SEATTLE — Multiple sclerosis may be caused by a herpes virus, according to PathoGenesis Corp., a Seattle biotechnology company. The cause is not proven, but the finger of suspicion points to HHV-6, the virus that causes the babies’ disease known as roseola.

Like the chicken pox, roseola is something that almost everyone gets early in life — usually a fever so mild that the baby never sees a doctor. The chicken pox virus stays in the spine, and late in life can emerge and cause shingles. Likewise, the hypothesis is, the roseola virus can come back and cause MS — but only in one person in a thousand.

The company’s scientific team, headed by Dr. Glenna Burmer, lays out its findings in a paper published Monday in the Proceedings of the National Academy of Sciences.

Multiple sclerosis is a disease in which the neural wires that connect brain cells lose their insulation, ceasing to function. This loss occurs in patches called plaques that grow over decades. Symptoms typically begin with numbness, stumbling and partial blindness, and progress to total disability as the plaques knock out more of the brain function. MS is a disease with no known cause and no cure.

Joseph Brown, vice president for laboratory research at PathoGenesis, said that clusters of victims suggested an unknown infectious agent. Burmer’s team went to look for it by using a technique called representational difference analysis, which is able to spot the foreign DNA in a tissue sample by removing all the known DNA. Brown likens it to “finding a needle in a haystack by removing all the hay.”

Using autopsied brains of MS patients, they found the DNA for HHV-6. They also found it in nonpatients. What was eye-opening was when they went to look at exactly where it was: In slices of tissue stained with a monoclonal antibody to the virus, they found that in MS patients the virus was concentrated in bright stains at the edges of the plaques. In non-MS patients, there wasn’t enough of the virus to cause staining.

Such results aren’t enough to prove cause, but they suggest it. They also suggest that MS might be arrested if drugs could be used to control the virus.

“We hope that that proves to be the case,” said Brown. “We’re looking very seriously at the possibility of developing a drug.” His team is testing against HHV-6 drugs that work on other herpes viruses.

The company is not alone in its work on MS. Research at the University of Wisconsin has fingered HHV-6 for infections in AIDS and organ-transplant patients that has led to MS-like brain pathologies.

Founded in 1991, PathoGenesis has raised $63 million and remains privately held. It has two compounds in human trials, an aerosol for cystic fibrosis in phase 3 (final) trials, and a tuberculosis drug in phase 1 trials. It has 75 employees. It has filed for a patent on its MS work.