Two years ago, Jill Escher found evidence that she thinks might explain why two of her three children have severe autism.
Ms. Escher, a onetime lawyer married to a former Silicon Valley marketing executive, learned that when her mother was pregnant with her, she took a whole host of fertility and anti-miscarriage drugs, including some that are no longer marketed.
Doctors had called her mother a "habitual aborter" because she had suffered two miscarriages before she gave birth to Ms. Escher, and so they gave her mother two fertility drugs to boost her ovulation, and then, once she had conceived, they gave her a steroid and two synthetic sex hormone combinations to prevent miscarriage.
While she has no proof that the drugs led to her children's autism, she knows that girls are born with all the eggs they will have for life, and she wonders if the medications could have affected her eggs in such a way that it triggered autism in her children, a 14-year-old boy and a 7-year-old girl.
This "transgenerational" effect may sound unlikely, but it has become a hot topic in autism research as part of an emerging field known as epigenetics.
Mysteries of the Mind
- First part: Dealing with the different worlds of autism
- Second part: Researchers work to unravel causes of autism
"Epi" means above or around, and in this case, it suggests that environmental factors -- infections, drugs, diet, toxins, or stress -- could affect the fetus in ways that would change the child's DNA in the womb and lead to illnesses or disorders ranging from diabetes to heart disease to mental disabilities.
There are intriguing hints from animal and human studies that these environmental influences can sometimes have effects on future generations by affecting a fetus's eggs or the precursor cells for sperm. In those cases, you wouldn't know for sure that an epigenetic change had been passed along until the grandchildren's generation, because it would be the first offspring not exposed to the original stressor.
A key part of the epigenetics concept is that these factors will influence the way a person's DNA is activated, but won't change the structure of the DNA itself, which makes it different from a mutation that scrambles the letters of the DNA alphabet.
While no one has made a convincing case yet that autism is one of those transgenerational disorders, several researchers have said it is worth investigating the possibility that pregnant women exposed to environmental toxins from the 1950s to the 1970s, when air and water pollution were worse, might have given birth to children whose eggs and sperm were altered by those substances, who in turn gave birth in the next generation to the autistic children we see today.
There are cases where environmental factors seem to have influenced future generations.
In one frequently cited human study, Swedish scientist Lars Olov Bygren examined children born in a remote Arctic region during years when there were either crop failures or crop surpluses. Surprisingly, he found that the children and grandchildren of men born in the abundant crop years had much shorter lifespans than those born to men who had grown up in the lean years.
He has said that he does not know why prosperous years would cause shorter life spans in future generations, but he noted that the grandfathers' diets could have affected their sperm when those cells first formed between the ages of 9 and 12, and those changes could have affected the health of their sons and grandsons. He noted that grandmothers in the abundant crop years did not pass along similar disadvantages to female offspring.
There also have been animal studies showing epigenetic effects, and in March, Ms. Escher's family foundation helped finance a conference at the MIND Institute at the University of California at Davis that featured some of those studies.
One researcher who spoke there, Michael Skinner of Washington State University, said he has found transgenerational effects in laboratory rodents exposed to dioxin, an industrial toxin sometimes found in meat, fish and dairy products, and vinclozidin, an antifungal agent used in the wine-making industry.
His group's 2012 dioxin study showed that when pregnant rats were exposed to the pollutant, their offspring had higher rates of prostate and ovarian diseases through the third generation, and that in that third generation of rats, 50 regions in the DNA controlling sperm production had been altered by an epigenetic process known as methylation.
In the vinclozidin experiment, his team exposed pregnant rats to the antifungal agent and found that even four generations later, the male offspring still had poor sperm production and similar DNA methylation changes to those found in the dioxin rats.
When an environmental insult causes DNA methylation, he said, it affects the way the animal's genes are expressed, and "when that animal becomes an adult, it has this shifted epigenome which gets transferred to the next generation through the sperm or egg, and you now have a shift that's generated in every single cell type, and when there's a tissue that is susceptible to that shift, you get a disease."
He and other scientists at the California conference said they often run into criticism and even disdain from other genetic researchers when they suggest that these kinds of epigenetic changes might be linked to disorders like autism.
"It's not surprising if you have studied classic genetics for 30 or 40 years and someone comes along and says that's not the whole story, there will be a knee jerk reaction" against it, he said.
Sometimes, DNA methylation changes can be positive.
In a 2003 study at the University of North Carolina, Randy Jirtle and his colleagues took a breed of mice that was genetically prone to be overweight and have yellow fur, known as Agouti mice, and fed pregnant mothers supplements of B vitamins. They gave birth to lean, brown mice, and the Jirtle team was able to show that the B vitamins had silenced the gene that normally made the mice fat and yellow.
Even though there isn't evidence yet of environmental stresses causing brain disorders in multiple generations, there are studies showing they can have an impact in children whose mothers had certain illnesses.
Alan Brown, a psychiatrist and epidemiologist at Columbia University, found with colleagues at Kaiser Permanente that when they looked at a large group of babies born in California in the 1950s and 1960s, children whose mothers had been exposed to the flu during pregnancy had a threefold greater risk of getting schizophrenia decades later.
He is now working with a large birth registry in Finland. Along with researchers at Turku University and the National Institute for Health and Welfare there, he has found that mothers with higher levels of a liver substance known as C. reactive protein, which can be caused by infectious or inflammatory illnesses, had about a 50 percent greater rate of giving birth to children with autism.
In an interview at the epigenetics conference, Dr. Brown said that the flu virus itself doesn't usually cross over from mother to fetus, but inflammatory substances known as cytokines in the developing baby might create damaging inflammation in the baby's brain.
"I take a fairly agnostic approach," on what triggers autism, he said. "I'm interested in understanding the causes, whether they are natural or imposed. If that exposure is toxic and there is strong evidence that it is related to autism, efforts need to be made to reduce or eliminate it from the population, whether it's due to man-made causes or natural causes."
Finally, in a study published just last month in the journal Molecular Psychiatry, a group at Johns Hopkins University in Baltimore autopsied the brains of 21 people with autism and of 19 people without autism, and found four regions of the autistic people's brains that had a sharply different pattern of DNA methylation.
Because people have been taught for years that genes govern all our biological functions and only change when there are rare mutations or inherited diseases, the idea that illnesses or diet could alter our genes within a generation can be hard for some to accept.
But Scott Selleck, a genetics researcher at Penn State University, said that scientists who have worked with fruit flies, as he has, don't find the idea strange at all.
Fruit flies are used to study how genes control development and behavior, and researchers are acutely aware of how environmental conditions such as crowding, temperature and diet can change the way the flies' genes function.
In one case, he remembered, his lab had identified a strain of flies whose nervous system genes were so abnormal that they didn't survive to adulthood. But when a Japanese doctoral student took the strain of flies to his home country, where the labs used a much richer type of fly food, the flies routinely lived to adulthood, even though they often had major defects like an extra wing or missing leg.
If environmental conditions can affect the lowly fruit fly so easily, it's possible they might alter the development of humans, too, Jill Escher believes.
When she discovered her mother's pregnancy medication history, she also learned that she and her mother had been part of a study on the impact of those drugs on children's health.
The woman heading up that study was June Reinisch, who was then a doctoral student and went on to become head of the Kinsey Institute for sexual research. Now retired from that position, she is continuing her investigation into possible links between prenatal drugs and disorders like autism, working with a large Danish birth registry.
After the March conference, Ms. Escher said: "I think if there was a major message from these sessions, it is that we cannot look at genetics in a vacuum."
"Genetics has to be looked at in the context of the environmental impacts that affect genetic expression. This is not a genetics vs. environment question. A lot of this is a genetics and environmental effect question, and because of the germline effect [on eggs and sperm], we can see impacts long after the initial exposure."
Just in the past year, she said, "I find that people are very open to this idea now. There is an increasing appreciation that our germline is not this imperturbable little marble of genes, but it's a vulnerable and sensitive tissue that is susceptible to damage by exogenous influences."health
Mark Roth: firstname.lastname@example.org or 412-263-1130. Twitter: @markomar. First Published October 7, 2013 8:00 PM