Dna: The Secret of Life

by Watson, James

  Robert Winston at the Hammersmith Hospital in London is a leading gynecological microsurgeon, an expert in such procedures as the correction of Fallopian tube defects that prevent a woman from conceiving. He has also become one of British television's leading popularizers of science and biomedical research and even finds time, as Lord Winston of Hammersmith, to sit in Parliament advising the government on such matters. By combining two state-of-the-art technologies – in vitro fertilization (IVF)* and PCR-based DNA diagnosis – Winston pioneered a method for checking the genetic status of an embryo before it is implanted in a woman's uterus and begins to develop. After in vitro fertilization, the several conceptuses are grown in the laboratory until each fertilized egg has divided three or four times to produce a ball of eight to sixteen cells. One or two cells are carefully removed from each for DNA extraction and then PCR is used to amplify the relevant sequences to determine in each case whether or not a mutation is present. It is the astonishing capacity of PCR to amplify even the tiniest quantities of target DNA that makes possible this method of ultra-early diagnosis. Parents are then free to implant only those embryos that test negative for genetic disease.

  * IVF is a method of assisted reproduction in which sperm and egg are fused in a laboratory dish. The resulting embryo—rather ominously called a "test-tube baby" in the early days of the technology—is then transferred to the uterus to develop naturally.

  The first preimplantation tests performed in 1989 screened for the sex of the fetus – important information when the risk is for a sex-linked disorder such as DMD. A mother who is a carrier may select only female embryos on the premise that they will not be affected by the disorder, though they may be carriers. It was Winston's colleague, Alan Handyside, and others who subsequently extended preimplantation diagnosis beyond simple sex determination to the detection of specific mutations: in 1992, they first applied the technique to screen for cystic fibrosis, which is not sex-linked.

  As we have seen, despite being sex-linked, fragile X can affect both males and females. The disorder is therefore a natural target for gene-specific preimplantation diagnosis, but it still took impassioned parents familiar with the difficulties of raising a fragile X child to mobilize doctors to do it. Debbie Stevenson, a CNBC television news reporter, has a son, Taylor, whose fragile X was only diagnosed after the birth of a second son, James. Though James fortunately beat the 50-50 odds of being affected, the Stevensons were unwilling to trust their third child to fate. They decided to seek preimplantation diagnosis: "Some people think it's unethical to select for healthy embryos," says Debbie Stevenson, "but I think it's better than having to make the wrenching decision about whether to terminate or continue a pregnancy after learning that your baby has a significant disorder." In 2000, at the end of the family's frustrating yearlong search for a lab willing to carry out the procedure, the newest member of the Stevenson family was conceived and, just days later, tested for fragile X. Like James, Samantha is free of Taylor's debilitating disorder.

  In our culture human reproductive biology seems an inexhaustible source of controversy, and a procedure involving the manipulation of human embryos for any purpose is sure to become a lightning rod. Preimplantation diagnosis has been no exception. Ethical considerations aside, however, the procedure still has two major drawbacks: it requires a huge commitment on the part of the couple undertaking it, and, like all forms of IVF, it is very expensive. But the method is so powerful in principle, and so much less traumatic than abortion, one can only hope that with time we shall see improved techniques, and with them diminished cost – the usual pattern with developing technology. Preimplantation diagnosis has the potential to become an extremely important weapon in our war on genetic disease.

  The disorders we have discussed so far are all "simple" in the genetic sense: they are caused by a mutation in a single gene, and environment has no bearing on whether or not you will get one of these diseases. The situation is rather more complicated in the case of illnesses like cancer, which, as we've seen, may be triggered by a combination of hereditary and environmental influences. But even with cancer, there are some individual genes that have a major effect, regardless of the environment. Though BRCA1, one of the genes implicated in breast cancer, only accounts for about 5 percent of all cases, women with mutations in the gene have an estimated 90 percent chance of developing the disease by the age of sixty.

  In the early 1990s, Francis Collins, then at the University of Michigan, joined forces with Mary-Claire King at UC Berkeley in the hunt for BRCA1. They took the standard approach: collecting families, preparing DNA samples, and testing markers, all with a view to homing in on the gene. One family of more than fifty members included multiple cases of breast cancer – a clear instance of an inherited predisposition to the disease. In September 1992, one member of that family – I'll call her Anne – revealed to Barbara Weber, an associate of Collins's, that she had scheduled a bilateral mastectomy the following week, even though there was no sign that she had cancer. Anne had decided that she could no longer tolerate the uncertainty, the question mark hanging over her future, and preferred to take this drastic preemptive step. Weber, however, had concluded from the DNA analysis that Anne was actually not in particular danger: her risk of breast cancer was no greater than that of a woman without a family history of the disease. But this inference was made in the context of a research project and it had been agreed long in advance that as a rule such preliminary data should not be used for clinical diagnosis.

  Weber and Collins, however, decided that Anne's plight outweighed the rule book: they informed Anne that her risk was low, and with great relief she canceled the surgery. But having disclosed their findings to one member of the family, the researchers felt obliged to offer the same benefit to others who asked for it; and so Weber and Collins set up an ad hoc breast cancer genetic counseling program. One family member who also proved to be at no special risk had already undergone a prophylactic bilateral mastectomy five years before. She received the belated diagnosis philosophically: the surgery, she figured, had bought her five years' peace of mind. But had she tested positive for the mutation, the radical course might in fact have bought her more than peace of mind. For years, prophylactic mastectomy had been recommended by clinicians, even though no surgery can feasibly remove all the breast tissue and there were no solid data showing that the measure was saving lives. Today, however, there is proof that the extreme approach does indeed reduce mortality rates among women at high risk; in one group of 639 who had the surgery, only 2, rather than the 20 to 40 statistically expected, actually died of breast cancer. Similarly, removal of the ovaries before age forty (but after a woman has finished having children) reduces the risk of both ovarian and breast cancers. Genetic analysis can give women the power to make decisions that can literally make the difference between life and death.

  But the keyhole view into the future that DNA analysis permits can also create opportunities to defeat breast cancer by less extreme means, as another story from the Michigan study reveals. A cousin of Anne's was told that she was in all likelihood carrying the BRCA1 mutation that was devastating her family. Since she had not had a mammogram in years – a fear-based negligence ironically not uncommon in high-risk families – she panicked. Weber scheduled one for later that day and a tiny incipient tumor was found; it was easily removed but would almost certainly have been missed in a routine examination. Self-examination and regular mammography have doubtless saved many lives, but the campaign to universalize these procedures may have had the unintended consequence of creating a false sense of security in some cases. Screening for genetic risk allows us to find those individuals whose imaging examinations merit extra-high scrutiny. Greater risk demands greater surveillance. And in the long run more needles will be found the smaller we make the haystack.

  Nancy Wexler, as a member of a Huntington family, and Anne, as a member of a breast cancer family, are both part of a new generation for whom newly available screenin
g tests can provide glimpses of genetic destiny. And as we learn more about the genetic basis of relatively common adult afflictions, from diabetes to heart disease, the biological crystal ball will become ever more powerful, telling the genetic fortunes relevant to us all.

  In the past decade, few diseases have struck terror in as many hearts as Alzheimer, which each year draws ever greater numbers into its grip of awful mental and physical debilitation – the disease affects more than 4 million Americans. Family and friends of sufferers may notice first some minor lapses of memory – trouble recalling recent events or finding the right word – which they might hopefully attribute to the ordinary effects of getting on in years. The afflicted may then begin to show evidence of mood swings, also not altogether unnatural among the elderly. But as the disease progresses, the symptoms become more pronounced and unmistakable; the memory loss soon grows unnaturally severe, making the familiar challenges at work and even simple household tasks unmanageable. Speech becomes more labored; sentences go unfinished as the victim loses the train of thought. And the person's awareness of these changes may lead to depression, which in turn intensifies the effect of other increasingly distressing changes in personality. Advanced Alzheimer patients do not know who or where they are; they cannot recognize even their closest kin. With the inexorable erosion of memory and personality, their very essence as individuals is gradually destroyed.

  Alzheimer typically first appears at around age sixty, but a rarer form, accounting for about 5 percent of all cases, strikes individuals in their forties. This early-onset form of the disease puts families through the same kind of hell as Huntington disease does, seizing its victims in the prime of life and gradually, relentlessly destroying them. One family with multiple affected members over several generations was described as having been struck by its own "biological Holocaust." Following the argument first advanced by Mary-Claire King in her breakthrough study of breast cancer, that any early-onset version of a disease is likelier than the ordinary form to have a clear genetic basis, most initial Alzheimer research focused on the early-onset form. By 1995 three genes had been found, all of them involved in some way with the processing of amyloid protein deposits, whose accumulation in the brains of patients was noted as early as 1906, in Dr. Alois Alzheimer's original description of the disease. Early-onset Alzheimer is, then, clearly inherited. But what of the more common variety?

  Allen Roses at Duke University preferred to ignore the wisdom of the majority, and set out straightaway to tackle the much more familiar late-onset form, which only occasionally runs in families. Ronald Reagan, for instance, who announced his affliction in 1994, lost his brother Neil to late-onset Alzheimer two years later. Their mother had died of it as well.

  With training as a neurologist and a background in muscle disorders like DMD, Roses began his search in 1984. His claim in 1990 that a gene on chromosome 19 appeared to correlate with the disease was met by skepticism. Nothing, however, gives Roses more pleasure than an opportunity to prove everyone else wrong. Two years later, he had actually identified the critical gene. It turned out to code for apolipoprotein E (APOE), a protein involved in processing cholesterol. The gene comes in three forms (alleles), APOEε2, APOEε3, and APOEε4, but it was APOEε4 that proved the crucial one: a single copy of that variant increased fourfold one's risk of developing Alzheimer. And individuals with two copies were at a risk ten times greater than that of persons with no APOEε4 allele. Roses found that 55 percent of those with two copies of APOEε4 will have developed Alzheimer by age eighty. Could this correlation be the basis for a genetic test? Probably not. Dcspite being correlated with the disease, the APOEε4 allele is common and is not a good enough predictor of Alzheimer for testing purposes: though their risk is higher, plenty of people with two APOEε4 alleles never develop Alzheimer. But the use of APOEε4 screens in conjunction with clinical evaluations does improve the accuracy of Alzheimer diagnosis. And perhaps once we understand the correlation in causal terms, the genetic analysis can be refined. Recent work that induced Alzheimer– like symptoms in mice has suggested APOE is involved in the metabolism of the protein that causes nerve cell death in human Alzheimer sufferers.

  What of treatment? Most genetic diseases present us with much the same heartrending frustration that comes with Huntington: we know enough to diagnose them, perhaps to evade them, but not to treat them. Happily, there are a few cases in which our genetic understanding has taken us the rest of the way, providing therapies that work. Unfortunately, few of these remedies are as simple and effective as that for PKU, from which a normal life can be retrieved through a few dietary restrictions.

  Too often genetic disorders result in the cell-by-cell decimation of particular tissues: muscles in DMD, nerve cells in Huntington and Alzheimer. There is no quick fix to this kind of insidious decay. But though these are early days yet, I think there is a realistic chance that we will eventually be able to treat diseases like these using stem cells. Most cells in the body are capable only of reproducing themselves – a liver cell, for instance, produces only liver cells – but stem cells can generate a variety of specialized cell types. In the simplest case, a newly fertilized egg – the stem cell with maximum potential – will ultimately give rise to every one of the 216 recognized human cell types. Stem cells are accordingly most readily derived from embryos; they can also be found in adults, but such cells tend to lack that embryonic ability to differentiate into any cell type. We are beginning to learn how to induce stem cells to produce particular cell types, and someday, I hope, we will be able to replace the lost brain cells in people with Huntington and Alzheimer with new healthy cells. But I caution that we have a long way to go before we have a thorough understanding of the molecular triggers that cause a cell to develop in one direction rather than another. It will take ten years or so of grappling with this fundamental problem in developmental biology before we are in a position to explore properly the therapeutic value of stem cells. I think it would be a tragedy for science, and for all people who may eventually benefit from stem-cell therapy, if research is hindered by religious considerations. Polls consistently show that the majority of Americans favor research using embryonic stem cells, and yet politicians continue to pander to the outspoken religious minority that opposes it. The result is restrictive legislation in the United States that is hampering efforts to develop this potentially valuable technology.

  For now, treating genetic disorders does not extend to the wholesale replacement of cells a la stem-cell therapy, but it may involve replacement of a missing protein. Striking 1 in 40,000 individuals, Gaucher disease is a rare condition resulting from a mutation in the gene for glucocerebrosidase, an enzyme that helps break down a particular kind of fat molecule, which otherwise accumulates harmfully in the body's cells. The disorder can be devastating, with a suite of symptoms including bone pain and anemia. Initial attempts to supply the missing enzyme directly were made as early as 1974. The results were promising but the logistics were nightmarish: the replacement enzyme had to be extracted from human placentas, and twenty thousand placentas were needed to furnish a year's supply for a single patient. A big breakthrough came in the early 1990s, when researchers synthesized a modified form of the enzyme, one that was taken up more efficiently by the cells that most needed it. In 1994, Genzyme, a biotech company, started to produce the modified form using recombinant methods. Treatment of Gaucher does not combat the genetic root of the disorder but rather the effect of the mutation: it provides the patient with the vital protein that the faulty gene cannot.

  Righting the genetic abnormality via this biochemical route is evidently feasible and effective. But even given the remarkable efficiency of recombinant methods, the treatment is expensive – $175,000 per year – and the need for continual infusions imposes a burden on patients. Naturally, therefore, geneticists have long dreamed of a practical way to fix the cause of the problem rather than compensate for its effects. The ideal treatment for genetic disorders would be a fo
rm of genetic alteration, a correction of the genes that cause the problem. And the benefit of such gene therapy would last the patient's whole life; once fixed, it's fixed for good. There are, at least in principle, two approaches: somatic gene therapy, by which we change the genes within a patient's body cells; or germ-line therapy, by which we alter the genes in a patient's sperm or egg cells, preventing the transmission of the harmful mutation to the next generation.

  Such solutions to the ravages of genetic defects might be obvious, but the idea of gene therapy has not met with the warmest of professional or public receptions. Such reactions are not altogether surprising: a culture wary of genetic modification in a corn plant might be expected to be averse to transgenic people – GM humans, if you prefer – however great the potential benefit. And more vociferous objections are made, also not unexpectedly, to the germ-line approach, because of the risk of causing genetic damage when manipulating the DNA. In somatic gene therapy, such damage may be limited in its effect; in germ-line therapy the possibility exists of accidentally producing impaired people. Even its proponents – of whom I count myself one – would never suggest that such a procedure should be carried out until our techniques are good enough for us to be confident that we will not inadvertently cause damage. Many scientists are convinced, though, that we should never attempt germ-line gene therapy. Whether based in ethics or unfounded fears of the unknown, such arguments are ultimately not compelling in my judgment. Germ-line therapy is in principle simply putting right what chance has put horribly wrong. But for now the controversy is academic: germ-line therapy is still way beyond our technical powers. Until it is in reach, we should concentrate our efforts on making somatic gene therapy a powerful tool in its own right.