Errors in DNA don't seem to follow pattern

As if the science of how genetics leads to disease isn't already complex enough, researchers in Seattle and Long Island, N.Y., say individuals appear to develop schizophrenia from a varying smorgasbord of bad genes rather than common genetic flaws.

Scientists at the University of Washington and Cold Spring Harbor Laboratory report in Friday's edition of the magazine Science that multiple errors or deletions in a person's genetic code, or DNA, can lead to schizophrenia -- a psychiatric illness characterized by delusions and disordered thinking that today affects one of every 100 people.

The finding that multiple genes are involved is, by itself, not surprising, since other diseases or disorders are, or strongly appear to be, the result of many flaws rather than just a single bad gene. That fits nicely within the standard dogma of genetics.

What is surprising, challenging to the dogma and perhaps confusing to many experts who study the interplay between genetics and neuroscience, is that the UW-Cold Spring Harbor team found strong evidence that it's usually not the same set of genes going bad in people who develop schizophrenia.

"It's different genes in different people," said Dr. Jon "Jack" McClellan, a UW psychiatrist and a co-author of the report. This is a big challenge to the conventional wisdom, McClellan said, adding that he believes this could turn out to be the same for most other complex psychiatric diseases -- if not for all diseases that arise without a simple genetic flaw.

"The standard dogma is that any complex trait (such as mental illness) is going to be caused by the cumulative effect of multiple, common defects," said Dr. Mary-Claire King, a world-renowned geneticist and also one of the UW co-authors on the report.

"But that's not what we found."

It's long been clear, she said, that there had to be some sort of genetic contribution to this debilitating illness, since schizophrenia tends to run in families.

But the familial tendency still seemed kind of sporadic, King said, and there is the mystery of why such a self-defeating genetic disorder should exist at all. Considering the evolutionary process of natural selection, she said, schizophrenia genes should not persist because most of those who are afflicted tend to not produce children.

"The genes should have been selected out," King said. "There have been a lot of these paradoxes with schizophrenia."

King and McClellan worked with geneticist Jonathan Sebat and his Cold Spring Harbor colleague Shane McCarthy, along with the UW's Dr. Tom Walsh, Evan Eichler and others.

They employed some powerful, relatively new techniques of computerized genetic analysis to resolve these mysteries.

The goal was to examine the genome (an individual's entire genetic code, in humans more than 3 billion units of DNA) for patterns that might reveal some answers. The scientists examined the DNA of 150 people with schizophrenia (most of them patients at Western State Hospital) and compared what they found with the genetics of 268 people without the illness.

Assuming that the genes of most interest would be those involved in neural development and brain function, they looked for any differences. All individuals have some level of genetic errors or mutations in their genome. The researchers found that people with schizophrenia had flaws in brain-related genes 15 percent of the time compared with 5 percent in healthy people.

But it was never the same set of genes going bad.

"They were all different," King said. "The only way we could have found this was to look at the overall (genetic) profile."

While this certainly complicates the exploration of the genetics of schizophrenia, McClellan said it does reveal some common "neurological pathways" that may lead to better treatments for those with this mental illness. The many different malfunctioning genes in schizophrenics, he said, all play their parts within a limited number of brain-related functions.

Instead of chasing after individual genes to identify and develop new drugs for schizophrenia, McClellan said, this research suggests the focus should be on fixing the problems that arise -- owing to the widely varying genetic flaws -- on the biochemical pathways that govern brain development and function.

"This certainly shows why we should treat patients as individuals," he said. This shouldn't just be a marketing mantra followed by some generic, uniform therapeutic approach to illness, McClellan said, especially if -- as this study indicates -- individual variation occurs even at the most basic, genetic level of illness.

King said she recognizes that not everyone will be happy with, or even accept, the conclusions that they've drawn from this peculiar finding (even though the study was replicated by a team at the National Institutes of Health). It's not just about challenging dogma, she noted; it's also a challenge to a lot of established, funded studies already well under way.

"A lot of those approaches, based on the assumption that they are looking for common or shared genetic mutations, aren't going to work," King said.

P-I reporter Tom Paulson can be reached at 206-448-8318 or