BBC Science Correspondent Pallab Ghosh: "Scientists have created the world's first supermouse"

Humans also have the added gene, although it is not yet known whether it has the same function in people. But the scientists believe their results show that one day it may be possible to boost human intelligence.

The breakthrough will ignite debate about whether such a feat would be ethical but Ira Black, chairman of neuroscience at Rutgers University says: "It's very exciting and holds the hope of not only making animals smarter but also, ultimately, of having a human gene therapy for use in areas such as dementia."

Side effects

However, Dr Tim Bliss, head of neurophysiology at the National Institute for Medical Research in London, said: "When you insert a gene at random into the genome you don't know what might happen. These animals seem to be OK, but there might be all sorts of hidden down sides to having this extra protein.

"I think it's very unlikely that tinkering with one gene is going to increase intelligence and do nothing else. In my view talk of genetically enhancing human intelligence is nonsense."

However, he said new drug or gene therapy treatments might emerge from the research. "It certainly raises that possibility, but we are very far away at the moment."

Smart young things

The research team from Princeton, Washington and MIT universities found that adding a single gene to mice significantly boosted the animals' ability to solve maze tasks, learn from objects and sounds in their environment and to retain that knowledge.

Joe Tsien: "They're smarter"

The intelligence-boosting effect results from the mice retaining into adulthood certain brain features of juvenile mice. Like young humans, young mice are widely believed to be better than adults at grasping large amounts of new information.

This new strain of mice is named Doogie, after a precocious character on the US television show Doogie Howser, MD.

One key feature of learning is the ability to associate one event with another - associating fire with the sensation of pain is a simple example. The research shows that the gene used, called NR2B, is very important in controlling this ability.

It is the blueprint for a protein that spans the surface of neurons and serves as a docking point, or receptor, for certain chemical signals. This receptor, called NMDA, is like a double lock on a door; it needs two keys or events before it opens.

Studies have shown that in young animals the NMDA receptor responds even when the two events happen relatively far apart, so it is easier to make connections between events and to learn. After adolescence, the receptor becomes less responsive, making learning more difficult. But introducing the NR2B gene kept the Doogie mice's brains "young".

Testing times

Showing that the Doogie mice were more intelligent required a number of tests:

• Object recognition - The mice explored two objects for five minutes. Several days later, one object was replaced with a new one. When the mice returned, the Doogie mice remembered the old object, and devoted their time to exploring the new one. The normal mice, however, spent an equal amount of time exploring both objects. The Doogie mice remembered objects four to five times longer.
• Emotional memory - The animals received mild electric shocks to their feet. When the mice were put back in the chamber up to 10 days later, the Doogie mice showed much more fear than the control mice.
• Learning response - The mice again received shocks and were then placed back into the fear-causing environment, but without the shocks. The Doogie mice were much quicker to resume normal behaviour: they learned faster.
• Spatial learning - The mice were put into a pool of water that had a hidden platform where they could climb out of the water. The Doogie mice learned to find the platform after three sessions, while the control mice required six.