June 27, 2000

READING THE BOOK OF LIFE

Most Ills Are a Matter of More Than One Gene

By ERICA GOODE

	Susan Spann for The New York Times
	Dr. Neil Risch, in his office at Stanford, says scientists have learned that the likelihood of finding a single gene that causes major disorders like schizophrenia is very small.
	Related Articles • The Human Genome Project Video • The Composition of Life narrated by Nicholas Wade, science reporter, The New York Times. (Requires Macromedia Flash Plugin) 3D Interactive Images • DNA Replication Process (Requires Hypercosm Player) • The Scale of DNA (Requires Hypercosm Player) Biographies • Francis Collins: Dedicated Researcher, Family Man Heads Public Project • Craig Venter: A Maverick Making Waves Chronology • Timeline: Journey to the Genome Glossary • Genetic Terms Forum • Join a Discussion on DNA Research
	ADD YOUR THOUGHTS
	The Genome Project What are the implications of a decoded human genome? Add your thoughts and see what other readers are saying in Abuzz. Predict the uses humankind will find for this knowledge. Compare this achievement to other milestones in science.

n Feb. 26, 1987, a group of scientists made a thrilling announcement: they had succeeded in locating a gene for manic depression in a large Amish family in Lancaster, Pa.

The researchers' finding, published in the journal Nature, made headlines around the world and was hailed as the "first proof" of a specific genetic mutation underlying psychiatric illness. Mental health experts predicted that the discovery would revolutionize the diagnosis and treatment of severe mental disorders. And many patients and their families felt a new sense of hope.

Two years later, however, the scientists were forced to concede that the rejoicing had been premature. They had pinpointed the gene's probable location near the tip of the short arm of Chromosome 11, one of the 23 pairs of chromosomes that make up the human genetic library. But a re-analysis of the data, including new evidence from several family members who had since fallen ill, cast serious doubt on the results, making it highly unlikely that a gene was on Chromosome 11.

"We are sort of back to square one," said Dr. Kenneth Kidd, a co-author of the original Nature report, when the re-analysis was published.

More than a decade has now passed since the Amish study. But mental illnesses have proved as stubborn and as protective of its genetic secrets as the most challenging and complex medical illnesses. And scientists still have not conclusively identifying a gene for manic depression, schizophrenia, or, for that matter, any common psychiatric disorder except Alzheimer's, though no one doubts that heredity plays a strong role in many such diseases.

It is not for lack of trying. Within a year of the Amish study's publication, for example, two other teams announced equally promising discoveries: a strong link for schizophrenia on Chromosome 5, and one for manic depression on the X chromosome. But neither finding held up.

Over the last 10 years, scientists studying mental illness in the United States and other countries have published suggestive -- though not conclusive -- results implicating many different chromosomes. Most recently, investigators at Rutgers University and the University of Toronto reported, in the journal Science, a strong link to schizophrenia on Chromosome 1, while a Scottish research team also announced a schizophrenia link on Chromosome 1 -- in a different region.

Yet schooled by experience, geneticists no longer expect to discover a single, powerful gene for schizophrenia, manic depression or other disorders. Instead, most now agree that such illnesses are most likely "oligogenic," mediated not by a single gene, but by multiple genes of smaller effect acting in combination with one another and the environment.

And in hindsight, the Amish study seems less a fluke than an early object lesson in the barriers to uncovering the genetic underpinnings of such complex illnesses.

"These are turning out to be very difficult diseases," said Dr. Neil Risch, professor of genetics and statistics at Stanford University. "The original approach in psychiatry was to find one gene that by itself can cause the disease and we will understand more about it. What we've learned in the last 15 years is that the likelihood of the existence of such single genes is very small."

In many ways, the search for a gene for schizophrenia or manic depression, geneticists say, is no more challenging than efforts to decipher the genetics of other complex illnesses like diabetes, hypertension, ulcers or multiple sclerosis, which also involve an interplay among multiple genes and environmental influence.

Researchers still argue, for example, over how tightly to define many psychiatric disorders for the purposes of genetic studies, debating, for instance, whether to include patients with less severe symptoms. But similar debates can be found among scientists studying hypertension. And investigators in other areas have had their share of false starts and unreproducible findings.

Yet the challenges offered by psychiatric disorders are in some ways unique. While the presence of diabetes can be determined with blood tests, scientists have yet to find reliable biological markers for schizophrenia or manic depression.

And for some mental illnesses, there remains significant disagreement about how big a role heredity plays. Dr. Steven Moldin, chief of the genetics research branch at the National Institute of Mental Health, for example, said that 100 psychiatric researchers gathered in a room would all probably agree that schizophrenia has a strong genetic basis: the illness has long been known to run in families, and studies of identical twins indicate that heredity accounts for 50 to 60 percent of disease risk.

The case for genetic influence in manic depression and autism, Dr. Moldin said, is similarly strong.

But move on to other illnesses like obsessive compulsive disorder, eating disorders, phobias, alcoholism and some forms of depression, and the same 100 researchers, he said, would be split into camps, some arguing for major hereditary influence, others that evidence for genetic susceptibility is lacking.

"We don't yet have sufficient data in hand to know the magnitude of the genetic effects in those diseases," Dr. Moldin said.

Nevertheless, success in finding a gene linked to any mental disorder would have a great impact. In the ideal case, for instance, it might lead to the development of new drugs. That potential, said Dr. Kenneth Kendler, the Banks distinguished professor of psychiatry at Virginia Commonwealth University, "has not been lost on all the major pharmaceutical companies," especially in an era when psychiatric medications are fixtures on the drug industry's best-seller list.

A gene shown to be associated with many cases of schizophrenia or manic depression might also reduce stigma, aid in diagnosis, or be used as a screening tool, an application bound to be controversial, said Dr. Kendler, who with colleagues has reported linkage findings for schizophrenia on Chromosomes 5, 6, 8 and 10.

Yet there are less optimistic possibilities, as well. A gene, if found, could turn out to be associated with only a tiny percentage of cases of an illness. Or the gene could be involved in a way that affords no new avenues for treatment. And because heredity only sets the stage for manic depression and schizophrenia, while environment opens the curtain, identification of even a central gene will provide limited answers.

Meanwhile, for all complex illnesses, whether in psychiatry or other medical fields, the challenge remains the same: How can scientists isolate the influence of each of a number of minor genes acting together?

Dr. Risch of Stanford and others argue that the hunt for such genes may have been limited by the methods scientists have had at their disposal. The most popular technique for investigating disease genetics over the last two decades has been an approach called linkage analysis, which involves looking at the relationship between inheritance patterns and genetic markers on chromosomes in families affected by an illness.

Linkage studies have proved enormously proficient in locating genes for conditions like Huntington's and cystic fibrosis, where there is a one-to-one relationship between a mutation in a single gene and the development of the disease.

In such illnesses, the gene responsible sends out a strong signal, the mode of inheritance following the orderly laws of Gregor Mendel, the Bohemian monk who started modern genetics with his experiments on garden peas. Typically, each successive linkage study of a single gene illness offers more conclusive results, until the gene's location becomes clear. The cloning of the Huntington's gene, for example, was one in a rapid series of such discoveries for a variety of Mendelian illnesses.

But Dr. Risch and others argue that linkage is more problematic in complex diseases, where each of multiple genes exerts a small effect, sending out a weaker signal that may be much harder to detect, particularly without huge numbers of research subjects. Under these circumstances, adding more subjects to an analysis sometimes wipes out the finding, as happened in the Amish study.

As a result, it is anyone's guess how to interpret the combined findings of psychiatric linkage studies. Are there, for example, genes on each of the at least 10 different chromosomes identified in schizophrenia studies since 1987? Or are many findings false positives, artifacts of a limited number of subjects, problems in diagnosis, or other methodological complexities?

The last decade of research, said Dr. T. Conrad Gilliam, co-director of the Columbia Genome Center at Columbia University, "has been a great mental exercise."

"But in terms of what it's actually delivered, we may look back 15 years from now and say, 'Not a lot!' or we may look back and say: 'You know those little blips everyone obsessed about? Those were the signals,' " he said.

"The jury is out."

Scientists can allow themselves to hope that 15 years will make a difference in large part because of the sequencing of the human genome, whose effective completion was announced Sunday, and the advances in technology that have accompanied it.

"It's like anything where you have a complex reality," said Dr. Nelson Freimar, director of the center for neurobehavioral genetics at the University of California at Los Angeles, who studies manic depression in large Costa Rican families. "The more windows you can look in, the better your approximation of that reality will be."

With the genome in hand, researchers will for the first time be able to analyze groups of genes working together, looking at how they might be altered in schizophrenia, manic depression or other disorders. Scientists conducting genetic studies will also have a clearer idea of exactly which genes are in the region they are studying. And when, sometime in the future, a detailed map of the variations in human DNA -- individual point mutations, called single nucleotide polymorphisms or SNP's, in the chain of nucleotides that forms a strand of DNA -- becomes available, investigators will be able to directly compare specific chromosomal variations in people affected by an illness and in healthy subjects.

In addition, knowing where each gene lies, and ultimately what it does, will offer new possibilities to scientists trying to study genetic influence by breaking down mental illnesses into their biological components. Suggestive brain imaging findings -- shifts in transmission patterns of messenger chemicals, for example -- could be linked to genes, on the one hand, and to the overt symptoms of patients on the other.

Yet even the genome carries no guarantees. Genetic researchers may dream of finding a gene and boarding the plane for Stockholm to collect a Nobel Prize. But in their worst nightmares, they discover that the genes contributing to mental illness are so manifold, so tiny in their effects and so complex in their interplay that identifying them is a hopeless task.