Thompson & Thompson Genetics in Medicine, 8th Edition

CHAPTER 11. The Molecular Basis of Genetic Disease

General Principles and Lessons from the Hemoglobinopathies

The term molecular disease, introduced over six decades ago, refers to disorders in which the primary disease-causing event is an alteration, either inherited or acquired, affecting a gene(s), its structure, and/or its expression. In this chapter, we first outline the basic genetic and biochemical mechanisms underlying monogenic or single-gene disorders. We then illustrate them in the context of their molecular and clinical consequences using inherited diseases of hemoglobin—the hemoglobinopathies—as examples. This overview of mechanisms is expanded in Chapter 12 to include other genetic diseases that illustrate additional principles of genetics in medicine.

A genetic disease occurs when an alteration in the DNA of an essential gene changes the amount or function, or both, of the gene products—typically messenger RNA (mRNA) and protein but occasionally specific noncoding RNAs (ncRNAs) with structural or regulatory functions. Although almost all known single-gene disorders result from mutations that affect the function of a protein, a few exceptions to this generalization are now known. These exceptions are diseases due to mutations in ncRNA genes, including microRNA (miRNA) genes that regulate specific target genes, and mitochondrial genes that encode transfer RNAs (tRNAs; see Chapter 12). It is essential to understand genetic disease at the molecular and biochemical levels, because this knowledge is the foundation of rational therapy. In this chapter, we restrict our attention to diseases caused by defects in protein-coding genes; the study of phenotype at the level of proteins, biochemistry, and metabolism constitutes the discipline of biochemical genetics.

By 2014, the online version of Mendelian Inheritance in Man listed over 5500 phenotypes for which the molecular basis is known, largely phenotypes with autosomal and X-linked inheritance. Although it is impressive that the basic molecular defect has been found in so many disorders, it is sobering to realize that the pathophysiology is not entirely understood for any genetic disease. Sickle cell disease(Case 42), discussed later in this chapter, is among the best characterized of all inherited disorders, but even here, knowledge is incomplete—despite its being the first molecular disease to be recognized, more than 65 years ago.

The Effect of Mutation on Protein Function

Mutations involving protein-coding genes have been found to cause disease through one of four different effects on protein function (Fig. 11-1). The most common effect by far is a loss of function of the mutant protein. Many important conditions arise, however, from other mechanisms: a gain of function, the acquisition of a novel property by the mutant protein, or the expression of a gene at the wrong time (heterochronic expression) and/or in the wrong place (ectopic expression).


FIGURE 11-1 A general outline of the mechanisms by which disease-causing mutations produce disease. Mutations in the coding region result in structurally abnormal proteins that have a loss or gain of function or a novel property that causes disease. Mutations in noncoding sequences are of two general types: those that alter the stability or splicing of the messenger RNA (mRNA) and those that disrupt regulatory elements or change gene dosage. Mutations in regulatory elements alter the abundance of the mRNA or the time or cell type in which the gene is expressed. Mutations in either the coding region or regulatory domains can decrease the amount of the protein produced. HPFH, Hereditary persistence of fetal hemoglobin.

Loss-of-Function Mutations

The loss of function of a gene may result from alteration of its coding, regulatory, or other critical sequences due to nucleotide substitutions, deletions, insertions, or rearrangements. A loss of function due to deletion, leading to a reduction in gene dosage, is exemplified by the α-thalassemias(Case 44), which are most commonly due to deletion of α-globin genes (see later discussion); by chromosome-loss diseases (Case 27), such as monosomies like Turner syndrome (see Chapter 6(Case 47); and by acquired somatic mutations—often deletions—that occur in tumor-suppressor genes in many cancers, such as retinoblastoma(Case 39) (see Chapter 15). Many other types of mutations can also lead to a complete loss of function, and all are illustrated by the β-thalassemias(Case 44) (see later discussion), a group of hemoglobinopathies that result from a reduction in the abundance of β-globin, one of the major adult hemoglobin proteins in red blood cells.

The severity of a disease due to loss-of-function mutations generally correlates with the amount of function lost. In many instances, the retention of even a small percent of residual function by the mutant protein greatly reduces the severity of the disease.

Gain-of-Function Mutations

Mutations may also enhance one or more of the normal functions of a protein; in a biological system, however, more is not necessarily better, and disease may result. It is critical to recognize when a disease is due to a gain-of-function mutation because the treatment must necessarily differ from disorders due to other mechanisms, such as loss-of-function mutations. Gain-of-function mutations fall into two broad classes:

• Mutations that increase the production of a normal protein. Some mutations cause disease by increasing the synthesis of a normal protein in cells in which the protein is normally present. The most common mutations of this type are due to increased gene dosage, which generally results from duplication of part or all of a chromosome. As discussed in Chapter 6, the classic example is trisomy 21 (Down syndrome), which is due to the presence of three copies of chromosome 21. Other important diseases arise from the increased dosage of single genes, including one form of familial Alzheimer disease due to a duplication of the amyloid precursor protein (βAPP) gene (see Chapter 12), and the peripheral nerve degeneration Charcot-Marie-Tooth disease type 1A(Case 8), which generally results from duplication of only one gene, the gene for peripheral myelin protein 22 (PMP22).

• Mutations that enhance one normal function of a protein. Rarely, a mutation in the coding region may increase the ability of each protein molecule to perform one or more of its normal functions, even though this increase is detrimental to the overall physiological role of the protein. For example, the missense mutation that creates hemoglobin Kempsey locks hemoglobin into its high oxygen affinity state, thereby reducing oxygen delivery to tissues. Another example of this mechanism occurs in the form of short stature called achondroplasia(Case 2).

Novel Property Mutations

In a few diseases, a change in the amino acid sequence confers a novel property on the protein, without necessarily altering its normal functions. The classic example of this mechanism is sickle cell disease(Case 42), which, as we will see later in this chapter, is due to an amino acid substitution that has no effect on the ability of sickle hemoglobin to transport oxygen. Rather, unlike normal hemoglobin, sickle hemoglobin chains aggregate when they are deoxygenated and form abnormal polymeric fibers that deform red blood cells. That novel property mutations are infrequent is not surprising, because most amino acid substitutions are either neutral or detrimental to the function or stability of a protein that has been finely tuned by evolution.

Mutations Associated with Heterochronic or Ectopic Gene Expression

An important class of mutations includes those that lead to inappropriate expression of the gene at an abnormal time or place. These mutations occur in the regulatory regions of the gene. Thus cancer is frequently due to the abnormal expression of a gene that normally promotes cell proliferation—an oncogene—in cells in which the gene is not normally expressed (see Chapter 15). Some mutations in hemoglobin regulatory elements lead to the continued expression in adults of the γ-globin gene, which is normally expressed at high levels only in fetal life. Such γ-globin gene mutations cause a benign phenotype called the hereditary persistence of fetal hemoglobin (Hb F), as we explore later in this chapter.

How Mutations Disrupt the Formation of Biologically Normal Proteins

Disruptions of the normal functions of a protein that result from the various types of mutations outlined earlier can be well exemplified by the broad range of diseases due to mutations in the globin genes, as we will explore in the second part of this chapter. To form a biologically active protein (such as the hemoglobin molecule), information must be transcribed from the nucleotide sequence of the gene to the mRNA and then translated into the polypeptide, which then undergoes progressive stages of maturation (see Chapter 3). Mutations can disrupt any of these steps (Table 11-1). As we shall see next, abnormalities in five of these stages are illustrated by various hemoglobinopathies; the others are exemplified by diseases to be presented in Chapter 12.

TABLE 11-1

The Eight Steps at Which Mutations Can Disrupt the Production of a Normal Protein


LDL, Low-density lipoprotein; mRNA, messenger RNA.

The Relationship between Genotype and Phenotype in Genetic Disease

Variation in the clinical phenotype observed in an inherited disease may have any of three genetic explanations, namely:

• allelic heterogeneity

• locus heterogeneity, or

• the effect of modifier genes

Each of these types can be illustrated by mutations in the α-globin or β-globin genes (Table 11-2).

TABLE 11-2

Types of Heterogeneity Associated with Genetic Disease

Type of Heterogeneity



Genetic heterogeneity


 Allelic heterogeneity

The occurrence of more than one allele at a locus



 Locus heterogeneity

The association of more than one locus with a clinical phenotype

Thalassemia can result from mutations in either the α-globin or β-globin genes

Clinical or phenotypic heterogeneity

The association of more than one phenotype with mutations at a single locus

Sickle cell disease and β-thalassemia each result from distinct β-globin gene mutations

Allelic Heterogeneity

Genetic heterogeneity is most commonly due to the presence of multiple alleles at a single locus, a situation referred to as allelic heterogeneity (see Chapter 7 and Table 11-1). In many instances, there is a clear genotype-phenotype correlation between a specific allele and a specific phenotype. The most common explanation for the effect of allelic heterogeneity on the clinical phenotype is that alleles that confer more residual function on the mutant protein are often associated with a milder form of the principal phenotype associated with the disease. In some instances, however, alleles that confer some residual protein function are associated with only one or a subset of the complete set of phenotypes seen with a missing or completely nonfunctional allele (frequently termed a null allele). As we will explore more fully in Chapter 12, this situation prevails with certain variants of the cystic fibrosis gene, CFTR; these variants lead to a phenotypically different condition, congenital absence of the vas deferens, but not to the other manifestations of cystic fibrosis.

A second explanation for allele-based variation in phenotype is that the variation may reflect the specific property of the protein that is most perturbed by the mutation. This situation is well illustrated by Hb Kempsey, a β-globin allele that maintains the hemoglobin in a high oxygen affinity structure, causing polycythemia because the reduced peripheral delivery of oxygen is misinterpreted by the hematopoietic system as being due to an inadequate production of red blood cells.

The biochemical and clinical consequences of a specific mutation in a protein are often unpredictable. Thus no one would have foreseen that the β-globin allele associated with sickle cell disease would lead to the formation of globin polymers that deform erythrocytes to a sickle cell shape (see later in this chapter). Sickle cell disease is highly unusual in that it results only from a single specific mutation—the Glu6Val substitution in the β-globin chain—whereas most disease phenotypes can arise from any of a number or many substitutions, usually loss-of-function mutations, in the affected protein.

Locus Heterogeneity

Genetic heterogeneity also arises when mutations in more than one locus can result in a specific clinical condition, a situation termed locus heterogeneity (see Chapter 7). This phenomenon is illustrated by the finding that thalassemia can result from mutations in either the β-globin or α-globin chain (see Table 11-2). Once locus heterogeneity has been documented, careful comparison of the phenotype associated with each gene sometimes reveals that the phenotype is not as homogeneous as initially believed.

Modifier Genes

Sometimes even the most robust genotype-phenotype relationships are found not to hold for a specific patient. Such phenotypic variation can, in principle, be ascribed to environmental factors or to the action of other genes, termed modifier genes (see Chapter 8). To date, only a few modifier genes for human monogenic disorders have been identified, although one anticipates that there will be numerous examples as our understanding of the basis for disease increases. One example described later in this chapter is seen in β-thalassemia homozygotes (carrying mutations at the β-globin locus) who also inherit an α-thalassemia variant at the α-globin locus.

The Hemoglobins

To illustrate in greater detail the concepts introduced in the first section of this chapter, we now turn to disorders of human hemoglobins—referred to as hemoglobinopathies—the most common single-gene diseases in humans. These disorders cause substantial morbidity, and the World Health Organization estimates that more than 5% of the world's population are carriers of genetic variants for clinically important disorders of hemoglobin. They are also important because their molecular and biochemical pathology is better understood than perhaps that of any other group of genetic diseases. Before the hemoglobinopathies are discussed in depth, it is important to briefly introduce the normal aspects of the globin genes and hemoglobin biology.

Structure and Function of Hemoglobin

Hemoglobin is the oxygen carrier in vertebrate red blood cells. Each hemoglobin molecule consists of four subunits: two α-globin chains and two β- (or β-like) globin chains. Each subunit is composed of a polypeptide chain, globin, and a prosthetic group, heme, which is an iron-containing pigment that combines with oxygen to give the molecule its oxygen-transporting ability (Fig. 11-2). The predominant adult human hemoglobin, Hb A, has an α2β2 structure in which the four chains are folded and fitted together to form a globular tetramer.


FIGURE 11-2 The structure of a hemoglobin subunit. Each subunit has eight helical regions, designated A to H. The two most conserved amino acids are shown: His92, the histidine to which the iron of heme is covalently linked; and Phe42, the phenylalanine that wedges the porphyrin ring of heme into the heme “pocket” of the folded protein. See discussion of Hb Hammersmith and Hb Hyde Park, which have substitutions for Phe42 and His92, respectively, in the β-globin molecule.

As with all proteins that have been strongly conserved throughout evolution, the tertiary structure of globins is constant; virtually all globins have seven or eight helical regions (depending on the chain) (see Fig. 11-2). Mutations that disrupt this tertiary structure invariably have pathological consequences. In addition, mutations that substitute a highly conserved amino acid or that replace one of the nonpolar residues, which form the hydrophobic shell that excludes water from the interior of the molecule, are likely to cause a hemoglobinopathy (see Fig. 11-2). Like all proteins, globin has sensitive areas, in which mutations cannot occur without affecting function, and insensitive areas, in which variations are more freely tolerated.

The Globin Genes

In addition to Hb A, with its α2β2 structure, there are five other normal human hemoglobins, each of which has a tetrameric structure like that of Hb A in consisting of two α or α-like chains and two non-α chains (Fig. 11-3A). The genes for the α and α-like chains are clustered in a tandem arrangement on chromosome 16. Note that there are two identical α-globin genes, designated α1 and α2, on each homologue. The β- and β-like globin genes, located on chromosome 11, are close family members that, as described in Chapter 3, undoubtedly arose from a common ancestral gene (see Fig. 11-3A). Illustrating this close evolutionary relationship, the β- and δ-globins differ in only 10 of their 146 amino acids.


FIGURE 11-3 Organization of the human globin genes and hemoglobins produced in each stage of human development. A, The α-like genes are on chromosome 16, the β-like genes on chromosome 11.The curved arrows refer to the switches in gene expression during development. B, Development of erythropoiesis in the human fetus and infant. Types of cells responsible for hemoglobin synthesis, organs involved, and types of globin chain synthesized at successive stages are shown. SeeSources & Acknowledgments.

Developmental Expression of Globin Genes and Globin Switching

The expression of the various globin genes changes during development, a process referred to as globin switching (see Fig. 11-3B). Note that the genes in the α- and β-globin clusters are arranged in the same transcriptional orientation and, remarkably, the genes in each cluster are situated in the same order in which they are expressed during development. The temporal switches of globin synthesis are accompanied by changes in the principal site of erythropoiesis (see Fig. 11-3B). Thus the three embryonic globins are made in the yolk sac from the third to eighth weeks of gestation, but at approximately the fifth week, hematopoiesis begins to move from the yolk sac to the fetal liver. Hb F (α2γ2), the predominant hemoglobin throughout fetal life, constitutes approximately 70% of total hemoglobin at birth. In adults, however, Hb F represents less than a few percent of the total hemoglobin, although this can vary from less than 1% to approximately 5% in different individuals.

β-chain synthesis becomes significant near the time of birth, and by 3 months of age, almost all hemoglobin is of the adult type, Hb A (α2β2) (see Fig. 11-3B). In diseases due to mutations that decrease the abundance of β-globin, such as β-thalassemia (see later section), strategies to increase the normally small amount of γ-globin (and therefore of Hb F (α2γ2)) produced in adults are proving to be successful in ameliorating the disorder (see Chapter 13).

The Developmental Regulation of β-Globin Gene Expression: The Locus Control Region

Elucidation of the mechanisms that control the expression of the globin genes has provided insight into both normal and pathological biological processes. The expression of the β-globin gene is only partly controlled by the promoter and two enhancers in the immediate flanking DNA (see Chapter 3). A critical requirement for additional regulatory elements was first suggested by the identification of a unique group of patients who had no gene expression from any of the genes in the β-globin cluster, even though the genes themselves (including their individual regulatory elements) were intact. These informative patients were found to have large deletions upstream of the β-globin complex, deletions that removed an approximately 20-kb domain called the locus control region (LCR), which begins approximately 6 kb upstream of the ε-globin gene (Fig. 11-4). Although the resulting disease, εγδβ-thalassemia, is described later in this chapter, these patients demonstrate that the LCR is required for the expression of all the genes in the β-globin cluster.


FIGURE 11-4 The β-globin locus control region (LCR). Each of the five regions of open chromatin (arrows) contains several consensus binding sites for both erythroid-specific and ubiquitous transcription factors. The precise mechanism by which the LCR regulates gene expression is unknown. Also shown is a deletion of the LCR that has led to εγδβ-thalassemia, which is discussed in the text. SeeSources & Acknowledgments.

The LCR is defined by five so-called DNase I hypersensitive sites (see Fig. 11-4), genomic regions that are unusually open to certain proteins (such as the enzyme DNase I) that are used experimentally to reveal potential regulatory sites. Within the context of the epigenetic packaging of chromatin (see Chapter 3), these sites configure an open chromatin state at the locus in erythroid cells, the role of which is to maintain an open chromatin configuration at the locus, a configuration that gives transcription factors access to the regulatory elements that mediate the expression of each of the β-globin genes (see Chapter 3). The LCR, along with its associated DNA-binding proteins, interacts with the genes of the β-globin locus to form a nuclear domain called the active chromatin hub, where β-globin gene expression takes place. The sequential switching of gene expression that occurs among the five members of the β-globin gene complex during development results from the sequential association of the active chromatin hub with the different genes in the cluster as the hub moves from the most proximal gene in the complex (the ε-globin gene in embryos) to the most distal (the δ- and β-globin genes in adults).

The clinical significance of the LCR is threefold. First, as mentioned, patients with deletions of the LCR fail to express the genes of the β-globin cluster. Second, components of the LCR are likely to be essential in gene therapy (see Chapter 13) for disorders of the β-globin cluster, so that the therapeutic normal copy of the gene in question is expressed at the correct time in life and in the appropriate tissue. And third, knowledge of the molecular mechanisms that underlie globin switching may make it feasible to up-regulate the expression of the γ-globin gene in patients with β-thalassemia (who have mutations only in the β-globin gene), because Hb F (α2γ2) is an effective oxygen carrier in adults who lack Hb A (α2β2) (see Chapter 13).

Gene Dosage, Developmental Expression of the Globins, and Clinical Disease

The differences both in the gene dosage of the α- and β-globins (four α-globin and two β-globin genes per diploid genome), and in their patterns of expression during development, are important to an understanding of the pathogenesis of many hemoglobinopathies. Mutations in the β-globin gene are more likely to cause disease than are α-chain mutations because a single β-globin gene mutation affects 50% of the β chains, whereas a single α-chain mutation affects only 25% of the α chains. On the other hand, β-globin mutations have no prenatal consequences because γ-globin is the major β-like globin before birth, with Hb F constituting 75% of the total hemoglobin at term (see Fig. 11-3B). In contrast, because α chains are the only α-like components of hemoglobins 6 weeks after conception, α-globin mutations cause severe disease in both fetal and postnatal life.

The Hemoglobinopathies

Hereditary disorders of hemoglobin can be divided into the following three broad groups, which in some instances overlap:

• Structural variants, which alter the amino acid sequence of the globin polypeptide, altering properties such as its ability to transport oxygen, or reducing its stability. ExampleSickle cell disease(Case 42), due to a mutation that makes deoxygenated β-globin relatively insoluble, changing the shape of the red cell (Fig. 11-5).


FIGURE 11-5 Scanning electron micrographs of red cells from a patient with sickle cell disease. A, Oxygenated cells are round and full. B, The classic sickle cell shape is produced only when the cells are in the deoxygenated state. SeeSources & Acknowledgments.

• Thalassemias, which are diseases that result from the decreased abundance of one or more of the globin chains (Case 44). The decrease can result from decreased production of a globin chain or, less commonly, from a structural variant that destabilizes the chain. The resulting imbalance in the ratio of the α:β chains underlies the pathophysiology of these conditions. Example: promoter mutations that decrease expression of the β-globin mRNA to cause β-thalassemia.

• Hereditary persistence of fetal hemoglobin, a group of clinically benign conditions that impair the perinatal switch from γ-globin to β-globin synthesis. Example: a deletion, found in African Americans, that removes both the δ- and β-globin genes but leads to continued postnatal expression of the γ-globin genes, to produce Hb F, which is an effective oxygen transporter (see Fig. 11-3).

Hemoglobin Structural Variants

Most variant hemoglobins result from point mutations in one of the globin structural genes. More than 400 abnormal hemoglobins have been described, and approximately half of these are clinically significant. The hemoglobin structural variants can be separated into the following three classes, depending on the clinical phenotype (Table 11-3):

• Variants that cause hemolytic anemia, most commonly because they make the hemoglobin tetramer unstable.

• Variants with altered oxygen transport, due to increased or decreased oxygen affinity or to the formation of methemoglobin, a form of globin incapable of reversible oxygenation.

• Variants due to mutations in the coding region that cause thalassemia because they reduce the abundance of a globin polypeptide. Most of these mutations impair the rate of synthesis of the mRNA or otherwise affect the level of the encoded protein.

TABLE 11-3

The Major Classes of Hemoglobin Structural Variants


*Hemoglobin variants are often named after the home town of the first patient described.

Additional β-chain structural variants that cause β-thalassemia are depicted in Table 11-5.

AD, Autosomal dominant; AR, autosomal recessive; Hb M, methemoglobin; see text.

Hemolytic Anemias

Hemoglobins with Novel Physical Properties: Sickle Cell Disease.

Sickle cell hemoglobin is of great clinical importance in many parts of the world. The disease results from a single nucleotide substitution that changes the codon of the sixth amino acid of β-globin from glutamic acid to valine (GAG → GTG: Glu6Val; see Table 11-3). Homozygosity for this mutation is the cause of sickle cell disease(Case 42). The disease has a characteristic geographical distribution, occurring most frequently in equatorial Africa and less commonly in the Mediterranean area and India and in countries to which people from these regions have migrated. Approximately 1 in 600 African Americans is born with this disease, which may be fatal in early childhood, although longer survival is becoming more common.

Clinical Features.

Sickle cell disease is a severe autosomal recessive hemolytic condition characterized by a tendency of the red blood cells to become grossly abnormal in shape (i.e., take on a sickle shape) under conditions of low oxygen tension (see Fig. 11-5). Heterozygotes, who are said to have sickle cell trait, are generally clinically normal, but their red cells can sickle when they are subjected to very low oxygen pressure in vitro. Occasions when this occurs are uncommon, although heterozygotes appear to be at risk for splenic infarction, especially at high altitude (for example in airplanes with reduced cabin pressure) or when exerting themselves to extreme levels in athletic competition. The heterozygous state is present in approximately 8% of African Americans, but in areas where the sickle cell allele (βS) frequency is high (e.g., West Central Africa), up to 25% of the newborn population are heterozygotes.

The Molecular Pathology of Hb S.

Nearly 60 years ago, Ingram discovered that the abnormality in sickle cell hemoglobin was a replacement of one of the 146 amino acids in the β chain of the hemoglobin molecule. All the clinical manifestations of sickle cell hemoglobin are consequences of this single change in the β-globin gene. Ingram's discovery was the first demonstration in any organism that a mutation in a structural gene could cause an amino acid substitution in the corresponding protein. Because the substitution is in the β-globin chain, the formula for sickle cell hemoglobin is written as α2β2S or, more precisely, α2Aβ2S. A heterozygote has a mixture of the two types of hemoglobin, A and S, summarized as α2Aβ2A2Aβ2S, as well as a hybrid hemoglobin tetramer, written as α2AβAβS. Strong evidence indicates that the sickle mutation arose in West Africa but that it also occurred independently elsewhere. The βS allele has attained high frequency in malarial areas of the world because it confers protection against malaria in heterozygotes (see Chapter 9).

Sickling and Its Consequences.

The molecular and cellular pathology of sickle cell disease is summarized in Figure 11-6. Hemoglobin molecules containing the mutant β-globin subunits are normal in their ability to perform their principal function of binding oxygen (provided they have not polymerized, as described next), but in deoxygenated blood, they are only one fifth as soluble as normal hemoglobin. Under conditions of low oxygen tension, this relative insolubility of deoxyhemoglobin S causes the sickle hemoglobin molecules to aggregate in the form of rod-shaped polymers or fibers (see Fig. 11-5). These molecular rods distort the α2β2Serythrocytes to a sickle shape that prevents them from squeezing single file through capillaries, as do normal red cells, thereby blocking blood flow and causing local ischemia. They may also cause disruption of the red cell membrane (hemolysis) and release of free hemoglobin, which can have deleterious effects on the availability of vasodilators, such as nitric oxide, thereby exacerbating the ischemia.


FIGURE 11-6 The pathogenesis of sickle cell disease. SeeSources & Acknowledgments.

Modifier Genes Determine the Clinical Severity of Sickle Cell Disease.

It has long been known that a strong modifier of the clinical severity of sickle cell disease is the patient's level of Hb F (α2γ2), higher levels being associated with less morbidity and lower mortality. The physiological basis of the ameliorating effect of Hb F is clear: Hb F is a perfectly adequate oxygen carrier in postnatal life and also inhibits the polymerization of deoxyhemoglobin S.

Until recently, however, it was not certain whether the variation in Hb F expression was heritable. Genome-wide association studies (GWAS) (see Chapter 10) have demonstrated that single nucleotide polymorphisms (SNPs) at three loci—the γ-globin gene and two genes that encode transcription factors, BCL11A and MYB—account for 40% to 50% of the variation in the levels of Hb F in patients with sickle cell disease. Moreover, the Hb F–associated SNPs are also associated with the painful clinical episodes thought to be due to capillary occlusion caused by sickled red cells (Fig. 11-6).

The genetically driven variations in the level of Hb F are also associated with variation in the clinical severity of β-thalassemia (discussed later) because the reduced abundance of β-globin (and thus of Hb A [α2β2]) in that disease is partly alleviated by higher levels of γ-globin and thus of Hb F (α2γ2). The discovery of these genetic modifiers of Hb F abundance not only explains much of the variation in the clinical severity of sickle cell disease and β-thalassemia, but it also highlights a general principle introduced in Chapter 8modifier genes can play a major role in determining the clinical and physiological severity of a single-gene disorder.

BCL11A, a Silencer of γ-Globin Gene Expression in Adult Erythroid Cells.

The identification of genetic modifiers of Hb F levels, particularly BCL11A, has great therapeutic potential. The product of the BCL11A gene is a transcription factor that normally silences γ-globin expression, thus shutting down Hb F production postnatally. Accordingly, drugs that suppress BCL11A activity postnatally, thereby increasing the expression of Hb F, might be of great benefit to patients with sickle cell disease and β-thalassemia (see Chapter 13), disorders that affect millions of individuals worldwide. Small molecule screening programs to identify potential drugs of this type are now underway in many laboratories.

Trisomy 13, MicroRNAs, and MYB, Another Silencer of γ-Globin Gene Expression.

The indication from GWAS that MYB is an important regulator of γ-globin expression has received further support from an unexpected direction, studies investigating the basis for the persistent increased postnatal expression of Hb F that is observed in patients with trisomy 13 (see Chapter 6). Two miRNAs, miR-15a and miR-16-1, directly target the 3′ untranslated region (UTR) of the MYB mRNA, thereby reducing MYB expression. The genes for these two miRNAs are located on chromosome 13; their extra dosage in trisomy 13 is predicted to reduce MYB expression below normal levels, thereby partly relaxing the postnatal suppression of γ-globin gene expression normally mediated by the MYB protein, and leading to increased expression of Hb F (Fig. 11-7).


FIGURE 11-7 A model demonstrating how elevations of microRNAs 15a and 16-1 in trisomy 13 can result in elevated fetal hemoglobin expression. Normally, the basal level of these microRNAs can moderate expression of targets such as the MYB gene during erythropoiesis. In the case of trisomy 13, elevated levels of these microRNAs results in additional down-regulation of MYB expression, which in turn results in a delayed switch from fetal to adult hemoglobin and persistent expression of fetal hemoglobin. SeeSources & Acknowledgments.

Unstable Hemoglobins.

The unstable hemoglobins are due largely to point mutations that cause denaturation of the hemoglobin tetramer in mature red blood cells. The denatured globin tetramers are insoluble and precipitate to form inclusions (Heinz bodies) that contribute to damage of the red cell membrane and cause the hemolysis of mature red blood cells in the vascular tree (Fig. 11-8, showing a Heinz body due to β-thalassemia).


FIGURE 11-8 Visualization of one pathological effect of the deficiency of β chains in β-thalassemia: the precipitation of the excess normal α chains to form a Heinz body in the red blood cell. Peripheral blood smear and Heinz body preparation. A-C, The peripheral smear (A) shows “bite” cells with pitted-out semicircular areas of the red blood cell membrane as a result of removal of Heinz bodies by macrophages in the spleen, causing premature destruction of the red cell. The Heinz body preparation (B) shows increased Heinz bodies in the same specimen when compared to a control (C). SeeSources & Acknowledgments.

The amino acid substitution in the unstable hemoglobin Hb Hammersmith (β-chain Phe42Ser; see Table 11-3) leads to denaturation of the tetramer and consequent hemolysis. This mutation is particularly notable because the substituted phenylalanine residue is one of the two amino acids that are conserved in all globins in nature (see Fig. 11-2). It is therefore not surprising that substitutions of this phenylalanine produce serious disease. In normal β-globin, the bulky phenylalanine wedges the heme into a “pocket” in the folded β-globin monomer. Its replacement by serine, a smaller residue, creates a gap that allows the heme to slip out of its pocket. In addition to its instability, Hb Hammersmith has a low oxygen affinity, which causes cyanosis in heterozygotes.

In contrast to mutations that destabilize the tetramer, other variants destabilize the globin monomer and never form the tetramer, causing chain imbalance and thalassemia (see following section).

Variants with Altered Oxygen Transport

Mutations that alter the ability of hemoglobin to transport oxygen, although rare, are of general interest because they illustrate how a mutation can impair one function of a protein (in this case, oxygen binding and release) and yet leave the other properties of the protein relatively intact. For example, the mutations that affect oxygen transport generally have little or no effect on hemoglobin stability.


Oxyhemoglobin is the form of hemoglobin that is capable of reversible oxygenation; its heme iron is in the reduced (or ferrous) state. The heme iron tends to oxidize spontaneously to the ferric form and the resulting molecule, referred to as methemoglobin, is incapable of reversible oxygenation. If significant amounts of methemoglobin accumulate in the blood, cyanosis results. Maintenance of the heme iron in the reduced state is the role of the enzyme methemoglobin reductase. In several mutant globins (either α or β), substitutions in the region of the heme pocket affect the heme-globin bond in a way that makes the iron resistant to the reductase. Although heterozygotes for these mutant hemoglobins are cyanotic, they are asymptomatic. The homozygous state is presumably lethal. One example of a β-chain methemoglobin is Hb Hyde Park (see Table 11-3), in which the conserved histidine (His92 in Fig. 11-2) to which heme is covalently bound has been replaced by tyrosine (His92Tyr).

Hemoglobins with Altered Oxygen Affinity.

Mutations that alter oxygen affinity demonstrate the importance of subunit interaction for the normal function of a multimeric protein such as hemoglobin. In the Hb A tetramer, the α:β interface has been highly conserved throughout evolution because it is subject to significant movement between the chains when the hemoglobin shifts from the oxygenated (relaxed) to the deoxygenated (tense) form of the molecule. Substitutions in residues at this interface, exemplified by the β-globin mutant Hb Kempsey (see Table 11-3), prevent the normal oxygen-related movement between the chains; the mutation “locks” the hemoglobin into the high oxygen affinity state, thus reducing oxygen delivery to tissues and causing polycythemia.

Thalassemia: An Imbalance of Globin-Chain Synthesis

The thalassemias (from the Greek thalassa, sea, and haema, blood) are collectively the most common human single-gene disorders in the world (Case 44). They are a heterogeneous group of diseases of hemoglobin synthesis in which mutations reduce the synthesis or stability of either the α-globin or β-globin chain to cause α-thalassemia or β-thalassemia, respectively. The resulting imbalance in the ratio of the α:β chains underlies the pathophysiology. The chain that is produced at the normal rate is in relative excess; in the absence of a complementary chain with which to form a tetramer, the excess normal chains eventually precipitate in the cell, damaging the membrane and leading to premature red blood cell destruction. The excess β or β-like chains are insoluble and precipitate in both red cell precursors (causing ineffective erythropoiesis) and in mature red cells (causing hemolysis) because they damage the cell membrane. The result is a lack of red cells (anemia) in which the red blood cells are both hypochromic (i.e., pale red cells) and microcytic (i.e., small red cells).

The name thalassemia was first used to signify that the disease was discovered in persons of Mediterranean origin. Both α-thalassemia and β-thalassemia, however, have a high frequency in many populations, although α-thalassemia is more prevalent and more widely distributed. The high frequency of thalassemia is due to the protective advantage against malaria that it confers on carriers, analogous to the heterozygote advantage of sickle cell hemoglobin carriers (see Chapter 9). There is a characteristic distribution of the thalassemias in a band around the Old World—in the Mediterranean, the Middle East, and parts of Africa, India, and Asia.

An important clinical consideration is that alleles for both types of thalassemia, as well as for structural hemoglobin abnormalities, not uncommonly coexist in an individual. As a result, clinically important interactions may occur among different alleles of the same globin gene or among mutant alleles of different globin genes.

The α-Thalassemias

Genetic disorders of α-globin production disrupt the formation of both fetal and adult hemoglobins (see Fig. 11-3) and therefore cause intrauterine as well as postnatal disease. In the absence of α-globin chains with which to associate, the chains from the β-globin cluster are free to form a homotetrameric hemoglobin. Hemoglobin with a γ4 composition is known as Hb Bart's, and the β4 tetramer is called Hb H. Because neither of these hemoglobins is capable of releasing oxygen to tissues under normal conditions, they are completely ineffective oxygen carriers. Consequently, infants with severe α-thalassemia and high levels of Hb Bart's (γ4) suffer severe intrauterine hypoxia and are born with massive generalized fluid accumulation, a condition called hydrops fetalis. In milder α-thalassemias, an anemia develops because of the gradual precipitation of the Hb H (β4) in the erythrocyte. The formation of Hb H inclusions in mature red cells and the removal of these inclusions by the spleen damages the cells, leading to their premature destruction.

Deletions of the α-Globin Genes.

The most common forms of α-thalassemia are the result of gene deletions. The high frequency of deletions in mutants of the α chain and not the β chain is due to the presence of the two identical α-globin genes on each chromosome 16 (see Fig. 11-3A); the intron sequences within the two α-globin genes are also similar. This arrangement of tandem homologous α-globin genes facilitates misalignment due to homologous pairing and subsequent recombination between the α1 gene domain on one chromosome and the corresponding α2 gene region on the other (Fig. 11-9). Evidence supporting this pathogenic mechanism is provided by reports of rare normal individuals with a triplicated α-globin gene complex. Deletions or other alterations of one, two, three, or all four copies of the α-globin genes cause a proportionately severe hematological abnormality (Table 11-4).


FIGURE 11-9 The probable mechanism underlying the most common form of α-thalassemia, which is due to deletions of one of the two α-globin genes on a chromosome 16. Misalignment, homologous pairing, and recombination between the α1 gene on one chromosome and the α2 gene on the homologous chromosome result in the deletion of one α-globin gene.

TABLE 11-4

Clinical States Associated with α-Thalassemia Genotypes


The α-thalassemia trait, caused by deletion of two of the four α-globin genes, is distributed throughout the world. However, the homozygous deletion type of α-thalassemia, involving all four copies of α-globin and leading to Hb Bart's (γ4) and hydrops fetalis, is largely restricted to Southeast Asia. In this population, the high frequency of hydrops fetalis due to α-thalassemia can be explained by the nature of the deletion responsible. Individuals with two normal and two mutant α-globin genes are said to have α-thalassemia trait, which can result from either of two genotypes (−−/αα or −α/−α), differing in whether or not the deletions are in cis or in trans. Heterozygosity for deletion of both copies of the α-globin gene in cis (−−/αα genotype) is relatively common among Southeast Asians, and offspring of two carriers of this deletion allele may consequently receive two −−/−− chromosomes. In other groups, however, α-thalassemia trait is usually the result of the trans −α/−α genotype, which cannot give rise to −−/−− offspring.

In addition to α-thalassemia mutations that result in deletion of the α-globin genes, mutations that delete only the LCR of the α-globin complex have also been found to cause α-thalassemia. In fact, similar to the observations discussed earlier with respect to the β-globin LCR, such deletions were critical for demonstrating the existence of this regulatory element at the α-globin locus.

Other Forms of α-Thalassemia.

In all the classes of α-thalassemia described earlier, deletions in the α-globin genes or mutations in their cis-acting sequences account for the reduction of α-globin synthesis. Other types of α-thalassemia occur much less commonly. One important rare form of α-thalassemia is ATR-X syndrome, which is associated with both α-thalassemia and intellectual disability and illustrates the importance of epigenetic packaging of the genome in the regulation of gene expression (see Chapter 3). The X-linked ATRX gene encodes a chromatin remodeling protein that functions, in trans, to activate the expression of the α-globin genes. The ATRX protein belongs to a family of proteins that function within large multiprotein complexes to change DNA topology. ATR-X syndrome is one of a growing number of monogenic diseases that result from mutations in chromatin remodeling proteins.

ATR-X syndrome was initially recognized as unusual because the first families in which it was identified were northern Europeans, a population in which the deletion forms of α-thalassemia are uncommon. In addition, all affected individuals were males who also had severe X-linked intellectual disability together with a wide range of other abnormalities, including characteristic facial features, skeletal defects, and urogenital malformations. This diversity of phenotypes suggests that ATRX regulates the expression of numerous other genes besides the α-globins, although these other targets are presently unknown.

In patients with ATR-X syndrome, the reduction in α-globin synthesis is due to increased accumulation at the α-globin gene cluster of a histone variant (see Chapter 3) called macroH2A, an accumulation that reduces α-globin gene expression and causes α-thalassemia. All the mutations identified to date in the ATRX gene in ATR-X syndrome are partial loss-of-function mutations, leading to mild hematological defects compared with those seen in the classic forms of α-thalassemia.

In patients with ATR-X syndrome, abnormalities in DNA methylation patterns indicate that the ATRX protein is also required to establish or maintain the methylation pattern in certain domains of the genome, perhaps by modulating the access of the DNA methyltransferase enzyme to its binding sites. This finding is noteworthy because mutations in another gene, MECP2, which encodes a protein that binds to methylated DNA, cause Rett syndrome(Case 40) by disrupting the epigenetic regulation of genes in regions of methylated DNA, leading to neurodevelopmental regression. Normally, ATRX and the MeCP2 protein interact, and the impairment of this interaction due to ATRX mutations may contribute to the intellectual disability seen in ATR-X syndrome.

The β-Thalassemias

The β-thalassemias share many features with α-thalassemia. In β-thalassemia, the decrease in β-globin production causes a hypochromic, microcytic anemia and an imbalance in globin synthesis due to the excess of α chains. The excess α chains are insoluble and precipitate (see Fig. 11-8) in both red cell precursors (causing ineffective erythropoiesis) and mature red cells (causing hemolysis) because they damage the cell membrane. In contrast to α-globin, however, the β chain is important only in the postnatal period. Consequently, the onset of β-thalassemia is not apparent until a few months after birth, when β-globin normally replaces γ-globin as the major non-α chain (see Fig. 11-3B), and only the synthesis of the major adult hemoglobin, Hb A, is reduced. The level of Hb F is increased in β-thalassemia, not because of a reactivation of the γ-globin gene expression that was switched off at birth, but because of selective survival and perhaps also increased production of the minor population of adult red blood cells that contain Hb F.

In contrast to α-thalassemia, the β-thalassemias are usually due to single base pair substitutions rather than to deletions (Table 11-5). In many regions of the world where β-thalassemia is common, there are so many different β-thalassemia mutations that persons carrying two β-thalassemia alleles are more likely to be genetic compounds (i.e., carrying two different β-thalassemia alleles) than to be true homozygotes for one allele. Most individuals with two β-thalassemia alleles have thalassemia major, a condition characterized by severe anemia and the need for lifelong medical management. When the β-thalassemia alleles allow so little production of β-globin that no Hb A is present, the condition is designated β0-thalassemia. If some Hb A is detectable, the patient is said to have β+-thalassemia. Although the severity of the clinical disease depends on the combined effect of the two alleles present, survival into adult life was, until recently, unusual.

TABLE 11-5

The Molecular Basis of Some Causes of Simple β-Thalassemia


*One other hemoglobin structural variant that causes β-thalassemia is shown in Table 11-3.

mRNA, Messenger RNA.

Derived in part from Weatherall DJ, Clegg JB, Higgs DR, Wood WG: The hemoglobinopathies. In Scriver CR, Beaudet AL, Sly WS, Valle D, editors: The metabolic and molecular bases of inherited disease, ed 7, New York, 1995, McGraw-Hill, pp 3417-3484; and Orkin SH: Disorders of hemoglobin synthesis: the thalassemias. In Stamatoyannopoulos G, Nienhuis AW, Leder P, Majerus PW, editors: The molecular basis of blood diseases, Philadelphia, 1987, WB Saunders, pp 106-126.

Infants with homozygous β-thalassemia present with anemia once the postnatal production of Hb F decreases, generally before 2 years of age. At present, treatment of the thalassemias is based on correction of the anemia and the increased marrow expansion by blood transfusion and on control of the consequent iron accumulation by the administration of chelating agents. Bone marrow transplantation is effective, but this is an option only if an HLA-matched family member can be found.

Carriers of one β-thalassemia allele are clinically well and are said to have thalassemia minor. Such individuals have hypochromic, microcytic red blood cells and may have a slight anemia that can be misdiagnosed initially as iron deficiency. The diagnosis of thalassemia minor can be supported by hemoglobin electrophoresis, which generally reveals an increase in the level of Hb A2 (α2δ2) (see Fig. 11-3A). In many countries, thalassemia heterozygotes are sufficiently numerous to require diagnostic distinction from iron deficiency anemia and to be a relatively common source of referral for prenatal diagnosis of affected homozygous fetuses (see Chapter 17).

α-Thalassemia Alleles as Modifier Genes of β-Thalassemia.

One of the best examples in human genetics of a modifier gene comes from the fact that both β-thalassemia and α-thalassemia alleles may be present in a population. In such populations, β-thalassemia homozygotes may also inherit an α-thalassemia allele. The clinical severity of the β-thalassemia is sometimes ameliorated by the presence of the α-thalassemia allele, which acts as a modifier gene: the imbalance of globin chain synthesis that occurs in β-thalassemia, due to the relative excess of α chains, is reduced by the decrease in α-chain production that results from the α-thalassemia mutation.

β-Thalassemia, Complex Thalassemias, and Hereditary Persistence of Fetal Hemoglobin.

Almost every type of mutation known to reduce the synthesis of an mRNA or protein has been identified as a cause of β-thalassemia. The following overview of these genetic defects is therefore instructive about mutational mechanisms in general, describing in particular the molecular basis of one of the most common and severe genetic diseases in the world. Mutations of the β-globin gene complex are separated into two broad groups with different clinical phenotypes. One group of defects, which accounts for the great majority of patients, impairs the production of β-globin alone and causes simple β-thalassemia. The second group of mutations consists of large deletions that cause the complex thalassemias, in which the β-globin gene as well as one or more of the other genes—or the LCR—in the β-globin cluster is removed. Finally, some deletions within the β-globin cluster do not cause thalassemia but rather a benign phenotype termed the hereditary persistence of fetal hemoglobin (i.e., the persistence of γ-globin gene expression throughout adult life) that informs us about the regulation of globin gene expression.

Molecular Basis of Simple β-Thalassemia.

Simple β-thalassemia results from a remarkable diversity of molecular defects, predominantly point mutations, in the β-globin gene (Fig. 11-10; see Table 11-5). Most mutations causing simple β-thalassemia lead to a decrease in the abundance of the β-globin mRNA and include promoter mutants, RNA splicing mutants (the most common), mRNA capping or tailing mutants, and frameshift or nonsense mutations that introduce premature termination codons within the coding region of the gene. A few hemoglobin structural variants also impair processing of the β-globin mRNA, as exemplified by Hb E (described later).


FIGURE 11-10 Representative point mutations and small deletions that cause β-thalassemia. Note the distribution of mutations throughout the gene and that the mutations affect virtually every process required for the production of normal β-globin. More than 100 different β-globin point mutations are associated with simple β-thalassemia. SeeSources & Acknowledgments.

RNA Splicing Mutations.

Most β-thalassemia patients with a decreased abundance of β-globin mRNA have abnormalities in RNA splicing. More than two dozen defects of this type have been described, and their combined clinical burden is substantial. These mutations have also acquired high visibility because their effects on splicing are often unexpectedly complex, and analysis of the mutant mRNAs has contributed extensively to knowledge of the sequences critical to normal RNA processing (introduced in Chapter 3). The splice defects are separated into three groups (Fig. 11-11), depending on the region of the unprocessed RNA in which the mutation is located.

• Splice junction mutations include mutations at the 5′ donor or 3′ acceptor splice junctions of the introns or in the consensus sequences surrounding the junctions. The critical nature of the conserved GT dinucleotide at the 5′ intron donor site and of the AG at the 3′ intron acceptor site (see Chapter 3) is demonstrated by the complete loss of normal splicing that results from mutations in these dinucleotides (see Fig. 11-11B). The inactivation of the normal acceptor site elicits the use of other acceptor-like sequences elsewhere in the RNA precursor molecule. These alternative sites are termed cryptic splice sites because they are normally not used by the splicing apparatus if the correct site is available. Cryptic donor or acceptor splice sites can be found in either exons or introns.

• Intron mutations result from defects within an intron cryptic splice site that enhances the use of the cryptic site by making it more similar or identical to the normal splice site. The “activated” cryptic site then competes with the normal site, with variable effectiveness, thereby reducing the abundance of the normal mRNA by decreasing splicing from the correct site, which remains perfectly intact (see Fig. 11-11C). Cryptic splice site mutations are often “leaky,” which means that some use of the normal site occurs, producing a β+-thalassemia phenotype.

• Coding sequence changes that also affect splicing result from mutations in the open reading frame that may or may not alter the amino acid sequence but that activate a cryptic splice site in an exon (see Fig. 11-11D). For example, a mild form of β+-thalassemia results from a mutation in codon 24 (see Table 11-5) that activates a cryptic splice site but does not change the encoded amino acid (both GGT and GGA code for glycine [see Table 3-1]); this is an example of a synonymous mutation that is not neutral in its effect.


FIGURE 11-11 Examples of mutations that disrupt normal splicing of the β-globin gene to cause β-thalassemia. A, Normal splicing pattern. B, An intron 2 mutation (IVS2-2A>G) in the normal splice acceptor site aborts normal splicing. This mutation results in the use of a cryptic acceptor site in intron 2. The cryptic site conforms perfectly to the consensus acceptor splice sequence (where Y is either pyrimidine, T or C). Because exon 3 has been enlarged at its 5′ end by inclusion of intron 2 sequences, the abnormal alternatively spliced messenger RNA (mRNA) made from this mutant gene has lost the correct open reading frame and cannot encode β-globin. C, An intron 1 mutation (G > A in base pair 110 of intron 1) activates a cryptic acceptor site by creating an AG dinucleotide and increasing the resemblance of the site to the consensus acceptor sequence. The globin mRNA thus formed is elongated (19 extra nucleotides) at the 5′ side of exon 2; a premature stop codon is introduced into the transcript. A β+ thalassemia phenotype results because the correct acceptor site is still used, although at only 10% of the wild-type level. D, In the Hb E defect, the missense mutation (Glu26Lys) in codon 26 in exon 1 activates a cryptic donor splice site in codon 25 that competes effectively with the normal donor site. Moderate use is made of this alternative splicing pathway, but the majority of RNA is still processed from the correct site, and mild β+ thalassemia results.

Nonfunctional mRNAs.

Some mRNAs are nonfunctional and cannot direct the synthesis of a complete polypeptide because the mutation generates a premature stop codon, which prematurely terminates translation. Two β-thalassemia mutations near the amino terminus exemplify this effect (see Table 11-5). In one (Gln39Stop), the failure in translation is due to a single nucleotide substitution that creates a nonsense mutation. In the other, a frameshift mutation results from a single base pair deletion early in the open reading frame that removes the first nucleotide from codon 16, which normally encodes glycine; in the mutant reading frame that results, a premature stop codon is quickly encountered downstream, well before the normal termination signal. Because no β-globin is made from these alleles, both of these types of nonfunctional mRNA mutations cause β0-thalassemia in the homozygous state. In some instances, frameshifts near the carboxyl terminus of the protein allow most of the mRNA to be translated normally or to produce elongated globin chains, resulting in a variant hemoglobin rather than β0-thalassemia.

In addition to ablating the production of the β-globin polypeptide, nonsense codons, including the two described earlier, often lead to a reduction in the abundance of the mutant mRNA; indeed, the mRNA may be undetectable. The mechanisms underlying this phenomenon, called nonsense-mediated mRNA decay, appears to be restricted to nonsense codons located more than 50 bp upstream of the final exon-exon junction.

Defects in Capping and Tailing of β-Globin mRNA.

Several β+-thalassemia mutations highlight the critical nature of post-transcriptional modifications of mRNAs. For example, the 3′ UTR of almost all mRNAs ends with a polyA sequence, and if this sequence is not added, the mRNA is unstable. As introduced in Chapter 3, polyadenylation of mRNA first requires enzymatic cleavage of the mRNA, which occurs in response to a signal for the cleavage site, AAUAAA, that is found near the 3′ end of most eukaryotic mRNAs. Patients with a substitution that changes the signal sequence to AACAAA produced only a minor fraction of correctly polyadenylated β-globin mRNA.

Hemoglobin E: A Variant Hemoglobin with Thalassemia Phenotypes

Hb E is probably the most common structurally abnormal hemoglobin in the world, occurring at high frequency in Southeast Asia, where there are at least 1 million homozygotes and 30 million heterozygotes. Hb E is a β-globin variant (Glu26Lys) that reduces the rate of synthesis of the mutant β chain and is another example of a coding sequence mutation that also impairs normal splicing by activating a cryptic splice site (see Fig. 11-10D). Although Hb E homozygotes are asymptomatic and only mildly anemic, individuals who are genetic compounds of Hb E and another β-thalassemia allele have abnormal phenotypes that are largely determined by the severity of the other allele.

Complex Thalassemias and the Hereditary Persistence of Fetal Hemoglobin

As mentioned earlier, the large deletions that cause the complex thalassemias remove the β-globin gene plus one or more other genes—or the LCR—from the β-globin cluster. Thus, affected individuals have reduced expression of β-globin and one or more of the other β-like chains. These disorders are named according to the genes deleted, for example, (δβ)0-thalassemia or (Aγδβ)0-thalassemia, and so on (Fig. 11-12). Deletions that remove the β-globin LCR start approximately 50 to 100 kb upstream of the β-globin gene cluster and extend 3′ to varying degrees. Although some of these deletions (such as the Hispanic deletion shown in Fig. 11-12) leave all or some of the genes at the β-globin locus completely intact, they ablate expression from the entire cluster to cause (εγδβ)0-thalassemia. Such mutations demonstrate the total dependence of gene expression from the β-globin gene cluster on the integrity of the LCR (see Fig. 11-4).


FIGURE 11-12 Location and size of deletions of various (εγδβ)0-thalassemia, (δβ)0-thalassemia, (Aγδβ)0-thalassemia, and HPFH mutants. Note that deletions of the locus control region (LCR) abrogate the expression of all genes in the β-globin cluster. The deletions responsible for δβ-thalassemia, Aγδβ-thalassemia, and HPFH overlap (see text). HPFH, Hereditary persistence of fetal hemoglobin; HS, hypersensitive sites. SeeSources & Acknowledgments.

A second group of large β-globin gene cluster deletions of medical significance are those that leave at least one of the γ genes intact (such as the English deletion in Fig. 11-12). Patients carrying such mutations have one of two clinical manifestations, depending on the deletion: either δβ0-thalassemia or a benign condition called hereditary persistence of fetal hemoglobin (HPFH) that is due to disruption of the perinatal switch from γ-globin to β-globin synthesis. Homozygotes with either of these conditions are viable because the remaining γ gene or genes are still active after birth, instead of switching off as would normally occur. As a result, Hb F (α2γ2) synthesis continues postnatally at a high level and compensates for the absence of Hb A.

The clinically innocuous nature of HPFH that results from the substantial production of γ chains is due to a higher level of Hb F in heterozygotes (17% to 35% Hb F) than is generally seen in δβ0-thalassemia heterozygotes (5% to 18% Hb F). Because the deletions that cause δβ0-thalassemia overlap with those that cause HPFH (see Fig. 11-12), it is not clear why patients with HPFH have higher levels of γ gene expression. One possibility is that some HPFH deletions bring enhancers closer to the γ-globin genes. Insight into the role of regulators of Hb F expression, such as BCL11A and MYB (see earlier discussion), has been partly derived from the study of patients with complex deletions of the β-globin gene cluster. For example, the study of several individuals with HPFH due to rare deletions of the β-globin gene cluster identified a 3.5-kb region, near the 5′ end of the δ-globin gene, that contains binding sites for BCL11A, the critical silencer of Hb F expression in the adult.

Public Health Approaches to Preventing Thalassemia

Large-Scale Population Screening.

The clinical severity of many forms of thalassemia, combined with their high frequency, imposes a tremendous health burden on many societies. In Thailand alone, for example, the World Health Organization has determined that there are between half and three quarters of a million children with severe forms of thalassemia. To reduce the high incidence of the disease in some parts of the world, governments have introduced successful thalassemia control programs based on offering or requiring thalassemia carrier screening of individuals of childbearing age in the population (see Box). As a result of such programs, in many parts of the Mediterranean the birth rate of affected newborns has been reduced by as much as 90% through programs of education directed both to the general population and to health care providers. In Sardinia, a program of voluntary screening, followed by testing of the extended family once a carrier is identified, was initiated in 1975.

Ethical and Social Issues Related to Population Screening for β-Thalassemia*

Approximately 70,000 infants are born worldwide each year with β-thalassemia, at high economic cost to health care systems and at great emotional cost to affected families.

To identify individuals and families at increased risk for the disease, screening is done in many countries. National and international guidelines recommend that screening not be compulsory and that education and genetic counseling should inform decision making.

Widely differing cultural, religious, economic, and social factors significantly influence the adherence to guidelines. For example:

In Greece, screening is voluntary, available both premaritally and prenatally, requires informed consent, is widely advertised by the mass media and in military and school programs, and is accompanied by genetic counseling for carrier couples.

In Iran and Turkey, these practices differ only in that screening is mandatory premaritally (but in all countries with mandatory screening, carrier couples have the right to marry if they wish).

In Taiwan, antenatal screening is available and voluntary, but informed consent is not required and screening is currently not accompanied by educational programs or genetic counseling.

In the United Kingdom, screening is offered to all pregnant women, but public awareness is poor, and the screening is questionably voluntary because many if not most women tested are unaware they have been screened until they are found to be carriers. In some UK programs, women are not given the results of the test.

Major obstacles to more effective population screening for β-thalassemia

The principal obstacles include the facts that pregnant women feel overwhelmed by the array of tests offered to them, many health professionals have insufficient knowledge of genetic disorders, appropriate education and counseling are costly and time-consuming, it is commonly misunderstood that informing a women about a test is equivalent to giving consent, and the effectiveness of mass education varies greatly, depending on the community or country.

The effectiveness of well-executed β-thalassemia screening programs

In populations where β-thalassemia screening has been effectively implemented, the reduction in the incidence of the disease has been striking. For example, in Sardinia, screening between 1975 and 1995 reduced the incidence from 1 per 250 to 1 per 4000 individuals. Similarly, in Cyprus, the incidence of affected births fell from 51 in 1974 to none up to 2007.

*Based on Cousens NE, Gaff CL, Metcalfe SA, et al: Carrier screening for β-thalassaemia: a review of international practice, Eur J Hum Genet 18:1077-1083, 2010.

Screening Restricted to Extended Families.

In developing countries, the initiation of screening programs for thalassemia is a major economic and logistical challenge. Recent work in Pakistan and Saudi Arabia, however, has demonstrated the effectiveness of a screening strategy that may be broadly applicable in countries where consanguineous marriages are common. In the Rawalpindi region of Pakistan, β-thalassemia was found to be largely restricted to a specific group of families that came to attention because there was an identifiable index case (see Chapter 7). In 10 extended families with such an index case, testing of almost 600 persons established that approximately 8% of the married couples examined consisted of two carriers, whereas no couple at risk was identified among 350 randomly selected pregnant women and their partners outside of these 10 families. All carriers reported that the information provided was used to avoid further pregnancy if they already had two or more healthy children or, in the case of couples with only one or no healthy children, for prenatal diagnosis. Although the long-term impact of this program must be established, extended family screening of this type may contribute importantly to the control of recessive diseases in parts of the world where a cultural preference for consanguineous marriage is present. In other words, because of consanguinity, disease gene variants are “trapped” within extended families, so that an affected child is an indicator of an extended family at high risk for the disease.

The initiation of carrier testing and prenatal diagnosis programs for thalassemia requires not only the education of the public and of physicians but also the establishment of skilled central laboratories and the consensus of the population to be screened (see Box). Whereas population-wide programs to control thalassemia are inarguably less expensive than the cost of caring for a large population of affected individuals over their lifetimes, the temptation for governments or physicians to pressure individuals into accepting such programs must be avoided. The autonomy of the individual in reproductive decision making, a bedrock of modern bioethics, and the cultural and religious views of their communities must both be respected.

General References

Higgs DR, Engel JD, Stamatoyannopoulos G. Thalassaemia. Lancet. 2012;379:373–383.

Higgs DR, Gibbons RJ. The molecular basis of α-thalassemia: a model for understanding human molecular genetics. Hematol Oncol Clin North Am. 2010;24:1033–1054.

McCavit TL. Sickle cell disease. Pediatr Rev. 2012;33:195–204.

Roseff SD. Sickle cell disease: a review. Immunohematology. 2009;25:67–74.

Weatherall DJ. The role of the inherited disorders of hemoglobin, the first “molecular diseases,” in the future of human genetics. Annu Rev Genomics Hum Genet. 2013;14:1–24.

References for Specific Topics

Bauer DE, Orkin SH. Update on fetal hemoglobin gene regulation in hemoglobinopathies. Curr Opin Pediatr. 2011;23:1–8.

Ingram VM. Specific chemical difference between the globins of normal human and sickle-cell anaemia haemoglobin. Nature. 1956;178:792–794.

Ingram VM. Gene mutations in human haemoglobin: the chemical difference between normal and sickle cell haemoglobin. Nature. 1957;180:326–328.

Kervestin S, Jacobson A. NMD, a multifaceted response to premature translational termination. Nat Rev Mol Cell Biol. 2012;13:700–712.

Pauling L, Itano HA, Singer SJ, et al. Sickle cell anemia, a molecular disease. Science. 1949;110:543–548.

Sankaran VG, Lettre G, Orkin SH, et al. Modifier genes in Mendelian disorders: the example of hemoglobin disorders. Ann N Y Acad Sci. 2010;1214:47–56.

Steinberg MH, Sebastiani P. Genetic modifiers of sickle cell disease. Am J Hematol. 2012;87:795–803.

Weatherall DJ. The inherited diseases of hemoglobin are an emerging global health burden. Blood. 2010;115:4331–4336.


1. A child dies of hydrops fetalis. Draw a pedigree with genotypes that illustrates to the carrier parents the genetic basis of the infant's thalassemia. Explain why a Melanesian couple whom they met in the hematology clinic, who both also have the α-thalassemia trait, are unlikely to have a similarly affected infant.

2. Why are most β-thalassemia patients likely to be genetic compounds? In what situations might you anticipate that a patient with β-thalassemia would be likely to have two identical β-globin alleles?

3. Tony, a young Italian boy, is found to have moderate β-thalassemia, with a hemoglobin concentration of 7 g/dL (normal amounts are 10 to 13 g/dL). When you perform a Northern blot of his reticulocyte RNA, you unexpectedly find three β-globin mRNA bands, one of normal size, one larger than normal, and one smaller than normal.

What mutational mechanisms could account for the presence of three bands like this in a patient with β-thalassemia? In this patient, the fact that the anemia is mild suggests that a significant fraction of normal β-globin mRNA is being made. What types of mutation would allow this to occur?

4. A man is heterozygous for Hb M Saskatoon, a hemoglobinopathy in which the normal amino acid His is replaced by Tyr at position 63 of the β chain. His mate is heterozygous for Hb M Boston, in which His is replaced by Tyr at position 58 of the α chain. Heterozygosity for either of these mutant alleles produces methemoglobinemia. Outline the possible genotypes and phenotypes of their offspring.

5. A child has a paternal uncle and a maternal aunt with sickle cell disease; both of her parents do not. What is the probability that the child has sickle cell disease?

6. A woman has sickle cell trait, and her mate is heterozygous for Hb C. What is the probability that their child has no abnormal hemoglobin?

7. Match the following:


8. Mutations in noncoding sequences may change the number of protein molecules produced, but each protein molecule made will generally have a normal amino acid sequence. Give examples of some exceptions to this rule, and describe how the alterations in the amino acid sequence are generated.

9. What are some possible explanations for the fact that thalassemia control programs, such as the successful one in Sardinia, have not reduced the birth rate of newborns with severe thalassemia to zero? For example, in Sardinia from 1999 to 2002, approximately two to five such infants were born each year.