More MS news articles for Jan 2002

DNA detective work may find mystery killers

19:00 13 January 2002
Emma Young
Data from the Human Genome Project could be used to identify the bacteria and viruses suspected to be behind many unexplained human diseases, say researchers in the US.

The team, led by Matthew Meyerson at the Dana-Farber Cancer Institute in Boston, will tackle the problem via a process of elimination. They plan to compare the known human genome sequence with sequences of all the active genes in samples of diseased tissue.

Once human DNA is removed from the equation, the team can investigate whether what is left contains DNA belonging a particular microbe - or DNA that looks like it might belong to an unidentified virus or bacterium.

Initial tests on samples of healthy and diseased tissue suggest the approach will work. Meyerson hopes it can be used to prove whether, as suspected, microbes cause many common human diseases, including coronary artery disease, diabetes, inflammatory bowel disease, multiple sclerosis and various cancers.

"I think this will become the method of choice for finding the causes of unexplained diseases," Meyerson says.

Infectious agents

Meyerson's team started by "subtracting" the known human genome sequence from more than 3.2 million "expressed-sequence tags" (ESTs) of general tissue samples. An EST represents all the genes switched on in a sample.

"We took the ESTs and compared them to the human genome, and we found sequences that didn't match. The idea was that these sequences could include infectious agents causing disease," Meyerson says.

Further research revealed that some of the unmatched sequences did indeed belong to known disease-causing microorganisms. Some of the unmatched sequences were unidentifiable and are likely to include sequences of microbes currently unknown to science, the team say.

DNA filter

The team then tested their microbe-finding technique by examining ESTs of cervical cancer cells. The technique "filtered" out human and other DNA to reveal sequences that match a virus known to cause cervical cancer.

If the team had not already known that the virus was linked to cervical cancer, the technique would have revealed it as a prime candidate for further investigation, Meyerson says.

Meyerson's lab is now creating libraries of ESTs of a wide variety of diseased tissues. "Our goal is to use this method to find new organisms," he says. But it could take some time, he admits.

The "sequence subtraction" part of the procedure could be done in as little as a week. But identifying a new disease-causing organism using a sequence of mysterious DNA could take anything from weeks to years, he says.

Journal reference: Nature Genetics (DOI: 10.1038/ng818)

© Copyright Reed Business Information Ltd