In a study of nutrition's effects on development, the scientists showed they could change the coat color of baby mice simply by feeding their mothers four common nutritional supplements before and during pregnancy and lactation. Moreover, these four supplements lowered the offspring's susceptibility to obesity, diabetes and cancer.
Results of the study are published in and featured on the cover of the Aug. 1, 2003, issue of Molecular and Cellular Biology.
"We have long known that maternal nutrition profoundly impacts disease susceptibility in their offspring, but we never understood the cause-and-effect link," said Randy Jirtle, Ph.D., professor of radiation oncology at Duke and senior investigator of the study. "For the first time ever, we have shown precisely how nutritional supplementation to the mother can permanently alter gene expression in her offspring without altering the genes themselves."
In the Duke experiments, pregnant mice that received dietary supplements with vitamin B12, folic acid, choline and betaine (from sugar beets) gave birth to babies predominantly with brown coats. In contrast, pregnant mice that did not receive the nutritional supplements gave birth predominantly to mice with yellow coats. The non-supplemented mothers were not deficient in these nutrients.
A study of the cellular differences between the groups of baby mice showed that the extra nutrients reduced the expression of a specific gene, called Agouti, to cause the coat color change. Yet the Agouti gene itself remained unchanged.
Just how the babies' coat colors changed without their Agouti gene being altered is the most exciting part of their research, said Jirtle. The mechanism that enabled this permanent color change - called "DNA methylation" -- could potentially affect dozens of other genes that make humans and animals susceptible to cancer, obesity, diabetes, and even autism, he said.
"Our study demonstrates how early environmental factors can alter gene expression without mutating the gene itself," said Rob Waterland, Ph.D., a research fellow in the Jirtle laboratory and lead author of the study. "The implications for humans are huge because methylation is a common event in the human genome, and it is clearly a malleable effect that is subject to subtle changes in utero."
During DNA methylation, a quartet of atoms -- called a methyl group - attaches to a gene at a specific point and alters its function. Methylation leaves the gene itself unchanged. Instead, the methyl group conveys a message to silence the gene or reduce its expression inside a given cell. Such an effect is referred to as "epigenetic" because it occurs over and above the gene sequence without altering any of the letters of the four-unit genetic code.
In the treated mice, one or several of the four nutrients caused the Agouti gene to become methylated, thereby reducing its expression - and potentially that of other genes, as well. Moreover, the methylation occurred early during gestation, as evidenced by its widespread manifestation throughout cells in the liver, brain, kidney and tail.
"Our data suggest these changes occur early in embryonic development, before one would even be aware of the pregnancy," said Jirtle. "Any environmental condition that impacts these windows in early development can result in developmental changes that are life-long, some of them beneficial and others detrimental."
If such epigenetic alterations occur in the developing sperm or eggs, they could even be passed on to the next generation, potentially becoming a permanent change in the family line, added Jirtle. In fact, data gathered by Swedish researcher Gunnar Kaati and colleagues indicates just such a multi-generational effect. In that study of nutrition in the late 1800s, boys who reached adolescence (when sperm are reaching maturity) during years of bountiful crop yield produced a lineage of grandchildren with a significantly higher rate of diabetes. No cause-and-effect link was established, but Jirtle suspects epigenetic alterations could underlie this observation.
Humans and other animals are susceptible to epigenetic changes because of an evolutionary trait in which "junk" remnants of viral infections, called "transposons," inserted themselves randomly within the human and animal genomes. Transposons use the gene replication machinery to reproduce themselves. Cells use methylation as a means to inactivate these junk transposons and prevent their replication. Yet if the transposons have inserted themselves in or near a functional gene, the gene can be inadvertently methylated, too, thereby reducing its expression.
The scientists demonstrated that such inadvertent methylation occurred at the Agouti gene when the mice were fed the nutrients. The four nutrients encourage methylation because they possess chemicals that donate methyl groups within cells. Thus, they are primed to methylate susceptible sites in the genome. In fact, more than 40 percent of the human genome is comprised of transposons that are likely to be methylated, so any genes positioned near them could be at risk for inadvertent methylation.
"We used a model system to test the hypothesis that early nutrition can affect phenotype through methylation changes," said Jirtle. "Our data confirmed the hypothesis and demonstrated that seemingly innocuous nutrients could have unintended effects, either negative or positive, on our genetic expression."
For example, methylation that occurs near or within a tumor suppressor gene can silence its anti-cancer activity, said Jirtle. Similarly, methylation may have silenced genes other than Agouti in the present study - genes that weren't analyzed for potential methylation. And, the scientists do not know which of the four nutrients alone or in combination caused methylation of the Agouti gene.
Herein lies the uncertainty of nutrition's epigenetic effects on cells, said Jirtle. Folic acid is a staple of prenatal vitamins, used to prevent neural tube defects like spina bifida. Yet excess folic acid could methylate a gene and silence its expression in a detrimental manner, as well. The data simply don't exist to show each nutrient's cellular effects.
Moreover, methylating a single gene can have multiple effects. For example, the Agouti gene regulates more than just coat color. Mice that over-express the Agouti protein tend to be obese and susceptible to diabetes because the protein also binds with a receptor in the hypothalamus and interferes with the signal to stop eating. Methylating the Agouti gene in mice, therefore, also reduces their susceptibility to obesity, diabetes and cancer.
Hence, the researchers stress the importance of understanding the molecular effects of nutrition on cells, not just the outward manifestations of it.
"Diet, nutritional supplements and other seemingly innocuous compounds can alter the development in utero to such an extent that it changes the offspring's characteristics for life, and potentially that of future generations," said Waterland. "Nutritional epigenetics could, for example, explain the differences between genetically identical twins, or the disparities in the incidence of stroke between the South and the North. The possibilities are endless."