IT'S EASY TO GET THE IDEA THAT scientists are discovering genes for just about everything known to afflict or bless humankind--deadly diseases like colon cancer, psychological problems like manic depression, curious endowments like immunity to a particular carcinogen. But this gene-of-the-week impression, with its implication that nature has trumped nurture in the contest to control human fate, belies the real new direction in genetic research: a quest for understanding when, and why, DNA is not destiny. What a geneticist sees, warn biologists, isn't always what you get. Paradoxically, disease genes do not always make you sick.
That admission represents an abrupt turnabout. Geneticists have long admitted that they would never find "the gene" for such defining traits as intelligence, or personality, or aptitude. But they have been almost smug in their confidence that they would handily identify the snippets of DNA responsible for such "single gene" afflictions as sickle-cell anemia, Huntington's disease, cystic fibrosis, muscular dystrophy and others. On paper, the gene hunters have scored resounding successes: they have traced to a particular spot on the 23 pairs of human chromosomes no fewer than 2,600 genes, some 700 of them for a disease (most of the others are for various enzymes and other biochemicals that make cells run). But now nature is pulling a fast one. Some of the genes don't operate the way textbooks claim: they turn deadly only in the presence of, say, sunlight; or they become ever more dangerous with each generation; or they cause no damage at all in one person but kill another; or they're benign if inherited from the mother but fatal if they come from the father. Having a gene may be necessary, but not sufficient, for developing a disease. The new genetics is throwing a dizzying ambiguity into the connection between DNA and health. As a result, concludes a new report by the Institute of Medicine (part of the National Academy of Sciences), "Genetic tests are seldom perfect predictors of [health] risk."
Nothing puts the promise and pitfalls of genetic medicine in starker contrast than the gene for cystic fibrosis. Discovered in 1989, the gene lies on chromosome 7. In its mutated form, it causes CF, afflicting 30,000 newborns a year with a disease that makes sticky mucus accumulate in the pancreas and lungs. Or so the simple version of the story goes. But it turns out that there are more than 300 possible mutations in the CF gene. Current gene-screening technology can detect only about 20 of them. As a result, some couples are told that there is no possibility their fetus can have CF, when in fact it does. "If there's a mutation there's no test for, you miss it," says geneticist Arno Motulsky of the University of Washington, who chaired the IOM panel. Conversely, some mutations assumed to cause CF turn out not to, at least sometimes. Geneticist Barbara Handelin of Integrated Genetics in Framingham, Mass., has found several men who have the supposedly fatal CF gene. But they don't show the slightest symptom of the disease. Her off-the-cuff hunch is that another gene plays maestro: like a conductor pointing to the instruments that can join in, this gene would determine whether the mutations are "heard."
Handelin may not be far off. Pioneers in a new field called "ecogenetics" are finding gene after gene that acts as a bad seed only in the presence of some environmental factor. The classic case involves the gene for the fatal disease xeroderma pigmentosum: the gene makes someone supersensitive to ultraviolet light, and so causes almost-always-incurable melanoma. But if you avoid sunlight (don't be a wheat farmer or a lifeguard), even with this "fatal" gene you could survive to 90.
Even better, people who do not carry some bad gene might have nothing to worry about from environmental factors that are otherwise life-threatening. For instance, a particular gene on chromosome 15, discovered by Daniel Nebert and colleagues at the University of Cincinnati Medical Center, turns the compounds in cigarette smoke into carcinogens. If a smoker does not harbor this gene, be or she would likely be one of the 93 percent of smokers who do not get lung cancer. (They would still have to watch out for heart disease and emphysema.) The absence of two other genes might let some people have their steak and char it, too. The genes make enzymes that convert chemicals in food into carcinogens, report scientists led by Dr. Fred Kadlubar of the National Center for Toxicological Research; someone who carries the bad genes and eats meat cooked too long or too fast has 10 times the average risk of getting colon cancer. But the genes are innocuous if the person never eats charred meat. Someone without the genes who eats blackened meat would face no more than the normal risk. "What is a good gene and what is a bad gene depends on how you treat it," says geneticist Charles Sing of the University of Michigan. "Genes don't wake up until they are exposed to some environmental factor."
Sing counts himself among the growing number of conscientious objectors to the genes-as-destiny dogma. "We're overselling," he says of geneticists who, particularly in pushing for the $3 billion project to map the human genome, conjure a miracle age of genetic medicine. "We're not going to be able to deliver [on promises to predict and prevent disease]. We're being dishonest."
OR AT LEAST NAIVE. SINGLE-GENE illnesses--sickle-cell disease is the most common--account for an underwhelming 2 percent of total diseases. And not even these genes stick to the rule book. One on chromosome 4, if mutated, causes Huntington's disease, the neurological crippler whose most famous victim is folk singer Woody Guthrie. The healthy gene starts with a stutter: up to 34 repetitions of three chemicals. in the diseased gene, the stutter is downright pathological, growing 42 to 100 repeats as it passes from parent to child. The more repeats, the earlier Huntington's strikes and the worse it is. just a few years ago the notion of a gene growing with the generations would have been laughed out of the lab. But wait, it gets worse. The stutter grows faster in cells that become sperm than in those that become eggs. If the gene passes from fathers to sons, within three generations the stutter becomes so long that Huntington's may strike at the age of 20. The same gene passed through women produces later, less severe Huntington's. Mama's genes stutter less.
Huntington's is just one of several genes marked with the molecular version of a blue or a pink ribbon. Psoriasis, diabetes and some forms of mental retardation all manifest themselves differently depending on whether the bad mutation comes from the mother or the father. Patients who inherit a damaged gene on chromosome 11 from the father, for instance, develop a neck tumor called paraganglioma; a child who inherits the exact mutation from the mother is fine. "It's the sort of stuff that's just not supposed to happen," says geneticist Judith Hall of the University of British Columbia. "But it does."
Not all genes are so tricky. Genetic tests given at birth for phenylketonuria and hypothyroidism, which can cause mental retardation if untreated, are close to foolproof Scientists have also nabbed genes for retinoblastoma (a blinding eye tumor) and for the rare Li-Fraumeni syndrome, in which mutations in a gene on chromosome 17 cause tumors in several organs. They can also detect genes that unambiguously cause Duchenne muscular dystrophy and a form of neurofibromatosis that covers patients in benign tumors. Earlier this year researchers led by Bert Vogelstein at Johns Hopkins University discovered a mutation in a gene on chromosome 2 that causes colon cancer in virtually 100 percent of the people who inherit it--even if they run marathons and eat like Zen Buddhists. The only salvation from this genetic predestination is to launch a pre-emptive strike: regular colonoscopies to detect precancerous polyps.
Cardiologists have the same hopes. Geneticists have already discovered genes for an inherited tendency to astronomically high cholesterol levels (carriers tend to die of heart attacks at the age of 20) and for Marfan syndrome, which kills its often athletic victims by stopping their hearts. But in most heart disease, the genetic picture is decidedly more complicated. Several abnormal genes on several chromosomes are at fault, and "many people with the same mutant genotype have no heart disease," says Dr. Richard Strohman of the University of California, Berkeley. "The presence of a genetic marker...is no guarantee that abnormality or disease will show up." In many people with disease, moreover, tracing the malady to a gene is impossible: as many as 20 genes may be related to hypertension, at least three to manic depression and more than 100 to cancer.
There is some good news, however, in the rubble of the simple model of genetics. Regardless of whether a gene works as Mendel said, regardless of whether it needs an environmental push to turn it on, physicians might still be able to fix it. in the first gene therapy three years ago, two little girls with an enzyme deficiency that devastates the immune system received genes for the healthy enzyme; both now have functioning immune systems. To help a 29-year-old woman with hypercholesterolemia--cholesterol levels in excess of 500--Dr. James Wilson of the University of Pennsylvania used viruses to ferry healthy cholesterol-regulating genes into her liver; 17 months later her cholesterol count is near normal. A virus that causes the common cold has carried a healthy version of the cystic-fibrosis gene into the airways of CF patients and shows signs of alleviating the disease.
But the best news has nothing to do with brave new cures. It is, quite simply, that our future is not written indelibly in our genes.