What genes remember

Many geneticists now think that the behaviour of our genes can be altered by experience—and even that these changes can be passed on to future generations. This finding may transform our understanding of inheritance and evolution
May 23, 2008

It has long been known that an organism's fate is not determined by genes alone. This much we can tell by observing identical twins, who over time tend to diverge both physiologically (developing differences in, say, height and posture) and psychologically (exhibiting different personality traits and even, sometimes, sexual orientations). Despite most identical twins having similar diets and lifestyles, subtle cultural and environmental distinctions appear to alter their phenotype—the sum of their nature and nurture. In 1942, Conrad Waddington coined the term "epigenetics" to describe this idea that an organism's experience may cause its genes to behave (or "express themselves") differently. Scientists have found striking examples of epigenetic behaviour in the animal kingdom—in the way, for example, honeybee larvae "decide" whether to become queens or workers depending upon their interaction with other larvae and the environment.

Until recently, it was assumed that the impact of epigenetics was confined to individual organisms, and was not passed on to their offspring. Epigenetics was thought of as the cross-talk between genes and environment, giving individuals some adaptive capability in their lifetimes, but not beyond. Recently, though, scientists have become convinced that there is a form of inheritance, called epigenetic inheritance, in which the behaviour of genes in offspring is affected by the life experience of parents. Furthermore, these epigenetic changes can, at least for a small minority of genes, extend beyond immediate offspring to further generations, although the effects do not appear to last indefinitely. This discovery has a number of potential implications, both good and bad. On the one hand, it may give renewed impetus to health authoritarians and revive the discredited theory of Lamarckism (the idea that how we live alters our genes). On the other hand, it could provide scientists with the means to fill in important gaps in the story of evolution.

There is also the possibility that epigenetic inheritance is implicated in the passing down of certain cultural, personality or even psychiatric traits. For instance, historical "insults," such as Oliver Cromwell's brutal reconquest of Ireland in 1649, have led to an "embedding" of attitudes within the affected communities that persist for generations. However, it has generally been thought that this phenomenon could be explained by Richard Dawkins's theory of memes, according to which cultural or intellectual traits are passed down via non-genetic mechanisms such as storytelling. The possibility raised by epigenetics is that such cultural transmission may, after all, have a genetic component. Could it be that historical traumas, such as transatlantic slavery, leave some kind of genetic mark on the descendants of their victims?

Evidence for epigenetic inheritance

Evidence for epigenetic inheritance in mammals emerged first in animals such as mice, whose short lifespans enabled gene expression changes occurring over several generations to be observed within about a decade. Most recently, the phenomenon has been detected in chickens, in response to stress caused by abnormal levels of light in their environment. Researchers at Linköping University in Sweden reared one group of chickens under normal conditions of day and night, while another was exposed to randomly varying light. The offspring of the latter group, the scientists found, had significantly impaired spatial learning abilities, but were also more aggressive and grew faster. These behavioural characteristics in the offspring were linked to changes in the activity, or expression levels, of 31 genes in the hypothalamus or pituitary gland areas. Elsewhere in the animal, activity of these genes was largely normal, but it was changed in the areas of the brain known to be responsible for behavioural traits such as spatial learning. This exemplified a fundamental characteristic of epigenetic inheritance, which is that the genes themselves are handed down as normal, but their ability to be expressed—and therefore affect some behavioural trait or function—is changed.

Such clear links between epigenetic inheritance and gene expression have not yet been found in humans; this would require multigenerational studies taking at least half a century. We simply live too long. Fortunately, though, there are historical records that provide striking indirect evidence of epigenetic inheritance surviving for at least two generations. One of them, in Britain, was the Avon Longitudinal Study, a survey of children born to 14,000 mothers that took place in the early 1990s. The survey found that of 5,000 fathers who took part, 166 had started smoking very early, in the so-called "slow growth" period before puberty, which for boys is usually between nine and 12. The sons of these fathers tended to be significantly overweight by the time they were nine, but there was no noticeable difference for the daughters. This established a statistically significant link between fathers who smoked during the slow growth period and the above average weight of their sons.

The Swedish data

article body image

These findings may seem unimportant, given that few boys smoke during their slow growth period today. However, more significant findings, over a longer timescale, have emerged from an unlikely source: records of annual harvests, going back 200 years, from an isolated community in northern Sweden. Three geneticists from Umeå University in Sweden have mined this valuable data in studies examining the impact of food availability during the slow growth period in boys on the susceptibility to heart disease and diabetes in children and grandchildren. Some of the results are surprising, and seemingly contradictory, but the studies have yielded the strongest evidence yet of epigenetic inheritance in humans.

The first finding of the Swedish researchers was that grandchildren of both genders descended from paternal grandfathers who experienced some malnutrition during the slow growth period were likely to live longer than average. Correspondingly, grandchildren were likely to die younger if their grandfathers experienced an abundance of food during slow growth. The suggestion that grandchildren are better off if their grandparents go hungry during their formative slow growth period might seem counter-intuitive, but it does have some resonance with research on the benefits of temporary shortages of food during an individual's lifetime. Although research on the benefits of fasting has not been conducted directly on humans, various studies have shown that mice and rats subjected to intermittent fasting live 10 to 15 per cent longer, and are more resistant to a number of diseases relating to metabolism, such as diabetes.

However, on digging deeper into the Swedish data, and homing in on the relationship between food availability and death from cardiovascular disease and diabetes, the researchers found that it was not just a case of scarcity being good and abundance bad. In the case of heart disease, the effect of fathers having abundant food during the slow growth period was still detrimental to children, but with mothers the opposite was true. The children of mothers who were well fed during that period seemed protected against heart disease. And with diabetes, children seemed protected against the disease if their father experienced food abundance during the slow growth period—the opposite of the situation when their paternal grandfathers were well fed, suggesting a seesawing effect across generations.

Out of this seemingly complicated picture, geneticists have been able to extract some clear pointers to the mechanisms and evolutionary significance of epigenetic inheritance. First, the link between paternal grandfathers and grandchildren indicates a link between gender and epigenetic inheritance. The second significant finding was the apparent seesawing effect across the generations in the case of diabetes. This appears to make no sense. Other things being equal, a given change in gene expression should have a similar outcome in all descendants, so the implication of the seesawing effect seems to be that one epigenetic change in turn triggers another. It would help if the identity of the genes involved in epigenetic inheritance were known, because then the actual changes in expression could be studied directly in cells, rather than just through their overall impact on the people concerned.

An important clue in this direction is offered by the related phenomenon of genetic imprinting, another subject that has increasingly interested scientists in the last decade. We have two copies of nearly all our genes, one inherited from each parent (each copy is known as an allele). It has been discovered that a small number of genes—80 are currently known to exist in humans—are imprinted, or "silenced," by one or other parent in such a way that we in effect inherit just one copy. Because we in effect have only one copy of imprinted genes, scientists suspect that they are particularly vulnerable to environmental influence. And it is known that many imprinted genes play pivotal roles in biochemical pathways controlling growth and important metabolic processes, which if tampered with can cause diabetes, cardiovascular disease and other disorders. This strongly suggests that, if scientists want to understand epigenetic inheritance, they should look more closely at genetic imprinting.

Filling in the evolutionary gaps

What are the implications of the existence of epigenetic inheritance in humans? Unfortunately, perhaps one effect will be to provide additional fodder for those eager to regulate our lifestyles. The existence of epigenetic inheritance extending at least two generations down the line may suggest to some that we should take more care of our genome, particularly at formative times such as the early growth period.

The discovery of epigenetic inheritance has also led some to revive the previously discredited theory of Lamarckism, which stated that animals could influence their genes by the way they live. Giraffes, for example, were thought by the theory's architect, Jean-Baptiste Lamarck, to have developed long necks by the very act of stretching for high branches, rather than just through the natural selection of genes conferring this reach advantage. Similarly, children of blacksmiths were thought to have inherited genes for the strong arm muscles developed by their fathers during their lifetime, rather than just building such muscles through adopting the same trade. But this is a red herring, because epigenetic inheritance clearly does not involve rewriting genes. Lamarckism implied that the actions or experiences of an organism could lead to the underlying genes being modified through rewriting of the DNA code. Epigenetic inheritance merely alters the ability of a gene to be expressed in offspring, but leaves the DNA, and the genes, intact. Epigenetic inheritance can readily be reversed, and there is as yet little or no evidence that it persists for longer than a few generations.

Yet it is precisely this short-term range coupled with its ability to respond immediately to environmental cues, that makes epigenetics an invaluable adaptive tool. As we have seen, epigenetic mechanisms enable individual organisms to adapt within their lifetimes, and some of these changes can be passed on to descendants. But equally, some research—including the studies based on the Swedish harvest data indicating a seesawing effect through the generations—suggests that these changes can be undone just as easily, perhaps again in response to environmental cues.

It could turn out that the discovery of epigenetic inheritance will help fill in some of the gaps in evolutionary theory that creationists have exploited to bash Darwinism, adding a third evolutionary mechanism to the two we already know about: mutation and natural gene selection. Mutation is the slowest mechanism of evolution, allowing genes to change and develop variability within a population, but not fast enough on its own for complex slow-growing organisms such as mammals to adapt readily to changing conditions. Mutation can arise through errors in the DNA copying mechanism during cell division, or through external interference such as exposure to radiation. While mutation is of fundamental importance for all life, it is only of direct value as an adaptive tool in rapidly reproducing single-celled organisms such as bacteria. In mammals, including humans, mutation is crucial in developing a versatile gene pool, but it is then up to natural selection to select the appropriate combinations of genes from this pool to suit varying environmental conditions.

Natural selection, then, can be seen as a mid-range mechanism of adaptation, working faster than mutation but not as quickly as epigenetic inheritance. Natural selection still does not help individual animals, and takes a number of generations to kick in, but over time it allows sub-groups to evolve desirable attributes for their environment. It is natural selection, rather than mutation, that has led to racial differences among humans, for example. But even natural selection cannot act fast enough to adapt to more transient environmental changes, and this is where epigenetic inheritance perhaps comes in.

These are early days in epigenetics, but it seems increasingly likely that it will lead to a major overhaul of evolutionary theory. For although epigenetic inheritance does not technically revive Lamarckism, it does in practice mean that we pass on attributes we have acquired through our experiences to our children, and even grandchildren. Above all, it reveals that biology does not just rely on genes for information that determines an organism's fate. At least temporarily, heritable information can reside at a level above the genes, providing a bypass around environmental obstacles.