One third of Britons have some kind of allergy

Researchers told Nature that blocking p110delta in mast cells - which cause allergic reactions - substantially reduced symptoms in tests on mice.

The study, by the Ludwig Institute for Cancer Research's London branch, could lead to drugs to treat the cause of allergies rather than the symptoms.

A third of people in the UK suffer from allergies at some point in their lives.

Allergies are inappropriate responses by the immune system to allergens such as pollen, dust, insects and animals, which can cause runny noses, itching eyes, coughing, skin rashes and wheezing.

Experiments on mice do not always transfer that well to humans Professor Andrew Wardlaw, of the British Society for Allergy and Clinical Immunology

The number of people with allergies has been increasing for the past 20 years and, in the UK, 9m people have hay fever, 6m eczema and 5m asthma.

The team, from the University College London branch of the Ludwig Institute, an international academic centre with bases in seven countries, believe the research could lead to a change in how allergies are treated.

At the moment therapies tend to treat symptoms rather than the cause.

Researcher Dr Bart Vanhaesebroeck said: "Our work points towards a promising future for developing inhibitors for allergic conditions."

But he warned the development of a treatment for humans was still a long way off.

The team also believe the protein may have implications for the treatment of cancer.

In the experiments, mutated mice lacking the gene which creates the protein had a substantially reduced allergic reaction.

'Important research'

Mice treated with an experimental drug, which inhibited the protein, had no allergic reaction at all, the Nature journal reported.

Professor Andrew Wardlaw, the president of the British Society for Allergy and Clinical Immunology, said it was an important piece of research.

"Mast cells are central to allergies and if you can switch these off then it will be a major breakthrough.

"But a treatment for humans is still some way off. This is just one brick in the wall.

"Experiments on mice do not always transfer that well to humans."