Featured Articles

Chubby Blonde? Slim and Dark? Lab Mice Take After Mom's Diet

By Sharon Begley, Wall Street Journal

14 August 2003: The baby mice looked as different as night and day.

Those in one litter were dirty blondes, while those in the other were, well, mousy brown. Yet the mice's genes for coat color were identical, down to the last A, T, C and G that make up the twisting strands of DNA.

The reason some animals were yellow and some were brown lay deep in their fetal past, biologists at Duke University Medical Center, Durham, N.C., reported this month: Some of the mothers consumed supplements high in very simple molecular compounds that zip around the genome turning off genes. One silenced gene was for yellow fur; when it is turned off, the mouse's fur color defaults to brown. For the mice, it wasn't just that "you are what you eat," but that you are what your mother ate, too.

The ink on the final draft of the complete human genome sequence is hardly dry, but scientists are seeing more and more instances in which the sequence of those celebrated A's, T's, C's and G's constituting the genome is only part of the story.

Biologists have long known that having a particular gene is no guarantee you will express the associated trait, any more than having a collection of CDs will fill your home with music. Like CDs, genes are silent unless they are activated. Because activating and silencing doesn't alter the sequence of the gene, such changes are called epigenetic.

"Epigenetics is to genetics as the dark matter in the universe is to the stars; we know it's important, but it's difficult to see," says geneticist Andrew Feinberg of Johns Hopkins University School of Medicine, Baltimore. "What we're thinking now is that, in addition to genetic variation, there may be epigenetic variation that is very important in human disease."

Epigenetic variation may explain such long-running mysteries as why identical twins are, in many ways, no such thing, including whether they have such supposedly genetic diseases as schizophrenia and cancer. Epigenetics may also help explain how the seeds of many adult diseases may be planted during fetal life. Studies suggest that the nutrition a fetus receives -- as indicated by birth weight -- might influence the risk of adult-onset diabetes, heart disease, hypertension and some cancers. The basis for such "fetal programming" has been largely an enigma, but epigenetics may be key.

There is no doubt that, in the case of the brown or yellow mice, the "you are what your mom ate" phenomenon reflects just such epigenetic influences. The Duke scientists fed female mice dietary supplements of vitamin B12, folic acid, betaine and choline just before and throughout their pregnancy. Offspring of mice eating a regular diet had yellowish fur; pups of the supplemented mothers, although genetically identical to the yellow mice, were brown.

When they grew up, the brown mice also had much lower rates of obesity, diabetes and cancer, Robert Waterland and Randy Jirtle of Duke's Department of Radiation Oncology report in the journal Molecular and Cellular Biology. Whatever the extra nutrients did to the fetal mice's genes didn't stop with fur color.

Actually, that "whatever" isn't quite fair. The Duke team knows exactly what the supplements did. All of the compounds contain a simple molecule called a methyl group, which is one carbon and three hydrogen atoms. For a little guy, methyl wields a big stick: It can turn genes off.

That's what happened in the brown mice. Methyl from the supplements switched off a gene called Agouti, which both gives a mouse a yellowish coat and makes it obese. The yellowish babies weren't suffering from any nutritional deficiency; it's just that their Agouti gene was still activated. "Nutritional supplementation to the mother can permanently alter gene expression in her offspring without mutating the genes themselves at all," says Prof. Jirtle.

That's the very essence of epigenetics.

The reason the Agouti gene was silenced is that it had the misfortune to lie next to an interloper. Mammalian genomes are riddled with bits of DNA that leap around like so many jumping beans. Called transposons, they sometimes wind up beside the on/off switch for an important gene, and are sitting ducks for those gene-silencing methyl groups. In the offspring of mouse moms eating methyl-rich dietary supplements, just such a jumping gene was silenced, with the result that the Agouti gene it had snuggled up to was also struck dumb.

This isn't just about yellow and brown mice. "About 40% of the human genome is transposons," notes Prof. Jirtle.

That means an awful lot of human genes could be targets of methylation, and so silenced. Whether that is good or bad depends on what the gene does. Silencing a gene that raises the risk of schizophrenia would be welcome. Silencing a tumor-suppressor gene wouldn't be. What's clear, he adds, is that "we, too, have genes -- including those influencing susceptibility to cancer, obesity and diabetes -- that can be turned off or on by epigenetic factors triggered by early nutrition and exposure to chemical agents."

Next week: How epigenetics might explain certain puzzles from cancer to birth defects.