Gregor Mendel (1822-1884) is heralded as the Father of Genetics, although he would not recognize that title, since the term “gene” was not coined until 1909. At the time of Mendel’s experiments on plant hybrids (Figure 6-1), a prevailing theory of trait transmission was blending inheritance, in which traits are mixed and altered in the offspring. His training in physics and mathematics was unusual for a biologist of his time. It gave him a quantitative outlook on natural laws that enabled him to detect relationships in his results that other biologists had overlooked.
Figure 6-1. Gregor Mendel established a controlled breeding plan to trace the transmission of simple traits using plants like the garden pea. (Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
Mendel carried out carefully controlled experiments in which he cross-pollinated true-breeding strains (we would now call them “homozygous”) and protected the experimental plants from accidental pollination by insects or the wind. Strains of the garden pea, Pisum sativum, were available with distinctly differing traits like yellow versus green seeds and white versus purple flowers. He kept careful records of results, from which he was able to identify predictable patterns in the frequency of traits among large numbers of offspring. The proportions he found required an assumption that challenged the contemporary idea of blending inheritance. Mendel’s assumption of “unit factor inheritance” was that the determinants of traits (today’s “genes”) were distinct factors that occur in pairs in each individual (today’s “diploid”). Furthermore, only one of the two copies was passed by each parent to an offspring. Building on this theoretical foundation, he was able to deduce from his data the three rules of transmission that are often called the Mendelian Laws. The mathematical logic behind his assumption of unit factor inheritance and the significance of the regular patterns of factor transmission were not fully understood by his colleagues, and his work went unnoticed until his publications were independently rediscovered in 1900 by Carl Correns, Hugo de Vries, and Erich von Tschermak. Experimental exploration into the rules of hereditary transmission dates from that rediscovery. Now extensive datasets can be created, even from purely historical data, to show the hereditary relationships among related individuals.
The Mendelian Laws of genetic transmission are, however, not laws in the scientific sense of that word. A natural law is a tested theory that has been shown to be universally true. But, in contrast to this highest level of authority, each of the Mendelian “Laws” has important exceptions that affect the way patterns of inheritance are expressed. In this chapter we will discuss the mechanisms behind the three classical Mendelian Rules of transmission and some of the important ways in which a gene’s contributions to a phenotype can alter the appearance of the underlying patterns of inheritance.
Part 1: Background and Systems Integration
Mendelian Rules of Transmission
Mendel’s key experiments on plant hybridization were published in 1866. The quality of biological microscopy and nuclear staining techniques did not improve to the point where chromosomes could be seen and studied until the 1880s. Thus, the relationships that Mendel described were deduced from patterns he saw in his data. But at least in the case of two of his principal rules, segregation and independent assortment, he was actually describing the behavior of chromosomes during meiosis.
From our earlier discussion of meiosis, you will remember that each chromosome replicates during the S phase of interphase. Chromosomes then coil, and homologous pairs synapse during prophase I of meiosis (Figure 6-2). Each synapsed pair, or bivalent, becomes attached to kinetochore microtubules of the spindle. At anaphase I the homologous chromosomes separate, reducing chromosome number from the diploid to the haploid in each of the two resulting cells. This separation of homologous chromosomes in the first meiotic division is the mechanism behind Mendel’s rule of segregation.
Figure 6-2. An abbreviated summary of key events in meiosis. In this example, the homologous chromosomes carry different alleles for the Y gene. At metaphase I, they line up in the center of the spindle, and at anaphase I one homologue and its allele (y) moves to the left and the other homologue and its allele (Y) separates from the first and moves to the right. This separation is the basis of the Mendelian Rule of segregation. (Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
If we follow two pairs of homologous chromosomes that are heterozygous for genes Y and R (Figure 6-3), we can see that the way in which one pair lines up at metaphase I is independent of the way the other lines up. Half the time we expect the two dominant alleles to line up on opposite sides of the metaphase plate, and half the time they will be on the same side. When the homologous chromosomes separate from each other, the segregation of Y and y is at random with respect to the segregation of R and r. For that reason, gametes carrying the four possible allele combinations, YR, Yr, yR, and yr, are in equal frequency. This is the mechanism that explains Mendel’s rule of random, or independent, assortment.
Figure 6-3. Events of meiosis in which we follow the behavior of two different pairs of chromosomes. (Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
The third Mendelian rule, dominance and recessiveness, is unrelated to chromosome behavior. As mentioned earlier, dominance is simply a function of the way in which biochemical regulation of phenotypes can buffer against the deleterious influence of a recessive mutation. If one normal copy of a gene makes enough products that the phenotype is expressed normally, the mutant allele will be masked, i.e., it will be recessive to the dominant allele. Although it is true that most new mutations are recessive, there are many exceptions. Indeed, it is the exceptions to all three Mendelian rules that make the interpretation of genetic transmission such an interesting array of puzzles.
Transmission Probabilities and Recurrence Rates
When the genotypes of the parents and the nature of a gene’s expression are known, one can easily predict the probabilities of each genotypic and phenotypic outcome. The segregations of alleles from each parent are independent events. So, according to the product rule of probability, you simply multiply the independent probabilities to determine the overall likelihood of a given outcome. The likelihood of a particular allele being inherited from the father is multiplied by the likelihood of a particular allele from the mother. For example, since half of the alleles from a heterozygous Rr father or mother will be R, the overall probability of the first child being homozygous RR is ½ × ½ = ¼ (Figure 6-4). Such outcomes are often summarized in a table called a Punnett square after the pioneer geneticist, R.C. Punnett, who first used it.
Figure 6-4. Punnett square for a simple monohybrid cross, showing the basis of a 1:2:1 genotypic segregation pattern (¼ RR, ½ Rr, ¼ rr). If the R allele is fully dominant, the phenotypic ratio in this monohybrid cross is 3:1, i.e., ¾ R – and ¼ rr.
Being able to predict the probability of a typical genetic outcome is one of the most powerful foundations of genetics. If you are focusing on only one gene and both parents are heterozygous, as in Figure 6-4, then the genotypic ratio in the next generation is ¼: 
:¼, or 1:2:1. With complete dominance, the phenotypic ratio is ¾:¼ = 3:1. Adding a second segregating gene simply expands the probability calculation. For example, consider offspring from a dihybrid cross of Aa Bb × Aa Bb, where the dash indicates that the second allele in a genotype with one dominant could be either a dominant or recessive without changing the phenotype. According to the product rule, we simply multiply the probabilities of each gene combination:
For three independently-segregating genes, you simply add another cycle of multiplication by 3:1, giving eight different types in a ratio of 27:9:9:9:3:3:3:1. But the examples of most interest are those in which some other factor is influencing the expression of these genes. If one allele is not completely dominant over the other, for example, the phenotypic ratio will be the same as the genotypic ratio.
yields a 1:2:1:2:4:2:1:2:1 ratio, with the most common (at 4/16) being the dihybrid (2/4 Aa × 2/4 Bb = 4/16 Aa Bb). Of course, if the genes are located near each other on the same chromosome, they will tend to be inherited together and the patterns of transmission prediction must be modified.
On the surface, then, the rules governing gene transmission are fairly straightforward. It is one reason that genetics has such useful predictive power. But as we have already seen, genes code for products that can interact in complex ways with each other and with environmental conditions. A simple example is incomplete dominance seen, for example, in the pigmentation of some flowers (Figure 6-5). One allele does not produce enough active enzyme to catalyze as much red pigment in the heterozygote as two alleles do in the homozygote. Such incomplete dominance affects the ways in which genotypes are expressed phenotypically. It is a common exception to the classical Mendelian Rule of dominance. Thus, while inheritance patterns form the working foundation, each trait’s phenotypic assessment is often the primary focus in day-to-day applications of medical genetics.
Figure 6-5. Flower color provides a classical example of incomplete dominance. In this case, let us call the genotype that produces red flowers RR and the genotype for white flowers rr. The heterozygote, Rr, produces less red pigment, so flower color is an intermediate, pink. The intensity of flower pigmentation and copies of the functional (R) allele show a dosage effect. (© Robert Calentine/Visuals Unlimited.)
Special Cases: An Overview
For traits with medical implications, recurrence rates are an important consideration. These are based on both the pattern of Mendelian inheritance and on the way the alleles are expressed during development of the phenotype. But even if the genes are not contributing to development with equal weight, gene interactions can influence the phenotypic outcome in complex ways. In the following sections, we will introduce the genetic and biochemical basis underlying some common ways in which genetic outcomes can be expressed. Specific medical examples will be the focus of Parts 2 and 3.
Exceptions to the Rule of Dominance and Recessiveness
The concept of dominance was conceived before it was possible to measure the biochemical events underlying the development of a trait. As we saw with the example of flower color in Petunias, some alleles contribute to the phenotype with a dosage effect. This yields incomplete dominance of one allele over another. Each allele adds a certain amount to the trait’s intensity, and the heterozygote can, therefore, be distinguished from either homozygote. Indeed, many genes have a small, cumulative phenotypic effect like this. Furthermore, dominance is sometimes influenced by the techniques we apply to measure phenotypes. An example of this idea is seen in starch production in peas (Figure 6-6). On a superficial phenotypic level, both the dominant homozygous and the heterozygous seeds have the same round appearance. But biochemically, the heterozygous seeds have significantly less starch. At the visual level, round is dominant to wrinkle. But at the biochemical or histological level, it is easy to distinguish the heterozygotes from either of the homozygous genotypes, and the trait is incompletely dominant. Indeed, at the biochemical level, almost all traits are probably incompletely dominant to some extent.
Figure 6-6. The distinction between complete and incomplete dominance is often a function of the level of magnification or the precision of information we have about a phenotype. In this example, microscopic examination of pea seeds that are heterozygous for the r allele have markedly fewer starch grains than those that are homozygous RR. (Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
Gene Interactions
Although individual genes can have a large phenotypic effect, they seldom if ever act in complete developmental isolation. Gene interactions are more the rule than the exception. But the outcomes are no less predictable. This can be illustrated by a few simple examples based on dihybrid phenotypic patterns. When two genes affect the same or a closely-related trait, the 9:3:3:1 dihybrid segregation pattern is modified in predictable ways. As shown by the example in Figure 6-7, interactions merge segregating categories, depending on how the genes work together on the same trait. Some common examples are summarized in Figure 6-8.
Figure 6-7. Dihybrid Punnett square showing: (a) 9:3:3:1 dihybrid ratio for coat color and spot number. (b) 12:3:1 modified dihybrid ratio due to dominant epistasis, when the dominant A allele masks segregation at the B locus. Consider, for example, an animal that is normally light brown with yellow spots. A 12:3:1 phenotypic ratio is produced if the A mutant produces complete melanism, so no yellow spots can be seen. Melanism masks the expression of any segregating gene that would otherwise influence yellow appearance.
Figure 6-8. Modified dihybrid ratios due to some common kinds of gene interaction.
The typical dihybrid ratio will be modified as a function of how the genes interact in forming the phenotype. For example, epistasis (literally, to “stand above”) is a situation in which one gene masks the expression of a second gene. By analogy, having a completely shaven head masks the expression of genes that determine hair color, so complete baldness is epistatic to, it “stands above,” hair color (Figure 6-9). Similarly, Mexican hairless dogs do not express the genes that would otherwise define hair texture or curliness. By masking the segregation of other genes affecting the same trait, the overall phenotypic ratios will be altered. The expected phenotypic ratios will depend on whether the epistatic gene acts through a dominant or a recessive phenotype.
Figure 6-9. The absence of hair, whether due to a hair stylist or a mutation, is epistatic to (it masks) the genes coding for hair color and curliness. The Sphynx is a hairless breed of cat. (Photograph by M. Minderhoud. Licensed under GFDL, http://www.gnu.org/copyleft/fdl.html or CC-BY-SA-3.0, http://creativecommons.org/licenses/by-sa/3.0/, via Wikimedia Commons).
Genes in the same, or a closely related, pathway also interact in ways that can modify the predicted phenotypic ratios. Consider the situation in which a trait is affected by a dominant mutation of either the A or the B gene. Only in the 1/16th of the offspring that are homozygous recessive aabb will the alternative trait appear. The modified ratio is, therefore, 15:1. On the other hand, if the trait can be blocked by homozygous recessive genotypes at either locus, the resulting phenotypic ratio will be 9:7. Other combinations of interaction are also possible. The key idea, therefore, is to think about genetic applications on two levels: the underlying genetic segregations and the ways the resulting allele combinations influence development.
Genotype × Environment Interactions
Just as some genes will affect the phenotypic expression of other genes, environmental influences also play a key role in developmental expression. Indeed, genotype ×environmental interactions are probably more often the rule than the exception, especially for environmental variables like temperature that can affect molecular stability or the rate of a biochemical reaction. A classic example is the pigmentation pattern seen in Himalayan rabbits and cats (Figure 6-10). In this case, the enzyme for synthesizing melanin is inactivated in the warmer parts of the body so they are lightly pigmented. Only in the cooler areas like the ears, tip of the nose, and feet does the enzyme work normally. Not surprisingly, genotype × environment interactions can be limited to a specific time period or developmental stage if they are associated with genes that are only transcribed for a defined period.
Figure 6-10. The pigment pattern in Siamese cats is caused by a gene that codes an enzyme which is most active at cooler temperatures. (Photograph by Telekokopelli. Licensed under CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0), via Wikimedia Commons.).
In an earlier chapter, we briefly discussed sickle cell anemia as an example of a simple Mendelian trait. It is also the basis of a phenotype that is affected by environmental conditions. Sickling of the red blood cells (RBCs) takes place when oxygen (O2) concentrations in the blood are low, and hemoglobin begins to crystallize in the cells. This can happen during high physical activity in which O2 levels in the muscles decline or during travel by air or to high elevations when atmospheric pressure is lower. But by carefully monitoring conditions like these, individuals with sickle cell anemia can live normally. Similarly, phenylketonuria (PKU), one of the first human traits shown to follow Mendelian transmission, has a severity that is affected by environmental factors. Individuals with PKU lack functional phenylalanine hydroxylase. If their dietary phenylalanine levels are not restricted, its buildup can cause a variety of serious phenotypic effects including mental retardation. But by following a diet restricted in phenylalanine, their development can be normal. Understanding how an individual’s genotype interacts with their environment is critical to many treatment protocols.
Penetrance and Expressivity
We saw earlier that most, if not all, traits actually show incomplete dominance if one measures the level of biochemical activity in each genotype. There is an underlying, sometimes hidden, stepwise gradation of cellular and biochemical genetic effects. Dominance is essentially a phenomenon caused by the architecture of biochemical pathways that allows variation among common genotypes so that the heterozygote is still effective enough to produce a normal outcome. That same gradation in gene action underlies a separate, but related, pair of phenomena: incomplete penetrance and variable expressivity(Figure 6-11).
Figure 6-11. This graphical representation shows the relationships between inheriting the genotype for a trait (top), exhibiting that trait if it is incompletely penetrant (middle), and the severity when it is expressed (bottom).
A trait is completely penetrant if each individual expresses his or her own individual genotype. But sometimes a particular genotype is inherited, but the trait it determines is not expressed. In those instances, the genotype is partly masked, or incompletely penetrant into the phenotype. To quantify the degree of penetrance, it is necessary to evaluate a large number of individuals who are known to have the appropriate genotype. Given such population data, the percent penetrance can become a tool, i.e., an independent probability of expression, in predicting the expected outcome in a birth or a pedigree. For example, assume that two parents are heterozygous for a recessive condition that has 70% penetrance. What is the probability that their first child will show the trait? From two heterozygous parents, there is a ¼ chance that the child will be homozygous. But among such homozygotes only 0.7 will actually express the trait. The overall probability of a phenotypically affected child in this case is, therefore, ¼ (0.7) = 17.5%.
Phenotypic variability goes even further. Not all individuals will express a trait with equal severity. In some cases a trait will be strongly expressed and in other cases it can be mild. Thus, among those showing a trait, variation is seen in the range of expressivity. Indeed, penetrance and expressivity may only be different faces of the same phenomenon—variable outcomes in the development of a sensitive characteristic. Polydactyly (Figure 6-12), for example, is an incompletely penetrant trait that is also variable in its expressivity. An individual inheriting this dominant mutation may have five normal fingers and toes (incomplete penetrance). But if the mutation is shown phenotypically, the degree of expression can range from a small additional digit on only one hand or foot to well-formed extra digits on all of them (variable expressivity).
Figure 6-12. Polydactyly is a dominant trait with variable expression. This foot shows postaxial polydactyly with duplication of the 5th toe.
Lethal Alleles
When it can be traced to a genetic cause, “death” is a phenotype. In fact, gene mutations that cause death in homozygotes are the most common type of genetic change. At first this may seem hard to believe. But if we take a moment to think about all of the biological processes of cells and our general physiology that are indispensable to us, it begins to become less surprising. Although dominant lethal mutations might occur, they immediately kill the carrier, so they are not inherited. Thus, only recessive lethal mutations play a role in genetic assessment. The most relevant are probably those with both homozygous lethality and some separate physical or developmental phenotype in the heterozygote. In these instances, the physical or developmental phenotype will be dominant since it will be expressed in the heterozygous carriers, and lethality will be recessive because it is expressed in the homozygote.
For a recessive lethal mutation that has a phenotypic effect in heterozygotes, the expected Mendelian ratios among offspring will be modified. Consider the shortened tail in Manx cats (Figure 6-13). The tail is short in heterozygotes, and homozygotes for the tail mutation die early in development. If two Manx cats produce offspring, the resulting phenotypic ratio will be 2:1, i.e., 2/3 will have short tails and 1/3 will be normal. Thus, while the underlying Mendelian segregation ratio is unchanged, the data are changed by the elimination of one of the genotypic classes.
Figure 6-13. A tail-less Manx, named Silverwing, photographed in 1902. The shortened tail of a Manx cat is caused by a mutation that is homozygous lethal. The dihybrid 1:2:1 Mendelian ratio is, therefore, modified to a 2:1 ratio of shortened tail to normal tail. (A: Originally appeared in Cats and All About Them by Frances Simpson, published by Frederick A. Stokes Company Publishers, September 1902. B: Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
Multiple Alleles
In the examples discussed so far, we have focused on fairly direct connections between a genotype and a phenotype. Often, there are two fairly distinct phenotypes associated with normal versus abnormal development, even when a variety of alleles are present in a population. For that reason, it is common to simplify genotypic models to two alleles. But the increasingly available data about proteins and DNA sequences must often be interpreted with multiple allele models. Familiar examples of multiple allelism include the vast genetic diversity seen at the histocompatibility loci determining tissue types and the common polymorphism for cell surface antigens that is known as the ABO blood type (Figure 6-14). A key thing to keep in mind when analyzing traits like this is that a gene pool can carry several different alleles at some moderate frequency, but each individual in that population carries at most two of them.
Figure 6-14. ABO blood type is due to a surface antigen and the complementary serum antibodies. (a) Antigens on the RBC surfaces and the associated antibodies in the blood serum. (b) To predict offspring genotypes and phenotypes, one must relax the assumption that only two alleles are segregating in a mating. (c) Glycosyl transferase alleles encoded by the I gene recognize and bind different sugars to the carbohydrate tree. (Reprinted with permission from Brooker RJ: Genetics: Analysis & Principles, 3rd ed. New York: McGraw-Hill, 2008.)
Consider, for example, the paternity suit of a woman with blood type A against a man with blood type B. The child has blood type O. Can the man be excluded as being the child’s father? A and B blood types are detected with specific antibodies, and the inability to detect either of these yields a third alternative, the recessive O blood type. The A blood type is due to the IA allele, and the B blood type to the IB allele. The O blood type has neither and is represented as a recessive homozygote, i i. We can evaluate this problem using a Punnett square organized like the one in Figure 6-14. First, we realize that the blood types of the adults are only phenotypes and, thus, give us only partial information about their genotypes. For example, the man with blood type B could have either of two genotypes: IBIB or IBi. The mother must have at least one IA allele and the male must carry IB. But for the child to have the O blood type (i i), both parents must be heterozygous for the i allele. On the basis of the available information, the male cannot be excluded in this case. His genotype could possibly be IBi. But if additional evidence showed that his parents were both IAIB, for example, then he can be excluded, because he must be homozygous IBIB and could not have passed an i allele to the child.
Legal cases now frequently use multiple-allelic DNA markers. But until the advent of DNA markers as forensic evidence, blood type was often employed. One famous case in 1943 involved the actor Charlie Chaplin who was accused of fathering a child with the actress Joan Barry. She sued for child support, but blood tests definitively excluded him. When the court did not admit this evidence, he was forced to pay child support anyway. This eventually led to new laws, and the legal system continues to advance slowly as it recognizes the uses and limitations of new technologies to improve the value of evidence.
Contrasting Sex-linked, Sex-limited, and Sex-influenced Expression
Sex-linkage and the mechanisms we call sex-limited and sex-influenced expression are three quite different phenomena. The main thing they have in common is some kind of relationship with the distribution of genes on the sex-determining chromosomes or developmental effects that differ between females and males. Thus, a careful fitting of information about inheritance and mechanism of gene function is needed to interpret these superficially-related phenomena properly.
Authorities actually differ in the way they use the term “sex-linkage.” For some, it is synonymous to X-linkage, with the content of the Y chromosome providing a special case of its own. For others, “sex-linkage” applies to any gene on either sex chromosome and is then subdivided into the X-linked and the Y-linked categories. Since most genes are located on the X, this difference in terminology is almost never a source of confusion. For this discussion, however, the key point is that the X chromosome (and the Y chromosome for that matter) carries many genes that have nothing to do with gender differences per se.
In contrast, sex-limited and sex-influenced characters are clearly gender related. A sex-limited character is one that determines a phenotype that is found in only one of the two sexes. Genes responsible for sperm formation in males or a prolapsed uterus in females are examples. Sexually dimorphic animals like chickens show gender differences in traits like body proportions and feathering. The genes for such traits can be found on any chromosome. As expected, most are autosomal. Sex-influenced characters, on the other hand, are those that have a statistically higher frequency of being expressed in one sex than in the other. Examples include a higher frequency of cleft lip and gout in males and a higher frequency of cleft palate, spina bifida, and breast cancer in females. Pyloric stenosis is more common in males, especially firstborn males, than in females. Although the reason for these differences in frequency is often not known, the traits are not gender-specific.
When analyzing traits like these, and indeed anytime one is exploring the genetic mechanisms behind an observation, always remember the rules of probability. Simply having several males with a certain trait appearing in a family pedigree does not necessarily tell us anything concrete about the mode of inheritance or its expression. But patterns of expression in the pedigree can offer the key. Still, rare occurrences cannot be ignored. As is true for many aspects of medicine, the process of diagnosis combines information about both probability and people.
Part 2: Medical Genetics
Introduction
The work of Gregor Mendel in the late 1800s ushered in the modern era of the science of genetics. His work on segregation identified the basic principles of uni-factorial inheritance. His description of the patterns of inheritance provided the basis for what we now understand to be single gene transmission of traits. In fact, conditions exhibiting single gene inheritance patterns are often referred to as Mendelian traits. In this era of medical practice, every practicing health care provider should have a solid working knowledge of basic inheritance patterns. The standard of care is that a 3-generational pedigree is part of the medical records of each and every patient (see Chapter 9, Family History). Every practitioner should readily be able to look at a patient’s pedigree and identify familial conditions and determine the most likely mode of inheritance of that condition. In the section above, the basic principles of single gene inheritance have been reviewed. Now we will discuss the clinical characteristics, special considerations, and exceptions to the rules associated with each inheritance pattern.
Before we move into a discussion of specific inheritance patterns, a few definitions and concepts should be discussed. A genetic locus (plural loci) refers to the specific location of a particular segment of DNA, i.e., on which chromosome and where on that chromosome does a segment reside. When tracking traits through a family, the most common application of the term locus refers to the location of a particular gene. For example, the CFTR gene which is associated with the condition cystic fibrosis is found on the long arm of chromosome 7. Thus, the locus for the cystic fibrosis gene is designated 7q31.2 (Figure 6-15). Still, it should be emphasized that, although the typical use of the term locus in medical practice is for the location of a functioning gene, all other segments of DNA still have a physical position which is the locus for that designated segment.
Figure 6-15. The chromosomal location (locus) for the CFTR gene on chromosome 7q31.2. Mutations in this gene cause the genetic condition cystic fibrosis. (From Genetics Home Reference. US National Library of Medicine. Handbook: Help Me Understand Genetics. Available at: http://ghr.nlm.nih.gov/handbook. Accessed August 21, 2012.)
Each gene has an expected (normal or wild type) sequence. Although humans demonstrate tremendous variability in their phenotypes, it is important to note that over 99.9% of genetic code among people is exactly the same. For most genes, an alternative form is simply not compatible with viability. If a polymorphism (literally “multiple forms”) does exist for a given gene, the different sequence is termed an allele. It is estimated that there is 1 single nucleotide polymorphism (SNP) per 1000 base pairs in the human genome. This translates into approximately 3 million SNP’s per individual person. Most SNP’s are inconsequential as far as clinical expression. But these “benign” SNP’s are very important as markers of variation in people and can be used for linkage assessment. The small subgroup of SNP’s that do have clinical implications account for much of the observed human diversity.
A person is said to be homozygous for a specific gene if both copies of that gene have the same sequence; i.e., the two copies are of the same allele. Likewise, if the two copies of a gene have different sequences, the person is heterozygous for that gene; i.e., there are two different alleles. In describing the association of the genotype to the person, a reference could be made to the person being a homozygoteor a heterozygote for that condition. Genetic homogeneity for a particular disorder means that a single polymorphism (mutation) only causes that condition. In actuality, genetic homogeneity is rare. In fact, we can only think of one example of a human disorder that exhibits genetic homogeneity. A single nucleotide polymorphism (SNP) in the beta hemoglobin gene on chromosome 11 causes sickle cell anemia (SSA). The SNP that causes SSA is a change from an adenosine (A) to a thymidine (T) at nucleotide position 334 (this SNP has been designated rs334). This nucleotide change results in an amino acid switch in the hemoglobin molecule where a valine amino acid is substituted for a glutamate at the sixth position in the protein chain. Only this change in the hemoglobin molecule results in sickle cell. This makes sickle cell anemia one of the very few conditions in humans that exhibit genetic homogeneity (Figure 6-16).
Figure 6-16. (A, B) A single unique nucleotide polymorphism (SNP) in the beta hemoglobin gene on chromosome 11 causes sickle cell anemia (SSA).
When discussing disease-causing mutations, almost all other human conditions exhibit genetic heterogeneity. Genetic heterogeneity means that different mutations cause an identical, or very similar, phenotype. Two different types of genetic heterogeneity can occur. Allelic heterogeneity refers to different mutations within the same locus producing the same condition. While the CFTR gene is the only gene known to be associated with cystic fibrosis (CF), several different mutations of this one gene can all produce the condition. About 75% of all CF mutations are a specific change designated deltaF508 (a deletion of a phenylalanine at amino acid position 508). Analysis of the different mutations seen in persons with CF has led to the development of CF gene “panels.” These panels test for mutations selected for their relative frequency in a given population. For example, a panel of the 12 most common mutations will identify 85% of alleles. A panel of 34 common mutations will detect 90% of alleles. These panels offer an alternative to the more expensive and cumbersome process of sequencing the entire gene. For now this is a desirable first approach given that it is quicker and significantly less expensive. But advances in sequencing techniques are likely to change this situation in the very near future.
Alternatively, some conditions exhibit locus heterogeneity. Locus heterogeneity refers to mutations at completely different loci producing the same phenotype. Spinocerebellar atrophy (SCA) is an inherited disorder characterized by a progressive degeneration of the brainstem and cerebellum (Figure 6-17). Most of the families reported with SCAs demonstrate autosomal dominant inheritance. In the mid-1980s, linkage studies suggested that dominant SCA was linked to the major histocompatibility (HLA) locus on chromosome 6p. Further investigation revealed that this was true for only some families. Subsequently, other loci were discovered that also caused SCA. At the present time, over 30 different loci have been identified as being linked to dominant SCA (Table 6-1). Thus, dominant SCA demonstrates striking locus heterogeneity.
Figure 6-17. Sagittal MRI scan of the brain in a patient with spinocerebellar atrophy. The arrows point out the small size of the brainstem and cerebellum associated with atrophy of these structures.
Table 6-1. Locus Heterogeneity in Spinocerebellar Atrophy (SCA)
Clinical Aspects of Mendelian Inheritance
In these times, most students have had significant exposure to the concepts of single gene inheritance. Often biology classes in junior high school (or even earlier) have discussed the basic concepts of Mendelian transmission of traits. The introductory part of this chapter provides a targeted overview of the basics of Mendelian principles. As with most things, further understanding of the basic principles usually leads to “exceptions to the rules.” Such is the case with single gene inheritance. All health care providers should first have a solid understanding of Mendelian principles. Yet equally important, they will need to appreciate all of the potential nuances that can be introduced in the clinical realm. In the sections below we take each major single gene inheritance pattern and provide information about it in three categories:
1. Classic characteristics.
2. Recurrence risks.
3. Special inheritance considerations.
Autosomal Inheritance
Mendelian traits that are transmitted by genes that lie on the autosomes (the numbered 22 pairs of non-sex chromosomes) demonstrate autosomal inheritance. Autosomal inheritance is characterized by the transmission of the two alleles segregating independently in the same manner by both males in females. It is important to note that transmission is not the same as expression. Thus, while autosomal conditions are transmitted in the same manner in males and females, expression may be different between the two sexes (see more detailed discussion in Part 1). Depending on the pathogenesis of the genetic change (see Chapter 16), a condition may demonstrate autosomal dominantor autosomal recessive inheritance.
Autosomal Dominant (AD) Inheritance
1. Classic characteristics. For an allele to be characterized as dominant, only one copy of the abnormal gene is required for the individual to be affected. Both males and females are affected and may transmit the gene to offspring of either sex. Dominant conditions demonstrate “vertical transmission.” A review of the pedigree shows the trait of interest passing from one generation to the next; or in the vernacular, “down the line.”
Most people understand the concept of a trait being genetic if it “runs in the family.” If a trait is found in a parent and is subsequently seen in their child, it tends to make sense. What is often very confusing to families is the concept of a condition being genetic when there is no other affected person in the family. This is particularly true for dominant conditions. For dominant conditions, affected offspring typically have affected parents. There are several clinical scenarios, however, in which a person is seen with a known autosomal dominant condition without an affected parent. In these settings, careful review of the family history and examination of key persons in the family may be necessary to sort things out.
In some cases, there is no affected parent, because there is a new mutation in the child (truly the first occurrence in the family). The occurrence of new (sporadic) dominant mutations has been reported in association with an advanced paternal age in many human genetic conditions. For example, achondroplasia is a rare skeletal dysplasia (Figure 6-18). The condition is associated with a disproportionately short stature described as rhizomelic shortening (the upper segments are more shortened) of the limbs. Persons with achondroplasia also have a relative macrocephaly. Achondroplasia is known to be an autosomal dominant condition. It is caused by mutations in a gene called fibroblast growth factor receptor 3 (FGFR3). When it is familial, typical vertical transmission in keeping with autosomal dominant inheritance is observed. But, 80% of children born with achondroplasia do not have an affected parent. In this setting it has been shown that this is due to the occurrence of new, sporadic mutations of the FGFR3 gene in the offspring. The occurrence of new mutations in achondroplasia is now known to be correlated with an advanced paternal age. Achondroplasia is not the only autosomal dominant (AD) condition shown to have a paternal age affect (Table 6-2). In fact, it is likely that sporadic mutations at all loci are increased with aging fathers. It is just that the AD conditions are more readily observed events.
Figure 6-18. (a) Young girl with achondroplasia; (b) Female infant with homozygous achondroplasia.
Table 6-2. Examples of Dominant Conditions With Documented Paternal Age Effect∗
Achondroplasia
Apert syndrome
Marfan syndrome
Neurofibromatosis
Treacher–Collins syndrome
Crouzon syndrome
Progeria
∗This phenomenon is probably true for all loci, but is most easily observed in dominant syndromes; Thompson, J.N., G.B. Schaefer, M.C. Conley, and C.G.N. Mascie-Taylor: Parental Age Can Affect the Severity of an Inherited Human Trait. (Letter) New Engl. J. Med., 314(8):521, 1986.
The increased occurrence of new mutations with an advanced paternal age bears additional discussion. First, it is important to note that the new occurrence of abnormal genetic events increases with an advanced parental(both sexes) age. The association of an increased chance of chromosomal non-disjunction seen with an advanced maternal age was discussed in Chapter 5, Cytogenetics (Figure 6-19). Here we have discussed the increased chance of single gene mutations in association with an advanced paternal age (Figure 6-20). Simply put, as the parent get older, the chances of genetic abnormalities in the offspring increases. In reality, the occurrence of birth defects as it relates to parental ages is a somewhat “J” shaped curve (Figure 6-21). There is an increased incidence of birth defects in particularly young parents—presumably due to different mechanisms. The lowest incidence of congenital anomalies is seen with parents who are between 18 to 30 years old. As the ages of the parents go beyond 30, the chances of birth defects starts to increase. Beyond 40 years old, the incidence rises asymptotically.
Figure 6-19. Increasing incidence of births with infants with all trisomies correlated with mother’s age at the time of delivery.
(Reprinted with permission from Crow JF, Nat Rev Genet 2000;1:40.)
Figure 6-20. Graphs of increasing incidence of Apert syndrome and neurofibromatosis correlated with father’s age at the time of delivery. O/E = observed / expected. (Reprinted with permission from Crow JF, Nat Rev Genet 2000;1:40.)
Figure 6-21. Overall incidence of birth defects with advancing parental age.
As we discussed earlier, the nature of the difference in genetic changes seen in men versus women lies in the physiology of gamete formation (Figures 6-22 and 6-23). A newborn female infant has only a few hundred thousand in her ovaries. These oocytes have started into meiosis, but have stopped the process in meiosis I. Meiosis I is not completed until ovulation, and meiosis II is not accomplished until fertilization. Thus, as mothers age their oocytes have remained in this state of suspended chromosome division for long periods of time. This presumably leads to a greater chance of meiotic error. Alternatively, men do not make mature sperm until puberty. Spermatozoa are made in the millions with a turnover rate around 60 days. Thus, the more sperms that are made, the greater the chance of a copy (transcription) error.
Figure 6-22. Diagram of ontogeny of male and female germ lines.
Figure 6-23. Cell divisions and gametogenesis. (Reprinted with permission from Crow JF, Nat Rev Genet 2000;1:40.)
Besides the occurrence of spontaneous mutations, there are other reasons that a child affected with an AD condition may not have an affected parent. Possibilities include the parent carrying the gene but not expressing it (see incomplete penetrance below), or a parent having gonadal (germ line) mosaicism for a dominant mutation (discussed further in Chapter 12, Atypical Inheritance).
2. Recurrence risks. Consider the matings of an affected heterozygote for an autosomal dominant condition and a normal homozygote. There is a 50% chance that any given offspring will be affected and a 50% chance that it will be unaffected. Since the trait is autosomal, both sexes have the same probabilities.
For the rare matings of two affected heterozygotes, the recurrence risks for each offspring are as follows:
• 25% chance affected homozygote
• 50% chance affected heterozygote
• 25% chance normal homozygote
• Total recurrence risk for an affected child: 75%
3. Special inheritance considerations. In general, affected individuals of autosomal dominant conditions are heterozygotes. Since homozygous affected individuals are typically quite rare, the usual mating in dominantly inherited diseases is between a homozygous normal and a heterozygous affected individual. But, if the allele is sufficiently common in the population, matings between two heterozygous affected parents resulting in homozygous affected offspring are seen. True (complete)dominance implies that an identical phenotype is seen in those who are heterozygous or homozygous for the mutation. In other words, having only one copy of the abnormal allele produces the “full” phenotype. In the realm of human genetic conditions, few actually demonstrate complete dominance. Huntington disease (HD) is an autosomal dominant neurodegenerative condition. It is characterized by adult onset degeneration of the basal ganglia (Figure 6-24a). Correlated with these brain changes, patients with Huntington disease exhibit progressive neurologic symptoms that include involuntary and abnormal (choreiform) movements. Over time the progressive neurologic problems result in an early death. Studies in unique populations in which HD occurs in a high frequency have shown that HD demonstrates a true dominance pattern of inheritance. Another condition that has been shown to exhibit true dominance in its inheritance is Best disease—also known as vitelliform macular dystrophy. Best disease (Figure 6-24b) is associated with progressive accumulation of lipofucsin in the pigmentary retinal epithelium with resultant vision loss. Evaluation of a large Swedish kindred dating back to the 17th century has also confirmed true dominance–with no identified differences in expression seen between heterozygotes and homozygotes.
Figure 6-24. Two examples of conditions exhibiting ‘true’ dominant inheritance. (a) Brain MRI of a patient with Huntington disease. Arrow is pointing to area of marked degeneration of the basla ganglia. (b) Retinal photograph of a patient with Best disease (vitelliform macular degeneration). Arrow points to characteristic retinal abnormality—deposits of fatty material in the sub-retinal space, which creates a characteristic lesion resembling an egg yolk.
The above examples, however, are the exceptions. In humans, affected homozygotes for dominant conditions typically have a more severe phenotype than do the heterozygotes. Achondroplasia, for example, is an autosomal dominant skeletal dysplasia briefly described earlier (Figure 6-18a). Heterozygotes with achondroplasia have the bony changes described for the condition, but have normal intelligence and a normal life expectancy. In the rare event that two individuals with achondroplasia mate, there is a 1 in 4 chance that the conception will result in homozygosity for the FGFR3 mutation for this condition. In homozygosity for achondroplasia, the bony changes are much more severe (Figure 6-18b) and are most often associated with death in infancy. In the strictest sense, then, achondroplasia is not a true dominant condition, but would be better described as semi-dominant. Since almost all human dominant conditions are transmitted this way, convention is simply to call them autosomal dominant.
Another variation on the theme of dominant inheritance is that of co-dominance. Co-dominance refers to the simultaneous expression of both alleles in a compound heterozygote. In this setting it is again important to remember the distinction between genotype and phenotype and to be aware of the level at which the phenotype is defined. Some of the better examples of co-dominance in plants and animals involve different genes that control coloration. In humans, examples of co-dominance are rare. In fact, the ABO and MN blood types as phenotypes are among the few conditions that have been shown to be co-dominant phenotypes in people.
Dominant conditions typically show some degree of variable expression. There are many possible causes of variable expression including environmental factors, modifier genes, varying genetic backgrounds, and so forth. Variable expression (or variability) simply refers to the severity of the condition, i.e., to what degree is the person affected with the condition or “how much” the condition is manifest. The degree of variability differs from condition to condition. Some show a wide range of variability, others much less. A given condition may also differ in the variability seen within a family (intra-familial variability) as compared to between different families (inter-familial variability).
As a side note, it is important to distinguish variable expression from pleiotropism. Although the terms refer to two distinctly different concepts, they are often confused. Pleiotropism (or pleiotropic effects) refers to having several different clinical manifestations of a single genetic change. For instance, Marfan syndrome is a heritable disorder of connective tissue (Figure 6-25). Marfan syndrome is caused by mutations in the gene that codes for the protein fibrillin. Fibrillin is a microfibrillar protein that is a major component of the connective tissue of structures such as bones, eyes, skin, and larger blood vessels. As such, persons with Marfan syndrome can have problems with all of these tissues. They have a disproportionately tall stature, with the excessive length being in the limbs as compared to the trunk. They can have ocular changes, such as severe myopia, retinal detachments, and dislocations of the lenses. Their skin is hyperextensible, bruises easily, and exhibits poor wound healing. The most serious complication of the condition is progressive dilation of the root of the aorta, which if severe enough can lead to rupture and sudden death. In fact, in patients with untreated Marfan syndrome, the average life expectancy is under 30 years. All of the different problems listed above are due to the one genetic change (fibrillin mutation) and represent the pleiotropic effects seen from this mutation.
Figure 6-25. Young girl with Marfan syndrome.
An individual who has the genotype for a disease may not exhibit the disease phenotype at all, even though he or she can transmit the disease gene to the next generation. As discussed in an earlier section, this is referred to as incomplete penetrance. Thus, incomplete penetrance represents the far extreme of variable expression, i.e., it is so mildly expressed that it is not even detectable. Some conditions are completely penetrant meaning that if you have the genotype you will definitely express the condition. But if a trait is incompletely penetrant, the degree of penetrance cannot be reasoned out or inferred. It must be determined by examining a large number of families to calculate what proportion of known heterozygotes (autosomal dominant, AD) or homozygotes (autosomal recessive, AR) develop the disease phenotype. With enough family data, a penetrance rate can be determined for a particular condition. A condition that is found to be 60% penetrant, for example, means that 60% of persons with the mutant genotype will actually show clinical expression. The remaining 40% will have the same genotype without any clinical signs or symptoms of the condition.
Van der Woude syndrome (Figure 6-26a) is an autosomal dominant condition that has lower lip pits and clefting as major features. Both features are variably expressed. From extensive family data, the condition is also reported to be 80% penetrant. Figure 6.26b shows a pedigree of a family with van der Woude syndrome. First inspection of the pedigree suggests autosomal dominant transmission, which is in fact correct. But closer inspection of the pedigree shows an interesting phenomenon. Individual II.2 is an unaffected person but has two affected children. This person must then carry the van der Woude mutation but not be expressing it, i.e., showing incomplete penetrance.
Figure 6-26. (a) Young child with van der Woude syndrome. Note the bilateral cleft lip and the lower lip pits. (b) Pedigree of a family with van der Woude syndrome. See text for details about incomplete penetrance.
In this particular family, an interesting question then came from individual II.1 (noted by the arrow on the pedigree). This lady wanted to know what her chances were of having a child affected with van der Woude syndrome. It would have been tempting to give her a recurrence risk of zero since she was unaffected. However, what she observed in her brother and his children made her realize that this is, in fact, was not the correct answer. So what is correct? To get to the correct answer, you need to apply logic, knowledge of inheritance patterns, and all available information about the condition. Since her father (I.1) is affected, she has a 50% chance of inheriting the gene. Since she is unaffected, she could only have an affected child if she has the mutation and is non-penetrant. Thus, the chance that she is an unaffected person with the mutation is the product of the probability that she inherited the mutation times the probability that she failed to express it: 0.5 × 0.2 = 0.1 (note, if the condition is 80% penetrant, it is 20% non-penetrant). If she carries the gene, then there is a 50% chance that any of her children would inherit the mutant gene, giving an overall probability of 0.1 × 0.5 = 0.05. Finally, even if the child inherits the gene, the condition is only 80% penetrant. The actual risk of her having an affected child is 0.05 × 0.8 = 0.04. So her risk for an affected child is 4%, not zero!
In order to provide families with the best available information, counseling must be based on the knowledge of all of the above parameters for the condition, although we grant that, for rare conditions, this information may be incomplete. Still, when counseling with families, information about expression and penetrance needs to be included. Each condition has its own specific “profile” of how it is typically transmitted. Table 6-3 gives just a few examples.
Table 6-3. Genetic Profiles (Descriptions) of Inheritance Characteristics of a Few Select Conditions
Autosomal Recessive (AR) Inheritance
1. Classic characteristics. As a group, recessive conditions are less common and show less variability in their expression than do dominant conditions. In contrast to dominant conditions, recessive conditions typically show “vertical transmission.” This means that the condition may be found in multiple individuals in the same generation, but are noted to be passed down the generations. Because the risk is only 1 in 4 that a child will be born to two carrier parents, and because most American families are small (2-3 children), most affected individuals with AR disorders will appear to be sporadic cases (only 1 case in a kindred). The most common presentation is one where the parents of an affected individual are both unaffected. Assessment of the genotypes would show that both parents are heterozygotes, or carriers of the condition. The family history would be negative except for the possibility of affected siblings, or the possibility of affected relatives due to consanguinity in the family.
2. Recurrence risks. For two heterozygotes, the recurrence risk to have an affected child is 25% per conception. The phenotypically normal siblings of an affected child have a 2/3 chance of carrying the recessive allele. This probability may not be immediately evident, but you need to understand the answer completely. It is an important clinical concept, and it appears frequently on standardized tests. It may be helpful to visualize a Punnett square. The reason that the answer is 2/3, instead of 2/4, is that one of the possibilities (i.e., being affected and thus homozygous recessive) has already been eliminated, because of the information that the sib is phenotypically unaffected. Thus, of the three possible outcomes for each unaffected child, there is 1 out of 3 chance that the child will inherit neither abnormal allele and a 2 out of 3 chance of inheriting one copy, i.e., of being a carrier.
Occasionally a heterozygote (carrier) may mate with an affected homozygote. In this case each offspring has a 50% risk to be affected and a 50% risk to be a carrier. The pedigree in this situation would actually resemble autosomal dominant inheritance. This has been termed quasi-dominance. Quasi-dominant inheritance is more likely to occur with common autosomal recessive genes or in the case of parental consanguinity.
The mating of two affected homozygotes results in 100% of the offspring being homozygous affected. This situation is more likely to occur in situations of assortative mating, i.e., mating of phenotypically similar individuals. For example, the most common genetic cause of non-syndromic neurosensory hearing loss is an autosomal recessive mutation in any one of several genes. Persons that are deaf or hard of hearing often tend to find each other because of targeted social networking. As such, it is not at all uncommon for two deaf/hard of hearing persons to have children together. If the etiology of the hearing loss in both parents is caused by the same recessive gene, all of their children will have hearing loss as do their parents.
3. Special inheritance considerations. It is estimated that humans have approximately 22,000 functioning genes. It is also predicted that on average every person carries recessive mutations in 5 to 8 of these genes. In the case of random matings, then, the chance that both members of a couple happen to carry a recessive mutation of the same gene is relatively small. This then is the reason that recessive conditions tend to be seen less commonly than dominant conditions (Table 6-4). Still, there are a few situations in which recessive conditions occur more frequently than would be expected.
Table 6-4. Frequencies of a Few Selected Genetic Conditions
Consanguinity is defined as the mating of individuals with a close relationship by descent from a common ancestor. In simpler terms, it refers to the mating of closely related individuals. There are many important issues to consider in regards to consanguineous matings. There is, and always has been, a strong social stigma against consanguineous matings. Although there is probably no such thing as a universally accepted moral or ethical principle, consanguinity comes close. It is probably the nearest thing to a universal taboo that exists. Given the sensitive nature of the situation, the clinician must remain non-condemning if such information is revealed. Consanguinity as an event does not occur in an evenly-distributed manner. “Pockets” exist across the United States and the world in which consanguinity occurs at a significantly increased rate. Often geographic, cultural, or religious factors limit the pool of possible mates, making a consanguineous union much more likely.
Intuitively, it can be reasoned that the closer the degree of relationship exists for a couple, the more likely that a recessive condition might occur in their offspring. In defining the relative risk of consanguinity, one must first determine how close the relationship is. How closely persons are related can essentially be described by the number of intervening meioses between them (degrees). Thus, first degree relatives have only one meiosis between them. Figure 6-27 provides a graphic representation of degrees of relationship. This will be discussed in more detail in Chapter 9 on Family History and Pedigree Analysis. Mathematically, the degree of relationship can be expressed in several different ways. The coefficient of inbreeding (F) is the probability that an individual is homozygous at a given locus that is inherited from a common ancestor. For a pair of individuals, the kinship coefficient (?) is equal to the F of their offspring. The coefficient of relationship (R)estimates the proportion of genes shared by individuals with at least one common ancestor. Figure 6-28 demonstrates these concepts. While it is unlikely that the non-geneticist will spend time calculating these coefficients, it is helpful for any practitioner to understand the basic concepts of relatedness.
Figure 6-27. Sample pedigree demonstrating degrees of relationship to the proband (arrow).
Figure 6-28. Examples of (a) the degree of relationship, (b) coefficient of relationship (R), and (c) coefficient of inbreeding (F) for selected consanguineous matings.
The major biological impact of consanguinity is to increase the risk of genetic abnormalities in the offspring. When defining relative risk, one must of course start at the baseline. For most purposes, a baseline risk of 4% can be used for the occurrence of congenital anomalies and 3% for mental retardation. In the event of a consanguineous union, published risks over these baselines are:
1. Third degree relative (e.g., first cousins): There is a 2% to 3% risk of mental retardation or a serious genetic condition in the offspring. These numbers come from recent data. This risk is significantly lower than the 10% estimates that were previously reported.
2. Second degree relative (e.g., half-siblings, uncle-niece): There is a 5% to 15% chance of “genetic abnormalities.”
3. First degree relative (e.g., father-daughter, brother-sister): There is a:
• 40% chance of an offspring with any “significant abnormality”
• 12% chance of an autosomal recessive disorder
• 16% chance of a congenital anomaly
• 10% to 15% chance of significant cognitive impairments
Some recessive conditions occur at a higher frequency than would be predicted, because of a heterozygote advantage. A heterozygote advantage is seen when the carriers of a recessive disorder have a selective reproductive advantage, i.e., they are more likely to reproduce. Thus, even though the recessive disorder seen in the homozygotes may be serious or even fatal, the condition is perpetuated at a relatively high frequency in the population. The most commonly purported heterozygote advantage is that of sickle cell anemia (SSA). Patients who are affected with sickle cell (homozygote recessive) have a serious medical condition associated with abnormal hemoglobin molecules that result in an altered shape of the RBCs from the normal “donut” shape to a curvilinear or sickle-shaped cell (hence the name). The abnormally shaped RBCs do not flow smoothly through the small capillaries and produce multiple micro-infarcts in the tissues. Clinical symptoms due to these vascular occlusions include severe, incapacitating pain, congestion and destruction of the spleen, multiple somatic symptoms and poor growth. Sickle cell anemia occurs at a high frequency in people of African and Mediterranean descent. There is good evidence to suggest that the higher incidence of SSA in these groups is due to a heterozygote advantage. Although SSA is a serious condition, the heterozygous carriers (who typically have no symptoms) appear to have an increased resistance to malaria and, thus, are more likely to reproduce in locations where malaria is endemic. Some other examples of heterozygote advantage that have been suggested are listed in Table 6-5. Recessive disorders may also occur at a frequency higher than would be predicted because of a founder effect. The founder effect refers to the over-representation of a particular allele seen in a population that originates from a relatively small group in which the allele frequency is not representative of the overall population.
Table 6-5. Proposed Examples of Heterozygote Advantage (see text for explanation)
The founder effect may also apply to dominant conditions. All of these concepts will be discussed in more detail in Chapter 15 in the discussions on population genetics.
Sex-Linked Inheritance
Sex-linked inheritance refers to the inheritance patterns that are different than those described earlier for autosomal inheritance, due to the fact that the locus in question resides on either the X or the Y chromosome. These differences are due to variant structure and functions of these two chromosomes as compared to the 22 autosomes (see Chapter 5). Y-linked inheritance, also referred to as holandric inheritance, is of limited clinical significance. The Y chromosome contains few functioning genes. Most of these are genes that have something to do with sex determination. That is, they are genes which, if present, will shift gonadal differentiation from the default ovarian development to testicular development. The nature of this is, therefore, that a transmittable trait typically does not occur. To date there has not been a true Y-linked phenotypic feature seen in humans. The often purported example of “hairy ears” is more likely an autosomal trait with sex-limited expression. If a Y-linked condition were identified, the hallmark inheritance pattern would be that of exclusive male-to-male transmission.
X-linked inheritance is seen when the locus in question resides on the X chromosome. The X chromosome is one of the larger chromosomes and contains hundreds of functioning genes (estimated at 900–1400). In humans, the fact that females have two X chromosomes and males only one leads to a difference in expression of X-linked conditions between the sexes. Mutant alleles on the X chromosome are fully expressed in males, who have only a single X chromosome, i.e., are hemizygous for X-linked genes. Thus, for almost all X-linked conditions, males exhibit a more severe phenotype than females carrying the same mutation. This phenomenon is also part of the reason that certain conditions, such as mental retardation, occur much more often in males than in females (males have four times the incidence of mental retardation than females).
Lyonization (X-chromosome inactivation, see Chapter 5) is an important determinant of the degree of phenotypic expression of X-linked mutations in females. Functionally, women are mosaic for the expression of X-linked genes. For each cell the expression is dependent on which X chromosome is active and which is not. In general, the process of X-inactivation occurs early enough in embryogenesis that pattern of expression is actually clonal (in patches). Depending on the genetic change in question, the pattern of X-chromosome inactivation can sometimes actually be mapped (Figure 6-29). An example of this is hypohydrotic ectodermal dysplasia, a disorder caused by mutations of an X-linked gene EDA1 (Xq12-13). In this condition, changes in the ectodermal derivates cause problems with the skin, hair, teeth, and nails. The most serious medical consequence of this condition is that affected males will have a paucity of sweat glands. Inadequate sweating can lead to dramatic overheating of the core body temperature resulting in brain injury or death. Using a simple staining technique, sweat glands can be visualized (Figure 6-30). Female heterozygotes of this condition can be shown to have clonal patches of normal and abnormal density of sweat glands (the normal being around 250 pores per square centimeter).
Figure 6-29. Schematic showing an expression map of glucose-6-dehydrogenase deficiency in female carriers of this X-linked disorder.
Figure 6-30. (a) Iodine-starch painting of the skin in a normal female. The black “dots” represent individual sweat glands. Note the well developed and evenly distributed placement of the sweat glands. (b) Iodine-starch painting of the skin in a female heterozygous for a mutation in the gene EDA1. Note the clonal patches of skin showing decreased numbers of sweat glands.
In theory, Lyonization should be random. If so, the ratio of expression of the two X chromosomes should be on average 1:1. As discussed in Chapter 5, certain situations can produce a skewed X-chromosome inactivationratio (i.e., significantly deviating from the expected 1:1 ratio). If a female heterozygote happens to have a significantly skewed inactivation ratio such that a larger proportion of the X chromosome containing the normal allele is inactivated, she will tend to exhibit clinical findings beyond what is usually seen. The higher the degree of skewing that occurs, the closer the phenotype approaches that of an affected male. Duchenne muscular dystrophy (DMD) is an X-linked muscle disease. Boys with this condition show a progressive degeneration of muscle that leads to weakness (Figure 6-31). Worsening weakness leads to problems with ambulation, then respiration, and ultimately an early death with an average life expectancy that is currently around 20 years old. One of the hallmark clues to the diagnosis is a marked elevation in a serum enzyme known as creatine phosphokinase (CPK or CK). Elevations of this enzyme correlate with the degree of muscle deterioration. Many men with DMD will also have a dilated cardiomyopathy; the heart being an additional muscle involved. Duchenne dystrophy has been known for decades as exhibiting X-linked recessive inheritance. Traditionally it has been characterized as being asymptomatic or non-expressing in carrier females. Clinical experience in following these families over time has, however, shown that partial expression can occur in carrier females. Some may exhibit a cardiomyopathy with middle age onset. Others may have milder elevations in their CK levels with a few even demonstrating muscle weakness or wasting. The partial expression of signs and symptoms in female carriers of DMD is explained, at least in part, by skewed X-inactivation.
Figure 6-31. (a) Young boy with Duchenne muscular dystrophy. Note muscle wasting and hypertrophied calves. (b) Muscle biopsy of Duchenne muscular dystrophy. Muscle shows “dystrophic” changes with variable fiber size and staining. Centrally located nuclei are also seen.
A few other important principles of sex-linked inheritance are worth mentioning here. When reviewing a pedigree in attempting to define the likely mode of inheritance, one of the first things to note is the pattern of transmission related to the sexes of the parents and offspring. Fathers must transmit their Y chromosome to their sons; thus, there is no male-to-male transmission of X-linked genes. Again, Y-linked inheritance is highly unlikely to present as a clinically significant issue. As with autosomal traits, X-linked conditions may be recessive or dominant depending on the phenotypic threshold that is defined. One important clinical question that arises has to do with the new occurrence of a condition that is known to be X-linked. In other words, what about the situation of a male in the family affected with an X-linked condition in which there are no other affected individuals? Where did the mutation originate? One obvious possibility is that the mutation is a new event in the egg responsible for his conception.
Alternatively, the mutation could be present in the mother and she is asymptomatic. In this case the mutation could have started with her, or even in the generation(s) before her. Empiric data have shown that in the event of a new X-linked mutation in an affected male in the family, there is a 2/3 chance that the mother is a carrier, and, of course, a 1/3 chance that the affected male represents a new (sporadic) mutation.
X-linked Recessive (XLR) Inheritance
1. Classic characteristics. As noted above, the nature of X-linked conditions is such that expression in females typically is more severe than for males. There are, however, a few notable exceptions as will be discussed in the Clinical Correlation section. X-linked recessive conditions, then, are those in which the threshold for expression is such that the condition is rarely expressed clinically in heterozygous females, but full expression is seen in males; i.e., only males are affected.
2. Recurrence risks. In the usual mating between a heterozygous carrier female and a normal male, the risks for offspring are as follows:
• 25% chance affected male
• 25% chance unaffected male
• 25% chance carrier female (unaffected)
• 25% chance noncarrier female (unaffected)
• Total risk for an affected child: 25%
Another type of mating that might occur is between an affected male and a homozygous normal female. The risks for offspring of this mating are as follows:
• All males unaffected
• All females obligate carriers
• Total: 50% chance normal male, 50% chance carrier female
3. Special inheritance considerations. A classic characteristic of X-linked recessive disorders is that females are unaffected. There are, however, special circumstances under which a female may actually express such a condition. There are several known reasons for the observation of the rare expression of X-linked recessive traits in females.
• If the abnormal allele is common enough in the population, female homozygosity for an X-linked mutation may occur.
• Female hemizygosity of an X-linked mutation can occur in women with Turner syndrome.
• In the event of an X-chromosome/autosome translocation with resulting deleted X-chromosome material, there is preferential inactivation of the normal X. This results in the woman being hemizygous for the deleted region.
• If there is random significant skewed Lyonization of the X bearing a normal allele, a female heterozygote may show variable degrees of expression that correlates with the relative proportion of normal X’s expressed.
• A female may appear to be affected due to an autosomal phenocopy of the X-linked disorder (locus heterogeneity).
X-linked Dominant (XLD) Inheritance
1. Classic characteristics. In X-linked dominant conditions, males and females can both be affected. Within kindreds, the number of affected individuals should be equal between the two sexes. Typically the clinical expression is more consistent and severe in hemizygous males than in heterozygous females. In heterozygous females, the variability of expression is often quite broad. Depending on the condition, the severity of expression in males can be that of lethality (incompatible with postnatal life). In these kindreds no affected males are seen, but a disproportionate number of live-born females are seen. For example, Aicardi syndrome (Figure 6-32) is an X-linked dominant condition. The condition demonstrates male lethality. Affected females present with no dysmorphic features, marked neurodevelopmental delays, and difficult to control seizures. Evaluation of these girls shows agenesis of the corpus callosum and markedly abnormal ocular findings. Reported ocular findings include optic atrophy, optic nerve coloboma, chorioretinopathy, chorioretinal lacunae, retinal detachment, cataract, and nystagmus.
Figure 6-32. Young girl with Aicardi syndrome. She has severe developmental delays and difficult to control seizures. (a) Normal facies (i.e., not dysmorphic). (b) Retinal photograph showing multiple retinal abnormalities. (c) Sagittal MRI of brain showing agenesis of the corpus callosum.
2. Recurrence risks. X-linked dominant inheritance segregates according to the X chromosome with dominant (single abnormal allele) inheritance.
In the mating of an affected male with a normal female, assuming no male lethality:
• All daughters will be affected.
• All sons will be normal.
In the mating of an affected female with a normal male:
• Each daughter and each son has a 50% chance of being affected.
3. Special inheritance considerations. The distinction between dominant and recessive inheritance patterns may not always be clear. This is particularly true for X-linked conditions. In some X-linked conditions, mild expression in carrier females is common, and is sometimes referred to as “semi-dominance.” Hypohydrotic ectodermal dysplasia is a disorder of the derivatives of the embryonic ectoderm (skin, hair, nails and teeth). The condition is genetically heterogeneous, the most common form being due to mutations in an X-linked gene, EDA1. The condition is typically classified as an X-linked recessive disorder. Males with this condition exhibit full expression with poor growth and function of the affected tissues. If female carriers of this condition are carefully evaluated, subtle findings of this condition are often detected. They may be noted to sweat less than expected (somewhat subjective) and often have dental abnormalities. This makes the classification of the condition as XLR debatable. It might better be described as XLD (X-linked dominant), but it is not at all clear where to “draw the line”!
Part 3: Clinical Correlation
Craniofrontonasal dysplasia (CFND) is a disorder primarily seen in females. The phenotype is quite dramatic (Figure 6-33). The fully expressed phenotype is pleiotropic with multiple somatic manifestations. As the name of the condition implies, the craniofacial features are a key set of findings. Affected females have a short skull base (brachycephaly), premature fusion (synostosis) of the coronal cranial sutures, extremely wide-spaced eyes (hypertelorism), a protruberant forehead (frontal bossing), a “widow’s peak” configuration to the frontal hairline, down-slanting angle to the eyes, and a wide or bifid nasal tip. Other reported anomalies include oro-facial clefting and peripheral skeletal changes (short neck, sloping shoulders, Sprengel anomaly, brachydactyly and/or syndactyly [short and/or fused digits]). Patients with CFND have normal intelligence.
Figure 6-33. (a – c) Adolescent female with craniofrontonasal dysplasia demonstrating classic facial and digital changes. This young lady has a documented EPHB1 mutation. (d) Patient and her affected mother.
Although there may be genetic heterogeneity for this condition, most of the cases can be shown to be due to mutations in a gene designated as Ephrin-B (EPHB1). This gene is a member of the Eph family of receptor protein-tyrosine kinases. The gene is on the X chromosome at location Xq21.
A fascinating aspect of this condition is that even though this is due to an X-linked gene, males who have a (hemizygous) mutation of the Ephrin-B gene have a phenotype that is actually milder than that seen in females. Indeed, males with EPHB1 mutations may have no somatic features whatsoever. Some may have subtle facial features similar to, but milder than, females. Interestingly, there are a few features seen in males that do not occur in the females, such as short stature, pectus excavatum, anomalies of the clavicles, minor genital anomalies, and diaphragmatic hernias.
The nature of the X-linked basis of this is often evident on review of the pedigrees of some larger families (Figure 6-34). Apparently unaffected males who do have an EPHB1 mutation will show the expected outcome of having all of their daughters affected and none of their sons affected.
Figure 6-34. Pedigree of a kindred with craniofrontonasal dysplasia. Note that individual II.1 is an obligate hemizygous carrier. He has no discernable phenotypic features. The outcomes in his children are completely reflective of the gene being X-linked. All of his daughters and none of his sons were affected.
To date, no known explanation has been put forth to explain this unique pattern of inheritance of X-linked dominant inheritance with female predominant expression.
Chapter 6 Board-Format Practice Questions
1. In chickens there is a known gene that determines the type of tail and neck plumage. Cock-feathering is more long and curved, while hen-feathering is more short and rounded. This gene is not on a sex chromosome. The genotype–phenotype relationship is described below:
In this example, the inheritance that most likely exhibits cock feathering is
A. recessive with sex-limited phenotype.
B. dominant with sex-limited phenotype.
C. semi-dominance.
D. co-dominance.
E. mitochondrial.
2. Parental age effects on the incidence of genetic disorders include
A. lower incidence of genetic disorders in the offspring of older parents.
B. increased incidence of chromosome aneuploidy with advanced paternal age.
C. increased incidence of single gene mutations with advanced maternal age.
D. increased incidence of non-disjunction with advanced maternal age.
E. decreased incidence of transcription errors with advanced paternal age.
3. Patients with Smith-Lemli-Opitz (SLO) syndrome exhibit many clinical features including mental retardation, an unusual facial appearance, genital abnormalities, and syndactyly (fusion of the digits). Patients with SLO tend to be similar in their features. SLO is an autosomal recessive condition. All heterozygotes will have expression. There is only one gene known to be responsible for this condition. This condition shows
A. highly variable phenotype.
B. locus heterogeneity.
C. pleiotropism.
D. thresholding.
E. Lyonization.
4. The one feature that is most helpful in determining in a pedigree if a particular condition, which was occurring in multiple generations of a family is likely to be X-linked dominant rather than autosomal dominant is
A females affected more severely than males.
B. lack of female-to-female transmission.
C. more males affected than females.
D. all daughters of affected males are affected.
E. condition not appearing in any descendants of affected males.
5. Although there are exceptions to every rule, certain generalizations do apply to inheritance patterns. When comparing disorders that have dominant inheritance compared to those that have recessive inheritance, it is generally true that
A. dominant conditions occur less frequently.
B. dominant conditions show more variability of expression.
C. incomplete penetrance is a characteristic of recessive disorders.
D. pleiotropism only occurs in dominant conditions.
E. dominant conditions are often holandric in nature.