Reading the Book of Jim
It would be a mistake to think that reaching the age of 79 has mellowed James Watson. Fifty-four years after he discovered, with Francis Crick, the structure of DNA, and 45 years after sharing the Nobel Prize for it, he delights in provocation just as much as when he made his reputation as the bad boy of molecular biology, bulldozing colleagues and competitors (and using crucial data generated by one, Rosalind Franklin) in his headlong race to the double helix. In the years since, Watson built Cold Spring Harbor Laboratory in New York into a biology powerhouse, briefly led the Human Genome Project—and endorsed designer babies, genetic engineering to make "all girls pretty" and curing "stupidity" through genetics. Which makes his words this rainy May morning at the lab all the more surprising.
Two years ago Watson agreed to become the first person to have his genome sequenced and made public. A biotech company, 454 Life Sciences, has now determined, from a blood sample, every one of the 6 billion chemical "letters" (designated A, T, C and G) that make up the DNA in Watson's cells. He will see his genetic blueprint on May 30. The next day it will be posted in a National Institutes of Health database for all the world to look at and, in the case of experts, deduce whether his genes have spellings (ATTCGT ... ) associated with diseases, intelligence, neuroticism, risk-taking, belief in God, shyness and all the other traits that biologists have linked to genes. "I always wanted to be a hero," Watson says almost apologetically; he's doing this to encourage others to have their own genome sequenced. He thinks it will "make people healthier" by giving them information that could prevent disease. But he has another hope. If personal- genome sequencing becomes widespread, he says, "it will make people more compassionate.
"We'll understand why people can't do certain things," he continues. "Instead of asking a child to shape up, we'll stop having unrealistic expectations." If a child's genome shows that his awkwardness or inattention or limited intelligence has a genetic basis, "we'll want to help rather than be mad. If a child doesn't finish high school, we treat that as a failure, as his fault. But knowing someone's full genetic information will keep us from making him do things he'll fail at." For a clue to this softer side of Watson, one need look no farther than his office next door to the room where he is speaking. There, along with the Nobel citation and other honors, hangs a poster-size framed photo of his two sons when they were little. One suffers from a mental illness that causes symptoms of both autism and schizophrenia.
It remains to be seen whether society will look more kindly on people, like his son, who are different if those differences are traced to DNA. Homophobia didn't exactly vanish after sexual orientation was shown to have a genetic basis. And the notion that genes are destiny raises possibilities more disturbing than Watson's impish suggestion that we use genetic engineering to beautify the female half of the species. Most obviously, people may believe that what is written in their code of life determines not only their health but also their intelligence, character, talents and personality. "Will our genetic profiles make us self-limiting, and will we allow them to?" asks Elaine Ostrander of the National Human Genome Research Institute, part of the National Institutes of Health.
Ready or not, personal-genome sequencing is just around the corner. "Project Jim," as 454 calls it, took 67 days of sequencing time and cost $1 million. But with new technology "we are on our way to the $10,000 genome and soon the $1,000 genome," says Jonathan Rothberg, 454's founder and chairman. And unlike other medical tests, which must be done regularly to detect changes, your genome need be sequenced only once.
If Rothberg is right about the $1,000, that would be cheaper than a battery of genetic tests, which cost between $300 and several thousand dollars each. A personal genome offers another advantage. Most conventional genetic tests probe for known mutations. "If yours is one of the rare ones you're out of luck," says geneticist Richard Gibbs of Baylor College of Medicine, who first suggested that 454 approach Watson about sequencing his genome. A full sequence, in contrast, can be compared to the benchmark genome—sort of an average of the genomes of the people who ponied up DNA samples for the human genome project—so if you have any misspelling at all it will be detected.
The biggest potential benefit of having your genome fully sequenced is likely to come from the fact that no gene is an island. With the exception of diseases that arise from defects in a single gene, such as Huntington's and cystic fibrosis, the effect of one gene depends on your other genes. "We know it's not just your genes or your environment" that determines your health, says 454's Rothberg, "but the other genes in your genome." After all, the fact that identical twins—who have identical genomes—are identical for disease only 60 percent of the time (on average) shows that having a particular DNA sequence doesn't mean you will definitely develop the disease. One reason not every woman with a BRCA1 mutation develops breast cancer (some 30 percent do not) is almost certainly that some carry "modifier" genes that weaken the effect of the disease genes. A full genome sequence would pick up these modifier genes. "I think if we can identify those modifiers," says Baylor's Gibbs, "we'll be able to tell if you're in the 30 percent or the 70 percent," which today's genetic tests cannot.
The value of a personal genome depends on its accuracy, of course. At this early stage the error rate is a huge unknown. Watson's biggest concern is that he'll be told incorrectly that he has one chemical letter instead of another—a C instead of a G, say—somewhere and that the misspelling is associated with a disease, raising alarms unnecessarily.
An equal problem is that gene-disease claims have a lousy track record. Of those for complex diseases involving multiple genes, notably mental illness, few have been confirmed. And when scientists have tried to validate a claim the results have been sobering. Geneticists led by Thomas Morgan of Washington University recently examined 85 variants (that is, "spellings" of A, T, C and G that are different from the norm) in 70 genes that studies had linked to an increased risk of cardiovascular disease. Exactly zero of the variants were more frequent in heart patients than in healthy people, they reported last month in the Journal of the American Medical Association. That means the variants do not increase risk of heart disease as claimed. Yet companies offer genetic tests for at least seven of them. "An unfortunate number of claims based on candidate [disease] genes have not held up," admits Francis Collins, who led the human genome project to the finish line in 2003 and is now director of the genome institute at NIH.
Even assuming the sequencing is accurate and the link to disease correct, the value of a personal genome sequence is debatable. At least 90 percent of the human genome is "junk DNA" that has no clear function. An even higher percentage seems to have nothing to do with health. Genetic variations linked to disease are sprinkled across 0.01 percent of the genome, estimates George Church of Harvard University, who has been pushing technology to make genome sequencing affordable. Those regions could be sequenced for $1,000, he estimates, "and would give you 95 percent of the heavy-hitting mutations" linked to health.
Watson hopes that if people learn they harbor genes that raise the risk of any diseases, they will take steps to minimize that risk starting from birth. That way, a kid at risk for type 2 diabetes, say, will keep her weight in check and thus her diabetes risk lower before she becomes even a pudgy toddler. (Watson, though, concedes that if he has a gene variant that raises the risk of some disease 20 percent, and that the risk can be lowered by giving up chocolate, count him out.) But it is not at all clear how people will react to knowing their genetic blueprints. Many are genetic fatalists, says Angela Trepanier, president-elect of the National Society of Genetic Counselors. That is, they believe that health and even intellectual and emotional destiny is written in their DNA. Rather than reacting to the news that they have genes that raise the risk of colon cancer by having regular colonoscopies, some will say, "To hell with it, I'm doomed anyway. Where's the cheesecake?"
That would be especially unfortunate given that many genes linked to disease raise the risk of that disease 20 percent or 40 percent or even 200 percent—but well short of "you will definitely get it." The overall chance of getting schizophrenia, for instance, is 1 percent; a 200 percent extra risk due to a genetic variation means your risk is still just 3 percent. And the uncertainty about the consequences of having a particular gene variant is even greater for traits such as aggression, neuroticism, shyness and intelligence. Fatalism when it comes to those could be tragic, making parents give up on kids who struggle academically or resign themselves to sociopathic behavior despite reams of evidence showing that DNA is not destiny. With the spectacular advances in genetics since the discovery of the double helix, says the genome institute's Lawrence Brody, "we've convinced the public that genetics is important and deterministic. Now we have to back off a little and say it's not that deterministic."
Not even experts are immune from the sovereignty of DNA, though. Watson remains scientifically active and intellectually engaged, only occasionally forgetting a name, he says. But there is one part of his genome he has asked 454 to keep private, even from him: whether he carries gene variants associated with Alzheimer's disease. He fears that if he knew he did, he would interpret each slip of memory as impending dementia.
Coincidentally, just as the personal genome arrives, so does another novel way to detect disease early. Called biomarkers, they are proteins in the blood that reveal the presence of disease before it even causes symptoms, offering a good chance for successful treatment. The PSA test for prostate cancer is a biomarker, but others promise to be much more useful (the PSA test has not been shown to save lives). "Genetic tests just tell you you have a chance of getting this condition sometime," says Mark Chandler, chairman and CEO of Biophysical Corp., which sells a $3,500 test that screens for 250 biomarkers indicative of scores of disease. "With biomarkers, it's not just potential anymore; it's real."
Despite the concerns over its usefulness, if 454 is right about the $1,000 or even the $10,000 cost, the personal genome sequence will likely become a must-have novelty—even if for now it's just "recreational genetics," as Brody calls it, amusing but not terribly useful. As he prepared to fly to Baylor for the official unveiling of his genome, Watson expressed nary a doubt that everyone should learn the 6 billion chemical letters that spell their own genetic blueprint. "It's going to be an extraordinary period," he says. For better or worse.