Chapter 17:  Genetic testing in individuals and populations

RNA has advantages over DNA, but is more difficult to obtain and handle

If a gene has to be scanned for unknown mutations, testing by RT-PCR (see Figure 6.5) offers several advantages. DNA testing usually involves amplifying and testing each exon separately, and this can be a major chore in a gene with many exons. Most of the mutation-scanning methods (see Table 17.5) can scan fragments larger than the average sized exon, so that an RT-PCR product can be examined using a smaller number of reactions. Also, only RT-PCR can reliably detect aberrant splicing, which is sometimes hard to predict from a DNA sequence change, or may be caused by activation of a cryptic splice site deep within an intron. However, RNA is much less convenient to obtain and work with. Samples must be handled with extreme care and processed rapidly to avoid degrading mRNA, and the gene of interest may not be expressed in readily accessible tissues. In addition, many mutations result in unstable mRNA (Section 9.4.6), so that the RT-PCR product from a heterozygous person may show only the normal allele.

Functional assays of proteins have a role in genetic testing

A protein-based functional assay might classify the products of a highly heterogeneous allelic series into two simple groups, functional and nonfunctional - which is, after all, the essential question in most diagnoses. The problem with functional assays is that they are specific to a particular protein. DNA technology by contrast is generic. This has obvious advantages for the diagnostic lab, but in addition it encourages technical development, since any new technique can be applied to many problems.

β-Thalassemia and cystic fibrosis illustrate different situations of limited mutational diversity

For both these conditions, a very large number of mutations in the relevant gene have been described, but in each case a handful of mutations account for the majority of cases in any particular population. With β-thalassemia different mutations are predominant in different populations (Table 17.2). DNA testing is not needed to diagnose carriers or affected people (orthodox hematology does this perfectly well) but it is the method of choice for prenatal diagnosis. Provided one has DNA samples from the parents and knows their ethnic origin, the parental mutations can often be found using only a small cocktail of direct tests, after which the fetus can be readily checked.

In cystic fibrosis, by contrast, the F508del mutation is the commonest in all European populations, and is believed to be of ancient origin. However, the proportion of all mutations that are F508del varies, being generally high in the north and west of Europe and lower in the south. Direct testing for cystic fibrosis mutations divides into two phases. First, a limited number of specified mutations, always including F508del, are sought using the methods described in Section 17.1.3. As Table 17.3 shows, there is no obvious natural cut-off in terms of diminishing returns on testing for specific mutations. If this phase fails to reveal the mutations, then if resources allow, a screen for unknown mutations may be instituted, using the methods described in Section 17.1.4, or alternatively gene tracking (see Figure 17.14) may be used. The impact of this diversity on proposals for population screening is discussed below (see Figure 17.18).

Some combination of founder effects and heterozygote advantage must be the cause of this relative mutational homogeneity in recessive disorders. However, surprisingly often, when a recessive disease is particularly common in a certain population, it turns out that more than one mutation is responsible. An example is Tay-Sachs disease among Ashkenazi Jews, where there are two common HEXA mutations (Table 17.1). It is difficult to explain this situation except by assuming there was substantial heterozygote advantage some time when the founding population was small.

Testing for the presence or absence of a restriction site

When a base substitution mutation creates or abolishes the recognition site of a restriction enzyme, this allows a simple direct PCR test for the mutation (Figure 6.6). Although hundreds of restriction enzymes are known, they almost all recognize symmetrical palindromic sites, and many point mutations will not happen to affect such sequences. Also, sites for rare and obscure restriction enzymes are unsuitable for routine diagnostic use because the enzymes are expensive and often of poor quality. Sometimes, however, a diagnostic restriction site can be introduced by a form of PCR mutagenesis using carefully designed primers. Figure 17.2 shows an example.

Use of allele-specific oligonucleotide (ASO) hybridization

Under suitably stringent hybridization conditions, these short synthetic probes hybridize only to a perfectly matched sequence (see Figure 5.10). Figure 5.11 demonstrates the use of dot-blot hybridization with ASO probes to detect the single base substitution that causes sickle cell disease. For diagnostic purposes, a reverse dot-blot procedure has often been used. A screen for a series of defined cystic fibrosis mutations, for example, would use a series of ASOs specific for each mutant allele, spotted onto a single membrane which is then hybridized to labeled PCR-amplified test DNA. Recently reverse dot-blotting has developed from manually-spotted arrays of small numbers of ASOs to very large ASO arrays on ‘gene chips’ that can potentially detect all possible mutations in a gene (Section 17.1.4).

Allele-specific PCR amplification (ARMS test)

The principle of the ARMS (amplification refractory mutation system) method was shown in Figure 6.9. Paired PCR reactions are carried out. One primer (the common primer) is the same in both reactions, the other exists in two slightly different versions, one specific for the normal sequence and the other specific for the mutant sequence. Additional control primers are usually included, to amplify some unrelated sequence from every sample as a check that the PCR reaction has worked. The location of the common primer can be chosen to give different sized products for different mutations, so that the PCR products of multiplexed reactions form a ladder on a gel. With careful primer design, the mutation-specific primers can also be made to give distinguishable products. For example, they can be labeled with different fluorescent or other labels, or given 5′ extensions of different sizes. Multiplexed mutation-specific PCR is well suited to screening large numbers of samples for a given panel of mutations (Figure 17.3).

The ARMS principle can be adapted to allow the course of a quantitative PCR reaction to be followed in real time, using TaqMan or related fluorescer-quencher methods (Section 6.2.3).

Oligonucleotide ligation assay (OLA)

In the OLA test for base substitution mutations, two oligonucleotides are constructed that hybridize to adjacent sequences in the target, with the join sited at the position of the mutation. DNA ligase will covalently join the two oligonucleotides only if they are perfectly hybridized (Nickerson et al., 1990). Various formats for the test are possible, for example as an ELISA reaction or, as in Figure 17.4, for analysis on a fluorescence sequencer.

Sequencing, particularly of RT-PCR product, is increasingly the method of choice for mutation scanning

As automated fluorescence sequencers become standard items of laboratory equipment, sequencing becomes more and more attractive as a means of mutation scanning (Figure 17.5). Mutations picked up by other methods are often confirmed by sequencing, so it is tempting to do the sequencing straight away. Nevertheless, sequencing is expensive, especially for scanning many exons of a gene in a genomic DNA sample, and much time can be wasted investigating artefacts, especially if the sequencing template is not of the highest quality.

Several methods rely on formation of heteroduplexes

Many tests use the properties of heteroduplexes to detect differences between two sequences. Most mutations occur in heterozygous form, and heteroduplexes can be formed simply by heating the test PCR product to denature it, and then cooling slowly. For homozygous mutations, or X-linked mutations in males, it is necessary to add some reference wild-type DNA. Several properties of heteroduplexes can be exploited:

• Heteroduplexes often have abnormal mobility on nondenaturing polyacrylamide gels (Figure 17.6, lower panel). Special gels (Hydrolink™, MDE™) are supposed to improve the resolution. This is a particularly simple method to use. If fragments no more than 200 bp long are tested, insertions, deletions and most but not all single-base substitutions are detectable (Keen et al., 1991).

• 

  

  Figure 17.7

  CF mutation detection by denaturing gradient gel (more...)

  

  Figure 17.7

  .

     CF mutation detection by denaturing gradient gel electrophoresis

  Each exon of the CFTR gene (except exon 9, which in this protocol is sequenced directly) is PCR-amplified in one or more segments and run on 9% polyacrylamide gels containing a gradient of urea-formaldehyde denaturant. Bands contain exons as labeled. The band from any amplicon that contains a heterozygous variant splits into usually four sub-bands (arrows). Individual 1 (left lane in each panel) has variants in amplicons 6, 10 and 14; individual B (right lane) has variants in amplicons 3, 10, 13, 16, 17 and 24. Characterization of the variants showed that individual 1 was heterozygous for F508del (exon 10), and individual 2 was a compound heterozygote E60X (exon 3) / R1070Q (exon 17). Other variants were nonpathogenic. Courtesy of Dr Hans Scheffer, University of Groningen, Netherlands.

  Heteroduplexes have abnormal denaturing profiles. This is exploited in denaturing gradient gel electrophoresis (DGGE, Figure 17.7) and denaturing high performance liquid chromatography (dHPLC). In both cases the mobility of a fragment changes markedly when it denatures. These methods require tailoring to the particular DNA sequence under test, and so are best suited to routine analysis of a given fragment in many samples. DGGE requires special primers with a 5′ poly(G;C) extension (a GC clamp). Once optimized, these methods have very high sensitivity. Two-dimensional DGGE gels have near 100% sensitivity for mutation detection (Dhanda et al., 1998). As with any method that finds every sequence variant in a large gene, it is necessary to sort out pathogenic from nonpathogenic changes (Figure 17.7).

• 

  

  Figure 17.8

  CF mutation detection by chemical cleavage of mismatch (more...)

  

  Figure 17.8

  .

     CF mutation detection by chemical cleavage of mismatch

  The test sample, PCR-amplified exon 13 of the CFTR gene, was denatured and allowed to renature. Because a mutation is present in heterozygous form, heteroduplexes are formed, and these are cleaved by treatment with OsO₄ (or KMnO₄) or hydroxylamine. Running on a fluorescence sequencer shows the full-length fragment (906 bp) plus cleavage products of 357 and 549 bp. Sequencing revealed the mutation 2184delA. Courtesy of Dr Julie Wu, St Mary's Hospital, Manchester.

  Mismatched bases in heteroduplexes are sensitive to cleavage by chemicals or enzymes. The chemical cleavage of mismatch (CCM) method (Figure 17.8) is a sensitive method for mutation detection, with the advantages that quite large fragments (over 1 kb) can be analyzed, and the location of the mismatch is pinpointed by the size of the fragments generated. Its disadvantages are that it uses very toxic chemicals, particularly osmium tetroxide (though this can be substituted by potassium permanganate), and some practice is needed before it works well. An alternative is enzymic cleavage of mismatches, which uses enzymes such as T4 phage resolvase or endonuclease VII to achieve the same result without the toxic chemicals. Unfortunately in most people's hands, the quality of the gels produced leaves much to be desired.

Single-strand conformational polymorphism (SSCP) analysis is one of the most popular methods for mutation scanning

Single-stranded DNA has a tendency to fold up and form complex structures stablilized by weak intramolecular bonds, notably base-pairing hydrogen bonds. The electrophoretic mobilities of such structures on nondenaturing gels will depend not only on their chain lengths but also on their conformations, which are dictated by the DNA sequence. For SSCP (Figure 17.6), amplified DNA samples (which may be RT-PCR products) are denatured and loaded on a nondenaturing polyacrylamide gel. Primers can be radiolabeled, or unlabeled products can be detected by silver staining. Control samples must be run, so that differences from the wild-type pattern can be noticed. SSCP is simple and adequately sensitive for fragments up to 200 bp long, but it does not reveal the nature or position of any mutation detected (Sheffield et al., 1993). SSCP and heteroduplex analysis can be combined on a single gel, as in Figure 17.6. The precise pattern of bands seen is very dependent on details of the conditions. An elaboration of SSCP, dideoxy fingerprinting, analyzes each band in a sequencing ladder by SSCP, and is claimed to give 100% sensitivity (Sarkar et al., 1992).

The protein truncation test (PTT) provides an efficient means to check for mutations that produce premature termination codons

The PTT (Figure 17.9) is a specific test for frameshifts, splice site or nonsense mutations that truncate a protein product (van der Luijt et al., 1994). Clearly, the strength and weakness of the PTT is that it detects only certain classes of mutation. It would not be useful for cystic fibrosis, where only a minority of mutations introduce premature termination codons. But in Duchenne muscular dystrophy, adenomatous polyposis coli or BRCA1-related breast cancer, missense mutations are infrequent, and any such change found may well be coincidental and nonpathogenic. For such diseases, the PTT has several advantages. It conveniently ignores silent or missense base substitutions, and (like mismatch cleavage methods, but unlike SSCP) it reveals the approximate location of any mutation. Large exons, such as exon 15 of the APC gene (6.5 kb), can be tested using genomic DNA rather than by RT-PCR. Several variants have been developed to give cleaner results, usually by incorporating an immunoprecipitation step.

High-density oligonucleotide (GeneChip®) arrays promise to make high-throughput mutation scanning feasible

Trials to date have involved custom-designed gene-specific arrays of 20–25-mer oligonucleotide probes, in either of two basic designs (Figure 17.10):

• Hybridization chips contain oligonucleotides matching all wild-type and single nucleotide substitution sequences in a gene. Test DNA is PCR-amplified, fluorescently labeled and hybridized to the array, either alone or (preferably) in competition with a reference wild-type sequence, labeled with a different color. Trials of these designs have achieved over 90% mutation detection in blind trials of BRCA1 and ATM samples (Hacia, 1999). In general these designs detect homozygous base substitutions well, but miss a few heterozygous substitutions and have major problems with insertion mutations.

• Minisequencing chips use arrayed oligonucleotide primers with free 3′-OH groups. Unlabeled PCR-amplified test DNA is hybridized to the array, and DNA polymerase plus four differently labeled dideoxynucleotides are added. The test DNA acts as template for addition of a single labeled dideoxynucleotide to each array primer. As in an ARMS reaction, addition will occur only if the 3′ end of the primer exactly matches the template. The array could be made with primers specific not only for the wild-type sequence but also for all possible mutations. This technology has been less extensively tested than hybridization array technology because the chemistry used to produce most large arrays to date anchors the 3′ end of the oligonucleotide to the support, leaving a free 5′ end.

The emerging technology of gene chips promises to revolutionize mutation scanning, as well as other branches of human molecular genetics, but the revolution has not yet happened. Mutation detection arrays face a more difficult problem than expression arrays - the anchored probes are shorter, so specific hybridization is more difficult to achieve, and the test DNA must be PCR-amplified, unlike expression arrays, which can use total cellular poly(A) RNA. The current state of the art has been well reviewed by Hacia (1999).

A deletion screen reveals the majority of mutations in affected males

Although dystrophin mutations are extremely heterogeneous, 60–65% of all mutations are deletions of one or more exons (see Figure 16.3), and these preferentially affect certain exons. In affected males, two multiplex PCR reactions (that shown in Figure 17.11 and one testing exons in the 5′ part of the gene) will reveal 98% of all deletions. Most deletions remove more than one exon; deletions that appear to affect noncontiguous exons and deletions of just a single exon may need confirming (because of the risk of spurious results due to PCR failure).

Detecting deletions in females, for carrier testing, is much more difficult because the normal X chromosome masks any deletion present on the other X. Sequences deleted on one X will show a 50% reduction in dosage compared to nondeleted sequences, and this can be detected by quantitative PCR or quantitative Southern blotting. Very careful work is needed to get a result sufficiently unambiguous that a woman could prudently base reproductive decisions on it. A principal requirement for quantitative PCR is to restrict the number of cycles, so that the reaction does not reach a plateau; methods that use competitive reactions are also more likely to be reliably quantitative. Fluorescence sequencing machines, with their inherently quantitative mode of operation and their ability to measure small quantities of product, are probably the best tool for detecting carriers of deletions and duplications (Yau et al., 1996).

If a deletion is segregating in a family, typing females for microsatellites mapping within the deletion may reveal apparent nonmaternity, where a mother has transmitted no marker allele to her daughter because of the deletion (see Figure 17.12). In such families nonmaternity proves a woman is a carrier, while heterozygosity (in the daughter or sister of a carrier) proves a woman is not a carrier. Several markers suitable for this purpose have been identified in the introns at deletion hotspots. A final option for detecting deletion carriers is fluorescence in situ hybridization to metaphase chromosomes (Section 10.1.4), using a probe that does not hybridize to the deleted chromosome. Both the FISH and microsatellite methods work best in families where there is an affected male in whom the deletion can first be defined.

Mutation scanning in nondeletion families

If the multiplex screen fails to show a deletion, then defining the mutation is difficult. With 79 exons, screening the DNA exon by exon using SSCP or other standard methods is excessively laborious. Missense mutations very seldom cause DMD, so the best method is probably to use the protein truncation test, as illustrated in Figure 17.9. Because the dystrophin gene consists of so many scattered small exons (averaging only 180 bp), PTT testing requires RNA. Dystrophin is primarily expressed in muscle, but in skilled hands low-frequency ectopic (‘illegitimate’) transcripts can be amplified from lymphocytes.

Gene tracking in nondeletion families

In the setting of a routine diagnostic service, if the multiplex deletion screen fails to show a deletion, or if there is no sample available from an affected male, the best answer may be gene tracking, as described below. DMD presents special problems for gene tracking because of the extremely high recombination frequency across the gene. Even intragenic markers show an average 5% recombination with the disease. Therefore it is prudent to use flanking markers for gene tracking, as in Figure 17.13.

New mutations and mosaicism in DMD

To complete the list of problems posed by DMD, there is a high frequency of new mutations. The mutation-selection equilibrium calculations in Box 3.5 show that for any lethal X-linked recessive condition (f = 0), 1/3 of cases are fresh mutations. Therefore the mother of an isolated DMD boy has only a 2/3 chance of being a carrier. This greatly complicates the risk calculations that are necessary for interpreting gene tracking results. The interested reader should consult the book by Bridge (Further reading) for example calculations. Moreover, as shown in Figure 3.7, the first mutation carrier in a DMD pedigree is very often a mosaic (male or female). This raises yet more problems, both for risk estimation and for interpretation of the results of direct testing.

Bayesian calculations

Bayes' theorem provides a general method for combining probabilities into a final overall probability. The theory and procedure are shown in Box 17.2, and a sample calculation is set out in Figure 17.15. A very detailed set of calculations covering almost every conceivable situation in DNA diagnostics can be found in the book by Bridge (Further reading), which the interested reader should consult.

For simple pedigrees, Bayesian calculations give a quick answer, but for more complex pedigrees the calculations can get very elaborate. Few people feel fully confident of their ability to work through a complex pedigree correctly, although the attempt is a valuable mental exercise for teasing out the factors contributing to the final risk. An alternative is to use a linkage analysis program.

Using linkage programs for calculating genetic risks

At first sight it may seem surprising that a program designed to calculate lod scores can also calculate genetic risks - but in fact the two are closely related (Figure 17.16). Linkage analysis programs are general-purpose engines for calculating the likelihood of a pedigree, given certain data and assumptions. For calculating the likelihood of linkage we calculate the ratio:

For estimating the risk that a proband carries a disease gene, we calculate the ratio:

As in Box 17.2, the vertical line | means ‘given’.

What would screening achieve?

The most important single function of any screening program is to produce some useful outcome. It is quite unacceptable to tell people out of the blue that they are at risk of something unpleasant, unless the knowledge enables them to do something about the risk. Proposals to screen for genes conferring susceptibility to breast cancer or heart attacks must be assessed stringently against this criterion. Predictive testing for HD might appear to break this rule - but it is offered only to people who are at high risk of HD and who are suffering such agonies of uncertainty that they request a predictive test, and persist despite counseling in which all the disadvantages are pointed out.

Ideally the useful outcome is treatment, as in neonatal screening for phenylketonuria. Increased medical surveillance is a useful outcome only if it greatly improves the prognosis. A special case is screening for carrier status, where the outcome is the possibility of avoiding the birth of an affected child. People unwilling to accept prenatal diagnosis and termination of affected pregnancies would not see this as a useful outcome, and in general should not be screened.

An ethical framework for screening

Ethical issues in genetic population screening have been discussed by a committee of distinguished American geneticists, clinicians, lawyers and theologians, and the reader is referred to their report for a very detailed survey (Andrews et al., 1994). It is in the nature of ethical problems that they have no solutions, but certain principles emerge.

• Any program must be voluntary, with subjects taking the positive decision to opt in.

• Programs must respect the autonomy and privacy of the subject.

• People who score positive on the test must not be pressured into any particular course of action. For example, in countries with insurance-based health care systems, it would be unacceptable for insurance companies to put pressure on carrier couples to accept prenatal diagnosis, or financial pressure or inducements to terminate affected pregnancies.

• Information should be confidential. This may seem obvious, but it can be a difficult issue - we like to think that drivers of heavy trucks or jumbo jets have been tested for all possible risks. Societies with insurance-based health care systems have particular problems about the confidentiality of genetic data, since insurance companies will argue that they are penalizing low-risk people by not loading the premiums of high-risk people.

Specificity of a test

Unexpectedly, perhaps, false positive test results can pose a more serious problem than false negatives. If the specificity is low, then a positive test result does not mean much. Even if the false positives can be filtered out subsequently by a diagnostic test, many people will have been worried unnecessarily. Table 17.8 shows that if a test does have a significant false positive rate, then the specificity is hopelessly low except when testing for very common conditions. DNA tests are potentially valuable for population screening because, compared with biochemical tests, they should generate very few false positives. The most likely causes of false positives in DNA testing are laboratory or clerical errors.

Sensitivity of a test

A test must pick up a reasonable proportion of its intended target (i.e. the sensitivity must be high). While the specificity of DNA tests looks encouraging, the sensitivity usually depends on the degree of allelic heterogeneity. Unless a disease is unusually homogeneous (Table 17.1), it is not practicable to test for every conceivable mutation, especially in a large-throughput population screening program. Normally only a subset of mutations will be tested for. Figure 17.18 shows how the choice of mutations can affect the outcome of a CF carrier screening program.

It is clear that simply testing for the commonest mutation, F508del, would not produce an acceptable program. Almost as many affected children would be born to couples negative on the screening program as would be detected by the screening. Whether or not such a program were financially cost-effective, it would surely be socially unacceptable. What constitutes an acceptable program is harder to define. One suggestion focuses on ‘+/-’ couples (i.e. couples with one known carrier and the partner negative on all the tests). The partner still might be a carrier of a rare mutation. Professor LP Ten Kate suggested that an acceptable screening program is one in which the risk for such +/- couples is no higher than the general population risk before screening. That would require a sensitivity of about 95%.

Neonatal screening: screening for phenylketonuria

All babies in the UK are tested a few days after birth for PKU. A blood spot from a heel-prick is collected on a card (the Guthrie card) during a home visit and sent to a central laboratory. The phenylalanine level in the blood is measured by chromatography or a bacterial growth test. This is the screening test. Babies whose level is above a threshold are called in for a definitive diagnostic test. Only a small proportion eventually turn out to have PKU. The lack of informed consent on the part of the infant is justified by the benefit it receives from dietary treatment (Smith, 1993).

Prenatal screening: screening for β-thalassemia

Carriers of β-thalassemia can be detected by conventional hematological testing, either before marriage or in the antenatal clinic. Carrier-carrier couples can be offered prenatal diagnosis by DNA analysis. Pre-implantation diagnosis or fetal stem-cell transplants may become alternatives to termination of affected pregnancies. Two ethnic groups in the UK have a high incidence of β-thalassemia: Cypriots and Pakistanis. Screening was quickly accepted by Cypriots in the UK but uptake has been slower among Pakistanis. The comparison illustrates the complex social questions surrounding genetic screening and the relevance of cultural background (Gill and Modell, 1998). Importantly, long-term studies of the Cypriot community show how the success of screening can be measured, not by counts of affected fetuses aborted, but by counts of couples having normal families. Before screening was available, many carrier couples opted to have no children; now they are using screening and having normal families (Modell et al., 1984).

Population screening for carriers: proposals for cystic fibrosis

It is now technically feasible and financially worthwhile to screen northern European populations to detect CF carriers. Surveys in the UK suggest that most carrier-carrier couples would opt for prenatal diagnosis and would value the opportunity to ensure that they did not have affected children. This view might change if treatment becomes more effective, for example using gene therapy.

If a screening program is to be introduced, two sets of questions must be considered. How many mutations should the laboratory test for, and who should be offered the test? The problems raised by allelic heterogeneity have been discussed above (see Figure 17.18). On the question of who to screen, Table 17.9 shows some possibilities considered in the UK. Naturally the way health care delivery is organized in each country will determine the range of possibilities. Preliminary results from controlled pilot studies suggest that none of the methods has had the negative effects (increased anxiety) sometimes predicted.

DNA fingerprinting using minisatellite probes

These probes contain the common core sequence of a hypervariable dispersed repetitive sequence GGGCAGGAXG, first discovered by Jeffreys et al. (1985) in the myoglobin gene (see Section 7.4.2). When hybridized to Southern blots they give an individual-specific fingerprint of bands (Figure 17.19). Their chief disadvantage is that it is not possible to tell which pairs of bands in a fingerprint represent alleles. Thus, when matching DNA fingerprints, one matches each band individually by position and intensity. Other hypervariable repeated sequences have been used in the same way, for example those detected by the synthetic oligonucleotide (CAC)₅ (Krawczak and Schmidtke, 1998).

DNA profiling using single-locus minisatellite markers

Minisatellite probes recognize single-locus variable tandem repeats on Southern blots. Each probe should reveal two bands in any person's DNA, representing the two alleles. Profiling is based on four to ten different polymorphisms. These probes allow exact calculations of probabilities (of paternity, of the suspect not being the rapist, etc.), if the gene frequency of each allele in the population is known. For matching alleles between different gel tracks, the continuously variable distance along the gel has to be divided into a number of ‘bins’. Bands falling within the same bin are deemed to match. It is imperative that the criteria used for judging matches in each profiling test should be the same binning criteria that were used to calculate the population frequencies of each allele. The binning criteria can be arbitrary within certain limits, but they must be consistent. Minor variations within repeated units of some minisatellites potentially allow an almost infinite variety of alleles to be discriminated, so that the genotype at a single locus might suffice to identify an individual (Jeffreys et al., 1991).

DNA profiling using microsatellite markers

Microsatellite polymorphisms (Section 7.4.3) are based on short tandem repeats, usually di-, tri- or tetranucleotides. They have the advantages over minisatellites that they can be typed by PCR and that discrete alleles can be defined unambiguously by the precise repeat number. This avoids the binning problem and makes it easier to relate the results to population gene frequencies.

The use of Y-chromosome and mitochondrial polymorphisms

For tracing relationships to dead persons, Y-chromosome and mitochondrial DNA polymorphisms are especially useful because of their sex-specific pattern of transmission. An interesting example was the identification of the remains of the Russian Tsar and his family, killed by the Bolsheviks in 1917, by comparing DNA profiles of excavated remains with living distant relatives (Gill et al., 1994).