Medicines have to undergo rigorous testing on animals before they can be tried on humans. Even so, researchers cannot fully anticipate how human patients will react to a drug. Now the results of drug testing on animals could be more conclusive, thanks to a 'humanized' mouse.

Ron Evans of The Salk Institute of Biological Studies, La Jolla, California, and colleagues switched a mouse gene for a human one to create the mouse, which responds to drugs like a human1.

When the human body recognises a drug, it initiates--or 'induces'--the production of enzymes to break it down. In humans, a large proportion of drugs is broken down by a group of enzymes known as 'CYP3A'.

Different species have different drug-recognition and induction systems, so even closely related species respond very differently to particular drugs. This creates problems for those trying test human therapies on animals.

Evans's team discovered a protein involved in the induction of CYP3A in humans. They call the protein 'SXR' and explain that it "may represent the critical biochemical interface between man and his environment."

Mice have a protein related to SXR, called 'PXR'. Evans and his group found that, although the proteins differ only slightly, SXR senses drugs that induce CYP3A only in humans, and PXR responds only to mouse-specific inducers.

So the researchers made the 'humanized' mouse by swapping the mouse PXR gene for the human SXR gene. The mice then induce CYP3A production only in response to human-specific inducers, rather than those drugs that normally induce the enzyme in mice.

The induction of CYP3A is responsible for many 'drug-drug interactions', where one drug affects the way another is broken down. Such interactions can have serious consequences. The researchers say their mice will be "invaluable tools" during the drug development process to screen drugs for harmful drug interactions and toxic side effects.

Also, individuals vary in how much CYP3A they produce. This means that their responses to drugs also vary, so different people require different doses of a particular drug for it to be effective with minimal side effects. The mice should also be useful in investigating this inter-individual variation.