Did you know, a mother’s love is so strong that the power of her kiss can be felt many generations after it happens? It is true of rat mother love, at least. In 2004 Michael Meaney's group published the results of a study showing that the nurturing behaviour of a mother rat brings about physical changes in her babies that are subsequently transmitted to grandchildren too. It is a fascinating example of an epigenetic effect – a change that is passed across cellular or organismal generations, even though there is no change to any DNA sequence.
It used to be conventional wisdom that such changes are impossible. According to ‘Weismann’s barrier’ you can only pass acquired traits on to your children if those traits are acquired in virtue of a change to your DNA. Mere phenotypic alterations, such as occur when a blacksmith increases his muscle mass through frequent use, are not passed on. Jean Baptiste de Lamarck was scorned by the architects of the ‘New synthesis’ of evolutionary theory for proposing that inheritance could be extra-genetic.
And yet rat mothers bring about heritable changes in their babies, without altering their DNA. There is natural variation in the extent to which rat mothers are affectionate to their offspring. Some mothers spend a great deal of time licking and grooming their pups, while others mostly ignore them. We know that the behaviour has important consequences for the phenotype of the adult progeny. Rats raised by attentive, affectionate mothers grow up into calmer adults, better able to tolerate stress, than rats raised by negligent mothers. What’s more, these rats will go on to birth pups that are themselves calmer, and more able to tolerate stress, than the pups of rats raised by negligent mothers, even when the second generation are all raised by negligent mothers. The mother passes on stress-resistance to her grandchildren, even when they are unable to enjoy the actual nurturing behaviour directly. What is going on?
Meaney’s investigations revealed that the mother’s nurturing behaviour brings about a particular pattern of ‘methylation’ to a part of their DNA that affects regulation of the stress response. DNA becomes methylated when certain molecules (‘methyl tags’) get attached to it. These tags affect the shape into which the chromatin (the triple helix) is curled up, which in turn affects which genes are accessible to the cellular machinery which transcribes them. In other words, methyl tags can change what the DNA does, without changing the actual sequence of the DNA. What's more, some of these tags survive the processes of meiosis and fertilisation. In other words, when a well-licked rat becomes pregnant with pups, the methyl tags get copied along with the mother’s genes.
‘Epigenetics’ is an umbrella term for the study of all extra-genetic inheritance mechanisms, although it usually refers to inheritance that is cellular (rather than, say, via culture or the transmission of symbionts). Jablonka and Lamb’s spectacular 2005 book ‘Evolution in Four Dimensions’ detailed various known examples of cellular mechanisms by which particular states of gene activity can be maintained and transmitted across generations, including self-sustaining feedback loops, histone modification, prions and RNA silencing, along with methylation. It’s gripping stuff. Prions, for example, (famous as the agent of CJD and recently implicated in Alzheimers) are simply proteins which have a very particular structure, such that if they come into contact with another protein of the right sort – if they touch it – they cause the other molecule to take on their own deformed shape. It is cellular transmission via a horrible game of ‘tag’.
Mechanisms of epigenesis make multicellularity possible: they allow our cells to specialise for completely different tasks, even though they all carry the same DNA. A 'maternal effect', is an epigenetic effect that is added to an organism's DNA by its mother, at some point during its development, and which may or may not be passed on to that organism's own offspring. Most epigenetic signals are bleached out, reset, by the time a viable zygote is produced. But some slip through the net, with serious potential implications for inheritance. And because mothers, in particular, have disproportionately many opportunities to affect the epigenetic signals sent to her offspring – by reinstating or adding new marks in the womb, by silencing signals left by the father, and by causing new marks to be produced by the way she interacts with her young after their birth – epigenetics has serious potential implications for mothers.
Emma Whitelaw achieved some of the foundational work in the field of epigenetic maternal effects. Her team discovered that female mice can pass on a phenotype known as 'agouti' to their offspring. The agouti condition, which causes mice to be yellow, obese, and susceptible to tumours, is transmitted to offspring via a methyl tag. But the heritability of the condition depends on the mother’s diet, specifically on whether it is high in the particular chemicals from which the methyl tag is constituted. In other words, what the mummy eats while pregnant determines whether her babies are fat, diabetic and get cancer. The advice to women to take folic acid before they conceive is based on similar phenomena. Spina bifida is affected by some genes whose action is modulated in normal foetuses by molecular tags made out of folic acid. If the tags don't go on, the spine doesn't get built properly.
Another study found that a female Mongolian gerbil who develops in a uterus alongside a majority of male siblings will exhibit higher than average levels of aggression and territoriality as an adult, as a consequence of the high levels of testosterone she is exposed to. Furthermore, she will give birth to male-biased litters herself, making this a self-perpetuating phenomenon.
There are proscriptions against experimenting on human mothers, but disasters have served as valuable natural experiments, providing correlational data about the effects of stress and nutritional deprivation on unborn children and their descendants. For example, 40,000 women were known to be pregnant during the Dutch famine of 1944. Under the Nazi blockade the average calories consumption in April was just 500 calories per day. By May supplies were restored. Subsequent studies found that adults who were gestated during this period were subsequently found to have lower methylation of a particular gene (insulin-like growth factor 2 – IGF2) than their siblings. This gene increases insulin resistance, which may be a good thing in times of scarcity, since it helps us to get more energy out of our food. But, raised on normal diets these adults showed elevated levels of obesity, diabetes and clogged arteries. What’s more, when the Dutch famine kids grew up and had children of their own, those babies were born at lower birth weight and were shorter than average, as if primed to grow up in famine conditions themselves.
Epigenetic effects are widespread across species and are widely credited with having adaptive effects. A developing plant can use epigenetic effects to fine tune its phenotype to local ecological conditions by paying heed to up-to-the minute cues provided by its parent, rather than depending on its slow-responding genetic storehouse of information about what was adaptive in past environments, as argued by Shea, Penn and Uller in 2011. However, this sort of plasticity will only be adaptive when the cues provided are accurate - when the organism is able to prepare itself appropriately.
The human analog of this has been termed the 'Fetal programming hypothesis'. First proposed by David Barker in 1998, this theory states that epigenetic marks are used by the gestating baby's developmental systems as cues, giving it advance notice about what the world is going to be like after it is born. If the mother is highly stressed then the baby anticipates a stressful world and calibrates its systems accordingly. If she is underfed then the baby anticipates that it will struggle to eat, and it calibrates its systems accordingly. The mismatch that occurs if the world experienced is very different from what was anticipated on the basis of the epigenetic cues can be problematic – and long lasting.
Particular attention has been focused on maternal diet and maternal mood, as predictors of disorders in the adult children. Correlational evidence connects high sugar maternal diets to obesity and heart disease in children, via epigenetic changes in the leptin gene. High fat maternal diets correlate with obesity, insulin resistance and chronic inflammation, while undernutrition correlates with cardiovascular disease and type 2 diabetes. Maternal deficiencies of iodine and vitamin D may lead to cognitive retardation and autonimmune disorders such as multiple sclerosis in offspring. Maternal stress, on the other hand, has been linked to decreased robustness to stress in children (here is a 2016 review). Children of women borne after their mother experienced a significant beareavement, or survived a war, are more likely to suffer schizophrenia in adulthood. Other proposed triggers of disorders in adult offspring include environmental toxins,domestic abuse, and even temperature fluctuations caused by weather shocks. Economic cohort studies have uncovered possible epigenetic effects on IQ, educational attainment, income, longevity and more.
Blame the mother
Did you know that Victorian doctors often put pregnant women on starvation diets in order to minimise the birth weight (and so facilitate the passage) of their babies? Or that, much more recently, women were encouraged to take up smoking, to the same end? Jena Pincott’s very readable ‘Do chocolate lovers have sweeter babies?’ catalogues various fascinating ways in which medical pronouncements on optimal pregnancy have see-sawed over the years. In recent years, there have been rumblings from the medical community about the dangers of maternal obesity, but go back a mere generation or two and women were ‘eating for two’, fed the choicest morsels and generally encouraged to lounge about like overweight walruses.
One thing seems to have remained constant throughout all this schizophrenic flip-flopping, however: it is always women being lectured to about what they ought or ought not be doing for the good of the baby. There is rarely so much discussion of what society, or governments, or fathers, or medical professionals ought or ought not be doing for the good of the baby. History is littered with now-falsified theories which lay the blame for some malaise or dysfunction solely at the feet of the mother: autism was said to be caused by callous mothering, miscarriage was said to be induced by insufficient maternal enthusiasm, the ‘Elephant Man’s unfortunate deformity was believed to be a consequence of his mother getting scared by an elephant while pregnant. The now discredited 'maternal impressions' theory held that what a pregnant woman thought about and even looked at would have profound consequences for her young, resulting in all manner of taboos and proscriptions of maternal activities.
In some quarters there is fierce resistance to this constant and disproportionate singling out of mothers for invasive and unsympathetic blame.Women are sick and tired of the way their behaviour while pregnant with or caring for young children is scrutinised, criticised and sermonised by one and all. Just witness the response to Jamie Oliver’s recent advocacy of breast feeding. Certainly it is time that we offset some of our constant reprimanding and lecturing about what mothers ought to do with some serious consideration about what governments and societies can do to make lighter the enormous burden that mothers bear.
Against this background the field of maternal epigenetics looks set to ignite a whole new chapter in the history of mothers being blamed for anything that is wrong with their children. I was lucky enough to have a childhood full of cuddles from two very affectionate parents, and there is much joy to be found in the thought that every cuddle I give to my son might bring about physical and long-lasting improvements to him, and even influence the grandchildren I might have one day.
But the flipside of this is terrifying. What long-term sentences am I inflicting on my grandchildren without even realising it? What are the possible ill effects of the weeks I spent worrying about a journal article while pregnant with my son? What damage have I done to my daughters’ health because I was suffering depression when I conceived her in the aftermath of a miscarriage? How many of my imperfections are my mothers fault, consequences of actions she could have done differently? And what fresh despair is heaped upon all those mothers who, for reasons beyond their control, have experienced conditions that we now know may shorten the life and reduce the quality of life of generations of their descendants? The door is open to whole new vistas of guilt inflicted on the mother who has ruined the lives of a whole branch of descendants by suffering grief at the wrong moment, or surviving a natural disaster, or failing to protect herself from a pollutant.
What is the proper reception, amidst so much mother-bashing, of the science of maternal effects? How can epigenetic research be used in a way that avoids laying more blame at the feet of women already over-burdened with guilt? Most mothers are genuinely anxious to get hold of any information that can help her to do what is best for her baby. And anyway few scientists would want to withhold information which could have such huge consequences for the health and well being of unborn children. Fortunately, epigenetic research need not emerge as yet another cosh with which to beat mothers.
One reason why the political left has tended to be opposed to genetic determinism is that if things like IQ are determined genetically then there is no point, as a society, in trying to intervene on them. Epigenetic determinism could leave us just as fatalistic, as powerless to undo the damage caused to life chances by a war during gestation or a drug-addict mother. Yet nobody thinks that epigenetic effects are deterministic. Like genes, epigenetic markers will only bring about particular effects in concert with a huge number of additional factors, many of which can be intervened upon to alter the effect. And, unlike genes, epigenetic effects can be modified and even removed. The effect of a high fat maternal diet on insulin resistance was found to disappear after three subsequent generations with normal diets. No epigenetic effect need be thought about as an immoveable life sentence, and a better understanding of their triggers and interactions will give us clues about how we can modify them.
Furthermore, the research has profound implications for ways in which people other than mothers bear responsibility for the health and wellbeing of future generations.
For example, careful cohort studies have revealed details about particular stages of gestation which are vulnerable to diferent epigenetic triggers. Some theorists have suggested that it is in the early stages of pregnancy that epignenetic changes are most significant, because they will be passed onto a greater number of descendant cells. Yet, high blood sugar has been found to be most dangerous during the perinatal period which begins at 22 weeks gestation. And data from the 2001 World Trade Centre Attacks suggest that maternal stress is most damaging when women are in their third trimester. So science can identify priority targets for famine relief, and for support during disasters.
Famine relief used to be targeted mainly at children. But it was found that providing enhanced nutrition to babies whose mothers were malnourished often resulted in significant health problems for those children later on, including diabetes, and cardiovascular disease. The fetal programming hypothesis accounts for this by explaining that such children have adapted to have 'thrifty phenotypes', with upregulation of genes that enable them to hoard calories. In conditions of abundance, these genes make a person fat and unhealthy. So underweight mothers end up with overweight offspring if the diet of the offspring doesnt match that of the mother. We're now trying to combat spiralling diabetes and heart disease in developing nations by targeting famine relief at expectant mothers rather than at babies.
Epigenetic science can also help governments to target health interventions at very specific groups of people -women in a particular stage of gestation - knowing that intervening for just a few months can have payoffs that last for multiple generations. Now that is cost-effective!
And its not just mothers. Epigenetic research is opening our eyes to some previously unsuspected effects on children of things their fathers did before conceiving them. In the Överkalix study, paternal (but not maternal) grandsons of Swedish men who were exposed during preadolescence to famine in the 19th century were less likely to die of cardiovascular disease.And epigenetic phenomena are likely to be behind the fact that sperm quality deteriorates as a man ages, with possible effects upon disorders such as autism in offspring.
Here is my take-home.
Instead of blaming the mother, its about time WE - by whom I mean we as a society - stopped subjecting pregnant women to air pollution. WE ought to do more to protect young women who are caught up in violent conflicts. WE ought to do more to protect women from domestic abuse. WE ought not tolerate it when formula manufacturers convince mothers in developing countries that their breast milk is a second-rate food source. WE ought to lavish funding on support of all kinds - nutritional, emotional, psychological - for expectant mothers. Of course, we should do all these things already, in the interests of the women themselves. But all of these issues should receive extra priority as a consequence of the science that shows the huge and long-lasting burdens these evils place on the health and welfare of our populations.
Killing me softly: the fetal origins hypothesis
Origins: How the nine months before birth shape the rest of our lives (Annie Murphy Paul)