epigenetic biomarkers raise many questions, including whether such data might ultimately be used to classify individuals in some deterministic fashion, actually restricting their horizons rather than opening them up. It would be ironic if ‘genetic determinism’ were to be replaced by ‘epigenetic determinism’ as something to avoid in public policy.
The relationship between genetics and epigenetics, similarities and distinctions, was a constant theme. Most acknowledge that these two causal factors in our biology will interact in complex ways and it will not be easy to disentangle their contributions to our health, disease and longevity, echoing the similarly complex interaction between ‘nature and nurture’. How we think about genetics and epigenetics in the ethical setting is also an interesting challenge. One such setting concerns traditional ethical objections to purposeful alterations to the human germ line; in particular, inheritable edits of the human nuclear genome. One workshop contributor noted that such objections are often based on concerns about the unpredictability of the consequences of such genomic interventions (broadly, safety), and on their persistence over multiple generations in individuals who did not consent to their introduction. But such concerns arise in all instances of human inheritance conceived more broadly. These could include changes to nutritional advice in the public health sphere, evolution of education policy or even planning of the built environment. In all these cases there is the possibility, at least according to some experts, of such environmental impacts causing inheritable effects on the epigenome of affected individuals that might persist across generations, with both direct and indirect negative consequences for individuals and society more generally. And, of course, in all cases there are highly likely to be consequences that are similarly persistent, i.e. inheritable, but that are transmitted culturally. When inheritance is considered in this broad sense, we can see that ‘germline genome editing’ does not obviously raise any special ethical concerns that are not already raised by a myriad of ways in which we attempt to influence or direct human futures.
There was a clear sense at this workshop that hype – ‘epigenohype’, perhaps – is something to avoid. This requires better communication, by scientists and ethicists. Much more research is required to determine whether trans-generational epigenetic inheritance is a reality in humans. More research is also needed to establish the robustness and reliability of epigenomic signatures as predictive biomarkers of human health and disease. Some expert scientists remain sceptical of their utility. But nevertheless, there is an epigenetics industry. We learned of ‘epigenetic pancakes’, ‘epigenetic yoga’ and ‘epigenetic shampoo’. So, it seems there may be a bigger conversation to be had about the societal impact of epigenetics, whether it is always based on sound science or not. Epigenetics could be a big deal.