Changes in chromosome structure, called rearrangements, encompass several major classes of events. A chromosome segment can be lost, constituting a deletion, or doubled, to form a duplication. The orientation of a segment within the chromosome can be reversed, constituting an inversion. Or a segment can be moved to a different chromosome, constituting a translocation. DNA breakage is a major cause of each of these events. Both DNA strands must break at two different locations, followed by a rejoining of the broken ends to produce a new chromosomal arrangement (Figure 17-19, left side). Chromosomal rearrangements by breakage can be induced artificially by using ionizing radiation. This kind of radiation, particularly X rays and gamma rays, is highly energetic and causes numerous double-stranded breaks in DNA.

635

To understand how chromosomal rearrangements are produced by breakage, several points should be kept in mind:

Another important cause of rearrangements is crossing over between repetitive (duplicated) DNA segments. This type of crossing over is termed nonallelic homologous recombination (NAHR). In organisms with repeated DNA sequences within one chromosome or on different chromosomes, there is ambiguity about which of the repeats will pair with each other at meiosis. If sequences pair up that are not in the same relative positions on the homologs, crossing over can produce aberrant chromosomes. Deletions, duplications, inversions, and translocations can all be produced by such crossing over (see Figure 17-19, right side).

There are two general types of rearrangements: unbalanced and balanced. Unbalanced rearrangements change the gene dosage of a chromosome segment. As with aneuploidy for whole chromosomes, the loss of one copy of a segment or the addition of an extra copy can disrupt normal gene balance. The two simple classes of unbalanced rearrangements are deletions and duplications. A deletion is the loss of a segment within one chromosome arm and the juxtaposition of the two segments on either side of the deleted segment, as in this example, which shows loss of segment C–D:

A duplication is the repetition of a segment of a chromosome arm. In the simplest type of duplication, the two segments are adjacent to each other (a tandem duplication), as in this duplication of segment C:

However, the duplicate segment can end up at a different position on the same chromosome or even on a different chromosome.

Balanced rearrangements change the chromosomal gene order but do not remove or duplicate any DNA. The two simple classes of balanced rearrangements are inversions and reciprocal translocations. An inversion is a rearrangement in which an internal segment of a chromosome has been broken twice, flipped 180 degrees, and rejoined.

637

A reciprocal translocation is a rearrangement in which two nonhomologous chromosomes are each broken once, creating acentric fragments, which then trade places:

Sometimes the DNA breaks that precede the formation of a rearrangement occur within genes. When they do, they disrupt gene function because part of the gene moves to a new location and no complete transcript can be made. In addition, the DNA sequences on either side of the rejoined ends of a rearranged chromosome are sequences that are not normally juxtaposed. Sometimes the junction occurs in such a way that fusion produces a nonfunctional hybrid gene composed of parts of two other genes.

The following sections consider the properties of these balanced and unbalanced rearrangements.

A deletion is simply the loss of a part of one chromosome arm. The process of deletion requires two chromosome breaks to cut out the intervening segment. The deleted fragment has no centromere; consequently, it cannot be pulled to a spindle pole in cell division and is lost. The effects of deletions depend on their size. A small deletion within a gene, called an intragenic deletion, inactivates the gene and has the same effect as that of other null mutations of that gene. If the homozygous null phenotype is viable (as, for example, in human albinism), the homozygous deletion also will be viable. Intragenic deletions can be distinguished from mutations caused by single nucleotide changes because genes with such deletions never revert to wild type.

For most of this section, we will be dealing with multigenic deletions, in which several to many genes are missing. The consequences of these deletions are more severe than those of intragenic deletions. If such a deletion is made homotyzygous by inbreeding (that is, if both homologs have the same deletion), the combination is always lethal. This fact suggests that all regions of the chromosomes are essential for normal viability and that complete elimination of any segment from the genome is deleterious. Even an individual organism heterozygous for a multigenic deletion—that is, having one normal homolog and one that carries the deletion—may not survive. Principally, this lethal outcome is due to disruption of normal gene balance. Alternatively, the deletion may “uncover” deleterious recessive alleles, allowing the single copies to be expressed.

Small deletions are sometimes viable in combination with a normal homolog. Such deletions may be identified by examining meiotic chromosomes under the microscope. The failure of the corresponding segment on the normal homolog to pair creates a visible deletion loop (Figure 17-20a). In Drosophila, deletion loops are also visible in the polytene chromosomes. These chromosomes are found in the cells of salivary glands and other specific tissues of certain insects. In these cells, the homologs pair and replicate many times, and so each chromosome is represented by a thick bundle of replicates. These polytene chromosomes are easily visible, and each has a set of dark-staining bands of fixed position and number. These bands act as useful chromosomal landmarks. An example of a polytene chromosome in which one original homolog carried a deletion is shown in Figure 17-20b. A deletion can be assigned to a specific chromosome location by examining polytene chromosomes microscopically and determining the position of the deletion loop.

638

Another clue to the presence of a deletion is that the deletion of a segment on one homolog sometimes unmasks recessive alleles present on the other homolog, leading to their unexpected expression. Consider, for example, the deletion shown in the following diagram:

If there is no deletion, none of the seven recessive alleles is expected to be expressed; however, if b and c are expressed, then a deletion spanning the b+ and c+ genes has probably occurred on the other homolog. Because recessive alleles seem to be showing dominance in such cases, the effect is called pseudodominance.

In the reverse case—if we already know the location of the deletion—we can apply the pseudodominance effect in the opposite direction to map the positions of mutant alleles. This procedure, called deletion mapping, pairs mutations against a set of defined overlapping deletions. An example from Drosophila is shown in Figure 17-21. In this diagram, the recombination map is shown at the top, marked with distances in map units from the left end. The horizontal red bars below the chromosome show the extent of the deletions listed at the left. Each deletion is paired with each mutation under test, and the phenotype is observed to see if the mutation is pseudodominant. The mutation pn (prune), for example, shows pseudodominance only with deletion 264-38, and this result determines its location in the 2D-4 to 3A-2 region. However, fa (facet) shows pseudodominance with all but two deletions (258-11 and 258-14); so its position can be pinpointed to band 3C-7, which is the region that all but two deletions have in common.

Clinicians regularly find deletions in human chromosomes. The deletions are usually small, but they do have adverse effects, even though heterozygous. Deletions of specific human chromosome regions cause unique syndromes of phenotypic abnormalities. One example is cri du chat syndrome, caused by a heterozygous deletion of the tip of the short arm of chromosome 5 (Figure 17-22). The specific bands deleted in cri du chat syndrome are 5p15.2 and 5p15.3, the two most distal bands identifiable on 5p. (The short and long arms of human chromosomes are traditionally called p and q, respectively.) The most characteristic phenotype in the syndrome is the one that gives it its name, the distinctive catlike mewing cries made by affected infants. Other manifestations of the syndrome are microencephaly (abnormally small head) and a moonlike face. Like syndromes caused by other deletions, cri du chat syndrome includes mental retardation. Fatality rates are low, and many persons with this deletion reach adulthood.

Another instructive example is Williams syndrome. This syndrome is autosomal dominant and is characterized by unusual development of the nervous system and certain external features. Williams syndrome is found at a frequency of about 1 in 10,000 people. Patients often have pronounced musical or singing ability. The syndrome is almost always caused by a 1.5-Mb deletion on one homolog of chromosome 7. Sequence analysis showed that this segment contains 17 genes of known and unknown function. The abnormal phenotype is thus caused by haploinsufficiency of one or more of these 17 genes. Sequence analysis also reveals the origin of this deletion because the normal sequence is bounded by repeated copies of a gene called PMS, which happens to encode a DNA-repair protein. As we have seen, repeated sequences can act as substrates for unequal crossing over. A crossover between flanking copies of PMS on opposite ends of the 17-gene segment leads to a duplication (not found) and a Williams syndrome deletion, as shown in Figure 17-23.

639

Most human deletions, such as those that we have just considered, arise spontaneously in the gonads of a normal parent of an affected person; thus, no signs of the deletions are usually found in the chromosomes of the parents. Less commonly, deletion-bearing individuals appear among the offspring of an individual having an undetected balanced rearrangement of chromosomes. For example, cri du chat syndrome can result from a parent heterozygous for a reciprocal translocation, because (as we will see) segregation produces deletions. Deletions may also result from recombination within a heterozygote having a pericentric inversion (an inversion spanning the centromere) on one chromosome. Both mechanisms will be detailed later in the chapter.

Animals and plants show differences in the survival of gametes or offspring that bear deletions. A male animal with a deletion in one chromosome produces sperm carrying one or the other of the two chromosomes in approximately equal numbers. These sperm seem to function to some extent regardless of their genetic content. In diploid plants, on the other hand, the pollen produced by a deletion heterozygote is of two types: functional pollen carrying the normal chromosome and nonfunctional (aborted) pollen carrying the deficient homolog. Thus, pollen cells seem to be sensitive to changes in the amount of chromosomal material, and this sensitivity might act to weed out deletions. This effect is analogous to the sensitivity of pollen to whole-chromosome aneuploidy, described earlier in this chapter. Unlike animal sperm cells, whose metabolic activity relies on enzymes that have already been deposited in them during their formation, pollen cells must germinate and then produce a long pollen tube that grows to fertilize the ovule. This growth requires that the pollen cell manufacture large amounts of protein, thus making it sensitive to genetic abnormalities in its own nucleus. Plant ovules, in contrast, are quite tolerant of deletions, presumably because they receive their nourishment from the surrounding maternal tissues.

640

The processes of chromosome mutation sometimes produce an extra copy of some chromosome region. The duplicate regions can be located adjacent to each other—called a tandem duplication—or the extra copy can be located elsewhere in the genome—called an insertional duplication. A diploid cell containing a duplication will have three copies of the chromosome region in question: two in one chromosome set and one in the other—an example of a duplication heterozygote. In meiotic prophase, tandem-duplication heterozygotes show a loop consisting of the unpaired extra region.

Synthetic duplications of known coverage can be used for gene mapping. In haploids, for example, a chromosomally normal strain carrying a new recessive mutation m may be crossed with strains bearing a number of duplication-generating rearrangements (for example, translocations and pericentric inversions). In any one cross, if some duplication progeny have the recessive phenotype, the duplication does not span gene m, because, if it did, its extra segment would mask the recessive m allele.

641

Analyses of genome DNA sequences have revealed a high level of duplications in humans and in most of the model organisms. Simple sequence repeats, which are extensive throughout the genome and useful as molecular markers in mapping, were discussed in earlier chapters. However, another class of duplications is based on duplicated units that are much bigger than the simple sequence repeats. Duplications in this class are termed segmental duplications. The duplicated units in segmental duplications range from 10 to 50 kilobases in length and encompass whole genes and the regions in between. The extent of segmental duplications is shown in Figure 17-24, in which most of the duplications are dispersed, but there are some tandem cases. Another property shown in Figure 17-24 is that the dispersion of the duplicated units is mostly within the same chromosome, not between chromosomes. The origin of segmental duplications is still not known.

Segmental duplications are thought to have an important role as substrates for nonallelic homologous recombination, as shown in Figure 17-19. Crossing over between segmental duplications can lead to various chromosomal rearrangements. These rearrangements seem to have been important in evolution, inasmuch as some major inversions that are key differences between human and ape sequences have almost certainly come from NAHR (non-allelic homologous recombination). It also seems likely that NAHR has been responsible for rearrangements that cause some human diseases. The loci of such diseases are at segmental-duplication hotspots; examples of such loci are shown in Figure 17-24.

We have seen that, in some organisms such as polyploids, the present-day genome evolved as a result of an ancestral whole-genome duplicating. When whole-genome duplication has taken place, every gene is doubled. These doubled genes are a source of some of the segmental duplications found in genomes. A well-studied case is baker’s yeast, Saccharomyces cerevisiae. The evolution of this genome has been analyzed by comparing the whole-genome sequence of S. cerevisiae with that of another yeast, Kluyveromyces, whose genome is similar to that of the ancestral genome of yeast. Apparently, in the course of the evolution of Saccharomyces, the Kluyveromyces-like ancestral genome doubled, and so there were two sets, each containing the whole genome. After doubling occurred, many gene copies were lost from one set or the other, and the remaining sets were rearranged, resulting in the present Saccharomyces genome. This process is reconstructed in Figure 17-25.

642

We have seen that, to create an inversion, a segment of a chromosome is cut out, flipped, and reinserted. Inversions are of two basic types. If the centromere is outside the inversion, the inversion is said to be paracentric. Inversions spanning the centromere are pericentric.

Because inversions are balanced rearrangements, they do not change the overall amount of genetic material, and so they do not result in gene imbalance. Individuals with inversions are generally normal, if there are no breaks within genes. A break that disrupts a gene produces a mutation that may be detectable as an abnormal phenotype. If the gene has an essential function, then the break point acts as a lethal mutation linked to the inversion. In such a case, the inversion cannot be bred to homozygosity. However, many inversions can be made homozygous, and, furthermore, inversions can be detected in haploid organisms. In these cases, the break points of the inversion are clearly not in essential regions. Some of the possible consequences of inversion at the DNA level are shown in Figure 17-26.

643

Most analyses of inversions are carried out on diploid cells that contain one normal chromosome set plus one set carrying the inversion. This type of cell is called an inversion heterozygote, but note that this designation does not imply that any gene locus is heterozygous; rather, it means that one normal and one abnormal chromosome set are present. The location of the inverted segment can often be detected microscopically. In meiosis, one chromosome twists once at the ends of the inversion to pair with its untwisted homolog; in this way, the paired homologs form a visible inversion loop (Figure 17-27).

In a paracentric inversion, crossing over within the inversion loop at meiosis connects homologous centromeres in a dicentric bridge while also producing an acentric fragment (Figure 17-28). Then, as the chromosomes separate in anaphase I, the centromeres remain linked by the bridge. The acentric fragment cannot align itself or move; consequently, it is lost. Tension eventually breaks the dicentric bridge, forming two chromosomes with terminal deletions. Either the gametes containing such chromosomes or the zygotes that they eventually form will probably be inviable. Hence, a crossover event, which normally generates the recombinant class of meiotic products, is instead lethal to those products. The overall result is a drastically lower frequency of viable recombinants. In fact, for genes within the inversion, the recombinant frequency is close to zero. (It is not exactly zero because rare double crossovers between only two chromatids are viable.) For genes flanking the inversion, the RF is reduced in proportion to the size of the inversion because, for a longer inversion, there is a greater probability of a crossover occurring within it and producing an inviable meiotic product.

644

In a heterozygous pericentric inversion, the net genetic effect is the same as that of a paracentric inversion—crossover products are not recovered—but the reasons are different. In a pericentric inversion, the centromeres are contained within the inverted region. Consequently, the chromosomes that have engaged in crossing over separate in the normal fashion, without the creation of a bridge (Figure 17-29). However, the crossover produces chromatids that contain a duplication and a deletion for different parts of the chromosome. In this case, if a gamete carrying a crossover chromosome is fertilized, the zygote dies because of gene imbalance. Again, the result is that only noncrossover chromatids are present in viable progeny. Hence, the RF value of genes within a pericentric inversion also is zero.

Inversions affect recombination in another way, too. Inversion heterozygotes often have mechanical pairing problems in the region of the inversion. The inversion loop causes a large distortion that can extend beyond the loop itself. This distortion reduces the opportunity for crossing over in the neighboring regions.

Let us consider an example of the effects of an inversion on recombinant frequency. A wild-type Drosophila specimen from a natural population is crossed with a homozygous recessive laboratory stock dp cn/dp cn. (The dp allele encodes dumpy wings and cn encodes cinnabar eyes. The two genes are known to be 45 map units apart on chromosome 2.) The F₁ generation is wild type. When an F₁ female is crossed with the recessive parent, the progeny are

645

In this cross, which is effectively a dihybrid testcross, 45 percent of the progeny are expected to be dumpy or cinnabar (they constitute the crossover classes), but only 12 of 508, about 2 percent, are obtained. Something is reducing crossing over in this region, and a likely explanation is an inversion spanning most of the dp–cn region. Because the expected RF was based on measurements made on laboratory strains, the wild-type fly from nature was the most likely source of the inverted chromosome. Hence, chromosome 2 in the F₁ can be represented as follows:

Pericentric inversions also can be detected microscopically through new arm ratios. Consider the following pericentric inversion:

Note that the length ratio of the long arm to the short arm has been changed from about 4:1 to about 1:1 by the inversion. Paracentric inversions do not alter the arm ratio, but they may be detected microscopically by observing changes in banding or other chromosomal landmarks, if available.

In some model experimental systems, notably Drosophila and the nematode Caenorhabditis elegans, inversions are used as balancers. A balancer chromosome contains multiple inversions; so, when it is combined with the corresponding wildtype chromosome, there can be no viable crossover products. In some analyses, it is important to keep stock with all the alleles on one chromosome together. The geneticist creates individuals having genomes that combine such a chromosome with a balancer. This combination eliminates crossovers, and so only parental combinations appear in the progeny. For convenience, balancer chromosomes are marked with a dominant morphological mutation. The marker allows the geneticist to track the segregation of the entire balancer or its normal homolog by noting the presence or absence of the marker.

There are several types of translocations, but here we consider only reciprocal translocations, the simplest type. Recall that, to form a reciprocal translocation, two chromosomes trade acentric fragments created by two simultaneous chromosome breaks. As with other rearrangements, meiosis in heterozygotes having two translocated chromosomes and their normal counterparts produces characteristic configurations. Figure 17-30 illustrates meiosis in an individual that is heterozygous for a reciprocal translocation. Note the cross-shaped pairing configuration. Because the law of independent assortment is still in force, there are two common patterns of segregation. Let us use N₁ and N₂ to represent the normal chromosomes and T₁ and T₂ the translocated chromosomes. The segregation of each of the structurally normal chromosomes with one of the translocated ones (T₁ + N₂ and T₂ + N₁) is called adjacent-1 segregation. Each of the two meiotic products is deficient for a different arm of the cross and has a duplicate of the other. These products are inviable. On the other hand, the two normal chromosomes may segregate together, as will the reciprocal parts of the translocated ones, to produce N₁ + N₂ and T₁ + T₂ products. This segregation pattern is called alternate segregation. These products are both balanced and viable.

646

Adjacent-1 and alternate segregations are equal in number, and so half the overall population of gametes will be nonfunctional, a condition known as semisterility or “half sterility.” Semisterility is an important diagnostic tool for identifying translocation heterozygotes. However, semisterility is defined differently for plants and animals. In plants, the 50 percent of meiotic products that are from the adjacent-1 segregation generally abort at the gametic stage (Figure 17-31). In animals, these products are viable as gametes but lethal to the zygotes that they produce on fertilization.

Remember that heterozygotes for inversions also may show some reduction in fertility but by an amount dependent on the size of the affected region. The precise 50 percent reduction in viable gametes or zygotes is usually a reliable diagnostic clue for a translocation.

Genetically, genes on translocated chromosomes act as though they are linked if their loci are close to the translocation break point. Figure 17-32 shows a translocation heterozygote that has been established by crossing an a/a; b/b individual with a translocation homozygote bearing the wild-type alleles. When the heterozygote is testcrossed, recombinants are created but do not survive because they carry unbalanced genomes (duplication-and-deletions). The only viable progeny are those bearing the parental genotypes; so linkage is seen between loci that were originally on different chromosomes. The apparent linkage of genes normally known to be on separate nonhomologous chromosomes—sometimes called pseudolinkage—is a genetic diagnostic clue to the presence of a translocation.

647

Let’s return to the family with the Down syndrome child, introduced at the beginning of the chapter. The birth can indeed be a coincidence—after all, coincidences do happen. However, the miscarriage gives a clue that something else might be going on. A large proportion of spontaneous abortions carry chromosomal abnormalities, so perhaps that is the case in this example. If so, the couple may have had two conceptions with chromosome mutations, which would be very unlikely unless there was a common cause. However, a small proportion of Down syndrome cases are known to result from a translocation in one of the parents. We have seen that translocations can produce progeny that have extra material from part of the genome, and so a translocation concerning chromosome 21 can produce progeny that have extra material from that chromosome. In Down syndrome, the translocation responsible is of a type called a Robertsonian translocation. It produces progeny carrying an almost complete extra copy of chromosome 21. The translocation and its segregation are illustrated in Figure 17-33. Note that, in addition to complements causing Down syndrome, other aberrant chromosome complements are produced, most of which abort. In our example, the man may have this translocation, which he may have inherited from his grandmother. To confirm this possibility, his chromosomes are checked. His unaffected child might have normal chromosomes or might have inherited his translocation.

648

Inversions and translocations have proved to be useful genetic tools; some examples of their uses follow.

Gene mapping Inversions and translocations are useful for the mapping and subsequent isolation of specific genes. The gene for human neurofibromatosis was isolated in this way. The critical information came from people who not only had the disease, but also carried chromosomal translocations. All the translocations had one break point in common, in a band close to the centromere of chromosome 17. Hence, this band appeared to be the locus of the neurofibromatosis gene, which had been disrupted by the translocation break point. Subsequent analysis showed that the chromosome 17 break points were not at identical positions; however, because they must have been within the gene, the range of their positions revealed the segment of the chromosome that constituted the neurofibromatosis gene. The isolation of DNA fragments from this region eventually led to the recovery of the gene itself.

Synthesizing specific duplications or deletions Translocations and inversions are routinely used to delete or duplicate specific chromosome segments. Recall, for example, that pericentric inversions as well as translocations generate products of meiosis that contain a duplication and a deletion (see Figures 17-29 and 17-30). If the duplicated or the deleted segment is very small, then the duplication-and-deletion meiotic products are tantamount to duplications or deletions, respectively. Duplications and deletions are useful for a variety of experimental applications, including the mapping of genes and the varying of gene dosage for the study of regulation, as seen in preceding sections.

Another approach to creating duplications uses unidirectional insertional translocations, in which a segment of one chromosome is removed and inserted into another. In an insertional-translocation heterozygote, a duplication results if the chromosome with the insertion segregates along with the normal copy.

Position-effect variegation As we saw in Chapter 12, gene action can be blocked by proximity to the densely staining chromosome regions called heterochromatin. Translocations and inversions can be used to study this effect. For example, the locus for white eye color in Drosophila is near the tip of the X chromosome. Consider a translocation in which the tip of an X chromosome carrying w⁺ is relocated next to the heterochromatic region of, say, chromosome 4 (Figure 17-34a, top section). Position-effect variegation is observed in flies that are heterozygotes for such a translocation. The normal X chromosome in such a heterozygote carries the recessive allele w. The eye phenotype is expected to be red because the wild-type allele is dominant over w. However, in such cases, the observed phenotype is a variegated mixture of red and white eye facets (Figure 17-34b). How can we explain the white areas? The w⁺ allele is not always expressed because the heterochromatin boundary is somewhat variable: in some cells, it engulfs and inactivates the w⁺ gene, thereby preventing its expression and thereby allowing the expression of w. If the positions of the w⁺ and w alleles are exchanged by a crossover, then position-effect variegation is not detected (see Figure 17-34a, bottom section).

649

Cancer is a disease of abnormal cell proliferation. As a result of some insult inflicted on it, a cell of the body divides out of control to form a population of cells called a cancer. A localized knot of proliferated cells is called a tumor, whereas cancers of mobile cells such as blood cells disperse throughout the body. Cancer is most often caused by a mutation in the coding or regulatory sequence of a gene whose normal function is to regulate cell division. Such genes are called proto-oncogenes. However, chromosomal rearrangements, especially translocations, also can interfere with the normal function of such proto-oncogenes.

There are two basic ways in which translocations can alter the function of proto-oncogenes. In the first mechanism, the translocation relocates a proto-oncogene next to a new regulatory element. A good example is provided by Burkitt lymphoma. The proto-oncogene in this cancer encodes the protein MYC, a transcription factor that activates genes required for cell proliferation. Normally, the myc gene is transcribed only when a cell needs to undergo proliferation, but, in cancerous cells, the proto-oncogene MYC is relocated next to the regulatory region of immunoglobulin (Ig) genes (Figure 17-35a). These immunoglobulin genes are constitutively transcribed; that is, they are on all the time. Consequently, the myc gene is transcribed at all times, and the cell-proliferation genes are continuously activated.

650

The other mechanism by which translocations can cause cancer is the formation of a hybrid gene. An example is provided by the disease chronic myelogenous leukemia (CML), a cancer of white blood cells. This cancer can result from the formation of a hybrid gene between the two proto-oncogenes BCR1 and ABL (Figure 17-35b). The abl proto-oncogene encodes a protein kinase in a signaling pathway. The protein kinase passes along a signal initiated by a growth factor that leads to cell proliferation. The Bcr1-Abl fusion protein has a permanent protein kinase activity. The altered protein continually propagates its growth signal onward, regardless of whether the initiating signal is present.

DNA microarrays (see Figure 14-19) have made it possible to detect and quantify duplications or deletions of a given DNA segment. The technique is called comparative genomic hybridization. The total DNA of the wild type and that of a mutant are labeled with two different fluorescent dyes that emit distinct wavelengths of light. These labeled DNAs are added to a cDNA microarray together, and both of them hybridize to the array. The array is then scanned by a detector tuned to one fluorescent wavelength and is then scanned again for the other wavelength. The ratio of values for each cDNA is calculated. Mutant-to-wild-type ratios substantially greater than 1 represent regions that have been amplified. A ratio of 2 points to a duplication, and a ratio of less than 1 points to a deletion. Some examples are shown in Figure 17-36.

651