The natural trajectory of human genome research is toward the identification of genes, genes that control normal biological functions and genes that create genetic disease or interact with other genes to precipitate hereditary disorders. Genes are being localized far more rapidly than treatments are being developed for the afflictions they cause, and the human genome project will accelerate this trend. The acquisition of genetic knowledge is, in short, outpacing the accumulation of therapeutic power -- a condition that poses special difficulties for genetic knowing.

Our expectation is that the characterization of a disease- instigating gene will greatly assist our understanding of how and why it causes a malfunction in the body. It makes good sense to go to the root of the problem. But to learn a gene's secret, first you must find it. And finding it is not so simple. It is much easier to locate the neighborhood in the genome where a gene resides than it is to determine its exact address.

Lilliput and Brobdingnag: Beyond Gulliver's Travels. The magnitude of the challenge arises from the vast amount of DNA contained in the diploid human genome, which includes all of a person's genetic material. If strung out, the DNA in a single human genome would stretch to about two meters, but the diameter of the strand would amount to only about two billionths of a meter, 20 angstroms, a span a hundred times smaller than a wavelength of light. If the DNA from a single cell from every human being on the planet -- 6 billion people -- were stitched end to end, the resulting string would girdle the earth about 300 times. If the genomes from every cell of the 6 billion people were laid out end to end, they would extend 700 billion, billion miles -- enough to wrap around our galaxy more than 700 times.

To understand the enormous problem of finding a gene somewhere on an individual's strand of DNA, imagine that a single human genome is long enough to circle the globe. On this scale, the amount of DNA in a chromosome would extend for a thousand miles. A gene would span just one twentieth of a mile, and a disease-causing defect -- a point mutation, a change in only one DNA base pair -- could run as short as one twentieth of an inch. What we are thus searching for is comparable to a fraction of an inch on the circumference of the globe! In this immense morass of DNA, finding the exact address of a gene and pinpointing its fault makes for extremely tough going, and it requires all of the creativity and ingenuity of everyone engaged in the quest.

Hunting for Huntington's. The spectacular difficulty of the problem has been made painfully clear by the search for the gene causing Huntington's disease (HD). Huntington's disease is a movement disorder -- causing uncontrollable jerking and writhing movements of all parts of the body, called chorea. Even more distressing to patients and families than the obvious movements, it is preceded or accompanied by cognitive changes leading to profound intellectual deterioration and frequently severe emotional disturbances, usually suicidal depression and occasionally hallucinations and delusions. The disease runs a course of about fifteen to twenty-five years and is inevitably fatal. Its usual onset is between the ages of thirty-five and forty-five, but it can start as early as two and as late as the early eighties, an age when it can be hard to detect. The later the onset, the milder the symptoms. If the diagnosis is missed in an elderly person, manifestation of the disease in the next generation may appear erroneously to be due to a new mutation. No treatments are known beyond some marginal and temporary palliation for the movements and anti-depressants for the psychiatric symptoms.

Huntington's disease is the product of a gene transmitted in an autosomally dominant inheritance pattern -- in other words, a gene that occurs on one of the twenty-two non-sex human chromosomes and whose effect dominates its normal partner. It is entirely penetrant, which means that if a gene-carrier lives long enough, the disease is inexorably expressed.

One peculiarity of Huntington's disease is that the sex of the parent transmitting the abnormal gene seems to play a role in determining the age of disease onset in offspring. Children, both male and female, who fall ill when twenty years old or younger almost invariably have inherited the disease from their fathers. Whether a gene is passed on through an egg or sperm sometimes affects its level of expressivity, a phenomenon called "imprinting." (One possible explanation for imprinting is that the number of methyl groups added to a gene vary, depending on the sex of the parent passing on the gene.) This differential expression may, in turn, influence the timing of disease onset. Or other modifying genetic factors may alter the timing and expression of the HD gene. The identification and manipulation of these factors may lead to early therapeutic measures: if disease onset could be pushed until later in life, the illness might not be so onerous.

The fact that the development of new therapeutics for Huntington's disease and other hereditary disorders may require the pursuit and characterization of many normal as well as abnormal genes underscores the need for a unified and concerted effort such as the human genome project. For Huntington's disease therapeutics, genes that determine critical chemical pathways affected by the gene defect may prove more amenable to intervention than the HD gene itself.

"Riflips" to the rescue. In looking for the fraction of an inch responsible for HD on the globe of DNA, we get some very clever help from restriction enzymes that identify small, normal variations in DNA called restriction fragment length polymorphisms -- the RFLPs that geneticists pronounce as "riflips." Whenever a restriction enzyme sees its unique recognition site, it cuts the DNA right at that spot, like a miniature pair of scissors. The locations of these sites vary among individuals, and, as a result, the DNA fragments between two sites differ in length. When DNA is cut with restriction enzymes, these differences in fragment sizes can differentiate one person from another, one chromosome from another, and they are inherited, just as genes are. "Riflips" act as markers in a person's DNA, a telltale indicator of genetic identity. (There are now many new kinds of very informative markers that do not require restriction enzymes but still serve the same function of identifying specific regions in the DNA.)

When we began the search for the HD gene, we were looking for a RFLP marker that was close to it. We can get an idea of the relative closeness of a marker and a gene on a chromosome because of a process called recombination -- the tendency of segments of paired chromosomes (one from the father, the other from the mother) to change places during the creation of gametes, a kind of genetic "do-si-do." The further apart the marker is from the gene, the more likely it is that one of these "recombination events" will separate them; the closer together, the less likely. For every one million base pairs of DNA, there is a 1 percent chance that a recombination event will take place. Counting the number of recombination events found gives you a fairly good estimate of genetic distance between two markers or a marker and a gene. (I explain recombination probabilities to myself by imagining an earthquake at the North Pole where thousands of penguins occupy a huge ice floe; when the ice breaks up, two penguins sitting next to each other are more likely to stay on the same little piece of ice while two penguins far away from each other will drift away on separate pieces down each half of the globe.) If one of these penguins is a DNA marker and the other the HD gene on the same floe, the two will travel together. If the "penguins" are close on the same chromosome, they will be transmitted to offspring in a Mendelian fashion with a high degree of regularity. So if a mother with Huntington's disease has a pattern-A RFLP next to her HD gene, and the father, who is not affected with Huntington's disease, has a pattern-B RFLP next to his normal gene, then their children with the B pattern will most likely not inherit HD and those who have inherited the HD gene will show the A pattern. (Because of the possibility that a recombination event will separate and rearrange markers and genes, we can only say "most likely.")

In 1979, when the search began for the HD gene, the idea of "mapping" genes using RFLP markers was totally new and thought to be whimsical, if not heretical. No one had actually located a gene using DNA markers, although genes had been found by virtue of the proximity to "traditional" markers, red blood cell antigens and proteins. When we began discussing using this recombinant DNA strategy, only one RFLP marker was known; the claim that the myriad of markers existed in the human genome required to place some near every gene of interest was only theoretical, an extrapolation from other species. We had to hope against hope that new markers could be fashioned quickly and that one of these markers would lie close to the HD gene on the chromosomes of persons with the disease.

Needless to say, several knowledgeable scientists told us that we were crazy to look for the gene in this haphazard, hit-or-miss fashion. They predicted it could take fifty years or longer to find our target. What we were proposing was equivalent to looking for a killer somewhere in the United States with a map virtually devoid of landmarks -- no states, cities, towns, rivers, or mountains, and certainly no street addresses or zip codes -- with absolutely no points of demarcation by which to locate the murderer. Our critics said "wait until a more detailed genetic map is available, one with many more regularly spaced markers." This is, of course, a much better strategy if you have the time to wait. But we are in a race against the Huntington's disease killer and have no time to spare.

Venezuela bound. In 1979, despite such sensible advice, we began hunting for the Huntington's disease gene. We knew that just finding the location of the gene would not tell us anything about the nature of the gene defect itself. But we reasoned that if we could close in on the gene from either direction, using markers more and more closely spaced until we finally honed in on the gene itself, we could then direct all our energies and resources toward identifying the gene defect and developing therapeutic interventions. If we were incredibly lucky, the markers could tell us, "your murderer is living in Red Lodge, Montana," and then we could continue the hunt door-to-door.

The only way you can tell if a marker is close to an unmapped gene is to observe if the two consistently "travel" together in a family. We know certain people have the HD gene because they are sick. We were looking for people who had the disease to have one form of a marker and their unaffected relatives to have another form of that same marker. We needed to study large families, as the HD gene itself might vary in its locale from family to family, and the markers, having nothing to do with disease itself, would certainly vary across families as to which form of the marker traveled with the gene.

So we were looking for a large extended multigenerational family in which we could observe many instances of the Huntington's disease gene or its normal counterpart being passed on -- and we knew of just such a family, although when we began we had no idea how huge and important this family would eventually turn out to be. Members of the kindred live in Venezuela in three rural villages -- San Luis, Barranquitas, and Laguneta -- on the shores of Lake Maracaibo. Because it is situated in the northern region of Latin America and Lake Maracaibo is actually a huge ocean gulf, Venezuela has long communicated directly with Europe, and many European genes have appeared in the local population. Story has it that some sailor with Huntington's disease came over to trade and left his legacy, but we do not know if this is apocryphal.

We have been able to trace the disease as far back as the early 1800s, to a woman appropriately named Maria ConcepciÛn. We know that Maria lived in the "pueblos de agua," villages built on stilts in the water next to shores too marshy, jungly, and inhospitable to accommodate human life. Laguneta, where many of Maria's descendants live, remains such a stilt village.

Maria was the founder of a kindred now numbering close to 11,000 people, living and deceased. In the pedigree, there are 371 persons with Huntington's disease, 1,266 at 50 percent risk and 2,395 at 25 percent risk for the disease. Of the 11,000, 9,000 are living and the majority are under the age of forty. In these small and impoverished towns, we estimate that there are over 660 asymptomatic gene carriers who are too young to show symptoms, but as years pass -- if no treatment is found -- they will surely die. It is crushing to look at these exuberant children full of hope and expectation, despite poverty, despite illiteracy, despite dangerous and exhausting work for the boys fishing in small boats in the turbulent lake, or for even the tiny girls tending house and caring for ill parents, despite a brutalizing disease robbing them of parents, grandparents, aunts, uncles, and cousins -- they are joyous and wild with life, until the disease attacks. Every year we add more people to the pedigree who will suffer, diagnose more new cases, and watch helplessly as more and more begin that sad journey toward deterioration and death. It is impossible to be immune to their plight. It is only possible to be passionate and driven and work desperately to save as many as we can before it is too late.

My original interest in this family was sparked by its pockets of consanguinity. The kindred is now so numerous and extensive, spanning eight generations, that it is not particularly inbred. It also has a tumult of Hispanic and other European genes intermingled with indigenous Indian genes to make for a very rich genetic mixture. But in some branches of the families, gene carriers inter-married and produced offspring with a 25 percent chance of inheriting the HD gene from both parents, that is, being homozygous for the disease. We were hoping to find more direct evidence of the biochemical cause of the disease by studying a homozygote, who would have no normal gene product to mask the workings of the defective gene. When I first went to Venezuela, in 1979, it was only to look for such families. We did find a large family in which both parents were affected, drew blood samples from these family members, and went home thinking we would have a very small study.

Anatomy of a gene search. By 1981, however, we had changed our rationale for research with the Venezuelan family from searching for homozygotes to a full scale genetic linkage project using DNA markers. I went to Lake Maracaibo with a small research team in March 1981 for what proved to be the first of an annual month-long expedition ever since. We were assisted in all aspects of our work by an extraordinary physician, Dr. AmÈrico Negrette, who first correctly diagnosed Huntington's disease in this population and began constructing the pedigree, and two of his students, Dr. RamÛn ¡vila GirÛn and especially Dr. Ernesto Bonilla, who continues to be an active investigator. Our first task was to extend the pedigree that our Venezuelan colleagues had begun. Most couples represented in the pedigree are not legally married. Our pedigree is entirely composed of oral record, but just asking people to name their relatives has worked out pretty well. We check for possible nonpaternity but have found the rate lower than in the United States because people readily identify the father, even if he is not the mother's current partner.

Mapping the Huntington's disease gene required obtaining blood samples from as many relatives as possible in families in which the disease was traveling, according to certain guidelines for collection. Our Venezuelan colleagues warned us that the family members might be reluctant to give blood or refuse altogether; for many it was the first time ever. We have been privileged to receive extraordinary cooperation, despite people's fears. Most gave on behalf of their children. Although they may not know the exact odds, kindred members are experts on this disease and have a keen sense of the threat to their offspring. As birth control has become more available to them, it is being used increasingly.

I felt it was important for the Venezuelan family members to know that Huntington's disease was also in my family and in many others in the United States but that our families were not large enough to offer the gift to research that their family was capable of providing. We needed their help to find a cure. At that time we were doing skin biopsies, which I also had done for research. The family members were dubious of my story until I pointed out my skin biopsy scar and my wonderful colleague and friend, Fidela Gomez, a Florida nurse, grabbed my arm and dragged me around the room shouting, "She has the mark, she has the mark!" My mark and I became something of a passport for our research team and its activities.

Since the blood had to be sent immediately to James Gusella's laboratory in Boston, we could draw blood only when a team member was leaving and could hand-carry the samples. Those days became known as "draw days" -- chaotic days in boiling hot, deafeningly noisy rooms jam-packed with people of all ages, days spent going to sweltering homes where throngs of children would shout out in gleeful horror the number of tubes of blood we were drawing. The men tended to be more recalcitrant than women, fearing they would lose some vital bodily fluid and be weakened, or unable to work or drink, if they gave blood.

Genetic jackpot. At Massachusetts General Hospital, DNA was extracted from blood samples from the Venezuelan family members. Jim Gusella was also studying a large American family with Huntington's disease from Iowa. He searched the DNA from these two families for a telltale marker, helping to develop what were to become standard laboratory procedures in such ventures. Jim sliced up each person's DNA with restriction enzymes. He then developed markers, RFLPs, which he made radioactive. These markers were called anonymous because he did not know on which human chromosome they were located, only that they were in one unique spot in the genome, just like a gene, and they came in several forms so that individuals could be differentiated from one another. The fragments of chopped-up DNA from the family members were put on a gel that separates fragments on the basis of size. The radioactive probe (denatured, or single- stranded) was then added. When the probe is radioactive, it would "light up" where it was stuck on the gel, revealing distinctive bands. One would then need to check if a certain pattern of bands appeared only in individuals who had the disease and another pattern in their relatives who were healthy. If this difference was true more often than would be expected by chance, it would be very likely that the marker and the gene were close together on the same chromosome.

We all expected that the detection of a marker linked to the Huntington's disease gene would require thousands of tests and probes, but the third probe that Gusella characterized and the twelfth one he tried hit the jackpot. He began with the Iowan family, whose samples were the first to be collected, and the probe, called G8, was weakly positive, but not significantly so.

This finding gave him the crucial push, however, to try G8 in the Venezuelan family -- and it was the only probe he needed! It immediately showed odds far better than 1,000 to 1 that it was very close to the HD gene. P. Michael Conneally at Indiana University performed the computer linkage analyses that definitely proved that this probe and the HD gene were close neighbors. Almost all the Venezuelan family members with HD had one form of the marker while their healthy relative had another. At the time linkage was discovered, the chromosomal location of the probe was unknown, but it was quickly mapped, using in situ hybridization and other techniques, to chromosome 4. By inference, the position of the gene was mapped as well. Out of 3 billion possible base pairs on 23 chromosomes, we now knew we were a mere 4 million base pairs below the culprit gene way on the very top of the short arm of chromosome 4. We triumphantly announced the feat in an article in Nature, in November 1983.1

It had taken us just three years -- an astonishingly short time -- to localize the HD gene. Our critics and even our supporters said, rightly, that we had been incredibly lucky. It was as though, without the map of the United States, we had looked for the killer by chance in Red Lodge, Montana, and found the neighborhood where he was living.

An elusive prey. Next we needed to find the exact location of the HD gene, isolate it, and learn its secret. Since January 1984, the Hereditary Disease Foundation has supported a formal collaboration of seven scientists around the country and the world searching for this gene: Francis Collins, at the University of Michigan; Anna Maria Frischauf and Hans Lehrach of Imperial Cancer Research Fund, London; Peter S. Harper, University of Wales College of Medicine; David Housman, Massachusetts Institute of Technology; James Gusella, Massachusetts General Hospital/Harvard Medical School; and John Wasmuth, University of California, Irvine. The task has been arduous in the extreme in this inhospitable terrain at the top of chromosome 4. It has been like crawling up Mt. Everest over the past eight years. First we thought the gene was at the telomere, the very end of the chromosome. Now recent work has indicated that we probably jumped over the gene in our rush to get to the top and it actually is not quite to the end. I used to say confidently that we would be right on it within six months for sure, but I don't say that anymore.

Homozygotes for Huntington's. I mentioned earlier that our first research interest in the Venezuelan kindred was to find a homozygote for the illness. Once a marker for the gene was found, we immediately used it to learn which offspring might have inherited the HD gene from both affected parents. This included the family which originally drew us to Venezuela, a family of fourteen children and more than seventy grandchildren and great-grandchildren. Over the last decade of work, we have also found two other families in which both parents have Huntington's disease and many more in which both are at risk or one affected, the other at risk. Eight probable homozygotes have been identified from these families and more undoubtedly exist.

Even though a dominate gene is defined as "dominating" its normal partner, the few homozygotes who have been found for human dominant genetic diseases have been described as more severely affected than heterozygotes for that same illness. This would suggest a dose effect, even for dominant disorders. The normal gene plays an ameliorative role, even in one dose, and two doses of the defective gene makes the illness worse.

Huntington's disease provides the only exception to this clinical experience: it is the first completely dominant human hereditary disorder that has been genetically documented. Those who are most likely homozygous are not different clinically from their heterozygote relatives. Some putative homozygotes are symptomatically normal, presumably too young yet to develop signs, others have minor neurologic abnormalities, while some have definitive Huntington's disease but no earlier or more severely than anyone else. One tragedy that homozygotes face uniquely: all of their children will be afflicted with HD, as the homozygous parent has no normal genes at the locus to contribute. This is especially agonizing in Venezuela, where the family sizes are so large. The cells of the homozygotes may hold clues that will unlock this devastating disorder but until an intervention is found their families barely sustain themselves in widening pools of suffering. Every year we examine more family members and every year our hearts sink.

The Venezuelan Reference Collections. Over the past decade during which we have been doing field work in Venezuela, it has become increasingly apparent that this excellent kindred, so helpful for studies of Huntington's disease because of its size, geographical proximity, and cooperativeness, among other qualities, is also uniquely superb for gene mapping in general, called reference mapping. Mapping the human genome requires that we track many markers and genes, just like the probe G8 and the HD gene, traveling through generations. The Venezuelan family is unparalleled for determining who is giving what to whom. We can follow eight generations (the grandmothers in this community are only in their thirties). Family studies from this population have been used to make a marker map of chromosome 21, which was helpful in locating genes for Alzheimer's disease and amyotrophic lateral sclerosis (Lou Gehrig's disease); maps of chromosome 17 and 22, where genes causing two forms of neurofibromatosis are located; and a map of chromosome 11, which was used to search for a possible site for manic depression. We are also trying to investigate other genetic or polygenic afflictions in the community -- for example, obesity, diabetes, and hypertension -- for which large study populations are needed.

In this work, it is, of course, imperative to get clinical diagnoses right, because if they are wrong, the genetic analyses are going to be totally incorrect. We are fortunate to be able to maintain contact with the family and return every year. A member of our team, a superb Venezuelan physician named Dra. Margot de Young, attends the family members year round. We try, as much as possible, to collect blood samples from an individual only once. But sometimes you find someone in whom a critical recombination event has taken place that can help localize the HD gene more precisely. It is essential to be able to return to that person to reconfirm the diagnosis and reanalyze a new DNA sample to eliminate the possibility of a laboratory error. Our continuing access to family members makes this reference collection additionally valuable; investigators specified that they would not recontact individuals who have contributed to other major family collections for genetic reference mapping and are thereby precluded from gathering clinical information or checking mistakes.

A new era: Prediction outstrips prevention. While the search for the Huntington's disease gene goes on, painstakingly, the discovery of markers linked to the gene has opened a new, exciting yet troubling era: presymptomatic and prenatal diagnosis of Huntington's disease with no cure in sight.

Immediately after localizing the Huntington's disease gene, we confronted the question of genetic heterogeneity: Is the HD gene in the same chromosomal locale in all families with Huntington's disease throughout the world? Many other genetic disorders manifest genetic heterogeneity -- the causative gene may be on several different chromosomes, even though phenotypically the symptoms of the illness appear to be the same in all the affected families. Was our chromosome 4 home for Huntington's disease unique to the Lake Maracaibo kindred and an Iowan family, or was it universal? Over one hundred families have been tested from throughout the world -- in Europe, North and South America, even Paupa, New Guinea -- and in all of them the HD gene is in the same chromosomal locale on the top of chromosome 4. The actual mutations at that spot may turn out to differ, but the region is the same. Given its universality, we can now use G8 and other markers subsequently found closer to the gene to test whether an individual has the HD gene before any symptoms appear, even before birth. So here we confront our worst fears: our scientific success puts us on the threshold of an era of unknown but imaginable dangers. We can predict the flood but cannot leave or stop the tide. We can tell people that they possess the gene and will eventually come down with the disease, but we have no cure or even therapy to offer to soften the blow.

Cystic fibrosis as a model. Whether or not a disease shows chromosomal heterogeneity (more than one chromosomal location), or allelic heterogeneity (more than one mutation in the same gene at the same chromosomal locus), makes a big difference in genetic counseling. A current case in point is cystic fibrosis (CF), the most common hereditary disorder of Caucasians. Those who suffer from it have pancreatic enzyme insufficiencies and severe lung abnormalities; they cannot clear fluids from their lungs, which become inviting parks for bacteria. Children with cystic fibrosis now often survive until early adulthood, but the disease eventually proves fatal. About one in twenty-five Caucasians carries a single abnormal gene for this condition but is not clinically affected. CF is a recessive illness; each child of two parents who both have one gene for the disorder has a one in four chance of inheriting two genes for cystic fibrosis and expressing the disease. There are about 30,000 individuals with CF in the United States. For a Caucasian person in the population with no family history of the illness, the likelihood of having a child affected is about one in two thousand. If a reliable test were available to detect carriers, people might choose to use it.

In September 1989, Francis Collins, Lap-Chee Tsui, and Jack Riordan isolated a mutation that is found in 70 percent of all individuals with cystic fibrosis.2 Screening people for this particular three-base-pair deletion known as delta 508 would identify 70 percent of all CF carriers. If, however, you test positive but your husband does not, you still will not know whether he really is negative or just has a different mutation, one of the remaining 30 percent not yet known. Calculations indicate that screening couples who are both CF carriers would detect one partner but not the other in more than half the cases. To put the matter another way, if you assume that because your husband or wife has tested negative he or she is free of the gene for cystic fibrosis, half the time you would be wrong. Eminent geneticists on behalf of the American Society of Human Genetics issued a statement, with which other experts convened by the National Center for Human Genome Research, the National Institutes of Health (NIH), the Department of Energy (DOE) Human Genome Program, and other Institutes of the NIH concurred, recommending against population screening for cystic fibrosis until additional mutations were found and the test was more accurate.3 Although testing could be beneficial for those with a known family history of CF, geneticists stated that any other testing was premature and certainly not the standard of care. They advised that widespread use of the test should await two essential elements: identification of a larger proportion of all mutations, and putting in place the service infrastructure to give the test with adequate counseling provided. (Over 100 new mutations have been found since 1989, and the requirement for a higher degree of accuracy of detection is being met. Adequate counseling services are still insufficient, however, and providing these may be even more difficult to implement than achieving the scientific goals.)

In the case of cystic fibrosis, it appears to be useful for parents to know when a child is born that it is affected. Earlier and more aggressive interventions with antibiotics, pancreatic enzyme therapy, and physical therapy can definitely assist the child. Prospective parents and prospective partners might also want this information for family planning purposes. When should you give it? After conception, when options are limited to keeping or terminating the pregnancy? Before conception? When a couple applies for a marriage license? Should testing be obligatory before marriage? Should screening on a large scale be done at the school level? At what age? Should all children be screened at birth for CF carrier status? Each of these scenarios have very different economic, medical, psychological, and social repercussions.

Genetic illiteracy. In all of these screening programs people must understand the difference between being a carrier of one abnormal gene for a recessive condition, in which the carrier usually has no symptoms, as opposed to an affected individual who has two copies of the abnormal gene. People must equally understand that carriers for a dominant disorder, a late-onset illness such as Huntington's disease or polycystic kidney disease, will, in fact, get sick. In recessive diseases the carrier will become a patient. How do we explain technically complex and emotionally charged information to ordinary people, many of whom never heard of DNA and barely of genes, who have hardly a clue about probability, and whose science education never equipped them to make choices regarding these matters? How do we ensure justice in access to counseling services and make them available to more than the white middle and upper classes who typically utilize them now? Genetic diseases cross ethnic and class boundaries, but access to services, unfortunately, does not.

How do we guarantee that the doctors who test individuals or populations provide adequate genetic counseling when doctors themselves have minimal training in genetics and often fundamentally misunderstand its principles? What should we do about doctors who say to a couple with one child affected by cystic fibrosis and who are contemplating having another, "Don't worry, lightning never strikes twice in the same place." Or -- the ultimate in confusion about a genetically dominant disease -- "Don't worry about Huntington's, just tell 'em to marry out into families that don't have it!" Such medical mistakes have increasingly been addressed through malpractice suits, including wrongful birth and wrongful life cases. In wrongful birth cases, parents of a seriously impaired child bring an action claiming that the child should never have been born. The parents argue that they were deprived through the negligence of a health care provider of the information that they needed to decide whether to initiate or continue a pregnancy. Had they known, they claim, they never would have had the child. Wrongful life is an action brought by the child claiming that it should never have been born.4 Must we resort to the threat of lawsuits to ensure that good medical practices will be followed? Or should we have sufficient ingenuity and imagination to be able to introduce new genetic findings into medical practice without increasing the litigiousness of our already embattled society? I believe we can figure out how to offer people genetic information in a way they can understand and assimilate. We can resolve these difficulties if we start working on them now, before the deluge of new tests the human genome program will bring.

All in the family. A major problem in presymptomatic and prenatal testing using linked DNA markers is that the whole family must be involved. When we have the gene in hand and can detect directly the specific mutation in the gene, we will only need to look at an individual's DNA. But for tests using linked RFLP's, the marker patterns of all the relatives must be traced to determine which patterns of the marker in that family is consistently traveling with the appearance of the HD gene. For example, in the Venezuelan family it is the C variant of the marker G8 that tracks with the HD gene, but in the Iowan family it is the A variant. Over time, the process of recombination will gradually change which particular pattern of a marker is near a gene, unless the marker is exceedingly close to the gene. If a gene and a marker are such close neighbors that they are virtually never separated, they are described as being in "linkage disequilibrium." Within a family, however, the same pattern of the marker will tend to travel with the gene and the few recombination events -- the random exchange of segments between two homologous chromosomes -- that may have taken place tend to be obvious. This is why diagnostic testing using linked markers must be done in families: it is imperative to determine what pattern of markers prognosticates HD in that particular family. It is a tedious way of doing diagnostic testing, but until the gene itself can be found it is the only way of doing it and it is how tests must be carried out right now, not only for Huntington's disease but also for polycystic kidney disease and others. (Anybody in a family with a genetic disease -- this probably includes everybody -- should think about storing samples of DNA from relatives whose genotype would be essential to know for diagnostic testing. This can easily be done by freezing a DNA sample extracted from blood. DNA can also be taken from the brain, skin fibroblasts, or almost any other tissue, even after it has been frozen for a long time. The most important relatives to you are those in the family with the illness and those clearly unaffected, parents of these individuals and your own parents. If you have a genetic disorder, banking your own DNA could be critical to your descendants. Each family might have its own genetic variation, its own "genetic fingerprint" of the gene in question, and it is best to preserve a sample of the particular gene that plagues your family rather than extrapolate from the genes from other families.)

There are many families in which an insufficient number of genetically informative people are living (or have banked DNA samples) to permit diagnostic testing for Huntington's disease. And many people would prefer not to know their own genetic status -- whether or not they will develop HD. Can anything be offered for these people? One kind of test -- called a nondisclosing prenatal test -- allows couples at risk to gather some information about a fetus. This test can tell, almost definitively, if a fetus is not going to have HD but cannot tell definitely if the fetus is carrying the gene.

A person at risk has one chromosome 4 from a parent with HD, the other from a parent without the disease. The chromosome 4 from the affected parent may or may not be carrying the HD gene. The at-risk individual will pass on only one chromosome 4 to the fetus; the other comes from the partner. If that chromosome is the one from the unaffected parent, the fetus has a negligible risk (there is always some risk due to the possibility of recombination). If that chromosome comes from the affected parent, the fetus has the same risk as the at-risk parent: fifty-fifty. In a nondisclosing prenatal test, the risk status of the at-risk parent is not altered at all. The only new information that is acquired is whether the fetus has the chromosome 4 that came from its affected grandparent, in which case it has a fifty-fifty risk, or from its nonaffected grandparent, in which case its risk is minimal. And all that is required for this test is a DNA sample from the fetus, obtained through amniocentesis or chorionic villus sampling, both its parents, and one or both parents of the person at risk -- a minimum of four people. (If the affected grandparent has died, his or her genotype can usually be inferred from the other individuals available.)

When we first began offering testing for HD, many of us involved in providing the test thought that nondisclosing prenatal tests would be a preferred option. It offers a chance to ensure that children would be free of the disease while at the same time protecting individuals at risk from learning potentially traumatic information. But comparatively few have utilized the test. Its worst aspect is the possibility of aborting a fetus with a 50 percent probability of not having Huntington's disease, the same risk as the at-risk parent. Just imagine -- you're pregnant or you've fathered a baby, you're attached emotionally, your fantasies are engaged, and now you're confronted with the choice of aborting a baby who might be perfectly normal. How easy will it be for you to become pregnant again? How fast is your biological clock ticking? What if it happens again? A one in two chance is high. Some people feel that aborting a fetus with a 50 percent risk of HD is equivalent to aborting themselves, a rejection of who they are and of their legitimate place in the world. This sentiment is sometimes voiced by those who are disabled, those who object to genetic testing on the grounds that it is designed to eliminate people like them. Because of these particular difficulties, nondisclosing prenatal testing must be offered in a context of intensive counseling and support. If the couple is willing and interested, it is extremely valuable for research purposes to study any tissues resulting from terminations, particularly with respect to learning more about the timing of the HD gene expression. It is possible that the gene is expressed only in utero.

Considerations for genetic counseling. There are many factors that influence the nature and quality of genetic services. An important problem is timing -- when should you give genetic information? For a late-onset disorder like Huntington's disease, timing issues are complex. We are often faced with providing presymptomatic test information to someone whose affected parent is in the last stages of the disease or has just died of it. Etched in their mind's eye is the disease at its devastating worst. And now you are telling someone who is perfectly healthy and normal, "You have a 96 percent probability of having the HD gene" -- which they hear as, "You are going to be just like your mother or father." Wrenching news.

An opposite problem can occur if people coming for testing have never seen the disease. The affected parent may have drifted away from the family, died in a remote hospital somewhere, and the children know only that they are at risk but have never encountered anyone with Huntington's disease. Or a parent may have just been diagnosed and has only minimal symptoms. Newly confronting the ambiguity of being at risk, these people find it intolerable and run to the nearest testing clinic. But what happens if you test these "disease-naive" individuals and shortly after receiving the information that they are probably gene-positive, they happen to turn on NOVA and there on television is a graphic display of Huntington's disease from beginning to death. They think, "Oh my God! I had no idea this is what people at the testing center were talking about!" If you decide to educate potential testees about the disease to enable them to make more informed choices with respect to being tested, how much education is appropriate? What if someone whose father has just been diagnosed comes to you for testing? You don't want to drag him or her down to the local state hospital, where many patients with HD live, or to a nursing home and say, "This is what is in store for your father, and maybe you." You cannot crack open a person's very beneficial shell of denial too radically, and yet you also cannot allow someone to be tested without having some fundamental appreciation of the meaning of the results. It is difficult to shatter denial and shoal it up at the same time; denial is a critical component of coping and must be treated with respect. Information should be carefully titrated, and intensive counseling over time is essential.

My fear is that when presymptomatic and prenatal laboratory tests become more rapid and accurate -- for example, when polymerase chain reaction (PCR) techniques can detect the mutation itself -- there will be a temptation to short- circuit the testing process, to make it faster and offer less counseling. But no matter whether the test is easy or not, the impact of the information is equally crucial for one's life. There is still no interdiction we can offer, no treatment or preventive. And even if testing can be done on a single individual without DNA samples from relatives, Huntington's disease is a family disease and the test results for one member reverberate through all.

At the moment, relatives must donate a DNA sample for linkage testing to be done. They sign an informed consent form when they give blood indicating that they give their permission for the sample to be used for the presymptomatic testing of someone in the family. Usually the at-risk person requesting testing must arrange with relatives for samples to be sent to the laboratory or for neurological examinations to be conducted on critical relatives whose clinical status must be accurate; with all these preparations going on, the at-risk individual's desire to be tested is generally known and discussed in the family. Relatives have an opportunity to convey their feelings and, in some instances, they try to dissuade the individual from continuing with the test. Some parents have gone even further to exert their influence and have refused to give a DNA sample, thereby halting the test. One parent refused because the testing program in the locale did not provide adequate counseling and follow-up. Many of the testing centers have encountered other situations in which parents might be willing to give a DNA sample for one offspring's test but not for another: "Jane can take the news, John can't." Of course, once you know what pattern of the marker the disease travels with in that family, you do not need to retest the parent's sample for each child. Whose rights take precedence -- John, who says "I can take it, and besides, I want to get married"; or Jane, who says "Your arguing is depriving me of my test, and besides, I want to have a baby"; or the father who says, "I own my genetic profile -- you can't rob me of my genetic information and use it without my permission for purposes of which I don't approve."

A similar problem arises when identical twins arrive at the testing center and one wants to be tested and the other does not. Now who should hold sway? One center said, "We'll test you but don't tell your twin." It doesn't work. If you are free of the disease it would be almost inconceivable not to run to your twin with the good news. And if the outcome is HD, it's hard to explain uncontrolled crying as a chronic cold to people who know you well. Other testing centers confronted with this predicament consulted ethicists who gave them pronouncements that autonomy is higher on the scale of ethical virtues than privacy, so the centers decided to proceed with the test. But to my mind, autonomy or privacy may be irrelevant if the twin who is not part of the testing process and has not even had the benefit of counseling learns the truth and commits suicide. The immediacy of each individual's psychological reality must take precedence over abstract, theoretical values and issues. One cannot consult a guidebook for who should be tested and under what circumstances. The professionals giving presymptomatic test counseling must be trained psychotherapeutically to help determine the best solutions for individuals and families as a whole.

Another factor that is insufficiently appreciated is that when one person in a family is tested, the entire family is tested and all must live with the outcomes. Many parents of persons at risk feel guilty about and responsible for their children's risk status, even though they may have known nothing about HD when the children were born. In some families, three or four children simultaneously may be diagnosed presymptomatically. A parent who has spent fifteen or twenty years caring for an ill spouse now has a grim preview of the future: the prospect of caring for children as well, or knowing that the children may have to depend on the mercy of strangers. One woman said, "When my husband died after twenty-five years of illness, I felt like a light had finally come on at the end of the tunnel. Now I watch my daughter and see her movements and the light has extinguished."

Testing minors. Because presymptomatic testing for Huntington's disease is difficult to undergo for all and potentially devastating for some, given the absence of any therapy, those professionals who are offering the test, including myself, have decided to restrict it to individuals who can give informed consent and who are age eighteen or older. This is not a legal requirement but it has been accepted as part of the HD testing protocol for centers throughout the world. Until we know more about the impact of having this information on adults who are choosing to learn it of their own free will, the testing center professionals are reluctant to test minors, at their own request, or to give information to parents about their minor children, with or without their children's knowledge. This conviction, on my part, was reinforced when a woman told me she wanted to test her two minor children because she only had enough money to send one to Harvard.

But parents do advance cogent arguments for testing their minor children -- they need the information for making financial and other life plans. It would certainly make a difference to know that any or all of your children are going to develop HD. Withholding this information from parents goes against the typical situation found in family case law, in which parents are entitled to medical information and the only instances in which the courts are likely to intervene is when parents are not providing medical attention for religious or other reasons.5 One of the complexities of providing nondisclosing prenatal testing is that it sometimes forces you to give information about a minor child despite your protocol. Prenatal testing for HD is not offered to couples who have no intention of terminating the pregnancy because there is no medical advantage to the parents to have this information, and because prenatal testing does pose a slight medical risk to the fetus and entails testing a minor without its consent. But if a couple finds that their fetus has a 50 percent risk of having HD, they still may change their minds about a termination and carry the pregnancy to term. If the at-risk parent later develops the disease, the genetic identity of the fetus is also revealed. This violation of the privacy of the minor must be endured because parents are entitled to change their minds regarding the termination of a pregnancy. But very careful counseling must be provided so that the couple knows exactly what the test involves and the nature of their options.

Another potential controversy among test providers is the decision made by those currently offering presymptomatic testing for Huntington's disease not to test children who are in agencies awaiting adoption. The adoption agency personnel requesting testing argue that children shown to be free of the disease will be more eligible for adoption. To date, test providers have responded that such testing still violates the privacy of a minor without informed consent and, furthermore, may consign those found to be gene- positive to permanent placement in an orphanage, eliminating any hope that a child still only at-risk might hold for adoption.

Some of these decisions regarding who and when to test may be altered as we learn more about the meaning of genetic diagnostic information to those who are receiving it. During the 1970s, in Canada, several studies by Charles Scriver and colleagues showed that secondary-school children who learned they were carriers for the recessive gene causing Tay-Sachs felt stigmatized and somehow inferior to their classmates even though being a carrier was in no way injurious to their health.6 It was an emotional stigmatization. Will such responses be common? Some people insist, "Make genetic testing mandatory when couples get married." Others advise, "Integrate it into the genetic services, so that couples can be tested when they are contemplating pregnancy or are already pregnant." However, people disinclined to choose abortion might want to have genetic information before selecting their mate. A sickle cell screening program in Orchemenos, Greece, undertaken in the early 1970s before prenatal diagnosis was possible, found that 23 percent of the population carried the trait.7 Those discovered to be carriers were stigmatized and, as a result, they sometimes concealed their carrier status in order not to jeopardize marriage prospects. The net consequence was that the same number of affected babies was born as before the screening program. In two of the four matings resulting in the birth of an affected child, the women hid their carrier status, and in the other two the couples had children although cognizant of the risks. Once prenatal testing for sickle cell disease and thalassemia became widely available, carrier status became less of an impediment to social acceptance, even in predominantly rural, Catholic countries like Sardinia.8

Genetic misunderstandings and their implications. I am always surprised by the imaginative ways in which people can misinterpret genetic information. One common and very understandable mistake is the belief that at least one person in every family will be sick. In the Huntington's disease testing programs, people often arrive with the conviction that whether they will or will not get the disease depends on the fate of siblings. If my siblings are sick, my risk goes down; if they are old and healthy, my risk goes up. This is a perfectly reasonable misinterpretation given the way in which genetic inheritance is usually explained. Most genetic textbooks and consumer pamphlets teach principles of inheritance by showing a family of four children and two parents in which two children are affected and two are not. And doctors often explain risks by saying "half your children" or "one-quarter of your children will become sick." You must always say, "Each child has a fifty-fifty or 25 percent risk, regardless of the rest of the family." The day people in Venezuela became really confused about inheritance was the day the newspapers published such a diagram.

It is difficult to teach someone that "chance has no memory" and that whatever may have happened in a previous conception has no bearing on whether or not a child carries the HD gene. Each person has his or her individual risk and whatever happens to the rest of the siblings does not matter. I often ask people to flip coins during the counseling sessions to see, concretely, how it is possible to flip ten heads in a row. If they flip a coin which says they will get HD, it also gives that dire possibility some hard reality.

The idea that one's life or death is controlled in such a random way as a flip of a coin is appalling for most people. We try to make sense, make meaning of our lives. We try magically to control that coin toss by inventing rules governing who gets sick and who does not, but the fact remains: it is totally a random accident of fate which gametes meet, and in that moment the future is sealed.

Gazing into the crystal ball. There are approximately 22 HD testing centers in the United States, 14 in Canada, 5 in Great Britain, and several more in Europe and Scandinavia. Probably fewer than 1,000 people have been tested in all these centers. This skewing of test results toward those who test negatively is probably a result of the arduous and intense counseling required by the test protocol. As people come to appreciate on an emotional level, not merely intellectually, the magnitude of a positive diagnosis on their lives, those who have some inkling that this is to be their fate may decide not to be tested.

It has also been the experience of most centers that some individuals coming for presymptomatic testing are clearly already symptomatic but show no awareness of these symptoms. They mostly want to learn if they will have HD in the future, not in the present. If they are ready psychologically for the information, they should be diagnosed clinically, not by DNA analysis, or else encouraged to come back at a later date.

One group of individuals who come for testing I call the "altruistic testees." These are people who would prefer not to be tested but are doing so to clarify the risk for their children who are getting to be dating or marrying age. The genetic risk for these people is lower because they are older, but many of them really do not want to be tested and would prefer not to "rock the boat." We know very little about the response of this group to a diagnosis of probably gene-positive.

Some clients who learn that they do not carry the disease gene find that this knowledge does not make all of their problems disappear. They may still have trouble finding the "right person" or managing their careers. Before the test, being at risk became a convenient excuse for putting decisions on hold, for postponing and avoiding issues that may have nothing to do with being at risk but become entangled in the risk situation. People at risk say, "Well, if I just weren't at risk, I could figure this all out." Then, suddenly, they are no longer at risk, but they still can't figure things out. The problems have become too entrenched, too much a part of their characters.

This is not to say that good news does not also lead to ecstasy and joy. Some people alter their lives -- have children, move, or change jobs -- and feel wonderful about the change. Others experience a kind of survival guilt and sense of unease with respect to other family members who may not know their genetic status or who tested positive for the gene.

Learning by experience in Venezuela. It is this latter group -- those who test most likely gene-positive -- who preoccupy us the most, who capture our imagination and concern. As there are only a very few such people, less than 100 people in the United States, we know little about how this new "presymptomatic group" will react to the bad news. Some clues are provided by our experience in Venezuela, where we are just determining the clinical status of the study population. Only in rare instances does anyone ask us, after a neurological examination, what we have learned. On one occasion, however, a woman in her twenties, mother of several children and pregnant again, came to be examined. We concluded that she did have Huntington's disease and were taken by surprise when she asked us what we thought. We asked her how she felt about herself and she said, "just fine -- no HD so far," which gave us the immediate clue to find out more about her and what this diagnosis would mean in her life. We told her we would like to get to know her better and see her over time -- we would be there all month and again the following year and encouraged her to spend time with us. A moment after she left, her friend came running into the clinic at top speed, looking panicked, and asked, "What did you say to her?" When we told her, she sank back in relief. "Oh, thank God," she said, explaining that her friend had announced, "I'm going to ask the American doctors if I have Huntington's and if I do I'm committing suicide." Just as in the United States, suicide attempts and completions do occur in the Lake Maracaibo community.

The advent of presymptomatic testing presented us with a dilemma: Should we try to make it available to the Venezuelan families? Scientifically it is essential for all of us working in the field to be blind to the genotype of the people whom we are following. It is impossible to do an unbiased clinical examination once you know the genotype, and those not doing examinations must remain blind to prevent nonverbal or accidental revelation of the information. It is particularly crucial not to know genotypes when assessing individuals with putative recombination events.

We were uncomfortable, however, with our research requirements depriving anyone of necessary and desired diagnostic information, so we visited with professionals at the University of Zulia in Maracaibo to see whether, if we were to perform the laboratory genotyping, they could arrange for faculty to provide the necessary genetic counseling that is indispensable to accompany genetic information. The "barrios" or neighborhoods where we work are quite poor and have a reputation for violence. To our concern, we felt that those providing this momentous information would give insufficient time and attention, and we were unwilling to release genotype data under these circumstances. We even worked with scientists in Caracas to try to set up the necessary laboratory and counseling structures we felt were imperative to have on site, but it turned out to be impossible.

We also arranged a meeting with family members who had heard that there had been a critical discovery and were disappointed that it was not the cure. We tried to explain what had been found and what it might mean to them. One man pointed to the bridge over the lake near where we were meeting and said succinctly, "If you tell me I am going to have this disease and I do not have someone to talk to about this, I am going to run to the nearest bridge and jump off!" We felt that if we gave diagnostic information and then left for a year, we would be acting like hit-and-run drivers. We are also limited by the options available to people who request genetic information: abortion is illegal in Venezuela, which eliminates the value of prenatal diagnosis for the HD gene since minors are not tested. We finally decided to provide clinical diagnoses, if people ask for them, and genetic counseling based on age of onset of HD in the family, but not to give genotype information.

Presymptomatic testing: preliminary outcomes. Our experience with genetic diagnostic testing for Huntington's disease in our own country suggests that, with very intensive counseling, the few people who have tested positively tend, by and large, to do pretty well. In Canada, one person who came for presymptomatic testing and discovered she was already symptomatic made a suicide attempt, and there has been one hospitalization in the United States that I know of for severe depression following a presymptomatic positive diagnosis. Most people who have tested most likely positive for the presence of the gene have had this news for only a year or two, and we have no idea how this group will respond once they find themselves becoming symptomatic. One woman told me she was often asked if she regretted her decision to be tested. She added, "You know, I don't think so, but I really can't afford to think about that question too long because I'm afraid it's going to take up housekeeping in my mind."

We are unable to tell people when the disease will start; we can just say that they most likely have the gene. In follow-up interviews some time after testing, people who have tested positive were asked if they think they will develop the disease; some reply, "I don't think so, because God will cure me, or science will cure me, or the test was wrong."9 It is traumatizing to be totally healthy and know with almost 100 percent certainty that Huntington's disease is in your future.

The gain/loss quandary. We know very little about how people decide whether or not to be diagnosed presymptomatically. Two psychologists, Daniel Kahneman and Amos Tversky, have studied how people assess risks and make decisions based on those assessments. One of their scenarios involves imagining that you command six hundred soldiers in battle and you must choose between two possible routes.10 If you take the first route, you will certainly save two hundred soldiers; if you take the second, the odds are one in three that all will be saved but there is a two- thirds chance that all six hundred will die. Another scenario is as follows: You are a commander of six hundred troops. If you take the first route four hundred of your soldiers will certainly die; if you choose the second, the odds are one in three that none of your soldiers will die and two in three that all six hundred will die. Of course, the two scenarios are exactly the same, but one is phrased in terms of most certainly saving two hundred soldiers and the other in terms of losing the lives of four hundred men. Kahneman and Tversky have found, by and large, most people are not risk-averse but loss-averse. In the first battle scenario, which emphasizes saving two hundred men, most people will take the first route, preferring the sure thing of saving some lives rather than gambling with all. But when the same choice is rephrased to emphasize that four hundred men will certainly die, more will choose to gamble for the possibility of saving all six hundred soldiers and will choose the second route. Faced with a certain gain, people tend to be conservative and maintain what they have, but faced with a certain loss, people are more willing to gamble. If you are given a certain amount of money, people tend to keep what they are given rather than gamble for the prospect of winning more. If you must give up money, you would be more inclined to gamble and even risk losing more for the chance of not losing any. Kahneman and Tversky argue that people dislike losses and will guard against them. But if the loss is certain, then people are willing to risk an even bigger loss if there is a chance of avoiding any loss at all.

Kahneman and Tversky's findings emphasize why it is so important to explain genetic information both in terms of gain and loss. Telling clients they have one chance in four of having an affected child conveys one psychological message, saying that they have three chances in four to have a normal baby conveys a different one -- even though the statistics are the same. Clients need to be given both constructions of the genetic information.

A person taking a genetic test makes a terrific gain-loss calculation. The gain, obviously, is to learn that you do not have the genes for Alzheimer's, cystic fibrosis, Huntington's, or any number of other diseases. The loss is to learn that you do. Is learning the good news worth risking hearing the bad? Many who come for testing already feel they are in a loss position; they consider being at risk just as bad as knowing they will be affected. They assume they cannot do one thing or another because they are at risk, even though they could do what they want. Nothing is really stopping them, but they are paralyzed by their risk situation: because certain things are unknown, everything becomes impossible. I asked a woman why she wanted to be tested. She replied, "If I find out I'm going to have HD, I'll take my son to Hawaii; but if I'm OK, then I'll wait." I said, "If you want to take your son to Hawaii, why are you waiting until you get diagnosed with HD to do it? Because by the time that you're ready to take him to Hawaii, he's going with his girlfriend, not his mother." Since many people at risk already feel themselves to be in a loss situation, they are more willing to take a test which may throw them for a greater loss -- learning that they do, in fact, carry the gene. If they see their lives more or less in a gain situation -- they have chosen their careers, had their children -- then they are more conservative about maintaining that gain and not willing to take the gamble. Some people will take the gamble for the sake of their children, because the only way to clarify their risk is to take the test.

What makes these problems so difficult is the absence of recourse to treatment. If you can do something about the disease, then there will be an incentive to undergo presymptomatic testing and the catastrophic nature of a positive diagnosis will be tempered. If the treatment is only marginal, the choice to be tested will still be difficult. Attitudes toward HIV testing changed when people learned that the drug AZT could retard disease onset in individuals positive for HIV.

The human genome project: Road map to health. The human genome project should eventually point the way to preventions and cures. The project should, within the next few years, establish an "index" map of the human genome, with markers spaced about every 10 million base pairs, placing at least one marker close enough to every gene of interest to locate it. This map should get us in the neighborhood of most disease-causing genes. Then will follow the construction of a "high-resolution" map, one with thousands of markers spaced every million or so base pairs. With this detailed map, it should be possible to localize genes more rapidly and precisely, and finding the genes should lead to their sequencing and characterization. This map will guide the "gene hunters" as they navigate along the genome.

Some people protest that simply knowing the molecular lesion in a gene does not guarantee the development of new treatments for the disease it causes. The mistake in the gene causing sickle cell disease has been known for twenty- five years and there is still no effective therapy or cure. But this discovery may have been ahead of its time; with newer technologies today, the defect may be more amenable to intervention. Of course, it may be true that knowing the cause of a disease at a molecular level may do nothing toward advancing palliation or cure. But it makes intellectual sense to go after the gene, the cause of the disease, as one possible avenue of interdiction. It is not the only route but it is a reasonable one in studying the etiology of a disorder. If you want to stop the damage of the Nile constantly overflowing its banks, you could either build protections along the length of the river shore or go to the source of the Nile and try to control the flow before damage occurs.

There is preliminary evidence that the identification of the abnormal genes causing cystic fibrosis, 1-antitrypsin deficiency, neurofibromatosis, and other hereditary disorders may lead to promising new avenues of research on therapeutics. Scientists have been able to place a normal human 1-antitrypsin gene in the epithelial lining of a rat lung, using a virus that causes the common cold, an adenovirus, to transfer in the gene.11 The rat lung tissue, both in the test tube and in the animal, produced normal human gene product for some time. Two other scientific groups have inserted the "normal" cystic fibrosis gene into lung and pancreatic tissues in culture taken from people with cystic fibrosis, also using a virus to transport in the gene. In tissue culture, the normal gene was able to reverse the effects of the disease and blocked channels functioned normally.12

Some investigators have contemplated using an aerosol spray to deliver the normal gene into diseased lung tissue. Dr. James Wilson of the University of Michigan, leader of one group that corrected CF in culture, commented to the New York Times, "My tendency is to be very conservative . . . at this point, it's impossible for me not to be optimistic about CF."13

These are all very preliminary findings and much work is ahead to demonstrate that genetic therapy, therapy based on using the normal gene itself as a treatment, is effective. But they are intriguing leads. And at a minimum, identifying the gene focuses all the energy and resources that had been scattered in analyzing irrelevant regions of the chromosome while the gene search was under way onto the aberrant gene itself.

Ethical, legal, and social issues. There are many social, psychological, ethical, legal, and economic problems awaiting us that I have not even mentioned. Once we have improved capacity to diagnose disorders presymptomatically, many more individuals and families may face the loss of health and life insurance. They may be exposed to discrimination from employers and stigmatization and ostracism from friends and relatives. Predictive information can be fraught with dangers to individuals and to society. To address these concerns, the National Center for Human Genome Research, the National Institutes for Health, and the Human Genome Program of the Department of Energy have established the Joint Working Group on Ethical, Legal, and Social Issues associated with mapping and sequencing the human genome. It is the mandate of this working group to support research in these critical arenas and develop policy recommendations for the necessary protections that must be put in place as new genetic tests are being developed.

If there are so many personal, social, and economic hazards and successful cures are not assured, some people ask, why proceed with the project? How can we not proceed? Many who suffer from hereditary diseases already make huge economic sacrifices, already pay exorbitant psychological and social costs. I could not go to Venezuela and say to those expectant people, "Sorry, we've called off the research for the Huntington's disease gene because having the gene in hand is too dangerous and there is no guarantee of a cure." I am an optimist. Even though I feel that this hiatus in which we will be able only to predict and not to prevent will be exceedingly difficult -- it will stress medical, social, and economic systems that were already under a severe strain before the advent of the human genome project -- I believe that the knowledge will be worth the risks. We are learning from our experience with Huntington's disease and other disorders about the power of clairvoyance and the need for caution. We are preparing for the future when tests for breast cancer, colon cancer, heart disease, Alzheimer's disease, manic depression, and schizophrenia might well be available. For a while we may have the worst of all possible worlds -- limited or no treatments, high hopes and probably unrealistic expectations, insurance repercussions -- everything to challenge our inventiveness and stamina. But these ingredients will be, I hope, catalysts for change. The stakes are high; the payoff is high. I am reminded of a line by the poet Delmore Schwartz: "In dreams begin responsibilities."

REFERENCES

1. James Gusella, Nancy Wexler, P. Michael Conneally, et al., "A Polymorphic DNA Marker Genetically-Linked to Huntington's Disease," Nature, 306 (1983), 234-238.

2. Johanna M. Rommens et al., "Identification of the Cystic Fibrosis Gene: Chromosome Walking and Jumping," Science, 245 (September 8, 1989), 1059-1065; John R. Riordan et al., "Identification of the Cystic Fibrosis Gene: Cloning and Characterization of Complementary DNA," ibid., pp. 1066-1073; Bat-Sheva Kerem et al., "Identification of the Cystic Fibrosis Gene: Genetic Analysis," ibid., pp. 1073-1080.

3. C.T. Caskey et al., "The American Society of Human Genetics Statement on Cystic Fibrosis Screening," American Journal of Human Genetics, 46 (1990), 393; B.S. Wilfond and N. Fost, "The Cystic Fibrosis Gene: Medical and Social Implications for Heterozygote Detection," Journal of the American Medical Association, 263 (May 23/30, 1990), 2777-2783; Workshop on Population Screening for the Cystic Fibrosis Gene, "Statement from the National Institutes of Health Workshop on Population Screening for the Cystic Fibrosis Gene," New England Journal of Medicine, 323 (July 5, 1990), 70-71.

4. Lori B. Andrews, "Legal Aspects of Genetic Information," Yale Journal of Biology and Medicine, 64 (1991), 29-40; N.S. Wexler, "Will the Circle Be Unbroken? Sterilizing the Genetically Impaired," in A. Milunsky, ed., Genetics and the Law (New York: Plenum Press, 1980); Nancy S. Wexler, "The Oracle of DNA," in L.P. Rowland, ed., Molecular Genetics in Diseases of Brain, Nerve, and Muscle (New York: Oxford University Press, 1989).

5. Symposium, "A Legal Research Agenda for the Human Genome Initiative," Arizona State University, Center for the Study of Law, Science and Technology, May 17- 18, 1991.

6. Reported in Marc Lappe, Genetic Politics (New York: Simon and Schuster, 1977).

7. G. Stamatoyannopoulos, "Problems of Screening and Counseling in the Hemoglobinopathies," A.G. Motulsky and W. Lenz, eds., Birth Defects (Amsterdam: Excerpta Medica, 1974), pp. 268-276.

8. A. Cao et al., "Prevention of Homozygous Betathalassemia by Carrier Screening and Prenatal Diagnosis in Sardinia," American Journal of Human Genetics, 33 (1981), 593-605.

9. Personal communication, Dr. Jason Brandt, Johns Hopkins Hospital, Baltimore, Md.

10. Amos Tversky and Daniel Kahneman, "The Framing of Decisions and the Psychology of Choice," Science, 211 (1981), 453-458.

11. Melissa A. Rosenfeld et al., "Adenovirus-Mediated Transfer of a Recombinant -Antitrypsin Gene to the Lung Epithelium in Vivo," Science, 252 (April 19, 1991), 431-434.

12. Mitchell L. Drumm et al., "Correction of the Cystic Fibrosis Defect in Vitro by Retrovirus-Mediated Gene Transfer," Cell, 62 (September 21, 1990), 1227-1233; D.P. Rich et al., "Expression of Cystic Fibrosis Transmembrane Conductance Regulator Corrects Defective Chloride Channel Regulation in Cystic Fibrosis Airway Epithelial Cells," Nature, 347 (September 27, 1990), pp. 358-363.

13. Natalie Angier, "Team Cures Cells in Cystic Fibrosis by Gene Insertion," New York Time, September 21, 1990, p. A1.