Advertisement
Mayo Clinic Proceedings Home

Ordering and Interpreting DNA Tests

  • Pamela S. Karnes
    Correspondence
    Address reprint requests to Dr. P. S. Karnes, Department of Medical Genetics, Mayo Clinic Rochester, 200 First Street SW, Rochester, MN 55905
    Affiliations
    Department of Medical Genetics, Mayo Clinic Rochester, Rochester, Minnesota
    Search for articles by this author
      DNA testing has become relatively common during the past few years. All humans have an estimated five to eight mutated genes and are at risk for a genetic disorder. Recognition of these facts and development of DNA technology have created substantial interest in DNA testing as a diagnostic and management tool. As with any emerging technology, the applications and limitations must be understood in order to use the new tests wisely. In this review, the basic molecular diagnostic techniques are discussed. In addition, the emerging ethical debate surrounding the use of this new technology is summarized. Caution is necessary when DNA tests are being used, especially in presymp-tomatic situations.
      FAP (familial adenomatous polyposis), PCR (polymerase chain reaction)
      The availability of new DNA tests has been staggering in light of the fact that few DNA diagnostic laboratories existed 7 years ago. Localization of new genes is often widely reported in the lay literature, and thus patients have considerable interest and ask primary-care physicians many questions. This review discusses basic concepts of molecular genetics, offers guidance in ordering DNA tests, and discusses ethical concerns about the use of DNA tests.

      GENE STRUCTURE

      More than 100,000 gene pairs are organized on 46 chromosomes in each nucleated cell of the body. On average, about 2,200 genes are on each chromosome. Genes are submicro-scopic units of inheritance and thus cannot be analyzed by using conventional laboratory techniques. They are composed of exons (coding regions) separated by introns (noncoding regions that account for about 90% of most genes). Genes may be millions of base pairs in length; the total human genome is about 3 billion base pairs long.
      • Rosenthal N
      DNA and the genetic code.
      Mutations can alter the structure of the gene, its regulatory regions, or its splice sites. Thus, researchers who are attempting to develop cost-effective DNA tests for genetic diseases have a technologically difficult task.
      • Korf B
      Molecular diagnosis (first of two parts).
      • Korf B
      Molecular diagnosis (second of two parts).
      The first DNA mutation tests offered clinically were for conditions in which only a few mutations were responsible for the disorder. Mutations in some genes occur at the same position or in the same exon in many patients. For example, about 70% of all cystic fibrosis mutations in humans occur at position 508 in the cystic fibrosis gene. Therefore, DNA analysis for cystic fibrosis can be targeted to detect the δF508 mutation and several other mutation hot spots within the gene. Consequently, the expense and effort involved in sequencing the entire gene can be avoided because most mutations can be detected by targeting a few hot spots. Alternatively, mutations in some genes occur at different positions in virtually every patient. For example, only about 5% of patients with type I neurofibromatosis have identical mutations. Because the neurofibromatosis gene is very large and virtually all patients have different mutations, performing a cost-effective DNA-based test for type I neurofibromatosis is currently not possible. Thus, detection of a substantial percentage of mutations would necessitate sequencing of the entire gene or development of a different method to detect mutations indirectly.

      LINKAGE ANALYSIS

      When specific mutation analysis is not feasible but the location of a gene is known, an indirect method of DNA testing can be used to determine whether family members at risk have inherited the mutated gene. This technique is called linkage analysis, and sometimes the Southern blot technique is used. As with any DNA test, DNA must be isolated from cells. DNA is fairly stable and can be obtained from any cell that has a nucleus. Occasionally, DNA can be extracted from cells that have been embedded in paraffin or from old blood on filter paper. Restriction enzymes that recognize specific sequences of DNA are used like “molecular scissors” to cut DNA at specific sites. This phenomenon reduces the DNA into pieces that can be handled. The size of the pieces varies from person to person because of benign differences in DNA sequences (polymorphisms). The DNA is then sorted by fragment size in an electrified gel; smaller pieces travel further in the gel. In order to facilitate detection of a specific fragment, the DNA is transferred from the gel to a nitrocellulose filter. A probe (often radioactive) is then used to search for a specific piece of a gene. The probe and gene bind together like a lock and key and are detected as a band on x-ray film. In linkage analysis, the pattern of bands is compared between various family members. Thus, the pattern of bands from one region of one chromosome should be identical between a parent and any offspring to whom the region is passed. Polymerase chain reaction (PCR) is another molecular technique that can be used to generate banding patterns that can be tracked through a family. PCR will be subsequently discussed in more detail.
      Linkage analysis is used to track a pattern of DNA markers (various sized bands) through a family. If some members of a family have familial adenomatous polyposis (FAP, an autosomal dominant disorder, the gene for which is located on chromosome 5), symptoms of the condition may not develop until midadulthood, and unless the colon is screened, the first sign of the disease may be colon cancer. One method for screening presymptomatic patients at risk is serial colonoscopy. Another method is indirect DNA testing. By analysis of a region of chromosome 5 from several family members, the pattern of DNA markers shared by the family members known to have the condition can be determined. For linkage analysis to be effective, DNA must be obtained from at least one affected family member when the disorder is inherited in an autosomal recessive fashion. DNA from two affected family members is necessary to perform linkage analysis of autosomal dominant disorders like FAP. Ideally, DNA from both parents of affected family members should be obtained in order to increase the specificity of the test. Nonpaternity will be identified, and families should be warned of this possibility before blood is withdrawn. Rarely, linkage analysis will not be helpful because of the inability to distinguish one family member's sample from another. In this situation, the banding pattern in some affected and unaffected family members is identical. If this occurs more than 5% of the time with any specific DNA test, the test should not be used clinically. In order to avoid these problems, laboratories must develop a battery of probes; thus, if one set of probes cannot distinguish a particular chromosome in various family members, then another set can be used. An example of Southern blot analysis of a family that has two children with cystic fibrosis is shown in Figure 1. The DNA in five family members has been digested by a restriction enzyme and run out on a gel. One probe has been added. In the father (lane 1), the probe detects a fragment of size A on one chromosome 7 and a fragment of size B on his other chromosome 7. In the mother (lane 2), the same two fragments are shown. This situation does not imply that the DNA around the fragments in the parents is identical. Their oldest child (lane 3) has cystic fibrosis and has received two “A” fragments—one from the mother and one from the father. On the basis of this information, the fragment A in each parent is tracking with a cystic fibrosis mutation. The second child (lane 4) is a carrier (having received an A from one parent). The third child (lane 5) is also affected. Of importance, this interpretation of fragments A and B is useful only for this specific family.
      Figure thumbnail gr1
      Fig. 1Southern blot analysis. Lanes 1 through 5 contain DNA from five family members. In this two-fragment system, “A” refers to fragment size A and “B” refers to fragment size B. Each family member has two fragments, one from one member of a chromosome pair and one from the other member of the chromosome pair. When only one fragment seems to be evident, this reflects the fact that the fragments from both members of the chromosome pair are the same and are “on top of each other” on the gel.
      When DNA tests are being ordered, the phenomenon of genetic heterogeneity must be recognized. For example, tuberous sclerosis is a genetically heterogeneous disorder. Mutations in several different genes are capable of causing the disorder clinically recognized as tuberous sclerosis. Of note, in linkage analysis, a mutation is not being identified. Rather, a specific segment of DNA is being tracked through a family. Knowing that the correct gene region is being tracked is important.

      MUTATION ANALYSIS

      Mutation analysis is intended to detect specific alterations in a gene and is maximally efficient if most people have one of a small number of mutations. Mutation analysis can be performed in several ways. Sometimes a mutation is known to remove a specific site that is usually cut by a restriction enzyme. Occasionally, the mutation is detected directly by analysis of a limited, specific sequence of DNA. For example, among patients with cystic fibrosis, the δF508 mutation accounts for about 70% of mutations. If two copies of this mutation are detected, the patient has cystic fibrosis. If only one δF508 mutation is detected, the patient could be a carrier (one cystic fibrosis mutation and one normal cystic fibrosis gene), or the patient could have cystic fibrosis (one identifiable and one unidentifiable mutation). The difference between carriers and affected persons is usually clinically apparent. Of more concern is the situation in which a person is identified as a noncarrier because none of the common mutations are identified. For example, if one member of a couple is known to be a carrier for an autosomal recessive condition such as Tay-Sachs disease and if the other member of the couple is also identified to be a carrier, the chance that the couple would have a child with Tay-Sachs disease is 1 in 4 with each pregnancy. If one partner is identified as a noncarrier based on mutation analysis, the possibility that the couple would have an affected child seems to be negligible; however, the actual risk, considering the possibility that that partner had an unidentified mutation, is about 0.8%. This percentage is of clinical significance to some couples and may be interpreted to differ substantially from “negligible” in the patient's mind. Thus, to be clinically helpful, mutation analysis should be capable of identifying most mutations (at least 90%). Calculations can then be made that allow a reassignment of risk.
      PCR is used in mutation analysis and sometimes in linkage testing. PCR allows amplification of very specific pieces of DNA from extremely small samples; thus, the previously mentioned tests can be performed.
      • Rosenthal N
      Tools of the trade—recombinant DNA.
      DNA tests can sometimes be performed when only a small amount of blood or tissue is available from the index patient. DNA does not have to be obtained from living cells; tests can occasionally be done on archival samples. PCR decreases the amount of time necessary to perform the test because the PCR products can be run out on a gel and be read directly. Because of its sensitivity, PCR-based testing is vulnerable to minute contamination by any DNA, even a few skin cells from laboratory personnel, cross-contamination from a previous sample, or amplified DNA present in the laboratory. Strict laboratory procedures must be followed to ensure accuracy of results. PCR is also vulnerable to infidelity during replication, leading to artifactual mutations. In summary, PCR can provide a specific answer about whether a particular mutation is present when the PCR procedure is specifically designed for this purpose.

      PRACTICAL ADVICE

      Common DNA tests should be performed in a service laboratory that is subject to quality assurance testing. Although clinical testing is sometimes performed in research laboratories, the results must be used cautiously. Confirming a clinical impression of a genetic disorder based on a research DNA test differs considerably from basing medical and prenatal decisions on a research-based DNA test. A geneticist will be helpful in performing and coordinating clinical and research-based DNA testing.
      Informed consent is necessary for most DNA tests, especially those performed on a prenatal or presymptomatic basis.
      • McEwen JE
      • Reilly PR
      State legislative efforts to regulate use and potential misuse of genetic information.
      • Wilfond BS
      • Nolan K
      National policy development for the clinical application of genetic diagnostic technologies: lessons from cystic fibrosis.
      To give informed consent, the patient and family must be aware of potential repercussions. As previously mentioned, nonpaternity could be disclosed by linkage testing. Patients can become uninsurable with respect to life or health insurance if results are positive.
      • Ad Hoc Committee on Genetic Testing/Insurance Issues
      Genetic testing and insurance.
      Although a few laws exist to protect patients, they vary from state to state and have not been legally challenged. Physicians must also be aware of potential psychologic repercussions such as survivor guilt, effect on future life plans, and the effect on the relationships among family members. Relatively few published articles have dealt with the psychologic effects of DNA testing, especially when presymptomatic DNA testing is being considered for adult-onset disorders. A positive test result may negatively affect self-esteem and may affect how the person is viewed within the family. Moreover, the minor has lost the “right not to know.” Until recently, little had been written about the rights of minors relative to presymptomatic DNA testing.
      • Wertz DC
      • Fanos JH
      • Reilly PR
      Genetic testing for children and adolescents: who decides?.
      • American Society of Human Genetics Board of Directors
      • American College of Medical Genetics Board of Directors
      Points to consider: ethical, legal, and psychosocial implications of genetic testing in children and adolescents.
      Physicians must also be aware of the need to maintain privacy of family members. This may be relatively easy with mutation analysis because family members can receive their test results individually and can choose whether to share the information with other family members. With linkage analysis, explaining test results can be difficult if one family member does not want his status disclosed. Finally, with widespread use of DNA testing, population-based DNA screening will be possible. This strategy may create potential for discrimination and stigmatization.
      • Markel H
      The stigma of disease: implications of genetic screening.

      CONCLUSION

      DNA-based testing will continue to expand and offer physicians an important opportunity to make diagnoses and develop more appropriate management plans for patients. Nonetheless, the potential pitfalls of this new and exciting technology must be recognized. Unintended information may be obtained, and patients may be at risk for discrimination in employment and may have difficulty in obtaining health and life insurance. Interpretation of DNA test results can be difficult and is affected by such issues as a person's ethnic background, availability of blood from various family members, and reliability of the diagnosis. Thus, the ordering of DNA tests may be more complicated than that of other types of diagnostic studies available to physicians.

      REFERENCES

        • Rosenthal N
        DNA and the genetic code.
        N Engl J Med. 1994; 331: 39-41
        • Korf B
        Molecular diagnosis (first of two parts).
        N Engl J Med. 1995; 332: 1218-1220
        • Korf B
        Molecular diagnosis (second of two parts).
        NEnglJ Med. 1995; 332: 1499-1502
        • Rosenthal N
        Tools of the trade—recombinant DNA.
        N Engl JMed. 1994; 331: 315-317
        • McEwen JE
        • Reilly PR
        State legislative efforts to regulate use and potential misuse of genetic information.
        Am J Hum Genet. 1992; 51: 637-647
        • Wilfond BS
        • Nolan K
        National policy development for the clinical application of genetic diagnostic technologies: lessons from cystic fibrosis.
        JAMA. 1993; 270: 2948-2954
        • Ad Hoc Committee on Genetic Testing/Insurance Issues
        Genetic testing and insurance.
        Am J Hum Genet. 1995; 56: 327-331
        • Wertz DC
        • Fanos JH
        • Reilly PR
        Genetic testing for children and adolescents: who decides?.
        JAMA. 1994; 272: 875-881
        • American Society of Human Genetics Board of Directors
        • American College of Medical Genetics Board of Directors
        Points to consider: ethical, legal, and psychosocial implications of genetic testing in children and adolescents.
        Am J Hum Genet. 1995; 57: 1233-1241
        • Markel H
        The stigma of disease: implications of genetic screening.
        Am JMed. 1992; 93: 209-215