Munich

Hearing aid: a mouse is tested for hereditary deafness, giving clues to its genetic origins.

A pilot study to systematically characterize 21 commonly used inbred — and thus genetically homogeneous — strains of mice has been launched by scientists in several US laboratories. The study is the first action of a project called the Mouse Phenotyping Initiative, which aims to create a publicly accessible database of the characteristics — known as traits or phenotypes — of inbred mouse strains.

According to members of the initiative's steering group, chaired by Ken Paigen, director of the Jackson Laboratory in Bar Harbor, Maine, such a database is urgently needed by mouse geneticists, as there is currently a lack of information about the strains that are routinely used to model human diseases.

Paigen is the driving force behind the initiative, which will set up the database at the Jackson Laboratory, the world's leading supplier of inbred mouse strains. The laboratory already hosts the Mouse Genome Database, which is accessible to researchers throughout the world.

With the Human Genome Project nearing completion, scientists are becoming increasingly interested in investigating the functions of the genes that have been sequenced, and mice have become a classic genetic tool in this endeavour.

Scientists around the world are due to finish sequencing the mouse genome by 2005. The speed of its completion was boosted by the creation earlier this month of the Mouse Genome Sequencing Network, funded by the US National Institutes of Health (NIH). The network involves ten collaborating laboratories which will share US$21 million over the next six months.

In addition, there are already hundreds of relevant inbred strains and thousands of mutant strains, created either by ‘knockout’ technologies, which target particular genes for deletion, or by chemical mutagenesis.

Geneticists use inbred strains to analyse complex physiological or behavioural traits, such as a propensity to a particular type of cancer, or a taste for alcohol. They can start to pin down the genes that determine the trait by crossing a strain that strongly displays a particular trait with one that displays it weakly, or not at all, and working back from genomic information. A more precise understanding of the phenotypes of inbred strains will allow experimenters to select the most appropriate strains for cross-breeding.

The Mouse Phenotype Initiative's steering group, which has two European, one Japanese and six US members, has drawn up a list of inbred strains and phenotypes that it thinks should be included in the database. It recommends that 20 animals from each strain be analysed in two different laboratories with expertise in a particular area, such as behaviour or clinical chemistry.

The steering group has suggested that data already published should not be included, because strains bred in different laboratories sometimes undergo ‘genetic drift’ and so can no longer be assumed to be identical to the reference strain from which they were derived. The Jackson Laboratory will provide all the mice that will be phenotyped for the database.

The pilot study, which will be conducted at the Howard Hughes Medical Institute at Northwestern University in Chicago, Baylor College of Medicine, and the Jackson Laboratory, will study three characteristics: circadian rhythms, haematology and the response to a high-fat, high-cholesterol diet.

Funding for the pilot study has come from donations from pharmaceutical companies. Rick Woychik, director of the Parke-Davis Laboratory for Molecular Genetics at Oakridge, California, one of the donors, says the database will be an “immensely powerful tool” for industry, which needs optimal mouse models to identify genes involved in human diseases. But long-term financing has not yet been secured.

Major funding agencies, including the NIH, Britain's Medical Research Council and France's INSERM, will be asked by the initiative's steering committee to fund laboratories involved in phenotype assessment and database management. Pharmaceutical companies will also be approached.

If such funding is secured, later phases of the project will include up to 50 different strains of mice and all phenotypes for which there is interest in the research community.

“The Mouse Phenotype Initiative is important because we need to develop comprehensive, common and systematic phenotyping tests,” says Steve Brown, director of the Mammalian Genetics Unit of Britain's Medical Research Council. “In particular, clinical vocabulary should be standardized to the mouse, since the mouse is used to model human disease.”

But mouse pathologists are rare, he points out. This was recognized last year by the NIH, which is set to fund several mutagenesis centres. These will focus on recessive genetic traits, to complement the European focus on dominant traits, and will be grouped into two areas: behaviour and development.

Maja Bucan, a geneticist at the Center for Neurobiology and Behavior at the University of Pennsylvania, who runs a mouse mutagenesis programme screening for behavioural traits, says “we all went into mutagenesis studies without information about how phenotyping should be standardized, and now we have to go back and do it. This is exactly why [the Mouse Phenotype Initiative] is so important.”