Bugs that cause stomach upsets in travellers are leading scientists to a vaccine for diabetes and arthritis.
By Jonathan Amos
BBC News Online science staff, in Salford
UK researchers will soon start human trials of a drug derived from the toxin of a bacterium that causes diarrhoea in globe trotters.
Could bacteria make insulin injections redundant?
It has already been tested in mice successfully, reducing the incidence of illness in model rodents from 80% to about 15%.
"And in arthritis, we saw two things," said Dr Neil Williams, of the University of Bristol. "We saw not only a reduction in the incidence but also in the mice where disease did occur, we saw a decrease in severity.
Dr Williams has been presenting details of his work at the British Association's annual science festival, which this year is in Salford, Greater Manchester.
The researchers are targeting so called auto-immune diseases - diseases which arise when the immune system goes wrong and starts to attack the body's own components.
The Bristol team's work focuses on a group of proteins which have the effect of resetting control mechanisms, preventing the body's own tissues from being damaged.
"We've discovered a molecule from an E. coli bacterium that seems to be able to change the way the immune system works - in a sense re-educating it.
"This provides an example of where vaccines may be used to turn off rather to turn on the immune system."
The team is using just a transport component of the bacterium's toxin molecule. Called ETxB, the component is separated off from the rest of the protein so there is no chance of a vaccine causing stomach upsets in patients.
Scientists have discovered that ETxB on its own is sufficient to stimulate so-called T regulatory cells, which in healthy patients act to suppress auto-immune diseases.
The Bristol group hope that by giving an ETxB vaccine to patients recently diagnosed with rheumatoid arthritis and type 1 diabetes they can significantly slow the progression of the diseases.
"We hope to be in healthy volunteer trials in the early part of next year and then if that goes well, we hope to be in patient groups towards the end of the year.
"Diabetes and arthritis are where we have focussed but in the longer term we would hope that a similar approach would work in other autoimmune disease, such as multiple sclerosis and Crohn's disease."