Th e gentle, rolling hills and beautiful farms of Lancaster County, Pennsylvania seem a most unlikely place to study the genetics of manic-depressive illness, but this lovely region is also the home of a special community of people, the Old Order Amish, a group whose way of life makes them of special interest to geneticists. For nearly 200 years, the Old Order Amish, descendants of devoutly religious, hardworking German immigrants, have farmed their land, clinging to most of their traditional practices even as modernity has surrounded them. Because all 12,000 of them are descended from about 30 persons who arrived in America in the early 1700s, and because they almost always marry within their small community, the Amish have a population profile that makes it easier to unravel the genetic component of any disorder. In addition, they have an increased prevalence of several rare genetic disorders, including a form of dwarfism, a problem that led them to Dr. Victor McKusick of Johns Hopkins University School of Medicine nearly 50 years ago. As gene mapping techniques improved, McKusick realized that they are an ideal group in which to try to map genes that predispose to complex disorders like mental illness. Mental illness is not really more common among the Amish; it is just easier to define the genetic influences.
During the early 1970s, one of McKusick's students, Dr. Janice Ege-land, now a psychiatrist at the University of Miami, decided to make the huge commitment of time and energy that it would take for an outsider to win the trust of the Amish so that she could study the problem of manic-depressive illness (also called bipolar illness or bipolar disorder) in their community. Dr. Egeland's relationship with the Amish, which took nearly a decade to establish, has been ongoing for nearly 30 years.
Serious depression, of which manic depression is one form, is among the most incapacitating, costly, and misunderstood illnesses in the Western world. A 1993 study estimated that each year 11 million Americans suffer bouts of clinical depression leading them to miss 290 million days of work. It estimated that this cost American business (in days lost from work and costs of care) more than $43 billion a year, ranking it second only to heart disease. Of the 11 million persons affected with clinical depression, the study estimated that 1.8 million had the manic-depressive form of the illness. More recent estimates have put the number at 2.5 million.
Everyone has had some experience with sadness and loss and can empathize with persons caught in the grip of severe depression. Manic-depressive illness is more mysterious. During the up or manic phase, patients may go with virtually no sleep for days on end. Gripped with what seems like euphoria, they spend hours planning and trying to carry out absurdly grandiose or just plain foolish tasks. One famous autobiographical description of mania comes from the writer, Clifford Beers. While at his most manic, he felt an irresistible compulsion to write letters. Regular stationary was much too confining. Given huge rolls of paper, he would write letters that stretched the length of corridors. One measured a full 100 feet. At times his output was measured at 1800 words an hour for hours on end. This bizarre case report resonates with an experience I had as a medical student. I became involved in the care of a man in a manic phase who had spent 48 consecutive hours just prior to his admission to the hospital repeatedly taking apart and reassembling a refrigerator that was in full working order.
Manic-depressive illness is not a classic Mendelian genetic disorder, but many individuals who are incapacitated by it also have relatives who suffer from it or from unipolar depression (depression without a manic side). The Amish are not at particularly high risk for manic-depressive illness, but there are good reasons why Dr. Egeland and other researchers would want to study the heritability of this complex disorder in them. They have large families who stay physically close through the generations, they shun alcohol and drugs (the regular use of which confounds efforts to diagnose clinical depression), and they are relatively inbred. This means that if several persons in the same family develop a disorder with a meaningful genetic component, it is far more likely that the hereditary contribution is due to the same gene in each case.
In the 1930s European psychiatrists began to study the prevalence of severe depression and manic-depressive illness in the general population. Studies in several different nations were remarkably congruent. They showed that about 1 in 200 adults suffers with a severe form of this mood disorder. To confirm their clinical impression that relatives of patients with manic depression were at higher than average risk for becoming ill with this disorder, many of the same psychiatrists conducted family studies. The key task in family studies is to compare the risk of having the same disease among close relatives of the index case (the first person to come to them for care) to the population at large. A dozen or so major studies conducted before 1960, again mostly in Europe, were in close agreement. First-degree relatives of persons with major mood disorders were about 20 times more likely than the general population to also be affected. Put another way, about 1 out of 10 of the parents, siblings, or children of persons with unipolar or bipolar depression also was afflicted.
During the 1970s and 1980s, more methodologically rigorous studies continued to find that close relatives of persons with bipolar disease were at much increased risk for the disorder. However, those studies found less evidence than had earlier investigations that close relatives of persons with unipolar disease were at higher than average risk for developing severe depression. Overall, by 1985 the various family studies strongly suggested that manic-depressive illness has a higher genetic load than does the more common unipolar depression. One important reason for the differences between the earlier and later studies was that the diagnosis of unipolar depression has in more recent times become more rigorous. Some of the persons who might have been given this diagnosis in 1940 would not be so labeled today.
In addition to family studies, another widely used research tool in psychiatric genetics is to determine the pair-wise concordance rate of disease among twins. This complex term describes a quite simple statistic that is arrived at by assembling a study group and dividing the number of twins both of whom have a disorder by that number plus the number of twin pairs in which only one is affected. A concordance rate of 1.0 indicates that the causes of the illness are completely genetic; a concordance rate of 0 indicates that there is no detectable genetic influence. By comparing concordance rates between fraternal twins who share half their genes and identical twins, researchers derive a crude measure of the heritability of a particular phenotype. Even allowing for environmental influences, high concordance rates among identical twins are thought to suggest a strong genetic component. Recently, much energy has been expended in studying traits in identical twins reared apart, all in an effort to refute arguments that many behavioral similarities among twins are due to the impact of common familial environments. Twin studies of risk for bipolar depression, most of which were done before 1960, found a much higher concordance rate among identical twins (about 0.8) than among fraternal twins (about 0.1 to 0.2), again suggesting that manic-depressive illness has a strong genetic contribution.
Another epidemiological tool used to study the heritability of disorders is adoption studies. The usual approach is to identify a group of children placed early for adoption who were later diagnosed with a particular disorder and then compare the presence of that disease among the adoptive parents with its presence among the biological parents. The genetic hypothesis supposes that the biological (non-rearing) parents will be significantly more likely to be affected than the rearing parents. There have only been a few such studies in manic-depressive illness, and they give only lukewarm support to the genetic hypothesis.
During the early and mid-1980s, molecular biologists made great strides in locating DNA markers along the various human chromosomes. These molecular mapmakers revolutionized the search for disease genes. Clinical researchers could now investigate the presence or absence of disease in extended families and correlate the diagnosis with the presence or absence of particular DNA markers. If affected individuals across several generations virtually always have the same pair of markers, it is highly likely that the disease is strongly influenced (if not directly caused) by a gene located between those two markers. The first and easiest chromosome on which to investigate the presence or absence of disease genes is the X, because genetic disorders originating here typically affect only males. There have been many studies testing the hypothesis that some cases of manic depression are due to X-linked genes. Although the issue is not settled, overall the evidence is not particularly impressive. One Israeli study did find strong evidence of an X-linked form of the illness in five Jerusalem families.
In 1987 Dr. Egeland and her colleagues published a paper in Nature which offered substantial evidence that a gene predisposing to manic-
depressive illness in the Amish was located in a small region of chromosome 11. The paper was based on the clinical and DNA-based study of 81 individuals in a single extended family. The psychiatrists concluded that 14 of the relatives had a definite mood disorder, including 11 with manic depression. The report, among the first efforts to scan the entire human genome to look for evidence of gene linkage, suggested that the illness was caused by a gene with a dominant mode of transmission and a moderately high degree of penetrance (chance of causing the disease to manifest). So powerful was the evidence that the scientists urged, "The demonstration by a linkage strategy that a simple genetic mechanism can account for the transmission of bipolar affective disorders in pedigree 110 should provide an impetus to analogous research on other common clinical conditions."
The report generated huge interest among scientists and the general public. The fact that two other papers published in the same issue of Nature did not find evidence of a gene for manic depression on chromosome 11 gave little pause. For such a complex and common disorder, it would not be surprising to find genetic heterogeneity (that changes in any one of several different genes could be causative). As predisposition to manic depression quite possibly arises due to one of several biochemical defects, it is likely that several different genes are involved. Some editorials hailed a new era in our ability to understand mental illness. On the other hand, as one put it, the news left the imagination "dangerously unfettered." Unfortunately, the scientific triumph was short-lived.
Less than two years later, further studies of the same Amish family forced the scientists to conclude that their original findings of linkage with a gene on chromosome 11 were spurious. In reanalyzing the family, the scientists (including some of the original team) extended the pedigree in two directions to add 37 individuals, diagnosed 2 more family members with bipolar disorder, and performed more detailed DNA marker studies. The crucial event was making the diagnosis of the condition in 2 persons who had DNA markers indicating they should not have the disease. The new findings now actually excluded chromosome 11 as a possible location for a predisposing gene! As Ken Kidd, a population geneticist at Yale who took part in both studies, put it for the New York Times, "It means we are sort of back to square one."
Twelve years have elapsed since Kidd's remark, and despite dramatic advances in gene mapping techniques, no one has yet found convincing evidence of a gene that predisposes to manic depression. Does this mean there is none? No. What it does mean is that investigating the genetics of mental illness is a devilishly difficult challenge, as will be the case for parsing the role of genes in any disorder that must often be influenced by strong environmental influences, has a wide range in age of onset and clinical manifestations, and about the diagnosis of which there is considerable disagreement even among experts. For now, we are left in the disconcerting situation that we know manic-depressive illness has in some cases an important genetic component, but we do not know what it is. In counseling families, we are reduced to quoting empirical risk figures that may or may not be relevant to them.
Ultimately, we will find and understand the genes that contribute to manic-depressive illness. One dramatic gesture made in the conviction that the genetic contribution is significant was the decision late in 1993 by the Charles A. Dana Foundation to commit $2.5 million to finding the culprit genes. The gift is being used by Johns Hopkins University, Stanford University, and Cold Spring Harbor Laboratory to conduct a painstaking search. The most important feature of the project was the task faced by the psychiatrists at Johns Hopkins of collecting 50 families in which there are at least three individuals with an unequivocal diagnosis of bipolar disorder. That took nearly four years. With this goal met, molecular biologists at Stanford began using hundreds of DNA markers to look for associations within the families of the presence or absence of disease with certain combinations of markers. The combined clinical and molecular database is stored and maintained at a laboratory in Cold Spring Harbor, New York.
One can make a fair argument that schizophrenia, not heart disease or cancer, is the most devastating illness in the western world. First well-characterized at the turn of the century by the German psychiatrists, Emil Kraepelin and Eugene Bleuler, who coined the term in 1911, the full-blown illness is marked by hallucinations, delusions, disorganized thought, neg ativism, and inappropriate, often flat, emotional responses. Too often forgotten today is the fact that Bleuler spoke of the disease in the plural. He recognized at least four types and speculated that heredity played an important role in each.
During the 1950s, schizophrenia accounted for more days spent in hospital in the United States than any other disease. Because of the development in the 1950s and 1960s of an array of antipsychotic drugs, hundreds of thousands of persons who in an earlier era would have spent their lives in institutions can now be cared for at home or are completely independent. Nevertheless, the disease still probably accounts for more days in hospital than any other. The background lifetime risk of developing the disorder, which typically first manifests during the teenage years, is about 1%.
Although schizophrenia typically appears as a single affected person in a family with no history of the disorder, there is a large and important subgroup of cases that are highly familial. Psychiatrists have repeatedly studied the familial clusters. They estimate that the offspring of couples in which one parent has schizophrenia have about a 10-15% risk of developing the disease. In the relatively few marriages in which both partners have the disorder, the risk to children approaches 40%. The risk to siblings of affected persons is about 5-10%. The risk for uncles and aunts, nieces and nephews, and first cousins is about half of that, nicely in keeping with a genetic hypothesis.
Twin studies have consistently found a much higher concordance rate among identical than among fraternal twin pairs. Identical twins have shown a lifetime concordance of 60-70%, whereas the lifetime risk of a fraternal twin whose co-twin is affected is about 10-15%. Of course, this is still much higher than the risk among the general population. The relatively few adoption studies of schizophrenia also generally support a genetic contribution to risk, but it is only modest.
As they had for bipolar disorder, advances in molecular biology also stimulated scientists to undertake linkage studies to look for a gene that predisposes its bearers to schizophrenia. In 1988, on the heels of the report on bipolar disorder, came a study, also published in Nature, asserting that there was a gene on chromosome 5 that conferred susceptibility to schizophrenia. The research was stimulated by the highly unusual case report of a Chinese man and his nephew who suffered from schizophrenia and had a small portion of a part of chromosome 5 stuck on to chromosome 1. One logical explanation for their schizophrenia was that the chromosomal rearrangement had damaged a gene, located at the point where the DNA had broken, which normally protected against this illness.
Focusing their mapping effort on chromosome 5, the scientists looked at five Icelandic and two British families burdened with schizophrenia across at least three generations. Psychiatrists interviewed 104 family members and diagnosed or confirmed the diagnosis in 39 persons using one standard, and replicated it in 31 of them using a more rigorous diagnostic standard. They also found many other persons who had obvious psychiatric problems, but who did not satisfy the research criteria. Using a statistic called LOD score (logarithm of the odds) analysis, the team found strong evidence that the risk of schizophrenia was compatible with having inherited a highly penetrant, dominantly acting gene on chromosome 5. They were quick to acknowledge that it was unlikely that the finding would be generalizable, as was apparent from a companion article that had found no linkage to 5 in a Swedish family.
The Nature paper generated wide interest among both scientists and the public and reignited the century-old debate about the relative contribution of genes and environment to the disease. The research, which was performed by a team of British scientists led by Dr. Hugh Gurling at the University of London, was sufficiently impressive that Eric Lander, one of the leading voices in gene mapping, concluded that it demonstrated "for the first time that at least some cases of schizophrenia are apparently monogenic." An editorial in Nature that accompanied the publication opined that "when the gene concerned has been located exactly, and perhaps its nucleotide sequence determined, it will also be possible to embark on studies of the mechanism of the disease, presumably biochemical in character, that may (with luck) be more generally applicable." Elsewhere, it described the schizophrenia in the Icelandic families as "genetically determined."
After the publication of the paper in Nature, many groups quickly looked for a linkage to chromosome 5 among the families they were study ing. None could replicate the linkage. Less than three years later, after conducting further research with the Icelandic families, Dr. Gurling and his colleagues were forced to concede that the original findings indicating the presence of a gene for schizophrenia on chromosome 5 were erroneous. Like the story with bipolar disorder, this work shows that powerful statistical evidence of linkage within even a large pedigree fades quickly if two or three individuals are found in whom the DNA analysis does not fit with clinical diagnosis. By 1992 some scientists were arguing that the original chromosome 5 linkage report had sent the research community on a wild goose chase that had for three years slowed progress in understanding the genetics of schizophrenia.
The search continued, and in late 1993 a team studying the genetics of schizophrenia in Irish families began to alert colleagues that they had found evidence of linkage to the short arm of chromosome 6. Even before the work on the Irish families was published, other teams were rushing to replicate the findings. By this time, so much progress had been made in adding reference points to the map of the human genome that scientists had been able to develop a new approach to search for genetic linkage called two-stage scanning. This brute force technique first looks for linkage in families singled out as most likely to have a predisposing gene, such as the Icelandic families. It uses them as a template to identify areas in the genome that give hints of linkage and then seeks to replicate the findings in a completely different set of families. In 1995 the results of a two-stage genome-wide search also gave tantalizing evidence of linkage to the short arm of chromosome 6 (as well as several other areas). The paper appeared in Nature accompanied by three shorter articles on the same topic. Two groups reported that they had not found evidence of linkage to chromosome 6 in their families; one group reported that it had.
In the fall of 1996 at the annual meeting of the American Society of Human Genetics in San Francisco, 60 of the world's foremost researchers on the genetics of schizophrenia, many of whom had traveled thousands of miles to attend, convened a special session on chromosome 6. The results were both intriguing and frustrating. For every researcher who summarized evidence in favor of linkage, there was another who reported that his or her group had been unable to find linkage in their patients. Despite longstanding and strong evidence from family, twin, and adoption studies and an immense effort using ever more powerful mapping techniques, we have not been able to isolate a gene that predisposes to schizophrenia. Chromosome 6 still looks like a fair bet, but it will take a few years yet to find the culprit genes and establish their role.
What about the future? Have no doubt: There are genes that predispose to schizophrenia. We will map, clone, and sequence them before the year 2005. Given the huge number of people afflicted with this disorder, the news that predisposing genes have been found will stimulate the pharmaceutical industry to direct immense resources to studying the proteins coded for by these genes. Several will compete in a billion dollar research race to gain new insight into how to treat these disorders. Such investment will be amply rewarded if it results in even a single new and more effective medication.
If it turns out that mutations in one or more genes strongly predispose individuals to be at risk for schizophrenia, the use of predictive tests to assess that risk will raise immensely difficult clinical and ethical issues. Schizophrenia often begins in adolescence. Imagine this not unlikely scenario. Researchers demonstrate that in about 10% of cases of schizophrenia, almost always those in which there is a family history, patients have a mutation in a particular gene that codes for a protein that is active in brain tissue. A genetic testing company develops a DNA-based test for this mutation. Clinical research shows that persons with the mutation have about a 30% chance of developing schizophrenia before age 18, but that the other 70% of carriers do not become ill. About the same time, a pharmaceutical company develops a new medicine that proves to be moderately effective in preventing the onset of the disease if given to children who are genetically at risk, but that has unpleasant, though not life-threatening, side effects.
Should children with a family history of schizophrenia be tested? When? Will a positive result alter the way their parents treat them? If teachers learn these kids are "carriers of the schizophrenia gene," will it shape how they treat them in the classroom? Should the children at risk be given the medication that reduces the risk of developing the disorder, but has serious side effects? Is it more dangerous to the child to be labeled a carrier than to not be tested at all? This sort of scenario is likely to unfold repeatedly as we learn more about the genetics of psychiatric disease. No one has the answers to the questions it poses, but somehow we must resolve them. Hopefully, well-crafted research will show us how to combine predictive tests with information that suggests the best intervention to maximize benefit for each individual.
Great Dane and pug, illustrating wide variation within a species. (© Jeanne White, Photo Researchers, Inc.)
Was this article helpful?