Dna: The Secret of Life

by Watson, James

  Sometimes it is almost more unbearable to care for a sufferer than to become one. Carol Carr of Hampton, Georgia, watched her husband, Hoyt, develop Huntington disease in his thirties. His sister Roslyn died of it, and his brother George committed suicide soon after being diagnosed. Carol quit her job and became Hoyt's full-time nurse for the next twenty years, as he fell apart. They already had three sons by the time Hoyt was diagnosed, and when he died in 1995, Carol was already nursing her two eldest, Randy and Andy, as she had her husband – feeding and bathing them, giving them their medicine, helping them to the bathroom. Soon James, her youngest, was also developing symptoms. In despair, Carol reluctantly placed Randy and Andy in a nursing home, where, on June 8, 2002, she shot them both dead. The New York Times reported James's opinion that Huntington had killed his brothers long before his heartbroken mother ever pulled the trigger.

  Not all genetic diseases are tragedies of medical helplessness. Perhaps the best example to the contrary is the disorder responsible for that strange fine-print warning appearing on some food products, especially soft drinks: "Contains Phenylalanine." Phenylalanine is an amino acid – a common component of proteins – that cannot be processed by people with a genetic disorder called phenylketonuria (PKU).

  The story starts in Norway in 1934. A young mother was determined to find out what was wrong with her two children, ages four and seven, both of whom had seemed perfectly normal at birth. The elder was not fully toilet trained and was barely capable of speaking a few words, let alone forming a complete sentence. The case came to the attention of Asbjørn Føiling, a biochemist and physician. After conducting a battery of tests, Følling found a biochemical abnormality he linked to their condition: they had too much phenylalanine in their urine. But he also learned that theirs was no isolated case: he discovered thirty-four others in twenty-two families across Norway, and realized that he had stumbled across a genetic disease.

  We know now that PKU is caused by a mutation in the gene for phenylalanine hydroxylase, an enzyme that converts phenylalanine to another amino acid, tyrosine. It is a rare disorder, affecting about 1 in 10,000 people in North America, and shows a recessive pattern of inheritance: you must have two mutated copies of the gene, one from each parent, to develop PKU. In the children affected, who lack a functioning enzyme, phenylalanine accumulates in the blood, impairing brain development and leading to severe mental handicap. Prevention is simple: PKU children raised from birth on a diet low in phenylalanine – with minimal protein and no artificially sweetened soft drinks, the two principal sources – grow up normal. Nutrition alone can make the difference between normal brain development and profound disability. Clearly it is important to know a child's PKU status as soon after birth as possible. Robert Guthrie devised a simple diagnostic test for blood levels of phenylalanine and tirelessly promoted its use until it became standard neonatal practice. Since 1966, a heel-prick blood sample has been taken from every newborn and analyzed for phenylalanine levels. Thus without ever examining a single base pair of DNA, the Guthrie test screens for a genetic disease in millions of babies every year. Prior to this testing program, as much as 1 percent of mental retardation in the United States was attributable to PKU; now there are only a handful of cases a year.

  The 1950s saw the development of cytogenetics, the study of chromosomes through the microscope. Employed diagnostically, this approach soon revealed that abnormalities in chromosome number – usually one too many or one too few – invariably cause profound dysfunction. These problems stem from an imbalance in the number of genes, a departure from the norm of two of each. Such conditions do not run in families like Duchenne muscular dystrophy (DMD) or cystic fibrosis (CF), but they are still very much genetic; they arise spontaneously through accidents in the cell divisions leading to the generation of sperm and egg cells.

  The best known is Down syndrome, named for John Langdon Down, who in 1866, as the medical superintendent of a home for the retarded, was first to describe its characteristic clinical features. He noted that 10 percent of the residents of his institution resembled one another: "So marked is this, that when placed side to side, it is difficult to believe that the specimens compared are not children of the same parents." But the first insight into the condition's biological basis did not come until ninety years later, when the French physician Jerome Lejeune found that children with Down syndrome have three copies of one chromosome, later shown to be chromosome 21. The normal condition, two copies of a chromosome, is called "disomy"; so Down syndrome is known in genetic parlance as "trisomy 21."

  The incidence of Down syndrome increases with the age of the mother. At age 20, a woman's chance of producing a Down baby is about 1 in 1,700; but at 35, it jumps to 1 in 400; and at 45 shoots to 1 in just 30. For this reason, many older pregnant women choose to have prenatal diagnosis performed on their fetus to determine whether it might possess 21 in triplicate. The test was first done in 1968, and today it is routinely offered to all pregnant women over 35.

  Because the developing fetus must be big enough to withstand safely the extraction of a tissue sample, such diagnosis cannot be performed in the very early stages of pregnancy. Typically it is done in the fifteenth to eighteenth week using amniocentesis, a procedure that entails drawing some amniotic fluid (which naturally contains cells from the fetus). An alternative test, which may be done as early as the tenth week, gathers cells from the chorionic villus, the part of the placenta that attaches to the uterine wall, but this method is less reliable. Because both procedures are mildly risky – amniocentesis results in a 1 percent rate of miscarriage, and chorionic villus sampling in a 2 percent rate – younger women are usually advised to avoid them: the probability that their fetus has a genetic defect is actually lower than the probability that it will be damaged by the procedure. At one time, extracted fetal cells had to be grown in petri dishes before being processed for chromosome analysis. Nowadays, a more rapid diagnosis can be done using fluorescence in situ hybridization (FISH); in this method, a small fluorescent molecule is attached to a stretch of DNA sequence specific to chromosome 21 and introduced to the sample, where it binds to the fetal chromosome 21 DNA. If two fluorescent patches appear in the nucleus of a cell, the fetus is normal; if three, the fetus has Down syndrome (see Plates 56 & 57).

  In Britain, 30 percent of Down pregnancies are detected by routinely testing the oldest 5 percent of women bearing children. This method boasts a clear efficiency in simple terms of detections per pound spent (Britain's National Health Service has been subject to such a calculus ever since Mrs. Thatcher's assault on health spending), but what of the remaining 70 percent of Down cases? Down is rarer in the babies of younger mothers, but these women account for the vast majority of all pregnancies. Since the standard tests are statistically not worth their attendant risks, there have been attempts to find alternative, noninvasive, indicators. It turns out that substances detectable in the mother's blood yield useful information. Low levels of alpha-fetoprotein and high levels of chorionic gonadotropin correlate to a significant degree with Down (though they are by no means ironclad indicators of trisomy). Modern practice, then, is to offer younger women the blood test, and, if it suggests the possibility of Down, they are then counseled to undergo amniocentesis or chorionic villus sampling for a definitive diagnosis.

  Sadly, today a woman who learns that her fetus has Down syndrome has only two choices: to become the mother of a Down baby or to abort the fetus. It is a painful decision that is made no easier by the variable severity of the affliction. People with Down all share the characteristic facial features identified by Dr. Down – a broad flat face, a small nose, and narrow slanting eyelids* – but they range considerably in IQ, scoring between 20 and 85 (i.e., from severely handicapped to low normal). They are especially prone to a range of ailments, including heart disease (which claims about 15 percent in the first year of life), gastrointestinal anomalies, leukemia, and, with increasing age, cataracts and Alzheimer; but it's also perfectly
possible that an individual will have relatively few health problems. With improved care, and better knowledge of medical hazards posed by possession of that extra chromosome, life expectancy has increased substantially: 50 percent of affected individuals today survive into their fifties. Despite typically acquiring over time what most would consider a depressing familiarity with the insides of hospitals, people with Down generally enjoy life, and have brightened many a family. The condition is perhaps tougher on their parents, who must adjust to caring for someone with special medical needs, as well as to the knowledge that their child will, in many ways, never really grow up.

  *Dr. Down originally called the disorder "mongolism" on the basis of these characteristics, entitling his 1866 paper "Observations of an Ethnic Classification of Idiots." Subscribing to the racist evolutionary views of his day, he believed that Down represented an evolutionary step backward from the exalted Caucasian state to the "inferior" Mongoloid one. To give him his due, though, he concluded that what he called "retrogression" undermined the claims of those who refused to accept that Caucasians and non-Caucasians were members of the same species.

  In general, women who learn that they are carrying a Down fetus choose to terminate the pregnancy.† As a result, in countries with routine prenatal screening, the number of Down babies born is declining. Statistically, however, this claim is more complicated than it sounds: the trend toward deferring motherhood – often for professional reasons – has actually increased the ranks of women at risk for a Down pregnancy. In Britain, therefore, the efficacy of screening programs is measured relative to the expected number of Down babies given the ages of the women having children that year. We are seeing an ever-decreasing proportion of Down babies; in 1994, for instance, screening programs reduced the incidence of Down by about 40 percent.

  †In the United Kingdom, 92 percent of fetuses diagnosed with Down are currently aborted. Typically only women who are willing to consider an abortion undergo prenatal testing (there's no point subjecting a fetus to the risks associated with testing if the mother intends to carry the pregnancy to term whatever the result), so we would expect this figure to be high.

  Trisomies can also occur for other chromosomes, but these result in abnormalities so severe that the pregnancies abort spontaneously in all cases except trisomies of chromosomes 13 and 18. But children with trisomy 13 seldom live more than a few weeks, and those with trisomy 18 usually die before their first birthday. Chromosomal abnormalities, trisomies included, are probably very common. While many are lethal – a current estimate is that as many as 30 percent of conceptions end in spontaneous abortion, and in about half of these there is some form of chromosomal aberration – some have little or no effect. Alterations may be far less drastic than the loss or gain of an entire chromosome, involving the rearrangement of segments within a chromosome or the transfer of part of one chromosome to another. If there has been a net loss or gain of genetic material, then, as in the case of a whole extra chromosome, the resulting imbalance will usually prove deleterious. Unfortunately, standard cytological analysis of fetal chromosomes can detect only gross imbalances, and yet even minor ones can have disastrous effects.

  After struggling to become pregnant for the first time at thirty-seven, Kathleen McAuliffe was relieved to learn that only two chromosome 21s had shown up in her amniocentesis. But what she had not realized was that the test could reveal other chromosomal abnormalities as well. The cytogeneticist had spotted an inversion in the fetus's chromosome 2: it was as though a segment had been popped out of the chromosome, flipped, and reinserted the other way around. The information was not accompanied by any useful advice: there was a chance that the inversion might create a problem – it might, for example, have resulted in a genetic imbalance – but then again it might have no effect. One way to find out more was to look at McAuliffe's own second chromosome, and that of her husband. If either parent had the inversion (i.e., if it was not a spontaneous alteration in their child), one could infer it would have little or no impact since both parents were normal. But neither McAuliffe nor her husband had an inverted chromosome 2, implying that it had arisen de novo in the sperm or egg. What would the inversion do to the baby? McAuliffe suddenly found herself confronting a life or death decision. After agonizing at length, she decided that the uncertainty was too great and she terminated the pregnancy. Despite a specific request that she not be informed of the autopsy results – she was sad and guilt-ridden over the loss of her fetus – by some administrative gaffe the report was sent to her home and she discovered that the fetus had indeed been profoundly abnormal. But this was cold comfort, and McAuliffe still keeps the ultrasound image tucked away in a drawer. Happily, subsequent pregnancies have met with no such complications, and McAuliffe is now blessed with two young children in, as she puts it, "ear-piercing good health."

  Genetic knowledge creates ethical dilemmas. McAuliffe had never been warned that her amniocentesis might detect problems other than trisomy 21; perhaps the cytogeneticist overstepped the bounds of duty and should have reported only the results for the test that had been ordered. Certainly there would have been no choice had the clinician used the FISH method, which reveals only the number of 21s present. As it grows more sophisticated, genetic testing becomes a Pandora's box, its consequences going far beyond the original issues motivating the test, sometimes spreading to lives beyond those of the tested individuals. Nowhere is this more evident than in genetic testing conducted among families with histories of an inherited condition like DMD, Huntington disease, or cystic fibrosis. In these cases, diagnosis is carried out not by a cytogeneticist but by a molecular biologist, who analyzes not chunks of chromosomes but specified stretches of DNA. DNA is extracted from a sample of tissue, which is obtained from a fetus by amniocentesis, or from a child or adult by taking blood or harvesting cheek cells from inside the mouth with the scrape of a spatula. These days tests usually involve PCR amplification of the critical region – the suspect gene – from the DNA sample, followed by sequence analysis to determine whether or not it carries the mutation. And the test results for any individual may tell us something about the genetic status of his or her relatives.

  Let's take, for example, a test for Huntington disease. In a recent case, a man in his twenties came into a genetics clinic to request that he be tested for Huntington. His paternal grandfather had died of the disease, and his father, in his forties, had decided not to be tested, preferring, like Nancy Wexler, to live with 50-50 uncertainty over knowing for sure. Because Huntington strikes relatively late in life, it was possible that the father was in fact carrying the mutation even though symptoms had not yet appeared. The young man knew that the probability of his having the mutation – and therefore the disease in his future – was 1 in 4.* But he wanted to know for sure. The problem is this: if he found out that he did indeed have the mutation, then he must have received it from his father, meaning that his father, too, would definitely develop the disease. The son's quest for genetic knowledge would directly contravene the father's desire to avoid it. A family feud developed, and in the end, only intervention by the young man's mother prevented him from proceeding with the test. His desire to know, she argued, surely paled beside her husband's right to be shielded from what may be a devastating death sentence. This dramatic example illustrates the difference between genetic diagnosis and any other kind. What I might learn about my genes has implications for my biological relatives, whether they care to know or not.

  *There is a 1 in 2 chance that the father had received the mutation from the grandfather, and then, if the father has it, another 1 in 2 chance of his having passed it on to the son. The probability for the son is the product of these independent events, a 1 in 4 or 25 percent chance.

  Sometimes the implications may not bear on the present generation but rather on generations to come. Fragile X is the commonest form of inherited mental retardation. (Down syndrome is more frequent, but, occurring spontaneously, it is not usually inherited.
) In addition to a low IQ, symptoms typically include a notably long face with an outsize jaw and ears, and a hyperactive, occasionally irritable temperament. Like DMD, it is a sex-linked disorder (the gene responsible is on the X chromosome), but, unlike DMD, it affects females as well as males. One normal copy of the gene is evidently not enough to render the effect of a mutated one negligible; still, women tend to suffer less severe symptoms, and their incidence is 1 in 8,300, compared with 1 in 5,000 for males. Fragile X is caused by a mutation similar to the one responsible for Huntington disease: a DNA triplet, CGG, is repeated over and over again. Normal individuals have about 30 while carriers of fragile X have at least 50 and sometimes as many as 90. For reasons we do not fully understand, the number of repeats tends to increase with each generation; and once there are about 230 CGG triplets, the gene can no longer make mRNA and therefore ceases to function. The condition gets its name from a discernible structural weakness in the X chromosome caused by all these repeats.

  As the number of repeats increases from one generation to the next, so the severity of the condition increases, and the age of onset decreases in each family line. The latest descendants in a fragile X pedigree have the largest numbers of repeats and are typically affected earlier in life and more severely than those from whom they inherited the mutation. Geneticists may therefore identify individuals carrying a "premutation" – too few repeats to cause problems at present, but sufficient to result in fragile X in subsequent generations, given the likely expansion next time around. We do not yet know exactly what the protein produced by the affected gene does, but it seems to bind to messenger RNA molecules in the connections – synapses – between nerve cells.