Published online 15 November 2007 | Nature | doi:10.1038/news.2007.250


When only mum or dad matters

Hundreds of our genes turn on just one copy, rather than both.

The random switching off of gene copies could create extra genetic variation between identical twins.Getty

The textbook rule that says activated human genes almost always express both of their copies — the one inherited from mum and that inherited from dad — seems not to be true. Instead, a good chunk of our genome could prefer the 'single life', according to new research.

Whether the maternal or paternal copy gets switched on in such cases seems to be random. But the result could have a big impact on disease susceptibility and other biological traits.

It had been thought that there are only a handful of situations in which just one of a pair of gene copies is used. But a new screen of 4,000 human genes has uncovered 371 that sometimes play favourites, suggesting that this phenomenon is far more pervasive than had been thought.

“What we’ve shown violates general rules about what we think about genes: if a gene is turned on you express both the maternal and paternal copy,” says Andrew Chess, a biologist at Harvard University in Cambridge, Massachusetts, who led the study along with colleague Alexander Gimelbrant.

This kind of selective gene expression could create an extra source of variation between people, even when some of their genes are identical.

“I like the idea that we're all mosaics, and this might contribute to differences,” says Steve Henikoff, a biologist at the Fred Hutchinson Cancer Research Center in Seattle, Washington.

Coalition of the expressed

Researchers have long known that some cells express just one version of certain genes, but this is for a specific purpose. Females have two X chromosomes but always shut down one copy, for example, to prevent making a double dose of some gene products. Immune cells designed to recognize the fingerprint of a specific disease turn off one set of immunoglobulin genes to optimize their specialist behaviour. And during development, embryos chemically flag one parent’s version of a set of known genes for silencing. This handful of instances accounts for hundreds of genes known to turn off a copy in a controlled way.

To hunt for more genes that behave similarly, Chess and Gimelbrant planted single human immune cells in dishes, and let the cells multiply. They analysed the millions of cell copies using a gene chip to distinguish between the maternal and paternal version of messenger RNA (mRNA) produced by 3,939 genes. They found that about 9% of the genes sometimes turned on just one copy. They publish their findings today in Science1.

Even though the cells were all genetically identical, some turned on maternal versions, some paternal, and others activated both. "Even identical twins would make different choices," says Gimelbrant.

The team noticed a similar pattern in a much smaller subset of genes that they examined from specimens of donated human placenta and white blood cells. “This is not something weird that happens only in cell culture,” says Chess. Nor does it happen only in immune cells, although the researchers' further work on other cell types has not yet been published.

Sloppy genome

These differences could potentially help an organism to regulate gene expression and affect disease risk. Higher expression of a gene called APP is linked to early onset of Alzheimer’s disease, for example. And Chess’s team found that cells in which just one copy of APP was turned on made about half as much mRNA as the cells with both copies turned on. At this point, the connection to Alzheimer’s remains speculative, he says.

The random gene inactivation that Chess's team has found could be a spillover from more controlled forms of silencing, such as shutting down an X chromosome, says Huntington Willard, a geneticist at Duke University in Durham, North Carolina. "I tend not to think of it as some grand evolutionary plan, but simply as the sloppiness of gene regulation."


Chess doesn’t know just when cells switch off a gene copy, but he suspects it happens during development. Unanswered, too, is how the cells turn off one copy while keeping the second copy active. Chess’s lab is now digging into this question.

The team next plans to survey the rest of humanity's 20,000 genes for signs of similar deactivation. If his team’s sample is representative of the whole genome, more than 1,000 genes should sometimes go it alone. 

  • References

    1. Gimelbrant, A. et al. Science 318, 1136-1149 (2007) | Article |
Commenting is now closed.