The genomes of lab mice can shift in subtle and unpredictable ways over generations of breeding. Credit: Anne Chadwick Williams/Sacramento Bee/ZUMA Press/Alamy

Adam and Eve, a pair of black mice, lived for less than two years and never left their home at the Jackson Laboratory (JAX) in Bar Harbor, Maine. But since they were bred in 2005, their progeny have spread around the globe: the pair’s living descendants, which likely number in the hundreds of thousands. They are members of the most popular strain of mice used in biomedical research, which was created nearly a century ago.

Now, researchers at JAX are reconstructing Eve’s genome in the hopes of better understanding — and compensating for — the natural mutations that occur in lab mice over the course of generations. These genetic changes can cause unanticipated physiological effects that can confound experiments. Related substrains of lab mice can differ in their taste for alcohol or their sensitivity to insulin, for example, and researchers suspect that such differences between supposedly identical mice lines have hampered some areas of research.

The scientists who founded JAX created Adam and Eve’s breed, which is called C57BL/6, in 1921. To keep the mice as genetically similar as possible, researchers have repeatedly bred brothers with sisters for nearly a century — and sold the resulting offspring to customers around the world. But this strategy created a genetic bottleneck: every generation, between 10 and 30 new mutations pop up and are passed down to offspring. This ‘genetic drift’ quickly accumulates over the years, says Laura Reinholdt, a geneticist at JAX. The genomes of the C57BL/6 mice that the lab sells today have thousands of genetic differences from the mouse reference genome, which was created in 2002 from three mice from the substrain C57BL/6J. The genome is used as a template for researchers developing genetically modified mice.

Other suppliers have inadvertently created divergent substrains of C57BL/6 mice when they’ve bought rodents from JAX and bred them over several generations. Although most mutations go unnoticed, some occur in genes that affect a mouse's appearance or physiology. In 2016, mouse supplier Envigo in Somerset, New Jersey, found that C57BL/6 mice at 6 of its 19 breeding facilities around the world had acquired a mutation in a gene related to the immune system. The company notified the researchers that bought these mice, and asked customers to specify which location they preferred to source mice from in the future, given that the company’s stocks were no longer identical.

Hidden changes

And although it is easy to spot a mutation that changes fur from black to white, for instance, some changes are discovered only if researchers are investigating a particular trait. A substrain of C57BL/6 mice that the US National Institutes of Health bred for 50 generations are uninterested in alcohol, whereas those bred at JAX’s facility display a preference for alcoholic beverages.

In 2005, a team at JAX decided to reset the genetic clock by selling only C57BL/6J mice descended from two chosen mice: Adam and Eve. The researchers froze hundreds of embryos of the duo's grandchildren, enough to last for 25-30 years. Every five generations, the company thaws some of these embryos and raises them to adulthood as new breeding pairs.

“In some ways, the changes that are acquired are insidious and unstoppable,” says Michael Wiles, the lab’s senior director of technology evaluation and development, who led the project. “We’ve not stopped general drift, but we’ve slowed it considerably.” Once the stockpiled embryos run out, however, JAX will have to start over with new breeding pairs from a much later generation.

Yet Eve's genome is very different from the 2002 mouse reference genome. In a presentation last month at the American Society for Human Genetics’ meeting in Orlando, Florida, JAX computational scientist Anuj Srivastava spoke about the company’s effort to reconstruct Eve’s genome in high detail, using three different sequencing methods. Wiles says that the genome will be finished by the end of November, and that JAX plans to publish it early in 2018.

Mouse trap

Other mouse breeders have started their own efforts to account for genetic drift. Taconic Biosciences, a mouse distributor in Hudson, New York, restarts its C57BL/6 line every ten generations from its stash of frozen embryos. Because Taconic has bred its line separately from the JAX line for decades, the Eve genome won’t necessarily reflect the genetic make-up of Taconic’s mice any more than the current mouse reference genome does.

Ana Perez, Taconic’s global director of genetic sciences and compliance, says that the company plans to publish the genome of its own Eve. “From my perspective, each particular breeder should have their own reference genome to follow,” she says. Buying mice from different breeders and expecting them to be the same is a fallacy, she adds.

But most researchers don’t think about the differences between the various substrains of C57BL/6 mice and how those disparities can affect reproducibility in research, says Cory Brayton, a pathobiologist at Johns Hopkins University in Baltimore, Maryland. “The vendors are pretty good about making the information available, but the awareness is still pretty low,” she says. It is impossible to quantify how often experiments or entire research programmes are wasted when researchers realize that their supposedly identical mice have genetically diverged from the ancestor they bought from a vendor, but Brayton suspects it is common.The Eve genome will be a useful addition for researchers who use animals from JAX, says Brayton, although it won’t solve all the reproducibility problems inherent to inbred mouse lines. “If you use [inbred mice] wisely, they can be highly informative,” she says. “If you use them stupidly, they may really confound your studies.”