Genetic Imprecision

As prenatal diagnosis becomes a rite of passage in many pregnancies, prospective parents expecting yes-or-no answersñif not guaranteesñfrequently are surprised when test results are ambiguous. From classical chromosome analysis to potential screening programs for cystic fibrosis, genetic tests today can give only peeks at an individualís genes. A great challenge today is the translation of that limited information into useful advice for families.

There are two basic levels of genetic tests. Cytogenetic testing produces a chart of a cellís chromosomes, a karyotype, arranged in size order. The source cell might be a fetal fibroblast sampled from amniotic fluid; a cell of a chorionic villus developmentally destined to become part of the placenta, representative of the fetus because it too derives from the fertilized egg; or perhaps a white blood cell from an anxious parentís peripheral circulation.

Some cytogenetic results are meaningful because of the severity of the condition detected. A third chromosome (called trisomy) 13 or 18, for example, nearly always results in a miscarriage, stillbirth, or severely ill newborn who will fail to reach his or her first birthday. However, some other results are less informative. Now, improved methods of staining chromosomes reveal nuances that may or may not affect health. For example, technicians who analyze karyotypes can spot subtle chromosomal rearrangements in a fetus. The health caregiver is then in the ethically disquieting position of having to tell a patient, "The amniocentesis result was abnormal, but we donít know what it means." Geneticistsí ability to detect deviations from the norm has outstripped their ability to predict symptoms, if any.

Getting down to the molecular level, the second area of genetic testing, presents problems in imprecision too, as current efforts to develop a meaningful screening test for cystic fibrosis illustrate. Geneticists once felt confident that a "single-gene defect" (when disease is caused by a defect in a single protein) would be simple to detect. Once the cystic fibrosis gene was discovered, they were sure a screening test was imminent. Such optimism was displayed at the fall 1990 meeting of the Cystic Fibrosis Foundation in a banner proclaiming that the organization would soon be out of business! Since that meeting, the discovery of many variants of the gene has tempered the initial excitement.

Just as ice cream comes in different flavors, a gene can come in several varieties, or alleles, representing differences in the DNA sequence. This variability sets the stage for misleading test results. A test based on detecting one allele, even the most prevalent one known to cause the disease, cannot rule out the presence of another allele that also might cause the disease. Conversely, the test may identify mutant alleles that do not alter the gene productís function, and therefore do not cause symptoms.

To those familiar with classical genetics, the recently revealed complexity of the cystic fibrosis gene is not a big surprise. "Genetics can be defined as a study of variability. This [protein behind cystic fibrosis] is just another big protein, and we should expect multiple mutations. Some will be ëprivate,í only seen within a particular family, and some will be common," said Michael Kaback, chairman of pediatrics at the University of California at San Diego and president of the American Society of Human Genetics, who spoke at the tenth annual conference of the National Society of Genetic Counselors in Cincinnati in October 1990. Kaback is well known in the field for heading the successful screening for Tay-Sachs disease carriers that began in the 1970s.

Cytogenetic testing has been with us for more than two decades; widespread molecular testing arose in the mid-1980s. The mapping of the human genome well under way in research laboratories certainly will augment the menu of conditions that can be detected prenatally. By considering the existing and expected dilemmas posed by ambiguous cytogenetic and molecular tests, researchers can perhaps find better ways to use the genetic information revealed in the human genome.

Preparing chromosomes for observation is moving from being an art to becoming a science. Efforts to display the human chromosomes date from the late nineteenth century. Until 1956, the total number of chromosomes could not be counted accurately and was variously cited as being anywhere from 37 to 48. The chromosomes are so numerous that obtaining a cell in which none of them touched each other was largely a matter of luck. A way had to be found to capture chromosomes in their most condensed stateñduring mitosisñand to spread them away from one another.

A combination of serendipity and ingenuity finally conquered the difficulties of displaying recalcitrant chromosomes. In 1951, a technician who mistakenly washed white blood cells in a salt solution less concentrated than the interiors of the cells found that the rush of water into the cells swelled the cells sufficiently to separate the tangled and overlapping chromosomes. Cytologists Joe-Hin Tjio and Albert Levan in 1953 contributed the technique of squashing a droplet of suspended cells between a cover slip and a glass microscope slide to spread the chromosomes.

As technological hurdles were overcome, the field of cytogenetics took off. In 1956, Tjio and Levan identified the correct diploid number of human chromosomes, 46, and cytologists C.E. Ford and J.L. Hamerton counted 23 chromosomes in human meiotic chromosomes. Within three years, three common chromosome anomalies were described: the extra chromosome 21 of Down syndrome, the XO chromosome constitution of Turner syndrome, and the XXY karyotype associated with Klinefelter syndrome. Persons with Turner or Klinefelter syndromes usually are infertile and develop diminished secondary sexual characteristics, if any.

Improved chromosome staining techniques were developed more slowly. In the 1960s, available stains could not distinguish among most chromosomes. The early staining revealed extra and missing chromosomes, large deletions and duplications, and the locations of centromeres (the major constrictions in chromosomes), but less extensive rearrangements of genetic material were not as obvious. Stains such as orcein and Feulgen enabled the chromosomes only to be grouped into seven classes, based on size. A person with Down syndrome in 1960 would have been described as having an extra G-group chromosome, rather than trisomy 21, the modern designation.

A second generation of stains became available in 1971, after T. Caspersson and his colleagues at the Karolinska Institute in Stockholm found that quinicrine mustard peppered the chromosomes with 320 dark bands. This "Q banding" soon was followed by the discovery of additional banding patterns. G bands were observed after treating cells with the protease trypsin and then staining with a dye called Giemsa. Light G bands correspond to euchromatin, which is active DNA rich in guanine and cytosine. Dark G bands are heterochromatin, unexpressed DNA rich in adenine and thymine. Most commercial cytogenetic labs today use G banding because the results are very consistent.

R bands, their name stands for "reverse Giemsa," are formed in an order opposite the G band pattern and are produced by using heated phosphate buffer along with Giemsa. R bands are not as consistent as G or Q bands. Dark R bands are thought to represent active genes. This idea is supported by the observation that chromosomes 13, 18, and 21 have the fewest R bands and the fewest light G bands. These three autosomes are the only ones that, when present in an extra copy, permit survival past birth. Survival may only be possible because the extra genetic material present in the three autosomes is deadweight DNA.

The availability of increasingly sophisticated banding techniques converged with the already developed technique of amniocentesis to usher in the era of prenatal testing. "Amniocentesis, available in the early 1960s, went hand in hand with chromosome analysis," says Kathleen Richkind, laboratory director at Vivigen, a cytogenetic testing facility in Santa Fe, New Mexico. Amniocentesis soon became routine for pregnant women older than 35, because at age 35 a pregnant womanís risk of conceiving a child with a chromosomal anomaly approximates the risk of amniocentesis damaging the fetus. (The risk of certain fetal chromosomal abnormalities rises with maternal age.)

Amniocentesis has provided useful information for the parents of many fetuses bearing extra or missing chromosomes. But the technique is not a crystal ball. Consider Down syndrome. Among children having identical karyotypes showing trisomy 21, one child may be healthy and only mildly retarded, another may have a serious heart defect, leukemia, or be profoundly retarded. The problem is that the diagnosis of trisomy 21 does not allow a clear prognosis. It only tells parents a range of what to expect.

Mosaicism also adds to the ambiguity. Some cells in a person can be normal and others abnormal. What if trisomy 21 is detected in two of twenty cells examined? Is this the same proportion of abnormal cells as will occur in the brain regions affected by the syndrome? There is no way to know. For example, if the affected cells are liver cells and not brain cells, the individual born will not be retarded.

Some chromosomal rearrangements are inherited from a parent. The fact that the affected parent is healthy provides some reassurance, but there are, again, no guarantees. "The fetus has a different genetic background," says Richkind. A rearrangement that is harmless in one person may have an effect in another. Vivigen estimates the risk of a birth defect resulting from such a "familial" chromosomal rearrangement to be approximately 3%; this percentage must be added to the 3% chance that any fetus has of having a birth defect.

Even more unsettling is a previously unreported anomaly that appears in a fetus but is not present in either parent. Such de novo rearrangements, as they are called in the cytogenetic lexicon, usually are either inversionsñwhen a reversed sequence of bands indicates that genetic material was flipped 180∞ñor translocations, in which a piece of a chromosome or an entire chromosome attaches to the end of a different (nonhomologous) chromosome (a Robertsonian or nonreciprocal translocation) or two chromosomes exchange parts (a reciprocal translocation). If these chromosomal contortions disrupt or delete active genes, a medical condition such as mental retardation can result. But there is no way to tell for certain if this has happened, other than spotting a significant physical abnormality on an ultrasound scan of the fetus. Unfortunately, many medical problems, such as mental retardation, do not show up on ultrasounds.

The result is tremendous pressure on the genetic counselor or physician who delivers the ambiguous news and on the recipient of the information. Said Janet Bylina, a dietitian who was told that her fetus had a chromosome with an inversion, "It was much worse than being in limbo. There was information, but not enough. We couldnít make an intelligent decision, not like for Downís or spine bifida. What if we aborted and things were fine?"

In cases of de nova rearrangements, about the best a caregiver can offer a patient is a probability of risk derived from relatively sparse empirical data. Lab reports are carefully worded. "When we detect a de novo rearrangement, we state that it appears, to the level of resolution that we can get in cytogenetics, that the genetic material is balanced [meaning all the genetic material is present]. But we know empirically from de novo balanced translocations (or inversions) that there is a slight risk of abnormality, about 5 to 10%, depending upon which report you cite," explains John Stone of Genetrix Inc., a cytogenetic testing company in Scottsdale, Arizona.

Many testing labs cite a 1987 study by Dorothy Warburton of Columbia University, who catalogued information from a survey of United States and Canadian cytogenetics laboratories to find out how abnormalities in genetic tests actually expressed themselves. She has just compiled a larger survey, which cytogenetics labs anxiously await. "Weíve looked at hundreds of cases now, and find as the numbers [of patients] get bigger, the percentage of abnormalities is stable, or even going slightly lower," Warburton says.

Trying to determine the risk that a de novo rearrangement will cause a medical problem is a statistical nightmare. First, fetal cells arriving at testing laboratories do not represent the general population of pregnant women. Some are collected after the patient has had abnormal biochemical test results or an abnormal ultrasound.

Second, many fetuses with de novo rearrangements are abortedñeither naturally or by parental choiceñso it is never determined whether the chromosomal flip or switch is associated with a defect. Third, many reports of a healthy baby are based on what the child looks like at birth, with no long-term follow-up, adds Warburton. Finally, a risk for a particular family, based on what happened in other families, is not definitive. Chromosomes from different families that stain in a similar manner may not have the identical variant at the molecular level.

The current ambiguities and limitations of amniocentesis may be resolved as data accumulate and molecular techniques refine chromosome analysis (see box page 291). Still, even single-gene defects are far from a simple matter, as the ongoing cystic fibrosis saga shows.

Cystic fibrosis is the most common genetic disease among US whites, of whom 1 in 25 are carriers and 1 in 2500 are ill. One early description of cystic fibrosis is a seventeenth-century English nursery rhyme that says a child that is "salty to taste" will die shortly after birth. Indeed, when parents report a salty taste on kissing a child, it is often an early sign of the disease, which may at first be noted as "failure to thrive." The first clinical description of cystic fibrosis, in 1938, cites a defect in exocrine gland ducts. This defect explains the symptoms: impaired breathing due to copious, viscous mucus clogging respiratory passages; poor digestion reflecting pancreatic and intestinal insufficiency; and the salty sweat. In 1960, the average patient lived 12 years, but by 1989, thanks to antibiotics, digestive enzyme supplements, and respiratory therapy, the average lifespan had risen to 28 years. A few patients have survived into their fifties.

Also in 1989, cystic fibrosis was put on the emerging genetic map. The elusive gene was identified by Francis Collins and his group at the University of Michigan at Ann Arbor and by Lap-chee Tsui and his colleagues at the Hospital for Sick Children in Toronto. They used a new technique, called chromosome jumping, to traverse the large genetic distances along chromosome 7 to reach the cystic fibrosis gene.

Most cases of cystic fibrosis stem from a single missing amimo acid, a phenylalanine at position 508 of a 1480-amino-acid-long protein dubbed the cystic fibrosis transmembrane conductance regulator. The defective protein resulting from the ∆F508 mutation apparently entraps chloride ions inside the epithelial cells that form exocrine gland ducts. In the respiratory tract, this entrapment draws water into the cells, leaving behind a thick mucus.

The initial euphoria at finding ∆F508 soon gave way to concern; it seemed that a new mutation in the gene was identified each week. The count as of January 1991 was 64. The mutations span the entire gene, and at least two of them, both amino acid substitutions, are silent, not producing detectable symptoms.

A carrier test to detect only the ∆F508 mutation will be falsely negative when a person carries a different mutation. The solution is to develop a panel of tests covering the most common alleles, which will improve the efficiency of the test.

For example, if 75% of all people carrying a cystic fibrosis mutation can be detected, then 56% (.75 x .75) of two-carrier couples will be identified. At a 95% carrier detection level, 90% of these couples will be found. (Each such couple has a one in four chance that each conceptus has cystic fibrosis, the odds for offspring of two carriers of an autosomal recessive trait.)

Explains Collins, "Testing for ∆F508 and, say, another three or four mutations will still only detect 80 to 85% of carriers. Thatís still missing a reasonably large fraction in the United States, if we screen the whole population. So I agree with the National Institutes of Health consensus statement that it is not yet time for population screening. But it is the time for pilot projects."

Several labs already offer testing for ∆F508 plus three or four relatively common mutations, bringing the efficiency to 80 to 85% for individuals in the general population. An added complication is that allele prevalence varies with ethnic group.

For now, cystic fibrosis testing is recommended only if an immediate family member is already affected. In these cases, ambiguities can be lessened by also conducting genetic marker tests, which require the participation of several family members to track the errant chromosome.

"I think the majority of genetic counselors are looking at family history, and, if there is none, help the family understand that it is not yet appropriate for mass screening," says Vickie Venne, genetic counselor at the Nichols Institute in San Juan Capistrano, California. She predicts that the genetic fine structure of the cystic fibrosis gene will ultimately prove so complexñwith many "personal" mutations unique to a single familyñthat researchers will go to the protein itself to develop a single type of screening test for adults wishing to know their cystic fibrosis carrier status.

Some geneticists and physicians such as Kaback and Collins question the very idea of identifying carriers of cystic fibrosis, an effort whose only outcome, at present, can be terminating a pregnancy or preparing the parents to care for an affected child. In contrast, after screening newborns for phenylketonuria, diagnosis is followed quickly by an effective treatment. Plus, as genetic diseases go, cystic fibrosis is not the worst.

ëëTay-Sachs disease, by anybodyís measure, is a horror. There is blindness, paralysis, a normal life for only the first three to six months. But a cystic fibrosis patient can be a 40-year-old who is active and intelligent. Therefore, the benefits of screening are less obvious. Eugenics of cystic fibrosis is an absurdity. Every one of us is a mutant," according to Kaback.

Which may be what the human genome project will ultimately tell us: that we all have genetic quirks. The incomplete information offered by amniocentesis, and the inefficiency of a screening test for a disease with many alleles, may be just the tip of a technological iceberg. Adequately using the genetic information expected to become available, will be the greatest challenge of the unprecedented effort to identify the genes that make us human.

A marriage between traditional karyotyping and molecular biology has produced a fast and powerful means of chromosome analysis: in situ hybridization with chromosome-specific DNA probes. Using this molecular cytogenetics technique, a technician can expose cells to short pieces of fluorescently tagged DNA that hybridize to specific parts of chromosome 21. Within hours, the technician can scan a cell nucleus under a fluorescent microscope. If three spots flash, a diagnosis of Down syndrome is made.

This technique is being pioneered by Oncor, Inc., of Gaithersburg, Maryland. It is based on the discovery in 1981 that modified DNA bases can be incorporated into DNA segments that make up the probe, so that when the probe zeros in on its complementary sequence on a specific chromosome, it can be detected by using standard histochemistry or fluorescence microscopy. The company uses dUTP (deoxyuridyl triphosphate) bonded to biotin, which in turn is bonded to avidin. (Avidin is a protein found in raw eggs that binds biotin and halts its absorption in the small intestine. People who eat many raw eggs can develop severe biotin deficiencies.) The avidin is bonded to fluorescein, which fluoresces, enabling the entire complementary structure to be detected. When the tagged dUTP (which base-pairs with adenine) is included in a probe for a sequence unique to chromosome 21, one has a test for Down syndrome. Two fluorescent spots indicate the normal pair of chromosomes, three spots indicate trisomy. Oncor offers such probes to researchers working on the clinically important chromosomes 13, 18, 21, X, and Y.

Oncorís probes target areas of repeated DNA sequences near the centromeres, the large constriction in each chromosome. The probes also may suggest potential reproductive outcomes for persons carrying a translocation. Although the carrier is healthy because all the genetic material is present, although rearranged, the chromosomes may not segregate normally at meiosis, resulting in gametes that are unbalanced. Therefore, some offspring may have too much of one chromosome and too little of another.

Distributing paired chromosomes into haploid gametes requires that the centromeres of homologous chromosomes repel one another. After a translocation, which chromosomes, if any, will be extra or missing in offspring depends on the source of the centromere of the translocated chromosomes. Banding analysis does not identify the centromere, but in situ hybridization with Oncorís probes can reveal from which chromosome a centromere hails, allowing the clinician to make firmer predictions about the likelihood of offspring abnormalities.

Another variant of molecular cytogenetics is chromosome painting, being developed by Amoco Technology Company in Naperville, Illinois. The company has developed from several genes unique to a chromosome something it calls a probe cocktail. The result: "An entire chromosome will light up under the fluorescent microscope. The technique literally paints the whole chromosome, rather than acting regionally. It is a replacement for banding," says Richard Benedict of the company.

Like Oncorís hybridization for repetitive sequences, Amocoís chromosome painting offers both speed and specificity. A cell with three chromosome 21s painted, for example, clearly spells Down syndrome. A cell in which chromosome 4 has been translocated would display painted portions of chromosome 4 on two unusual chromosomes.

So far, molecular cytogenetics in the medical laboratory has targeted only large stretches of DNA, but it has the potential to highlight individual genes. As the human genome project continues and optical equipment becomes more sensitive, in situ hybridization is expected to be used to reveal single genes. The art of cytogenetics and the science of molecular biology are fast becoming one and the same.

feedback form | permissions | international | locate your campus rep | request a review copy

digital solutions | publish with us | customer service | mhhe home

---

Copyright ©2001 The McGraw-Hill Companies.
Any use is subject to the Terms of Use and Privacy Policy.
McGraw-Hill Higher Education is one of the many fine businesses of the The McGraw-Hill Companies.