Animal welfare is very much at the top of our minds

Prof Steve Brown

Estimates vary, but there could be as may as 100,000 genes in the human genome.

Knocking out genes at random to discover their function cannot be done - for obvious reasons - in humans, but can be done in animals.

A consortium - including the Medical Research Council (MRC) Mammalian Genetics Unit in Harwell, near Oxford, SmithKline Beecham, Imperial College London, and Queen Mary and Westfield College, London - have now attempted to do this in a systematic way in mice, one of the most important animals used to model human health in the lab.

By injecting adult, male rodents with chemicals such as ethylnitrosourea (ENU), the scientists can randomly disrupt the DNA in the animals' sperm, leading to the birth of thousands of mutant individuals that are then screened for defects of interest.

Novel drugs

The 500 strains so far identified as potential models appear to mimic conditions as diverse as osteoporosis, visual impairment, renal failure, abnormal cholesterol metabolism, spina-bifida, hearing impairment and diabetes.

These will now be offered free of charge to academic groups to examine more closely. If the faulty gene responsible for a problem in a specific mouse can be identified, it may have a "mirror" in humans.

This could help scientists develop new genetic tests and novel drugs.

Professor Steve Brown, Director of the MRC Mammalian Genetics Unit and Mouse Genome Centre, said: "One of the major challenges for the 21st century is to understand the function of the thousands of human genes identified from the human genome project and their role in disease.

"The mouse is one of the most important organisms for studying the function of human genes and allowing treatments to emerge for human disease."

'Unfocused approach'

The fruit fly, one of science's other favourite lab models, has already been subjected to extensive chemical mutagenesis, providing researchers with a wealth of useful genomic data.

But a concerted effort to do the same with a higher animal such as the mouse concerns welfare groups. The British project parallels a similar programme centred mainly on Germany.

The Fund for the Replacement of Animals in Medical Experiments (Frame) which wants to moderate some of the uses and methods employed in labs has questioned the merits of the project with the UK's Animal Procedures Committee.

The APC advises the Home Office on matters relating to animal experimentation.

Dr Elizabeth Jenkins, a scientific officer with Frame, said: "It appears to be an unfocused approach to making disease models.

"If you want to create a disease model you need to understand the disease in humans first. To just knock out genes haphazardly in mice to find potential models seems to us to be the wrong way of approaching the problem.

"With these mutagenesis screens, they tend to find a lot of skeletal defects which are not particularly pleasant phenotypes to be generating."

Strict guidelines

Professor Steve Brown said that if people really wanted to see the human genome project bear fruit then mouse experimentation was the only way ahead.

"Animal welfare is very much at the top of our minds," he told the BBC. "In Britain, we have some of the strictest regulations in the world for licensing experiments that require very detailed and robust justification.

"I think that most people in the country accept the need for some animal experimentation.

"If they can see clearly the benefits that arise from that experimentation and that can be translated into better healthcare, better drugs and better therapeutic strategies for people who are ill.

"The challenge for us as scientists is to communicate to people better what those benefits and opportunities are."