The genetic battle of the sexes

Natural History, Feb, 1998 by Rosie Mestel

But behavior is not on Vrana's mind as, day after day, he does what every geneticist likes to do - put this female and that male together and wait for the inevitable to occur. Vrana is thinking about war: war between fathers and mothers, each wrestling to gain the upper hand, callously using their babies as genetic battlefields.

It's a bizarre and nasty-sounding concept, but one with logic and number crunching to back it. It's also a concept that might explain a strange phenomenon that molecular biologists have unearthed this last decade: genomic imprinting. Basically, genomic imprinting means that certain genes behave very differently depending on whether the mother or the father passed them down. This finding defies common, garden-variety genetic theory, which tells us that it really shouldn't matter whether father or mother gives us any particular gene. While for the most part this still holds true, scientists now know that the DNA of a few genes is chemically altered in cells that give rise to eggs and sperm. And these alterations make for dramatically different properties when the father, rather than the mother, donates the gene to the young.

Why should this situation have evolved? David Haig, an evolutionary biologist at Harvard University, thinks he knows. Imprinted genes, he says, are weapons that mothers and fathers use to bicker over how big their babies should grow.

Bicker? Shouldn't morn and dad both want the best for their fetus (with "want," in this context, being shorthand for whatever best furthers each individual's evolutionary ends)? Well, yes, says Haig, but he argues that mother and father define "the best" somewhat differently. Morn wants a healthy baby, but she doesn't want to end up so drained from feeding - or delivering - it that she can't bear healthy babies in the future. Dad, meanwhile, doesn't care about those future babies; who knows if they'll even be his? He wants this baby, his baby, to get all the mother can give. And therein lies the conflict, with paternal genes ever pushing for a larger, spunkier baby and maternal genes applying the brakes to reserve energy, for babies still to come.

Haig developed his theory of competing parental interests years before the first imprinted genes were discovered. How, he wondered, might this parental war of wills be orchestrated? The first critical clue came from experiments showing that the genetic contributions of mother and father are not the same. Scientists had engineered mouse embryos that contained (as do all placental mammal embryos) two copies of every gene, but with a crucial difference. In some of the embryos, both copies came from eggs. In others, both came from sperm. Either way, the development of the embryos went horribly wrong before they finally died.

Why did development go so awry? The explanation came in 1991, with the discover - in mice - of the first imprinted gene, one that codes for a protein called IGF-II (or insulin-like growth factor II). The gene is active - that is, able to make protein - in an embryo only if it comes from the father. And that's because the mother has chemically altered, or imprinted, her own copy of the gene, rendering it totally inactive. More than a dozen other imprinted genes - some active only in paternal DNA, some only in maternal - have since been found in mice and humans.

As yet, researchers don't know just how genes are turned off in the egg or the sperm. The process appears to involve enzymes, such as one called DNA methyltransferase, that add methyl groups to the DNA. Whatever the mechanism, the imprinting process is present in the embryo but wiped clean and fashioned anew in the offspring's own eggs or sperm (but not in other body cells) later on in life - in a male pattern if the offspring is a male, in a female pattern if it's a female.

To Haig, imprinting meant he had found a perfect way for a father's genes and a mother's genes to be pushing for different outcomes. "It was great. It was really pleasing," he says. "Within seconds of hearing about it, I said, 'Aha! I've got a theory to explain that.'" Haig wasn't the only scientist thinking along such lines; so was developmental biologist Tom Moore, now of the Babraham Institute in Cambridge, England. In 1991, the two joined forces and published their ideas together.

Is this evolutionary view of genomic imprinting correct? Certainly, the theory enjoys support among many researchers working to understand the molecular biology of imprinting and its contribution - when it goes wrong - to human disease.

"My feeling is that Haig's theory is the best explanation of imprinting around," says Wolf Reik, a developmental geneticist at the Babraham Institute. Shirley Tilghman, a molecular geneticist and Paul Vrana's adviser at Princeton, also finds the theory attractive. Others, though, say it's not yet time for an all-embracing theory: there simply isn't enough known about most imprinted genes. And while Laurence Hurst, an evolutionary geneticist at England's Bath University, praises the theory, there's a sting in the tail of that praise. "It's an absolutely delightful, elegant theory, and I'm not prepared to give up on it yet," he says. "But the bottom line is that while the early data on imprinting seemed to all go with the theory, just about all the data that have come out since then have not conformed to its predictions."