There are thousands of different ways to alter the sequence of a gene, each producing a mutant allele, although only some of these mutant alleles will appear in a real population. The known mutant alleles of a gene and its wild-type allele are referred to as multiple alleles or an allelic series.

One of the tests routinely performed on a new mutant allele is to see if it is dominant or recessive. Basic information about dominance and recessiveness is useful in working with the new mutation and can be a source of insight into the way the gene functions, as we will see in the examples. Dominance is a manifestation of how the alleles of a single gene interact in a heterozygote. In any experiment the interacting alleles may be wild type and mutant alleles (+/m) or two different mutant alleles (m₁/m₂). Several types of dominance have been discovered, each representing a different type of interaction between alleles.

The simplest type of dominance is full, or complete, dominance, which we examined in Chapter 2. A fully dominant allele will be expressed in the phenotype when only one copy is present, as in a heterozygote, whereas the alternative allele will be fully recessive. In full dominance, the homozygous dominant cannot be distinguished from the heterozygote; that is, at the phenotypic level, A/A = A/a. As mentioned earlier, phenylketonuria (PKU) and many other single-gene human diseases are fully recessive, whereas their wild-type alleles are dominant. Other single-gene diseases such as achondroplasia are fully dominant, whereas, in those cases, the wild-type allele is recessive. How can these dominance relations be interpreted at the cellular level?

217

The disease PKU is a good general model for recessive mutations. Recall that PKU is caused by a defective allele of the gene encoding the enzyme phenylalanine hydroxylase (PAH). In the absence of normal PAH, the phenylalanine entering the body in food is not broken down and hence accumulates. Under such conditions, phenylalanine is converted into phenylpyruvic acid, which is transported to the brain through the bloodstream and there impedes normal development, leading to mental retardation. The reason that the defective allele is recessive is that one “dose” of the wild-type allele P produces enough PAH to break down the phenylalanine entering the body. Thus, the PAH wild-type allele is said to be haplosufficient. Haplo means a haploid dose (one) and sufficient refers to the ability of that single dose to produce the wild-type phenotype. Hence, both P/P (two doses) and P/p (one dose) have enough PAH activity to result in the normal cellular chemistry. People with p/p have zero doses of PAH activity. Figure 6-1 illustrates this general notion.

How can we explain fully dominant mutations? There are several molecular mechanisms for dominance. A regularly encountered mechanism is that the wild-type allele of a gene is haploinsufficient. In haploinsufficiency, one wild-type dose is not enough to achieve normal levels of function. Assume that 16 units of a gene’s product are needed for normal chemistry and that each wild-type allele can make 10 units. Two wild-type alleles will produce 20 units of product, well over the minimum. But consider what happens if one of the mutations is a null mutation, which produces a nonfunctional protein. A null mutation in combination with a single wild-type allele would produce 10 + 0 = 10 units, well below the minimum. Hence, the heterozygote (wild type/null) is mutant, and the mutation is, by definition, dominant. In mice, the gene Tbx1 is haploinsufficient. This gene encodes a transcription-regulating protein (a transcription factor) that acts on genes responsible for the development of the pharynx. A knockout of one wild-type allele results in an inadequate concentration of the regulatory protein, which results in defects in the development of the pharyngeal arteries. The same haploinsufficiency is thought to be responsible for DiGeorge syndrome in humans, a condition with cardiovascular and craniofacial abnormalities.

218

Another important type of dominant mutation is called a dominant negative. Polypeptides with this type of mutation act as “spoilers” or “rogues.” In some cases, the gene product is a unit of a homodimeric protein, a protein composed of two units of the same type. In the heterozygote (+/M), the mutant polypeptide binds to the wild-type polypeptide and acts as a spoiler by distorting it or otherwise interfering with its function. The same type of spoiling can also hinder the functioning of a heterodimer composed of polypeptides from different genes. In other cases, the gene product is a monomer, and, in these situations, the mutant binds the substrate, and acts as a spoiler by hindering the ability of the wild-type protein to bind to the substrate.

An example of mutations that can act as dominant negatives is found in the gene for collagen protein. Some mutations in this gene give rise to the human phenotype osteogenesis imperfecta (brittle-bone disease). Collagen is a connective-tissue protein formed of three monomers intertwined (a trimer). In the mutant heterozygote, the abnormal protein wraps around one or two normal ones and distorts the trimer, leading to malfunction. In this way, the defective collagen acts as a spoiler. The difference between haploinsufficiency and the action of a dominant negative as causes of dominance of a mutation is illustrated in Figure 6-2.

Four-o’clocks are plants native to tropical America. Their name comes from the fact that their flowers open in the late afternoon. When a pure-breeding wild-type four-o’clock line having red petals is crossed with a pure line having white petals, the F₁ has pink petals. If an F₂ is produced by selfing the F₁, the result is

219

Figure 6-3 shows these phenotypes. From this 1:2:1 ratio in the F₂, we can deduce that the inheritance pattern is based on two alleles of a single gene. However, the heterozygotes (the F₁ and half the F₂) are intermediate in phenotype. By inventing allele symbols, we can list the genotypes of the four-o’clocks in this experiment as c⁺/c⁺ (red), c/c (white), and c⁺/c (pink). The occurrence of the intermediate phenotype suggests an incomplete dominance, the term used to describe the general case in which the phenotype of a heterozygote is intermediate between those of the two homozygotes, on some quantitative scale of measurement.

How do we explain incomplete dominance at the molecular level? In incomplete dominance, each wild-type allele generally produces a set dose of its protein product. The number of doses of a wild-type allele determines the concentration of a chemical made by the protein, such as pigment. In the four-o’clock plant, two doses produce the most copies of transcript, thus producing the greatest amount of protein and, hence, the greatest amount of pigment, enough to make the flower petals red. One dose produces less pigment, and so the petals are pink. A zero dose produces no pigment.

Another variation on the theme of dominance is codominance, the expression of both alleles of a heterozygote. A clear example is seen in the human ABO blood groups, where there is codominance of antigen alleles. The ABO blood groups are determined by three alleles of one gene. These three alleles interact in several ways to produce the four blood types of the ABO system. The three major alleles are i, I^A, and I^B, but a person can have only two of the three alleles or two copies of one of them. The combinations result in six different genotypes: the three homozygotes and three different types of heterozygotes, as follows.

In this allelic series, the alleles determine the presence and form of a complex sugar molecule present on the surface of red blood cells. This sugar molecule is an antigen, a cell-surface molecule that can be recognized by the immune system. The alleles I^A and I^B determine two different forms of this cell-surface molecule. However, the allele i results in no cell-surface molecule of this type (it is a null allele). In the genotypes I^A/i and I^B/i, the alleles I^A and I^B are fully dominant over i. However, in the genotype I^A/I^B, each of the alleles produces its own form of the cell-surface molecule, and so the A and B alleles are codominant.

The human disease sickle-cell anemia illustrates the somewhat arbitrary ways in which we classify dominance. The gene concerned encodes the molecule hemoglobin, which is responsible for transporting oxygen in blood vessels and is the major constituent of red blood cells. There are two main alleles Hb^A and Hb^S, and the three possible genotypes have different phenotypes, as follows:

220

Figure 6-4 shows an electron micrograph of blood cells including some sickled cells. In regard to the presence or absence of anemia, the Hb^A allele is dominant. In the heterozygote, a single Hb^A allele produces enough functioning hemoglobin to prevent anemia. In regard to blood-cell shape, however, there is incomplete dominance, as shown by the fact that, in the heterozygote, many of the cells have a slight sickle shape. Finally, in regard to hemoglobin itself, there is codominance. The alleles Hb^A and Hb^S encode two different forms of hemoglobin that differ by a single amino acid, and both forms are synthesized in the heterozygote. The A and S forms of hemoglobin can be separated by electrophoresis because it happens that they have different charges (Figure 6-5). We see that homozygous Hb^A/Hb^A people have one type of hemoglobin (A), and anemics have another (type S), which moves more slowly in the electric field. The heterozygotes have both types, A and S. In other words, there is codominance at the molecular level. The fascinating population genetics of the Hb^A and Hb^S alleles will be considered in Chapter 20.

Sickle-cell anemia illustrates the arbitrariness of the terms dominance, incomplete dominance, and codominance. The type of dominance inferred depends on the phenotypic level at which the assay is made—organismal, cellular, or molecular. Indeed, caution should be applied to many of the categories that scientists use to classify structures and processes; these categories are devised by humans for the convenience of analysis.

The leaves of clover plants show several variations on the dominance theme. Clover is the common name for plants of the genus Trifolium. There are many species. Some are native to North America, whereas others grow there as introduced weeds. Much genetic research has been done with white clover, which shows considerable variation among individual plants in the curious V, or chevron, pattern on the leaves. The different chevron forms (and the absence of chevrons) are determined by a series of seven alleles, as seen in Figure 6-6, which shows the many different types of interactions possible for even one allele. In most practical cases many alleles of a gene can be found together in a population, constituting an allelic series. The phenotypes shown by the allelic combinations are many and varied, reflecting the relative nature of dominance: an allele can show dominance with one partner but not with another. Hence, the complexity illustrated by the ABO blood type system is small compared with that in a case such as clover chevrons.

An allele that is capable of causing the death of an organism is called a lethal allele. In the characterization of a set of newly discovered mutant alleles, a recessive mutation is sometimes found to be lethal. This information is potentially useful in that it shows that the newly discovered gene (of yet unknown function) is essential to the organism’s operation. Essential genes are those without which an organism dies. (An example of an essential gene might be a ribosomal gene without which no protein would be made.) Indeed, with the use of modern DNA technology, a null mutant allele of a gene of interest can now be made intentionally and made homozygous to see if it is lethal and under which environmental conditions. Lethal alleles are also useful in determining the developmental stage at which the gene normally acts. In this case, geneticists look for whether death from a lethal mutant allele occurs early or late in the development of a zygote. The phenotype associated with death can also be informative in regard to gene function; for example, if a certain organ appears to be abnormal, the gene is likely to be expressed in that organ.

221

What is the diagnostic test for lethality? The test is well illustrated by one of the prototypic examples of a lethal allele, a coat-color allele in mice (see the Model Organism box below). Normal wild-type mice have coats with a rather dark overall pigmentation. A mutation called yellow (a lighter coat color) shows a curious inheritance pattern. If any yellow mouse is mated with a homozygous wild-type mouse, a 1:1 ratio of yellow to wild-type mice is always observed in the progeny. This result suggests that a yellow mouse is always heterozygous for the yellow allele and that the yellow allele is dominant over wild type. However, if any two yellow mice are crossed with each other, the result is always as follows:

Figure 6-7 shows a typical litter from a cross between yellow mice.

How can the 2:1 ratio be explained? The results make sense if the yellow allele is assumed to be lethal when homozygous. The yellow allele is known to be of a coat-color gene called A. Let’s call it A^Y. Hence, the results of crossing two yellow mice are

The expected monohybrid ratio of 1 :2 :1 would be found among the zygotes, but it is altered to a 2 :1 ratio in the progeny actually seen at birth because zygotes with a lethal A^Y/A^Y genotype do not survive to be counted. This hypothesis is supported by the removal of uteri from pregnant females of the yellow × yellow cross; one-fourth of the embryos are found to be dead.

The A^Y allele produces effects on two characters: coat color and survival. It is entirely possible, however, that both effects of the A^Y allele result from the same basic cause, which promotes yellowness of coat in a single dose and death in a double dose. In general, the term pleiotropic is used for any allele that affects several properties of an organism.

222

The tailless Manx phenotype in cats (Figure 6-8) also is produced by an allele that is lethal in the homozygous state. A single dose of the Manx allele, M^L, severely interferes with normal spinal development, resulting in the absence of a tail in the M^L/M heterozygote. But in the M^L/M^L homozygote, the double dose of the gene produces such an extreme abnormality in spinal development that the embryo does not survive.

The yellow and M^L alleles have their own phenotypes in a heterozygote, but most recessive lethals are silent in the heterozygote. In such a situation, recessive lethality is diagnosed by observing the death of 25 percent of the progeny at some stage of development.

Whether an allele is lethal or not often depends on the environment in which the organism develops. Whereas certain alleles are lethal in virtually any environment, others are viable in one environment but lethal in another. Human hereditary diseases provide some examples. Cystic fibrosis and sickle-cell anemia are diseases that would be lethal without treatment. Furthermore, many of the alleles favored and selected by animal and plant breeders would almost certainly be eliminated in nature as a result of competition with the members of the natural population. The dwarf mutant varieties of grain, which are very high yielding, provide good examples; only careful nurturing by farmers has maintained such alleles for our benefit.

223

Geneticists commonly encounter situations in which expected phenotypic ratios are consistently skewed in one direction because a mutant allele reduces viability. For example, in the cross A/a × a/a, we predict a progeny ratio of 50 percent A/a and 50 percent a/a, but we might consistently observe a ratio such as 55 percent : 45 percent or 60 percent : 40 percent. In such a case, the recessive allele is said to be sublethal because the lethality is expressed in only some but not all of the homozygous individuals. Thus, lethality may range from 0 to 100 percent, depending on the gene itself, the rest of the genome, and the environment.

We have seen that lethal alleles are useful in diagnosing the time at which a gene acts and the nature of the phenotypic defect that kills. However, maintaining stocks bearing lethal alleles for laboratory use is a challenge. In diploids, recessive lethal alleles can be maintained as heterozygotes. In haploids, heat-sensitive lethal alleles are useful. They are members of a general class of temperature-sensitive (ts) mutations. Their phenotype is wild type at the permissive temperature (often room temperature) but mutant at some higher restrictive temperature. Temperature-sensitive alleles are thought to be caused by mutations that make the protein prone to twist or bend its shape to an inactive conformation at the restrictive temperature. Research stocks can be maintained easily under permissive conditions, and the mutant phenotype can be assayed in a subset of individuals by a switch to the restrictive conditions. Temperature-sensitive dominant lethal mutations also are useful. This type of mutation is expressed even when present in a single dose but only when the experimenter switches the organism to the restrictive temperature.

Null alleles for genes identified through genomic sequencing can be made by using a variety of “reverse genetic” procedures that specifically knock out the function of that gene. These will be described in Chapter 14.

We now turn to the approaches that can be used to detect the interaction between two or more loci.