This article is reproduced with the permission of New Scientist for exclusive use by Nova users.

Rewriting Darwin: The new non-genetic inheritance
09 July 2008
From New Scientist Print Edition.
Emma Young

Half a century before Charles Darwin published On the Origin of Species, the French naturalist Jean-Baptiste Lamarck outlined his own theory of evolution. A cornerstone of this was the idea that characteristics acquired during an individual's lifetime can be passed on to their offspring. In its day, Lamarck's theory was generally ignored or lampooned. Then came Darwin, and Gregor Mendel's discovery of genetics. In recent years, ideas along the lines of Richard Dawkins's concept of the "selfish gene" have come to dominate discussions about heritability, and with the exception of a brief surge of interest in the late 19th and early 20th centuries, "Lamarckism" has long been consigned to the theory junkyard.

Now all that is changing. No one is arguing that Lamarck got everything right, but over the past decade it has become increasingly clear that environmental factors, such as diet or stress, can have biological consequences that are transmitted to offspring without a single change to gene sequences taking place. In fact, some biologists are already starting to consider this process as routine. However, fully accepting the idea, provocatively dubbed the "new Lamarckism", would mean a radical rewrite of modern evolutionary theory. Not surprisingly, there are some who see that as heresy. "It means the demise of the selfish-gene theory," says Eva Jablonka at Tel Aviv University, Israel. "The whole discourse about heredity and evolution will change" (see "Rewriting Darwin and Dawkins?").

That's not all. The implications for public health could also be immense. Some researchers are talking about a paradigm shift in understanding the causes of disease. For example, non-genetic inheritance might help explain the current obesity epidemic, or why there are family patterns for certain cancers and other disorders, but no discernible genetic cause. "It's a whole new way of looking at the inheritance and causes of various diseases, including schizophrenia, bipolar disorder and diabetes, as well as cancer," says Robyn Ward of the cancer research centre at the University of New South Wales in Sydney, Australia.

Lamarck's ideas about exactly how non-genetic inheritance might work were woolly at best. He wrote, for example, of the giraffe's neck becoming elongated over generations because of the animal's habit of stretching up to feed on leaves in high treetops. The recent research, by contrast, has a firm basis in biological mechanisms - in so-called "epigenetic" change.

Epigenetics deals with how gene activity is regulated within a cell - which genes are switched on or off, which are dimmed and how, and when all this happens. For instance, while the cells in the liver and skin of an individual contain exactly the same DNA, their specific epigenetic settings mean the tissues look very different and do a totally different job. Likewise, different genes may be expressed in the same tissue at different stages of development and throughout life. Researchers are a long way from knowing exactly what mechanisms control all this, but they have made some headway.

Inside the nucleus, DNA is packaged around bundles of proteins called histones, which have tails that stick out from the core. One factor that affects gene expression is the pattern of chemical modifications to these tails, such as the presence or absence of acetyl and methyl groups. Genes can also be silenced directly via enzymes that bind methyl groups onto the DNA. The so-called RNA interference (RNAi) system can direct this activity, via small RNA strands. As well as controlling DNA methylation and modifying histones, these RNAi molecules target messenger RNA - much longer strands that act as intermediaries between DNA sequences and the proteins they code for. By breaking mRNA down into small segments, the RNAi molecules ensure that a certain gene cannot be translated into its protein. In short, RNAi creates the epigenetic "marks" that control the activity of genes.

We know that genes - and possibly also non-coding DNA - control RNAi and so are involved in determining an individual's epigenetic settings. It is becoming increasingly apparent, though, that environmental factors can have a direct impact too, with potentially life-changing implications. The clearest example of this comes from honeybees. All female honeybees develop from genetically identical larvae, but those fed on royal jelly become fertile queens while the rest are doomed to life as sterile workers. In March this year, an Australian team led by Ryszard Maleszka at the Australian National University in Canberra showed that epigenetic mechanisms account for this. They used RNAi to silence a gene for DNA methyltransferase - an enzyme necessary for adding methyl groups to DNA - in honeybee larvae. Most of these larvae emerged as queens, without ever having tasted royal jelly (Science, DOI: 10.1126/science.1153069).

For honeybees then, what they eat during early development creates an epigenetic setting that has fundamental lifelong implications. This is an extreme example, but researchers are starting to realise that similar mechanisms are at play in other animals, and even in humans. And, as for honeybees, it seems there is a critical early period during which an individual's pattern of gene expression is "programmed" to a large extent. Environmental factors can feed into this programming, possibly with long-term health impacts.

In 2000, Randy Jirtle at Duke University in Durham, North Carolina, led a ground-breaking experiment on a strain of genetically identical mice. These mice carried the agouti gene, which makes them fat and prone to diabetes and cancer. Jirtle and his student Robert Waterland gave one group of females a diet rich in methyl groups before conception and during pregnancy. They found that the offspring were very different to their parents - they were slim and lived to a ripe old age. Though the pups had inherited the damaging agouti gene, the methyl groups had attached to the gene and dimmed its expression.

Jirtle then tried supplementing the diets of pregnant agouti mice with genistein, an oestrogen-like chemical found in soya. The dose was designed to be comparable to the amount consumed by a person on a high-soya diet, which is associated with a reduced risk of cancer and less body fat. These mice were also more likely to give birth to slim, healthy offspring which had less chance of becoming obese in adulthood. This change was associated with increased methylation of six DNA base-pair sites involved in regulating activity of the agouti gene.

These and other animal studies strongly suggest that a pregnant woman's diet can affect her child's epigenetic marks. So perhaps it is not surprising that the effect of certain nutrients is being called into question. Folate, for example, is a potent methyl donor. It is routinely recommended during pregnancy and added to cereal products in certain countries, including the US, because it reduces the risk of spinal tube defects if eaten around the time of conception. But Jirtle wonders whether it could also be inducing as-yet-unknown, damaging epigenetic effects.

The legacy of stress

Diet is not the only environmental factor that can influence the epigenetic setting of some genes. Michael Meaney at McGill University in Montreal, Canada, and colleagues have found that newborn mice neglected by their mothers are more fearful in adulthood - and that these mice show much higher than normal levels of methylation of certain genes involved in the stress response. On a brighter note, these mice also point the way to a possible way to reverse epigenetic changes (see "In sickness and in health").

In humans, too, there are troubling hints that damaging experiences early in life, while the brain is still developing, can affect epigenetic settings, perhaps with catastrophic consequences. In May, Meaney and his colleagues reported a study of 13 men who had committed suicide, all of whom had been victims of child abuse. They showed clear epigenetic differences in their brains, compared with the brains of men who had died of other causes. It is possible that the changes in epigenetic marks were caused by the exposure to childhood abuse, says the team. Could the changes have contributed to their suicides too?

There is recent evidence that abnormal epigenetic patterns play a role in mental health disorders. In March, Arturas Petronis at the Centre for Addiction and Mental Health in Toronto, Canada, and colleagues reported the first epigenome-wide scan of post-mortem brain tissue from 35 people who had suffered from schizophrenia. They found a distinctive epigenetic pattern, controlling the expression of roughly 40 genes (The American Journal of Human Genetics, vol 82, p 696). Several of the genes were related to neurotransmitters, to brain development and to other processes linked to schizophrenia. These findings lay the groundwork for a new way of understanding mental illness, says Petronis, as a disease with a significant epigenetic component.

As with the people who had committed suicide in Meaney's study, these epigenetic marks may have arisen during development. Yet there are also hints that the people with schizophrenia might instead have inherited them from their parents - and that they in turn might pass the marks on to their own children. In theory, epigenetic marks are wiped clear between generations in mammals. Intriguingly, though, the abnormalities in DNA methylation in Petronis's subjects were not restricted to their frontal cortex: they were also present in their sperm. "[This] suggests that it is possible that inherited epigenetic abnormalities may be contributing to the familial nature of schizophrenia and bipolar disorder," says team member Jonathan Mill at the Institute of Psychiatry at King's College London.

This work is only suggestive, but when it comes to cancer, the evidence is stronger. Some colorectal cancers are known to develop when a key DNA-repair gene called MHL1 becomes coated in methyl groups, preventing it from working. In 2007, Ward and her colleagues published a study of a woman with this type of cancer and her three children. The MHL1 gene was active in two of the children, but one son had a heavily methylated, silenced gene like his mother (The New England Journal of Medicine, vol 356, p 697).

The paper caused a sensation among cancer researchers because it suggested an entirely new way in which disease risk might be inherited. Of course the finding could have been a coincidence, or the son might have inherited a genetic propensity to methylation of this gene, rather than the epigenetic mark itself. Since the paper came out, though, direct inheritance is starting to look more likely. Other teams have identified similar families, and in all cases the effect seems to be transmitted down the maternal line via the egg. The MHL1 gene in the sperm of affected men appears normal.

Some epigenetic marks may also be inherited from fathers, however. In a now classic study published in 2005, Matthew Anway at the University of Idaho in Moscow and colleagues showed that male rats exposed to the common crop fungicide vinclozolin in the womb were less fertile and had a higher than normal risk of developing cancer and kidney defects. Not only were these effects transmitted to their offspring, they were passed from father to son through the three following generations as well (Science, vol 308, p 1466). The team found no DNA changes, only altered DNA methylation patterns in the sperm of these rats, suggesting that epigenetic factors were to blame.

The following year, a team at the University of Maryland in Baltimore found that male mice that had inhaled cocaine passed memory problems onto their pups. Again, their sperm showed no apparent DNA damage, but in the seminiferous tubules, where sperm are produced, the researchers found changes in the levels of two enzymes involved in methylating DNA.

In people, too, there is evidence that environmental impacts on fathers and mothers can produce changes in their children. This has led some researchers to consider a startling possibility. Could the current epidemic of type II diabetes and obesity in developed countries be related to what our parents and our grandparents ate?

Nutrition does seem to have some lasting effect, according to a study by Marcus Pembrey of the Institute of Child Health at University College London and his colleagues. They analysed records from the isolated community of Överkalix in northern Sweden and found that men whose paternal grandfathers had suffered a shortage of food between the ages of 9 and 12 lived longer than their peers (European Journal of Human Genetics, vol 14, p 159). A similar maternal-line effect existed for women, but in this case by far the biggest effect on longevity of the granddaughters occurred when food was limited while grandmothers were in the womb or were infants. It would appear that humans thrive on relatively meagre rations, and the team concluded that under these conditions some sort of key information - perhaps epigenetic in nature - was being captured at the crucial stages of sperm and egg formation, then passed down generations.

Pembrey's team also looked at more recent records from the UK, collected for the Avon Longitudinal Study of Parents and Children. They identified 166 fathers who reported starting smoking before the age of 11 and found that their sons - but not their daughters - had a significantly higher than average body mass index at the age of 9.

Also in 2006, Tony Hsiu-Hsi Chen at the National Taiwan University in Taipei and colleagues reported that the offspring of men who regularly chewed betel nuts had twice the normal risk of developing metabolic syndrome during childhood. Betel nuts are also associated with several symptoms of metabolic syndrome in chewers including increased heart rate, blood pressure, waist size and body weight.

The mother's nutrition might affect a child's risk of obesity, too. Women in the Netherlands who were in the first two trimesters of pregnancy during a famine in 1944 and 1945 gave birth to boys who, at 19, were much more likely to be obese.

All these results raise an important question. Why should factors like food intake or smoking around the time sperm or eggs are created, or at the embryo stage, have such an influence on a child's metabolism and weight?

Extended periods of too much or too little food might trigger a switch to a pattern of gene expression that results in earlier puberty and so earlier mortality, says Pembrey - and this might be heritable. "The reason why some people gain weight more easily is because their metabolic genes are used differently," says Reinhard Stöeger at the University of Washington in Seattle. He suggests that long before the emergence of modern humans, a network of metabolic genes evolved that was honed for a relative scarcity of food, but not feast or famine. "These genes have become epigenetically programmed during the early stages of life in response to adverse environmental conditions - such as feast. This might explain the current epidemic of type II diabetes and obesity in the west, where food is plentiful." Prolonged epigenetic silencing in response to the environment might also lead to a DNA change that "locks in" epigenetic marks, Stöeger suggests.

Out of the melting pot of recent findings, a host of fundamental questions are now being thrown up. If what we eat could affect our grandchildren, should we be more careful? If so, in what ways? Should we be more concerned about the long-term impact of war or child abuse? Could we choose a diet to reduce our own cancer risk, and that of our children? We are only starting to get an inkling about how to answer these, but one thing is clear: genes are only part of the story.

From issue 2664 of New Scientist magazine, 09 July 2008, page 28-33

Rewriting Darwin and Dawkins? The realisation that individuals can acquire characteristics through interaction with their environment and then pass these on to their offspring may force us to rethink evolutionary theory. While examples of this "transgenerational epigenetic inheritance" are only just emerging in mammals, there is long-standing and widespread evidence for it in plants and fungi. That may explain why botanists are much more ready to acknowledge and promote the idea that epigenetic inheritance has a significant role in evolution, whereas zoologists are generally reluctant to do so, says Eva Jablonka from Tel Aviv University, Israel. That looks set to change. "There was a trickle of findings of epigenetic inheritance in animals through the 20th century, and it is turning into a flood about now," says Russell Bonduriansky, at the University of New South Wales in Sydney, Australia. One of his favourite recent examples involves the water flea, daphnia. When predators are around, the fleas develop large, defensive spines. If they then reproduce, their offspring also develop these spines - even when not exposed to predators. For Bonduriansky, this suggests a possible adaptive function of epigenetic inheritance - the fine-tuning of an individual to short-term variations in its environment. "There's no lag time for the offspring to respond to the environment on their own," he says. The idea that epigenetic variation could be adaptive - rather than a form of random, non-directed variation - is very controversial, harking back as it does to the discredited theory of Lamarckian evolution. Nevertheless, this has not deterred some researchers from exploring the full implications of epigenetic inheritance. For example, there is evidence that epigenetic changes can affect mate preference. Last year, David Crews and Andrea Gore at the University of Texas at Austin published a study of male rats whose great-grandfathers had been exposed to the fungicide vinclozalin. Previous research has revealed that such exposure leads to increased infertility and higher risks of cancer even four generations later. Crews and Gore found that female rats tended to avoid these males. They could sense something was wrong, says Gore. The females seemed to select mates on the basis of an epigenetic pattern, as opposed to a genetic difference, she adds. Back to the future For Bonduriansky the accumulating evidence calls for a radical rethink of how evolution works. Jablonka, too, believes that "Lamarckian" mechanisms should now be integrated into evolutionary theory, which should focus on mechanisms, rather than units, of inheritance. "This would be very significant," she says. "It would reintroduce development, in a very direct and strong sense, into heredity and hence evolution. It would mean the pre-synthesis view of evolution, which was very diverse and very rich, can return, but with molecular mechanisms attached." That needn't necessarily mean an end to the idea of the gene as the basic unit of inheritance, or Richard Dawkins's selfish gene, according to some. "I don't think it violates the basic concept that Dawkins articulated," says Eric Richards, at Washington University in St Louis, Missouri. "Epigenetic marks can also be viewed as part of that basic unit in a more inclusive definition of a gene," he says. What does Dawkins himself think? "The 'transgenerational' effects now being described are mildly interesting, but they cast no doubt whatsoever on the theory of the selfish gene," he says. He suggests, though, that the word "gene" should be replaced with "replicator". This selfish replicator, acting as the unit of selection, does not have to be a gene, but it does have to be replicated accurately, the occasional mutation aside. "Whether [epigenetic marks] will eventually be deemed to qualify as 'selfish replicators' will depend upon whether they are genuinely high-fidelity replicators with the capacity to go on for ever. This is important because otherwise there will be no interesting differences between those that are successful in natural selection and those that are not." If all the effects fade out within the first few generations, they cannot be said to be positively selected, Dawkins points out.

In sickness and in health Epigenetic abnormalities have been found in nearly every type of cancer and in other diseases, such as cardiovascular disease. But the discovery that diseases can be caused by environmental factors influencing the expression of genes has an upside. "The beauty of any epigenetic modification is that it is reversible by drugs," says Robyn Ward from the University of New South Wales in Sydney, Australia. Take the epigenetic marks acquired by mice as a result of maternal neglect during infancy. Here, methyl groups become attached to genes involved in the stress response, resulting in heightened anxiety. But, using drugs, Michael Meaney at McGill University in Montreal, Canada, and his team have reversed the methylation of these genes and their associated behavioural responses in adulthood (Journal of Neuroscience, vol 25, p 11045). They injected the drugs directly into the brain although it is possible that a special diet could do the same trick, Meaney says. NEW ROLE FOR OLD DRUGS Other drugs that influence methylation are now in early-stage anti-cancer trials. Some of them are not new, but are being reassessed in the light of new knowledge about how they work. Azacytidine, for example, which was used years ago with limited success to treat a range of bone-marrow stem-cell disorders, is undergoing trials again on these very same disorders. Now that it has become clear the drug induces epigenetic changes, researchers are altering doses and redesigning trials with the aim of activating tumour-suppressor genes that have been silenced by methylation. This approach does have a major drawback - epigenetic drugs are not specific. Side effects, such as nausea and diarrhoea, are probably down to their broad range of action, says Ward. It might be possible to target drugs more specifically, but that is a very long way off. Still, the fact that it offers a whole new way of treating disease leads many to consider the epigenetics approach to be very promising.

For the latest from New Scientiist visit www.newscientist.com