Epigenetics: hope or hype?
The inaugural Huxley debate was held at the 2013 British Science Festival, to continue the Association’s tradition of capturing the spirit of public debate in an emerging area of science.
The original debate, at the Association’s meeting in 1860, pitted Thomas Huxley (‘Darwin’s Bulldog’) against Samuel Wilberforce (the Lord Bishop of Oxford and a Fellow of the Royal Society) in a debate about the still-tentative matter of evolution.
The 2013 debate saw Tim Spector and George Davey-Smith lock horns on the topic ‘Epigenetics: hope or hype?’
Until recently, we believed that the structure of our genes was by far the most important factor in determining our personality and risk of disease. However, studying identical twins (who are essentially clones, with the same DNA in every cell of their body) has changed that.
Although most traits are strongly heritable at a population level, twins rarely develop the same diseases over a lifetime. For example, Rheumatoid Arthritis is a common auto-immune disease but only 12 per cent of identical twins will develop the same disease as their diseased twin sisters. Low twin similarities ranging from 5-30 per cent are found for most common diseases.
The cause is probably a mechanism called epigenetics, by which chemical signals can alter the function of the genes without altering their basic structure. These signals are produced or altered by changes in our environment or lifestyle. They are usually reversible and can last many years. It used to be thought that they were wiped clean when the DNA was re-formatted to form the next generation, but some signals have been found to pass to the next generation. Epigenetics is one of the most exciting areas in science today and is changing how we view genes and our identity.
I disagree. It has been long recognised that it is environmental and behavioural factors that are responsible for the risk of disease.
In Britain the rise and fall of lung cancer mimics (with a time lag) perfectly the rise and fall of cigarette smoking, and indeed if no one smoked the rates of lung cancer would be very low.
Similarly, changes over time and differences between countries demonstrate that obesity, diabetes, coronary heart disease, stroke, infectious diseases, most cancers, asthma, allergies - indeed most health conditions - are influenced in this way.
Therefore it is no revelation to people who study the causes of disease that identical twins are usually discordant for most common diseases. Contrary to the notion that the ‘cause is probably a mechanism called epigenetics’, at a population level the causes of the differences in disease rates are environmental and behavioural factors that can be changed.
Epigenetic processes will mediate between such factors and the development of disease. Therefore, whilst there are exciting new technologies that allow the investigation of these epigenetic processes, and we may understand ways of preventing and treating disease through studying such processes, epigenetics does not represent in any sense a newly discovered cause of disease.
While it has long been recognised that environmental and behavioural factors MAY be responsible for the risk of disease – epidemiologists have in reality not managed to identify many precise agents that have large impacts. Most environmental agents have only small effects with the exception of smoking. Even in smoking in Europe – while reduction in lung cancer over the last four decades is mirroring reduction in smoking, the reduction in women is less, and around 30% of newly diagnosed lung cancer cases have actually never smoked.
Similarly recent epidemics in allergies, and autism and some autoimmune diseases cannot be explained by equivalent changes in any known environment. So in this state of ignorance it would be very unwise to dismiss so casually new ideas such as epigenetics, or even the possibility that we should be looking more widely for clues – even in previous generations.
The amazing differences in the diseases seen in identical twin pairs cannot be explained by genetics, and neither can it be explained by environment – epigenetic changes are the obvious answer and when we look we see abundant epigenetic differences that are like beacons for the disease.
Dismissing epigenetics because of pre-conceived ideas that we know ‘so-much’ is a big mistake.
Epigenetic processes are indeed likely to mediate between exposures (be they genetic, environmental or essentially stochastic - chance - events) and disease. Indeed in some circumstances this is almost tautologous since by definition processes which recapitulate cellular state across cell division are referred to as ‘epigenetic’.
The question is whether it is meaningful to consider these to be, in a general sense, epigenetic ‘causes’ of disease, since the epigenetic processes will themselves have underlying causes, except when purely stochastic. Thus the population-level causes of disease, which in many cases well account for its distribution, would be the appropriate target for intervention.
Much of the excitement about epigenetics relates to supposed cross-generational influences, for which there is little solid evidence, and much of the little that exists is based on epidemiological associations, with all the well-known problems of reliably identifying causes using observational data.
When new technologies become available there is always a cycle of initial hype and later more sober assessment, and transgenerational epigenetics is certainly currently in the hype phase of this cycle.
I’m glad we now agree that epigenetic changes are crucial in how a wide variety of exposures and even genetic influences can lead to disease. Where we still clearly differ is whether to consider these epigenetic changes of interest to those trying to prevent disease.
You seem to have muddled thinking here as you say (reasonably) that where the cause of the epigenetic change is known – say diet or smoking – we should ignore the epigenetics and focus on changing the exposure. But where chance is involved – which you argue is most of the time (as in identical twins) – you say this is of less interest. I disagree - understanding and reversing epigenetic change is crucial. Currently we can’t yet separate chance effects from unknown exposures without more research –although both are important.
Even not knowing why it occurred – detecting and reversing the epigenetic change, for example on a cancer gene, like BRCA1 in Angelina Jolie could prevent the need for mastectomy in breast cancer. There are already four epigenetic drugs on the market fighting cancer and 40 more in development by Pharma. Epigenetics is NOT just hype even if you disregard ALL the impressive trans-generational animal data.
As I said at the start, ‘we may understand ways of preventing and treating disease through studying such [epigenetic] processes’; this is not in dispute. But in this regard, epigenetics is in no way unique. The same may be said for many other new technologies, including applying such new technologies to the study of metabolites (‘metabolomics’) of micro-organisms (the microbiome) and of the spectrum of external exposures that may influence disease (‘exposomics’).
None of these, though, is hyped as leading to fundamental changes in how we consider biology, unlike the case of epigenetics, when we are repeatedly told that this ‘new approach’ (which, of course, it isn’t) fundamentally challenges how we should think of evolution and inheritance. The bottom line is that it doesn’t.
Quantitatively the inheritance of characteristics across generations generally follows the classical Mendelian schema and within genetically identical organisms differences in their characteristics are not passed down. Shiny new laboratory equipment and new technologies tend to set the pulse racing and such excitement is conveyed in expansive claims that fail to be realised, because of the basic truth that new technologies for studying how the world is do not themselves change how the world is, however much their promoters want us to believe so.
And I’ll stake the royalties on your book Identically different that in 10 years there will not be an ‘epigenetic test’ to individualise dealing with BRCA1 mutations (which increase a woman’s chance of developing breast cancer).
Perhaps the British Science Association could hold the stakes and adjudicate in a decade?