with the end of World War II still months away, millions of people in the German-occupied Netherlands were suddenly forced to go hungry. Upset over a railway strike, the Germans blocked food shipments to the region. Over the course of five or six months, thousands died of starvation. During the worst of the famine, survivors in the western part of the country were just barely getting by, consuming as few as 400 calories a day. To put it in perspective, that’s the equivalent of about four apples or one plain bagel. Most people start to lose weight on almost four times that number of calories.
Other countries pitched in, and by the spring of 1945, the so-called Dutch Hunger Winter was over, or so it appeared. But the winter of deprivation created a perfect opportunity for studies by scientists interested in the long-term effects of malnutrition and stress, because people were well fed both before and after the famine. Scientists have paid special attention to those who couldn’t tell them what the famine felt like — the children who were conceived and carried through that period of food scarcity inside the bellies of their pregnant mothers.
Those studies show that the effects of the Dutch Hunger Winter were anything but fleeting for some of the children born of those tough times. The developing unborn babies who were five or so months along when food supplies slowed to a trickle were smaller at birth and generally stayed small. But those who were conceived just as the famine began really paid a price, despite being of roughly normal size at birth. It must have looked to their parents as if their new babies had gotten off scot-free.
The evidence from later years presents a very different picture, however. The children who experienced the Dutch Hunger Winter from their very first days and weeks of life in the womb went on to show elevated rates of many unhealthy conditions and chronic diseases, such as obesity, altered lipid profiles, cardiovascular disease, diabetes, increased stress response, schizophrenia, and addiction. Although their cognitive function appeared normal at the age of 19, researchers have recently showed that, by their late 50s, men and women exposed to famine at the early stages of gestation were having more trouble paying attention. These discoveries suggest a pattern of accelerated aging.
Somehow, that very early hardship, experienced during a critical window of development, “programmed” those children for chronic diseases later in life. The question then is: How?
A paper published in the Proceedings of the National Academy of Sciences offered an important clue. While the children’s experiences with starvation so early in life had nothing to do with the genes they carry, it may have made lasting changes to the way some of their genes functioned, via changes at the epigenetic level.
Epigenetics refers to molecular processes that leave durable marks on the DNA, altering the gene’s function independent of the underlying DNA sequence. The presence or absence of those marks can affect the level of expression of genes, effectively silencing or activating them. Epigenetics explains how the cells in our bodies — muscle versus skin versus hair, for instance — can be so vastly different, even though they carry the same DNA.
When we exercise, it’s thanks to epigenetic reprogramming that our muscles bulk up, growing stronger and more resilient. A study published last year in the journal Epigenetics asked 23 healthy young people to do one-legged cycling for 45 minutes, four times a week, for three months. At the end of the training period, the researchers found differences in the chemical marks associated with more than 4,000 genes, many of them at regulatory sites responsible for controlling other genes.
Although these and other epigenetic marks within cells are clearly malleable, the epigenetic profile of some of our genes is established during gestation, while we are in our mother’s womb, and in our children around the time of their conception. This epigenetic programming determines which of our two copies of some genes work, the one we got from mom or the one from dad. About 100 of our genes are imprinted this way, although scientists haven’t defined all of them just yet.
Some liken epigenetics to punctuation marks that dictate how our genetic instructions are read. But epigenetics pioneer Randy Jirtle, who is affiliated with the University of Bedfordshire in the United Kingdom, the University of Wisconsin-Madison, and North Carolina State University, prefers to compare our genomes to computers, by which he means only the physical machines. Those chemical or protein marks on top of the genome — collectively known as the epigenome — are the software that tells our genomes what to do. As this analogy shows, to understand the roots of health and disease, it will be critical to consider not just our genomes but also the epigenomic programming.
“If you bring in a computer [that’s malfunctioning] and only look at the hardware of the computer, you might come to the conclusion that nothing’s wrong,” Jirtle says. “But the computer still doesn’t work right. That’s because if the problem is a bug in the software, you’re only looking at part of the problem.”
In biomedicine, there’s been a tremendous focus on genetic mutations, he says. By comparison, bugs in our software have largely flown under the radar. That brings us back to the study of people born just after the Dutch Hunger Winter ended. In addition to their many obvious health problems, those who were conceived around the time of the famine show a permanent decrease in the chemical methylation of one of the best-studied imprinted genes, known as insulin-like growth factor II (IGF2). The IGF2 gene is known to play a key role in human growth and development.
When the study was published in 2008, senior author L.H. Lumey, of the Columbia University Mailman School of Public Health, said it provided critical evidence that certain environmental conditions can result in changes in epigenetic information. “If there are indeed relationships between adverse conditions during development and adult health, then these epigenetic changes might provide a mechanism to explain the link,” he said.
Lumey’s team followed up that earlier study with a report published this year showing lasting methylation changes across the genome in children whose mothers were malnourished at famine levels during the first 10 weeks of pregnancy. Those changes were found in genes known to play a role in growth, development, and metabolism.
There are other intriguing examples in which the traumatic experience of groups of individuals has left lasting marks on their children. Rachel Yehuda and her colleagues at the James J. Peters Veterans Affairs Medical Center in the Bronx, New York, reported in 2014 that children whose mothers suffered from post-traumatic stress disorder after surviving the Holocaust inherit a heightened stress response. The findings pointed to a possible role for epigenetic programming, specifically of stress-related genes, in transmitting the effects of traumatic experiences from one generation to the next.
Moshe Szyf and his colleagues at McGill University took advantage of a natural disaster to ask a similar question. In January 1998, the people of Quebec experienced an ice storm, one of the worst natural disasters in Canadian history. Power lines toppled, leaving people in darkness for weeks. The personal and financial toll of the storm left a lasting impact on the city.
The researchers recruited 176 mothers who were pregnant at the time of the Quebec storm or who became pregnant very soon after and assessed the degree of hardship those women had experienced as a result of the storm —
including damage to their homes, the number of days spent without electricity, and time spent in a shelter — along with their personal feelings of distress. Thirteen years later, they conducted an epigenetic analysis of cells in the blood from 36 of the children born to those mothers, to find changes in the pattern of chemical methyl marks in their cells. They also compared those results to the DNA methylation measured from saliva samples provided earlier, eight years after the storm, from 34 of the same children. Those changes in DNA methylation were linked to the degree of prenatal hardship reported by the children’s mothers years before. According to the researchers, the findings offered the “first evidence in humans supporting the conclusion that prenatal maternal stress results in lasting, broad, and functionally organized DNA methylation signatures” in several tissues of their children.
Szyf says epigenetics offers a way for genetically identical cells to differ from one another during development, but it also provides cells with a form of memory. Epigenetics is a mechanism whereby experiences in one generation can be stored and passed on to the next. Those changes in epigenetic identity may lead to disease in two possible ways. On the one hand, he says, “accidents happen. If you don’t mark things properly, you get disease.”
If, for example, the tissue that makes insulin shuts down production, you’ll get sick. But, he says, epigenetics also shapes our genomes in utero and after birth to fit the “world that it sees. If it’s a ‘bad’ world, you have to be hyperanxious. You probably need an active immune response, and to take every piece of food and turn it to fat because you don’t know where the next meal will come from. If your mother sends you that signal by nutritional restriction, for example, or by high adversity early in life, your epigenome thinks this life will be really bad and adjusts itself.”
That’s not an automatic recipe for disease — in many cases it might be good — but sometimes the epigenetic programming turns out to be mismatched. If your programming is set up in anticipation of adversity and you find yourself instead in a middle-class city, he says, you may be in trouble.
It’s likely that disease states of all kinds involve epigenetic deregulation of one kind or another. There is evidence that epigenetics plays some role in cancer, diabetes, obesity, schizophrenia, Alzheimer’s disease, and possibly autism. Szyf has even proposed a role for epigenetics in the long-term consequences associated with chronic pain.
“Wherever people look, they find it,” Szyf says.
Although many of the studies to ferret out epigenetic effects in people have involved some form of major trauma or disaster, epigenetics certainly plays a role in more run-of-the-mill instances, too. In an effort to get a baseline understanding, Cathrine Hoyo, now at North Carolina State University, and her colleagues started enrolling pregnant women living in Durham, North Carolina, into the Newborn Epigenetic Study (or NEST) back in 2003, when she was in training with Jirtle at Duke University. The study includes the children of about 2,500 women who answered 55 pages of questionnaires on everything from diet to stress to assisted reproduction. The researchers collected umbilical cord blood samples of their children in the delivery room at birth, so they could establish their epigenetic profiles and follow them over time.
As early results of the NEST study are beginning to come in, these children are growing up. Hoyo says some of them are already becoming obese. Twenty-six of them have been diagnosed with autism. Meanwhile, she is busy enlisting colleagues from various disciplines to explore other conditions, such as attention deficit disorder, and the connections between epigenetics and exposures to everyday toxins, like lead, cadmium, and flame retardants.
So far, their studies suggest potential connections between DNA methylation and socioeconomic status. Another study reports a connection between maternal stress and epigenetic changes in a gene relevant to maternal care and obesity. There are signs that exposure to lead, which can be found at high levels in paint and dust in older buildings, may leave its mark on the epigenome as well.
Understanding exactly how it all works isn’t going to be as easy as Hoyo initially thought. Hoyo and Jirtle say that the idea for NEST was first scribbled on a restaurant napkin soon after a study of Jirtle’s made a big splash in the news. The study in mice showed that it was possible to silence a gene that would otherwise make a special strain of mice fat and yellow simply by giving their mothers extra vitamins. When fed the vitamin cocktail before, during, and after pregnancy, the animals gave birth to thin, brown pups because of increased methylation in the genomic region containing the Agouti gene. Otherwise, the young mice were fat and yellow as a result of reduced methylation in this region and inappropriate overexpression of the Agouti gene.
An epidemiologist and public health researcher by training, Hoyo says the result was the first to demonstrate “nutrition actually preventing disease … for me, that’s where the connection was.” In addition, it presented a new way to monitor nutritional status at the molecular level with epigenetic markers.
One of the challenges for researchers in human studies is figuring out where to look in the genome. Hoyo and her team have been focusing on a list of known imprinted genes. They have already seen that different factors seem to influence the epigenetic status of some of those genes and not others. Even when you find a change, it’s difficult to show what those changes in one type of cell really mean (if anything) for the person in question.
As Jirtle says, chuckling, it’s a “real good mess.” But “the marks themselves are there. That’s the thing you can put your hands on.”
In 2013, researchers Brian Dias and Kerry Ressler, of Emory University School of Medicine, added a compelling new piece of evidence to the puzzle, again via studies in mice. The researchers paired mild foot shocks with a scent similar to cherry blossom, to train male mice to fear that particular chemical scent. They then showed, in a study in Nature Neuroscience, that the children and grandchildren of those mice had inherited a sensitivity to the scent of cherry blossom. The findings highlight an important point: Epigenetic marks are inherited from fathers.
They went on to show that the sensitivity was similarly passed on through mothers trained in the same way, whether or not those mothers raised their pups. Dias says in vitro fertilization studies in the mice finally convinced him that these changes were really being transmitted through the gametes. They found that the very gene known to be important in sensing the cherry blossom scent had lost methyl marks in the sperm.
Although more details of the mechanism are needed, the findings help show how traumatic exposure, stress, and fear experienced by either parent — whether from famine, the Holocaust, or training in the lab — can be passed on to the next generation. Still, Dias sees reason for optimism.
“Really what this work suggests is that, just as you can lay down a mark, you can also strip away a mark,” he says. “There is a malleability to the epigenetic code which then affords individuals the ability to place themselves in certain contexts that will buffer them from ancestral trauma, if that’s what has occurred. We often focus on the pessimistic side — that our ancestors went through a detrimental experience. Perhaps more importantly, there’s the promise that not all is lost.”
With more investment in research and development, Szyf sees great promise in epigenetic drugs designed to target these chemical modifications very specifically. In fact, epigenetic drugs have already been approved for the treatment of certain types of cancer. He predicts that as researchers learn more about the role that epigenetics plays in the brain, a next wave of epigenetic drugs will target mental health conditions as well.
“Environment can change how our genes work, and it can cause disease sometimes,” Szyf says. “If you figure out how to play with it, you might cure some diseases. An advantage of doing it like this is you will cure the program that creates the disease and eliminate its foundation rather than its symptoms.”