	Huntington Disease	Myotonic Dystrophy
Inheritance	Autosomal dominant	Autosomal dominant
Chromosome locus	4p16.3	19q13.3
Trinucleotide repeat	CAG in 5′ translated region	CTG in 3′ untranslated region
Repeat sizes	Normal ≤26	Normal <37
	Mutable 27–35
	Reduced penetrance	Full mutation
	36–39	50–2000+
	Fully penetrant ≥40
Protein product	Huntingtin	MD protein kinase (DMPK)
Early-onset form	Juvenile	Congenital
	Usually paternally transmitted	Usually maternally transmitted

Explanations for this include the possibility that expansion is caused by slippage (p. 24) of DNA polymerase, simply reflecting the number of mitoses undergone during gametogenesis (p. 41). An alternative possibility is based on the observation that huntingtin is expressed in oocytes, so that there could be selection against oocytes with large expansions as a consequence of preferential apoptosis.

Clinical Applications and Future Prospects

Predictive genetic testing is part of routine clinical genetic practice, but there is universal agreement that this should be offered only as part of a careful counseling package. Experience to date indicates that more women than men come forward for this, and the psychological disturbance in those given positive results is low. Some 60% of candidates test negative (i.e., they receive good news), and the reasons for this departure from the expected 50% are not clear.

Prenatal diagnosis is possible for those couples who find this acceptable, although only about 25 such tests are performed in the United Kingdom annually. Obviously there are considerable emotional and ethical issues associated with termination of pregnancy; the condition is late in onset and the couple must consider the possibility of effective therapy being available in the foreseeable future. One appealing therapeutic approach is based on the observation that large CAG repeats result in intracellular accumulation of huntingtin ‘aggregates’, which are cleaved by a protease known as caspase to form a toxic product that causes cell death (apoptosis). Caspase inhibitors have been shown to have a beneficial effect in a HD mouse model. Another therapeutic approach under consideration is fetal neuronal cell transfer into regions of the brain, such as the caudate nucleus and putamen, which become atrophic in the early stages of the disease. This approach carries ethical considerations that will be difficult for some couples.

Myotonic Dystrophy

Myotonic dystrophy (MD) is the most common form of muscular dystrophy seen in adults, with an overall incidence of approximately 1 : 8000. It shares many features in common with HD (see Table 19.1)—both show autosomal dominant inheritance with anticipation, and an early-onset form with different clinical features. However, in MD the early-onset form is transmitted almost exclusively by the mother and presents at birth, in contrast to juvenile HD, which is generally paternally transmitted with an age of onset in the teens.

Clinical Features

In contrast to most forms of muscular dystrophy, clinical features in MD are not limited exclusively to the neuromuscular system. Individuals with MD usually present in adult life with slowly progressive weakness and myotonia. This latter term refers to tonic muscle spasm with prolonged relaxation, which can manifest as a delay in releasing the grip on shaking hands. Other clinical features include cataracts (Figure 19.1), cardiac conduction defects, disturbed gastrointestinal peristalsis (dysphagia, constipation, diarrhea), weak sphincters, increased risk of diabetes mellitus and gallstones, somnolence, frontal balding, and testicular atrophy. The age of onset is very variable and in its mildest form usually runs a relatively benign course. However, as the age of onset becomes earlier, so the clinical symptoms increase in severity and more body systems are involved. In the ‘congenital’ form, affected babies present at birth with hypotonia, talipes, and respiratory distress that can prove life threatening (see Figure 7.19). Children who survive tend to show a lack of facial expression (‘myopathic facies’) with delayed motor development and learning difficulties (Figure 19.2).

FIGURE 19.1 Refractile lens opacities in an asymptomatic person with myotonic dystrophy.

(Courtesy Mr. R. Doran and Mr. M. Geall, Department of Ophthalmology, General Infirmary, Leeds, UK.)

FIGURE 19.2 A mother and child with myotonic dystrophy. The child has clear features of facial myopathy and suffers from the congenital form; the mother has only mild facial myopathy. The marked generational difference in the severity of disease illustrates the phenomenon of anticipation.

The diagnosis of MD used to be based on the myotonic discharges seen on electromyography but mutation analysis is more reliable and much less painful.

Genetics

It follows autosomal dominant inheritance with increasing severity in succeeding generations—anticipation (p. 120). It used to be thought that this reflected ascertainment bias, caused by the greater likelihood of detecting a mildly affected parent with a severely affected child, rather than the other way round. However, studies in the 1980s confirmed that anticipation is a real phenomenon.

Mapping and Isolating the MD Gene

Linkage to the secretor and Lutheran blood group loci was established as long ago as 1971, with the locus mapped to chromosome 19 in 1982. The relevant region was cloned with great effort and the mutational basis shown in 1992 to be instability in a CTG repeat sequence, which is present in the 3′ untranslated region of a protein kinase gene—dystrophia myotonica protein kinase (DMPK).

Genotype–Phenotype Correlation in Myotonic Dystrophy

In unaffected persons the CTG sequence lying 3′ to the DMPK gene consists of up to 37 repeats (see Table 19.1). Affected individuals have an expansion of at least 50 copies of the CTG sequence. There is a close correlation between disease severity and the size of the expansion, which can exceed 2000 repeats. The severe congenital cases show the largest repeat copy number, with almost invariable inheritance from the mother. Thus, meiotic or germline instability is greater in the female for alleles containing large sequences. Curiously, expansion of a relatively small number of repeats appears to occur more commonly in the male, and most MD mutations are thought to have occurred originally during meiosis in the male. One possible explanation for these observations is that mature spermatozoa can carry only small expansions, whereas ova can accommodate much larger expansions.

Another puzzling feature of MD is the reported tendency for healthy individuals who are heterozygous for MD alleles in the normal size range to preferentially transmit alleles greater than 19 CTG repeats in size. This possible example of meiotic drive (p. 135) could explain the relatively high frequency of MD with constant replenishment of a reservoir of potential MD mutations.

The MD Protein Kinase

Perhaps surprisingly, it may be that DMPK is not directly responsible for muscle symptoms—mice with both overexpression and underexpression of Dmpk show neither myotonia nor other typical clinical features of MD. It is now appears that the RNA produced by expanded DMPK alleles interferes with the cellular processing of RNA produced by a variety of other genes. Expanded DMPK transcripts accumulate in the cell nuclei, and this is believed to have a gain-of-function effect through its binding with a CUG RNA-binding protein (CUG-BP) that has been identified. Excess CUG-BP has been shown to interfere with a number of genes relevant to MD; this is not surprising because CUG repeats are known to exist in various alternately spliced muscle-specific enzymes.

Clinical Applications and Future Prospects

Presymptomatic genetic testing and prenatal diagnosis can be offered to those families for whom it is appropriate and acceptable. This is particularly relevant for couples who have had a child with the severe congenital form, for whom the risk of recurrence is relatively high. As in HD, presymptomatic testing should not be undertaken without an offer of long-term support and medical care, and a discussion about possible difficulty in obtaining life and health insurance (p. 366).

Important components of the management of MD include regular surveillance for cardiac conduction defects and the provision of information about risks associated with general anesthesia. Logical approaches to gene therapy will almost certainly have to await a better understanding of the mechanism whereby expansion of the repeat sequence in the 3′ untranslated region of the DMPK gene causes such diverse and variable clinical abnormalities.

Type 2 MD

Some families with a variable presentation of similar features to MD, but without the (CTG)_n expansion of DMPK, have been shown to link to 3q21. Originally referred to as proximal myotonic myopathy (PROMM), these cases are designated type 2 MD to distinguish them from the more common type 1 MD. The molecular defect has been shown to be a (CCTG)_n expansion mutation in intron 1 of a gene called ZNF9 and its protein is thought to bind RNA. Most families are of German descent, and haplotype studies suggest a single founder mutation occurring between 200 and 500 generations ago.

Hereditary Motor and Sensory Neuropathy

Hereditary motor and sensory neuropathy (HMSN) comprises a group of clinically and genetically heterogeneous disorders characterized by slowly progressive distal muscle weakness and wasting. Other names for these disorders include Charcot-Marie-Tooth disease and peroneal muscular atrophy. Their overall incidence is approximately 1 : 2500.

HMSN can be classified on the basis of the results of motor nerve conduction velocity (MNCV) studies. In HMSN type I, MNCV is reduced and nerve biopsies from patients show segmental demyelination accompanied by hypertrophic changes with ‘onion bulb’ formation. In HMSN type II, MNCV is normal or only slightly reduced and nerve biopsies show axonal degeneration.

Clinical Features

In autosomal dominant HMSN-I—the most common form—there is onset of slowly progressive distal muscle weakness and wasting in the lower limbs between the ages of 10 and 30 years, followed later by the upper limbs in many patients, and associated ataxia and tremor. The appearance of the lower limbs has been likened to that of an ‘inverted champagne bottle’ (Figure 19.3). With age, locomotion becomes more difficult and the feet tend to show exaggeration of their normal arch, known as ‘pes cavus’. Despite these quite striking changes, many patients retain reasonable muscle strength and are not too seriously disabled. Other faculties such as vision, hearing, and intellect are not impaired. Palpable thickening of peripheral nerves can sometimes be detected.

FIGURE 19.3 Lower limbs of a male with hereditary motor and sensory neuropathy (HMSN) showing severe muscle wasting below the knees.

The clinical features in other forms of HMSN are similar but differ in the age of onset, rate of progression, and presence of other neurological involvement. For example, in HMSN-II onset is usually later than in HMSN-I, and the disease course is milder to the extent that some affected individuals are asymptomatic. By contrast, in HMSN-III, which is very rare, onset is in early childhood and there is severe delay in achieving motor milestones.

Genetics

HMSN can show autosomal dominant, autosomal recessive or X-linked inheritance, although autosomal dominant forms are by far the most common. More than 70% of cases of HMSN-I are due to a DNA duplication of 1.5 Mb chromosome 17p that encompasses the peripheral myelin protein-22 (PMP22) gene. The glycoprotein product is present in the myelin membranes of peripheral nerves, where it helps to arrest Schwann cell division. HMSN-I in humans is therefore thought to be the result of a PMP22 dosage effect, though point mutations can be found some patients. The duplication is generated by misalignment and subsequent recombination between homologous sequences that flank the PMP22 gene (Figure 19.4); this event usually occurs in male gametogenesis (rather than in the female, which is the case in Duchenne muscular dystrophy [p. 307]). The reciprocal deletion product of this misaligned recombination event, giving rise to haploinsufficiency, causes a relatively mild disorder known as hereditary neuropathy with liability to pressure palsies. Minor nerve trauma, such as pressure from prolonged sitting on a long-haul flight, causes focal numbness and weakness. The same misalignment recombination mechanism occurs in Hb Lepore and anti-Lepore (see Figure 10.3; p. 157), congenital adrenal hyperplasia (p. 174), and deletion 22q11 syndrome (p. 282), to name but a few.

FIGURE 19.4 Mechanism by which misalignment and recombination with unequal crossing over lead to formation of the duplication and deletion that cause hereditary motor and sensory neuropathy type I (HMSN-I) and hereditary neuropathy with liability to pressure palsies (HNPP). X and Y represent homologous sequences flanking the PMP22 gene.

Other Forms of Hereditary Motor and Sensory Neuropathy

A second HMSN-I locus is found on chromosome 1, which has led to the duplication chromosome 17 cases being referred to as HMSN-Ia and chromosome 1 cases as HMSN-Ib. The latter is caused by mutations in the gene that codes for another major myelin protein, known as myelin protein zero (MPZ). This plays a crucial role as an adhesion molecule in the compaction of myelin in peripheral nerves.

A more rare form of HMSN shows X-linked dominant inheritance, with males having typical HMSN-I features and females being more mildly affected—sometimes with HMSN-II characteristics. This is due to mutations in the gene that encodes a gap junction protein called GJB1 (previously Connexin 32).

HMSN-II is genetically heterogeneous and progress with clinical mutation testing has been considerably slower. However, a range of genes can now be analysed, including Mitofusin 2 (MFN2) and Neurofilament Protein, Light Peptide (NEFL).

Future Prospects

Genetic testing in HMSN has greatly improved diagnostic precision, though there is still a place for nerve conduction studies in clinical assessment, especially when genetic testing fails to yield a positive result. Treatment remains elusive but in HMSN-Ia efforts at gene therapy may be directed at trying to achieve a reduction in gene dosage by switching off, or ‘downregulating’, PMP22 expression.

Neurofibromatosis

References to the clinical features of neurofibromatosis (NF) first appeared in the eighteenth-century medical literature, but historically the disorder is most commonly associated with the name Von Recklinghausen, a German pathologist who coined the term ‘neurofibroma’ in 1882.

This is one of the most common genetic disorders in humans and gained public notoriety when it was suggested that Joseph Merrick, the ‘Elephant Man’, was probably affected. However, many now think he had the much rarer disorder known as Proteus syndrome.

There are two main types of neurofibromatosis, NF1 and NF2. Both conditions, especially NF2, could be included under familial cancer syndromes (see Chapter 14) but are covered in more detail here. NF1 has a birth incidence of approximately 1 : 3000; NF2 approximately 1 : 35,000 and a prevalence of around 1 : 200,000.

Clinical Features

The most notable features of NF1 are small pigmented skin lesions, known as café-au-lait (CAL) spots, and small soft fleshy growths known as neurofibromata (Figure 19.5). CAL spots first appear in early childhood and continue to increase in both size and number until puberty. A minimum of six CAL spots at least 5 mm in diameter is required to support the diagnosis in childhood, and axillary and/or inguinal freckling should be present. Neurofibromata are benign tumors that arise most commonly in the skin, usually appearing in adolescence or adult life, and increasing in number with age.

FIGURE 19.5 A patient with neurofibromatosis type I showing truncal freckling, café-au-lait spots, and multiple neurofibromata.

Other clinical findings include relative macrocephaly (large head) and Lisch nodules. These are small harmless raised pigmented hamartomata of the iris (Figure 19.6). The most common complication, occurring in a third of childhood cases, is mild developmental delay characterized by a non-verbal learning disorder. For many, significant improvement is seen through the school years. Most individuals with NF1 enjoy a normal life and are not unduly inconvenienced by their condition. However, a small number of patients develop one or more major complications, such as epilepsy, a central nervous system tumor, or scoliosis.

FIGURE 19.6 Lisch nodules seen in neurofibromatosis type I.

(Courtesy Mr. R. Doran, Department of Ophthalmology, General Infirmary, Leeds, UK.)

Genetics

NF1 shows autosomal dominant inheritance with virtually 100% penetrance by the age of 5 years. The effects are very variable and affected members of the same family can show striking differences in disease severity. The features in affected MZ twins are usually very similar, so the variability in family members with the same mutation may be due to modifying genes at other loci. Approximately 50% of cases of NF1 are due to new mutations, with the estimated mutation rate being approximately 1 per 10,000 gametes. This is around 100 times greater than the average mutation rate per generation per locus in humans.

There are a few reports of more than one affected child born to unaffected parents—the result of gonadal mosaicism (p. 121), usually paternal in origin. Somatic mosaicism in NF1 can manifest with features limited to a particular part of the body. This is referred to as segmental NF.

The Neurofibromatosis Type 1 Gene and its Product

The NF1 gene, neurofibromin, was successfully mapped to chromosome 17, adjacent to the centromere, in 1987. Its isolation was aided by the identification of two patients who both had a balanced translocation with a breakpoint at 17q11.2. A cosmid clone was identified containing both translocation breakpoints and a search for transcripts from this region yielded four genes, one of which was shown to be neurofibromin. It is large, spanning over 350 kilobases (kb) of genomic DNA and comprising at least 59 exons. The other three genes identified in this region were found to lie within a single intron of the neurofibromin gene, where they are transcribed in the opposite direction from the complementary strand (p. 14).

The neurofibromin protein encoded by this gene shows structural homology to the guanosine triphosphatase (GTPase)-activating protein (GAP), which is important in signal transduction (p. 214) by downregulating RAS activity. The place of neurofibromin in the RAS-MAPK pathway is shown in Figure 16.12, highlighting the link with Noonan syndrome (p. 254). Loss of heterozygosity (p. 215) for chromosome 17 markers has been observed in several malignant tumors in patients with NF1, as well as in a small number of benign neurofibromata. These observations indicate that the neurofibromin gene functions as a tumor suppressor (p. 214). It has been shown to contain a GAP-related domain (GRD), which interacts with the RAS proto-oncogene product. A mRNA editing site exists in the neurofibromin gene and edited transcript causes GRD protein truncation, which inactivates the tumor suppressor function. A higher range of editing is seen in more malignant tumors.

Other genes, including TP53 (p. 218) on the short arm of chromosome 17, are also involved in tumor development and progression in NF1. Conversely, it is also known that the neurofibromin gene is implicated in the development of sporadic tumors not associated with NF, including carcinoma of the colon, neuroblastoma, and malignant melanoma. These observations confirm that the neurofibromin gene plays an important role in cell growth and differentiation.

Genotype–Phenotype Correlation

Many different mutations have been identified in the neurofibromin gene, which include deletions, insertions, duplications, and point substitutions (p. 23). Most lead to severe truncation of the protein or complete absence of gene expression. To date, there is little evidence for a clear genotype-phenotype relationship with the exception of one specific mutation, a 3-bp inframe deletion in exon 17, which has recurred in different cases and families, and affected individuals do not appear to develop cutaneous neurofibromata. Generally, NF1 shows quite striking intrafamilial variation, suggesting the possibility of modifier genes. Patients with large deletions that include the entire neurofibromin gene tend to be more severely affected, with significant intellectual impairment, a somewhat marfanoid habitus, and a larger than average number of cutaneous neurofibromata.

Neurofibromatosis Type 2

In NF2 both CAL spots and neurofibromata can occur, but these are much less common than in NF1. The most characteristic feature is the development in early adult life of tumors involving the eighth cranial nerves—vestibular schwannomas (still sometimes called acoustic neuromas). Several other central nervous system tumors occur frequently, although more than half remain asymptomatic. Autosomal dominant spinal and peripheral schwannomas without vestibular schwannomas is an entity known as schwannomatosis. An ophthalmic feature seen in NF2, but not NF1, is cataracts, which are frequent but often subclinical.

The NF2 locus was mapped to chromosome 22q by linkage analysis in 1987. The gene, called schwannomin, was cloned in 1993, found to span 110 kb with 17 exons, and is thought to be a cytoskeleton protein that acts as a tumor suppressor.

Clinical Applications and Future Prospects

Mapping of the neurofibromin gene has provided a means of offering both presymptomatic and prenatal diagnosis, usually by direct mutation analysis but linkage if no mutation has been found. In practice very few families pursue either of these options, partly because NF1 is not perceived as a serious illness and also because mutation analysis does not help in predicting disease severity.

At present there is no cure for NF1. Drug therapy aimed at upregulating neurofibromin GAP activity or downregulating RAS activity could prove beneficial in the absence of effective gene therapy. However, it is difficult to envisage how this could be applied to diverse target tissues, including the central nervous system. Nevertheless, there is great interest in the whole RAS-MAPK pathway (see Figure 16.12, p. 256), especially its role in tumor formation, and whether drugs can effectively modify its activity.

Marfan Syndrome

The original patient described by the French pediatrician Bernard Marfan, in 1896, probably had the similar but rarer condition now known as Beal syndrome, or congenital contractual arachnodactyly (p. 301). In clinical practice physicians often consider the diagnosis of Marfan syndrome (MFS) for any patient who is tall with subjective features of long limbs and fingers. However, it is essential to be objective in clinical assessment because a number of conditions have ‘marfanoid’ features, and many tall, thin people are entirely normal. Detailed diagnostic criteria, referred to as the Gent criteria, are in general use by geneticists (Table 19.2).

Table 19.2 Revised (Gent) Criteria for Making a Diagnosis of Marfan Syndrome

System	Major Criteria	Minor Criteria
Skeletal	Four of these should be present:
	Pectus carinatum	Pectus excavatum
	Pectus excavatum requiring surgery	Joint hypermobility
	Reduced upper to lower segment body ratio or span : height ratio >1.05	High arched palate with dental crowding
	Hypermobility of wrist and thumbs Medial displacement of medial malleolus	Facial features, including down-slanting palpebral fissures causing pes planus

	Radiological protrusio acetabulae
Ocular	Ectopia lentis	Flat cornea Increased axial length of the globe Hypoplastic iris
Cardiovascular	Dilatation of the ascending aorta	Mitral valve prolapse
	Dissection of the ascending aorta	Dilatation or dissection of descending thoracic or abdominal aorta under 50 years
Pulmonary	None	Spontaneous pneumothorax Apical blebs
Skin/connective tissue		None
Dura	Lumbosacral dural ectasia	None
Family history/genetics	First-degree relative who meets criteria	None
	Presence of FBN1 mutation, or high-risk haplotype in MFS family	None

Clinical Features

MFS is a disorder of fibrous connective tissue, specifically a defect in type 1 fibrillin, a glycoprotein encoded by the FBN1 gene. In the classic presentation affected individuals are tall compared with unaffected family members, have joint laxity, a span : height ratio greater than 1.05, a reduced upper to lower segment body ratio, pectus deformity, and scoliosis (Figure 19.7). The connective tissue defect gives rise to ectopia lentis (lens subluxation) in a proportion of (but not all) families and, very importantly, dilatation of the ascending aorta, which can lead to dissection. The latter complication is obviously life threatening, and for this reason alone care must be taken over the diagnosis. Aortic dilatation may be progressive but the rate of change can be reduced by β-adrenergic blockade (if tolerated) and there is great hope for angiotensin-II receptor antagonists (similar properties to angiotensin-converting enzyme inhibitors), whose trials are under way. Surgical replacement should be undertaken if the diameter reaches 50 to 55 mm. Pregnancy is a risk factor for a woman with MFS who already has some dilatation of the aorta, and monitoring is very important.

FIGURE 19.7 A, An adolescent with Marfan syndrome showing disproportionately long limbs (arachnodactyly) and a very extreme example of chest bone deformity; he also has a dilated aortic root. B, Joint hypermobility at the wrist in a woman with Marfan syndrome; this appearance might also be seen in other joint-laxity conditions, such as Ehlers-Danlos syndrome.

A diagnosis of MFS requires careful clinical assessment, body measurements looking for evidence of disproportion, echocardiography, ophthalmic evaluation, and, in some doubtful cases, lumbar magnetic resonance imaging to look for evidence of dural ectasia (see Table 19.2). The metacarpophalangeal index, a radiological measurement of the ratio of these hand bone lengths, does not feature in the revised criteria. Where the family history is non-contributory, a positive diagnosis is made when the patient has a minimum of two major criteria plus involvement of a third organ system; for a person with a close relative who is definitely affected, it is sufficient to have one major criterion plus involvement of a second organ system.

Genetics

MFS follows autosomal dominant inheritance and the majority of cases are linked to the large FBN1 gene on 15q21, with 65 exons spanning 200 kb and containing five distinct domains. The largest of these, occupying about 75% of the gene, comprises about 46 epidermal growth factor repeats (see p. 190). Finding the causative mutations in affected patients was initially very difficult, but hundreds have now been reported. Most are missense and have a dominant-negative effect, resulting in less than 35% of the expected amount of fibrillin in the extracellular matrix. Mutations have also occasionally been found in related phenotypes such as neonatal MFS, familial ectopia lentis, Shrintzen-Goldberg syndrome, and the MASS phenotype (mitral valve prolapse, myopia, borderline aortic enlargement, non-specific skin and skeletal findings).

Loeys-Dietz Syndrome

Familial aortic aneurysm is not confined to MFS and a new, distinct, condition has been delineated. This also follows autosomal dominant inheritance and aneurysms can be aggressive and occur before major aortic dilatation. Additional findings include cleft palate or bifid uvula, craniosynostosis, mental retardation, and generalized arterial tortuosity with aneurysms occurring elsewhere in the circulation. Some individuals have features overlapping with MFS but they do not fulfill the accepted Gent diagnostic criteria. The condition is now known as Loeys-Dietz syndrome, and the gene was identified through a candidate approach. Transforming growth factor (TGF) signaling (p. 85) had been shown to be important in vascular and craniofacial development in mouse models; this led Loeys and colleagues to sequence the TGF-β receptor 2 (TGFBR2) gene in a series of families. Heterozygous mutations were found in most of these, and in the others missense mutations were found in the related gene, TGFBR1.

Congenital Contractural Arachnodactyly

Also known as Beal syndrome, this is probably the condition originally described by Marfan in 1896. Many features overlap with MFS, but there is less tendency to aortic dilatation and its catastrophic consequences. Individuals have congenital contractures of their digits, a crumpled ear helix, and sometimes marked scoliosis. It is due to mutated type 2 fibrillin, which shares the same organizational structure as fibrillin-1 and maps to 5q23.

Cystic Fibrosis

Cystic fibrosis (CF) was first recognized as a discrete entity in 1936 and used to be known as ‘mucoviscidosis’ because of the accumulation of thick mucous secretions that lead to blockage of the airways and secondary infection. Although antibiotics and physiotherapy have been very effective in increasing the average life expectancy of a child with CF from less than 5 years in 1955 to at least 30 years, CF remains a significant cause of chronic ill health and death in childhood and early adult life.

CF is one of the most common autosomal recessive disorders encountered in individuals of western European origin, in whom the incidence varies from 1 in 2000 to 1 in 3000. The incidence is slightly lower in eastern and southern European populations, and much lower in African Americans (1 in 15,000) and Asian Americans (1 in 31,000).

Clinical Features

The organs most commonly affected in CF are the lungs and the pancreas. Chronic lung disease caused by recurrent infection eventually leads to fibrotic changes in the lungs with secondary cardiac failure, a condition known as cor pulmonale. When this complication occurs, the only hope for long-term survival rests in a successful heart–lung transplant.

In 85% of people with CF, pancreatic function is impaired, with reduced enzyme secretion from blockage of the pancreatic ducts by inspissated secretions. This leads to malabsorption with an increase in the fat content of the stools but is satisfactorily treated with oral supplements of pancreatic enzymes.

Other problems commonly encountered in CF include nasal polyps, rectal prolapse, cirrhosis, and diabetes mellitus. Around 10% of children with CF present in the newborn period with obstruction of the small bowel from thickened meconium, known as meconium ileus. Almost all males with CF are sterile because of congenital bilateral absence of the vas deferens (CBAVD). There are some males with CBAVD, sometimes with rare mutations in the CF gene, who have virtually no other features of CF. Other rare presentations include chronic pancreatitis, diffuse bronchiectasis, and bronchopulmonary allergic aspergillosis.

Genetics

CF shows autosomal recessive inheritance. Other autosomal recessive disorders, such as hemochromatosis, which causes tissue iron overload, have a higher carrier frequency, but CF is by far the most serious autosomal recessive disorder encountered in children of western European origin. Possible explanations for this high incidence include a high mutation rate, meiotic drive, and heterozygote advantage. The latter explanation, possibly mediated by increased heterozygote resistance to chloride-secreting bacterially induced diarrhea, is often favored but does not explain why CF is rare in tropical regions where diarrheal diseases are common.

Mapping and Isolation of the Cystic Fibrosis Gene

The mapping and isolation of the CF was a celebrated milestone in the history of human molecular genetics and it is easy to forget how very difficult and time consuming such research was just 25 years ago. The CF locus was mapped to chromosome 7q31 in 1985 by the demonstration of linkage to the gene for a polymorphic enzyme known as paraoxonase. Shortly afterward, two polymorphic DNA marker loci, known as MET and D7S8, were shown to be closely linked flanking markers. The region between these markers was scrutinized for the presence of HTF or CpG islands, which are known to be present close to the 5′ end of many genes (p. 75). This led to the identification of several new DNA markers that were shown to be very tightly linked to the CF locus with recombination frequencies of less than 1%. These loci were found to be in linkage disequilibrium (p. 138) with the CF locus, and one CF mutation was found to be associated with one particular haplotype in 84% of cases, consistent with the concept of a single original mutation being responsible for a large proportion of all CF genes. The identification of loci tightly linked to the CF locus narrowed its location down to a region of approximately 500 kb. The CF gene was eventually cloned by two groups of scientists in North America in 1989 by a combination of chromosome jumping, physical mapping, isolation of exon sequences and mutation analysis. It was named the CF transmembrane conductance regulator (CFTR) gene, spans a genomic region of approximately 250 kb, and contains 27 exons.

The Cystic Fibrosis Transmembrane Conductance Regulator Protein

The structure of CFTR is consistent with a protein product containing 1480 amino acids with a molecular weight of 168 kDa. It is thought to consist of two transmembrane (TM) domains that anchor it to the cell membrane, two nucleotide binding folds (NBFs) that bind ATP, and a regulatory (R) domain, which is phosphorylated by protein kinase-A (Figure 19.8).

FIGURE 19.8 The cystic fibrosis locus, gene, and protein product, which influences closely adjacent epithelial sodium and outwardly rectifying chloride channels. R, Regulatory domain; NBF, nucleotide binding fold; TM, transmembrane domain.

The primary role of the CFTR protein is to act as a chloride channel. Activation by phosphorylation of the regulatory domain, followed by binding of ATP to the NBF domains, opens the outwardly rectifying chloride channel and exerts a negative effect on intracellular sodium absorption by closure of the epithelial sodium channel. The net effect is to reduce the level of intracellular sodium chloride, which improves the quality of cellular mucous secretions.

Mutations in the Cystic Fibrosis Transmembrane Conductance Regulator Gene

The first mutation to be identified in CFTR was a deletion of three adjacent base pairs at the 508th codon which results in the loss of a phenylalanine residue. This mutation is now known as Phe508del (previously deltaF508) and accounts for approximately 70% of all mutations in CFTR, the highest incidence of 88% being in Denmark (Table 19.3). The mutation can be demonstrated very simply by PCR using primers that flank the 508th codon (Figure 19.9).

Table 19.3 Contribution of Phe508del Mutation to All CF Mutations

Country	%
Denmark	88
Netherlands	79
UK	78
Ireland	75
France	75
USA	66
Germany	65
Poland	55
Italy	50
Turkey	30

Data from European Working Group on CF Genetics (EWGCFG) gradient of distribution in Europe of the major CF mutation and of its associated haplotype. Hum Genet 1990; 85:436–441, and worldwide survey of the Phe508del mutation—report from the Cystic Fibrosis Genetic Analysis Consortium. Am J Hum Genet 1990; 47:354–359

FIGURE 19.9 Polymerase chain reaction (PCR) amplification of 98- and 95-bp DNA fragments surrounding the Phe508del mutation site in the CFTR gene from a child with cystic fibrosis and her parents. The child, II₂, is homozygous for Phe508del. Her parents, I₁ and I₂, are heterozygous, and her brother, II₁, is homozygous for the normal allele. Heterozygotes are readily identified by the presence of heteroduplex bands formed between 95- and 98-bp products and electrophoresed on a non-denaturing gel.

More than 1500 other mutations in the CFTR gene have been identified. These include missense, frameshift, splice-site, nonsense, and deletion mutations (p. 23). Most of these are extremely uncommon, although a few can account for a small but significant proportion of mutations in a particular population. For example, the G542X and G551D mutations account for 12% and 3%, respectively, of all CF mutations in the Ashkenazi Jewish and North American Caucasian populations. Commercial multiplex PCR-based kits have been developed that detect approximately 90% of all carriers. Using these it is possible to reduce the carrier risk for a healthy individual from a population risk of 1 in 25 to less than 1 in 200.

Genotype-Phenotype Correlation

Mutations in CFTR can influence the function of the protein product by:

1 Causing a complete or partial reduction in its synthesis—e.g., G542X and IVS8-6(5T)

2 Preventing it from reaching the epithelial membrane—e.g., Phe508del

3 Causing it to function incorrectly when it reaches its final location—e.g., G551D and R117H.

The net effect of all these mutations is to reduce the normal functional activity of the CFTR protein. The extent to which normal CFTR protein activity is reduced correlates well with the clinical phenotype. Levels of less than 3% are associated with severe ‘classic’ CF, sometimes referred to as the PI type because of associated pancreatic insufficiency. Levels of activity between 3% and 8% cause a milder ‘atypical’ form of CF in which there is respiratory disease but relatively normal pancreatic function. This is referred to as the PS (pancreatic sufficient) form. Finally, levels of activity between 8% and 12% cause the mildest CF phenotype, in which the only clinical abnormality is CBAVD in males.

The relationship between genotype and phenotype is complex. Homozygotes for Phe508del almost always have severe classical CF, as do compound heterozygotes with Phe508del and G551D or G542X. The outcome for other compound heterozygote combinations can be much more difficult to predict. The complexity of the interaction between CFTR alleles is illustrated by the IVS8-6 poly T variant. This contains a polythymidine tract in intron 8 that influences the splicing efficiency of exon 9, resulting in reduced synthesis of normal CFTR protein. Three variants consisting of 5T, 7T, and 9T have been identified. The 9T variant is associated with normal activity but the 5T allele leads to a reduction in the number of transcripts containing exon 9. The 5T variant has a population frequency of approximately 5%, but is more often found in patients with CBAVD (40–50%) or disseminated bronchiectasis (30%). Curiously, it has been shown that the number of thymidine residues influences the effect of another mutation, R117H. When R117H is in cis with 5T (i.e., in the same allele) it causes the PS (pancreatic sufficiency) form of CF when another CF mutation is present on the other allele. However, in compound heterozygotes (e.g., Phe508del /R117H) where R117H is in cis with 7T, it can result in a milder but variable phenotype, ranging from CBAVD to PS CF. The milder phenotype is likely to result from the expression of higher levels of full-length R117H protein with some residual activity. The increasing number of CFTR mutations and variability of the associated phenotypes has led some authors to propose a spectrum of ‘CFTR disease’, recognizing that a label of CF may be inappropriate for patients with milder symptoms.

Clinical Applications and Future Prospects

Before the mapping of the CF locus and the subsequent isolation of CFTR, it was not possible to offer either carrier detection or reliable prenatal diagnosis. Now parents of an affected child can almost always be offered prenatal diagnosis by direct mutation analysis. Similarly, knowledge of one or both of the mutations in an affected child now permits the offer of carrier detection to close family relatives. In many parts of the world, it is now standard practice to offer cascade screening to all families in which a mutation has been identified. Population screening for carriers of CF (p. 318) and neonatal screening for CF homozygotes (p. 320) have been widely implemented.

CF is a prime candidate for gene therapy because of the relative accessibility of the crucial target organs (i.e., the lungs). Gene transfer studies carried out using adenoviruses and CFTR complementary DNA (cDNA)–liposome complexes have resulted in the restoration of chloride secretion in CF transgenic mice. Several clinical trials have been undertaken in small groups of volunteer patients with CF. Although there has been experimental evidence of CFTR expression in the treated patients, this has generally been transient. Problems have been encountered with poor vector efficiency and inflammatory reactions, particularly when adenoviruses have been used as the vector. Despite these initial problems, there is cautious optimism that effective gene therapy for CF will be developed eventually.

Inherited Cardiac Arrhythmias and Cardiomyopathies

In about 4% of sudden cardiac death in persons ages 16 to 64 years, no explanation is evident; this is enormously traumatic for the family left behind. In England this equates to about 200 such deaths annually. Understandably, there can be great anxiety when this is familial and affects young adults. Over the past few years, the term sudden adult death syndrome (SADS) has been applied, but also in use are sudden cardiac death (SCD) and inherited cardiac condition (ICC). This group of conditions includes the long QT syndromes (LQTS), Brugada syndrome, catecholaminergic (stress-induced) polymorphic ventricular tachycardia (CPVT), and arrhythmogenic right ventricular cardiomyopathy (ARVC). LQTS and Brugada syndrome are sodium and potassium ion channelopathies. CPVT and ARVC demonstrate overlap with the inherited cardiomyopathies and some cases are due to molecular defects affecting the calcium channel. In ARVC there is often pathological evidence of either a hypertrophic or a dilated myocardium.

Inherited Arrhythmias

Clinical Features

When sudden unexplained death occurs, a careful review of the post-mortem findings and an exploration of the deceased’s history, as well as the family history, are indicated. Most of those who die are young males, and death occurs during sleep or while inactive. In a proportion of cases, death occurs while swimming, especially in LQT1. Emotional stress can be a trigger, especially in LQT2, and cardiac events are more likely in sleep for LQT2 and LQT3. Careful investigation and questioning may reveal an antecedent history of episodes of syncope, palpitation, chest discomfort, and dyspnea, and these symptoms should be explored in the relatives in relation to possible triggers. If the deceased had a 12-lead electrocardiogram (ECG), this may hold some key evidence; however, a normal ECG is present in about 30% of proven LQTS and possibly a higher proportion of Brugada syndrome cases.

In LQTS, also known as Romano-Ward syndrome, the ECG findings are dominated by, as the name suggests, a QT interval outside the normal limits, remaining long when the heart rate increases. They are classified according to the gene involved (Table 19.4). The inheritance is overwhelmingly autosomal dominant but a rare recessive form exists, combined with sensorineural deafness, which is known as Jervell and Lange-Nielsen syndrome. The ECG changes may be evident from a young age and a cardiac event occurs by age 10 years in about 50%, and by age 20 years in 90%. First cardiac events tend to be later in LQT2 and LQT3. Predictive genetic testing, where possible, is helpful to identify those at risk in affected families, and decisions about prophylactic β-blockade can be made. β-Blockers are particularly useful in LQT1 but less so in LQT2 and LQT3; indeed, it is possible that β-blockers may be harmful in LQT3.

Table 19.4 The Inherited Cardiac Arrhythmias

Brugada syndrome also follows autosomal dominant inheritance and was first described in 1992. The cardiac event is characterized by a proneness to idiopathic ventricular tachycardia (VT), and there may be abnormal ST-wave elevation in the right chest leads with incomplete right bundle branch block. In at-risk family members with a normal ECG, the characteristic abnormalities can usually be unmasked by the administration of potent sodium channel blockers such as flecainide. The condition is relatively common in Southeast Asia; there is a male predominance of 8 : 1, and the average age of arrhythmic events is 40 years. The definitive treatment is an implantable defibrillator and exercise is not a particular risk factor. Mutations in the SCN5A gene are found in about 20% of Brugada syndrome patients, as well as some cases of LQT3 (see Table 19.4). In some families both arrhythmias occur.

ARVC, which follows mainly dominant inheritance, is characterized by localized or diffuse atrophy and fatty infiltration of the right ventricular myocardium. It can lead to VT and sudden cardiac death in young people, especially athletes with apparently normal hearts. The ECG shows right precordial T-wave inversion and prolongation of the QRS complex. ARVC appears to demonstrate substantial genetic heterogeneity (see Table 19.4) with five genes identified, one of which, encoding plakoglobin, is implicated in the rare recessive form found on the island of Naxos. The RYR2 gene, for ARVC2, is also mutated in catecholaminergic polymorphic ventricular tachycardia (CPVT), also known as Coumel’s VT. Individuals with CPVT present with syncopal events, sometimes in childhood or adolescence, and reproducible stress-induced ventricular tachycardia, without a prolonged QT interval; the heart is structurally normal.

Genetics

These are genetically heterogeneous conditions. Nearly all follow autosomal dominant inheritance; the genes and their loci are summarized in Table 19.4. In some cases, however, there is evidence for biallelic inheritance (p. 119)—i.e., patients require mutations at two different loci to have clinical symptoms and ECG changes. This poses very significant difficulties in relation to genetic testing strategies, interpretation of mutation tests, and the usefulness of predictive genetic testing based on the findings at one locus. The same problem may also apply to some cases of cardiomyopathy.

Inherited Cardiomyopathies

Dilated cardiomyopathy is characterized by cardiac dilatation and reduced systolic function. Causes include myocarditis, coronary artery disease, systemic and metabolic diseases, and toxins. When these are excluded the prevalence of idiopathic dilated cardiomyopathy is 35 to 40 per 100,000 and familial cases account for about 25%. As with the inherited cardiac arrhythmias, they are genetically heterogeneous but nearly always follow autosomal dominant inheritance. They are also very variable, and within the same family affected members may show symptoms in childhood at one end of the spectrum, whereas in other individuals the onset of cardiac symptoms may not occur until late in adult life. At least 10 different loci have been mapped in different family studies. One cause is the result of mutations in the LMNA gene (which encodes lamin A/C), noted for its pleiotropic effects (p. 112), of which dilated cardiomyopathy is one and may occasionally be isolated.

Hypertrophic cardiomyopathy is similarly genetically heterogeneous but the large majority follow autosomal dominant inheritance. The group includes asymmetric septal hypertrophy, hypertrophic subaortic stenosis, and ventricular hypertrophy. The most common single gene involved appears to be that which encodes the cardiac β-myosin heavy chain (MYH7) on chromosome 14q but, again, there are at least a further eight loci mapped for genes encoding different cardiac muscle proteins. Sudden death can occur, especially in young athletes. Notable is cardiomyopathy due to mutations in the gene encoding the ‘T’ isoform of cardiac troponin (TNNT2), located on chromosome 1q32. This isoform is not expressed in skeletal muscle but, when mutated, a mild and sometimes subclinical hypertrophy results. Unfortunately, there is a high incidence of sudden death.

Genetic testing is now available within clinical services, but the vast genetic heterogeneity means that the pick-up rate for mutations is low. After a diagnosis has been made in an index case, a detailed family history is indicated and investigation by ECG and echocardiogram should be offered. Screening may need to continue well into adult life. Among the causes of cardiomyopathy that can be detected relatively easily by a biochemical test is X-linked fabry disease, for which enzyme replacement is available (see Table 23.1; p. 350).

Spinal Muscular Atrophy

Spinal muscular atrophy (SMA) is the term used to describe a clinically and genetically heterogeneous group of disorders that are among the most common genetic causes of death in childhood. The disease is characterized by degeneration of the anterior horn cells of the spinal cord leading to progressive muscle weakness and ultimately death.

Three common childhood forms of SMA are recognized with a collective incidence of approximately 1 : 10,000 and carrier frequency of about 1 : 50. Of these, SMA type I is the most common and the most severe. Although three types are delineated here, it is now clear that the disease spectrum forms a continuum.

Clinical Features

Spinal Muscular Atrophy Type I (Werdnig-Hoffmann Disease)

This form of SMA presents at birth or in the first 6 months of life with severe hypotonia and lack of spontaneous movement. Sometimes the mother will have noticed that intrauterine fetal movements were reduced in both strength and frequency. These children show normal intellectual activity but their profound muscle weakness, which affects swallowing and respiratory function, leads to death within the first 2 years of life. The diagnosis is confirmed by electromyography and there is no effective means of treating the disorder or even delaying its rate of progression.

Spinal Muscular Atrophy Type II

This is less severe than type I, with an age of onset between 6 and 18 months. As with type I, muscle weakness and hypotonia are the main presenting features. These children can sit unaided but are never able to achieve independent locomotion. The rate of progression is slow and most affected children survive into early adult life.

Spinal Muscular Atrophy Type III (Kugelberg-Welander Disease)

This relatively mild form of childhood or juvenile-onset SMA presents after 18 months and all patients are able to walk without support. Slowly progressive muscle weakness results in many affected individuals having to use a wheelchair by early adult life. Long-term survival can be jeopardized by recurrent respiratory infection and the development of a scoliosis caused by weakness of the spinal muscles.

Genetics

All three types of childhood-onset SMA show autosomal recessive inheritance. Several other much rarer forms of SMA have been described, and the late, adult onset forms may follow autosomal dominant inheritance. SMA type I generally shows a high degree of intrafamilial concordance, with affected siblings showing an almost identical clinical course. In types II and III, intrafamilial variation can be quite marked. In all childhood-onset SMA, the predominant gene involved is SMN1.

Mapping and Isolating the Spinal Muscular Atrophy Gene

All three childhood forms of SMA were mapped to chromosome 5q in 1990 using linkage. Detailed mapping narrowed the locus to a 1000 kb consisting of a 500-kb inverted duplication (Figure 19.10). This region is noted for its high rate of instability, with several DNA duplications and a relatively large number of pseudogenes (p. 17).

FIGURE 19.10 The inverted duplication with the SMN and NAIP genes. SMA occurs when both copies of the SMN1 gene are mutated (autosomal recessive inheritance); in 95% to 98% this is a deletion of exons 7–8, and point mutations in the remainder. SMN, Survival motor neuron; NAIP, neuronal apoptosis-inhibitory protein.

Within the candidate region two distinct genes were isolated that show a high incidence of deletion in patients with SMA—SMN and NAIP—each present as two almost identical copies. The SMN genes are now referred to as SMN1 and SMN2 (the pseudogene of SMN1 that shares ~99% homology). SMN1 shows homozygous deletion of exons 7–8 in 95% to 98% of all patients with childhood-onset SMA. Point mutations in SMN1 have been identified in 1% to 2% of patients with childhood SMA who do not show the exons 7–8 deletion on one allele. The number of copies of SMN2, arranged in tandem in cis configuration on each chromosome, varies between zero and five. It produces a similar transcript to SMN1 but this is not sufficient to fully compensate. Nevertheless, the presence of copies of SMN2 modifies the phenotype, causing milder forms of SMA.

The other gene originally isolated, NAIP, codes for the neuronal apoptosis-inhibitory protein, and is deleted in ~45% of individuals with SMA type I and ~20% of those with SMA types II and III. However, for the purposes of clinical molecular genetics, it is no longer considered relevant.

Clinical Applications and Future Prospects

SMN1 is always mutated in SMA, in the vast majority by deletion of exons 7–8, and in the remainder by point mutation. Diagnostic testing is therefore very reliable and prenatal testing is an option for those couples who request it, assuming both parents are carriers. Carrier detection is based on determining the number of exon 7–containing SMN1 gene copies present in an individual. However, results can be difficult to interpret because some carriers have the normal number of SMN1 gene copies caused by the presence either of two SMN1 gene copies in cis configuration on one chromosome, or of a SMN1 point mutation. About 4% of the general population has two copies of SMN1 on a single chromosome. Furthermore, 2% of individuals with SMA have one de novo mutation, meaning that only one parent is a carrier. Because of these difficulties, SMA carrier testing should be provided in the context of formal, expert genetic counseling.

Duchenne Muscular Dystrophy

Duchenne muscular dystrophy (DMD) is the most common and most severe form of muscular dystrophy. The eponymous title is derived from the French neurologist Guillaume Duchenne, who described a case in 1861. A similar but milder condition, Becker muscular dystrophy (BMD), is caused by mutations in the same gene. The incidences of DMD and BMD are approximately 1 : 3500 males and 1 : 20,000 males, respectively. Great efforts are being directed at devising novel treatments for these serious disorders.

Clinical Features

Males with DMD usually present between the ages of 3 and 5 years with slowly progressive muscle weakness resulting in an awkward gait, inability to run quickly, and difficulty in rising from the floor, which can be achieved only by pushing on, or ‘climbing up’, the legs and thighs (Gowers’ sign). Most affected boys have to use a wheelchair by the age of 11 years because of severe proximal leg muscle weakness. Subsequent deterioration leads to lumbar lordosis, joint contractures, and cardiorespiratory failure, resulting in death at a mean age of 18 years without aggressive supportive measures.

On examination, boys with DMD show an apparent increase in the size of the calf muscles, which is actually due to replacement of muscle fibers by fat and connective tissue—referred to as pseudohypertrophy (Figure 19.11). DMD is sometimes known as pseudohypertrophic muscular dystrophy. In addition, approximately one-third of boys with DMD show mild-moderate intellectual impairment, with the mean IQ being 83.

FIGURE 19.11 Lower limbs of an adult male with Becker muscular dystrophy showing proximal wasting and calf pseudohypertrophy.

In BMD the clinical picture is very similar but the disease process runs a much less aggressive course. The mean age of onset is 11 years and many patients remain ambulant until well into adult life. Overall life expectancy is only slightly reduced. A few patients with proven mutations in the DMD/BMD gene have been asymptomatic in their fifth or sixth decade.

Genetics

Both DMD and BMD show X-linked recessive inheritance. Males with DMD rarely, if ever, reproduce. Therefore, as genetic fitness equals zero, the mutation rate equals the incidence in affected males divided by 3 (p. 133), which approximates to 1 : 10,000—one of the highest known mutation rates in humans.

Isolation of the Gene for DMD

The isolation of the gene for DMD—the dystrophin gene—represented a major scientific achievement at the time, because of a successfully applied positional cloning strategy. The initial clue to the site of the DMD locus was provided by reports of several females affected with DMD who had a balanced X-autosome translocation with a common X-chromosome breakpoint at Xp21. In these women, those cells in which the derivative X chromosome is randomly inactivated are at a major disadvantage because of inactivation of the autosomal segment (Figure 7.16; p. 117). Consequently, cells in which the normal X chromosome has been randomly inactivated are more likely to survive. The net result is that the derivative X autosome is active in most cell lines, and if the breakpoint has damaged an important gene, in this case dystrophin, the woman will be affected by the disease in question.

Further support for mapping came from affected males with visible microdeletions involving Xp21 and the next phase of gene isolation was the identification of conserved sequences in muscle cDNA libraries that were shown to be exons from the gene itself. The task was completed in 1987.

The dystrophin gene is huge in molecular terms, consisting of 79 exons and 2.3 Mb of genomic DNA, though only 14 kb are transcribed into mRNA. It is transcribed in brain as well as muscle, which explains why some boys with DMD show learning difficulties. The large size may explain the high mutation rate.

Mutations in the Dystrophin Gene

Deletions, which can be almost any size and location, account for two-thirds of all dystrophin gene mutations and arise almost exclusively in maternal meiosis, probably due to unequal crossing over. A smaller number of affected males with duplications have been described. Deletion ‘hotspots’ are found in the first 20 exons as well as around exons 45 through 53. One of the deletion breakpoint ‘hotspots’ in intron 7 contains a cluster of transposon-like repetitive DNA sequences that could facilitate misalignment in meiosis, with a subsequent crossover leading to deletion and duplication products.

Deletions that cause DMD usually disturb the translational reading frame (p. 20), but those seen in males with BMD usually do not alter the reading frame (i.e., they are ‘in-frame’). This means that the amino-acid sequence of the protein product downstream of the deletion is normal, explaining the relatively mild features in BMD. Mutations in the remaining one-third of boys with DMD include stop codons, frameshift mutations, altered splicing signals and promoter mutations. Most lead to premature translational termination, resulting in the production of little, if any, protein product. In contrast to deletions, point mutations in the dystrophin gene often arise in paternal meiosis, most probably because of a copy error in DNA replication. Full sequencing of the dystrophin gene is now available as a service, which has transformed molecular diagnosis of DMD and carrier detection.

The Gene Product Dystrophin

The 427-kDa dystrophin protein is located close to the muscle membrane, where it links intracellular actin with extracellular laminin. Absence of dystrophin, as in DMD, leads gradually to muscle cell degeneration. The presence of dystrophin in a muscle biopsy sample can be assessed by immunofluorescence. Levels less than 3% are diagnostic. In muscle biopsies from males with BMD, the dystrophin shows qualitative rather than gross quantitative abnormalities.

Dystrophin binds to a glycoprotein complex in the muscle membrane through its C-terminal domain (Figure 19.12). This glycoprotein complex consists of several subunits, abnormalities of which cause other rare genetic muscle disorders, including several different types of autosomal recessive limb girdle muscular dystrophy, as well as congenital muscular dystrophy.

FIGURE 19.12 Probable structure of the dystrophin protein molecule, depicted as a dimer linking intracellular actin with extracellular laminin.

(Adapted from Ervasti JM, Campbell KP 1991 Membrane organization of the dystrophin–glycoprotein complex. Cell 66:1121–1131.)

Carrier Detection

Before DNA analysis, carrier detection was based on pedigree information combined with serum creatine kinase (CK) assay (p. 314). CK levels are grossly increased in boys with DMD, and marginally raised in approximately two-thirds of all carriers (see Figure 20.1; p. 314). CK levels are only occasionally useful today, as DNA testing has become increasingly sophisticated. Linkage studies may still be useful in circumstances where no DNA is available from an affected male, but perhaps available from normal males in the same family; each situation has to be assessed individually. Care has to be taken when using linkage for carrier detection because of the high recombination rate of 12% across the DMD gene.

Prospects for Treatment

At present, there is no cure for DMD or BMD, although physiotherapy is beneficial for maintaining mobility and preventing muscle spasm and joint contractures.

Gene therapy offers the only realistic hope of a cure in the short to medium term. Several approaches have been tried experimentally in transgenic and naturally occurring mutant mice with dystrophin-negative muscular dystrophy. These include direct injection of recombinant DNA, myoblast implantation, and transfection with retroviral or adenoviral vectors carrying a dystrophin minigene containing only those sequences that code for the important functional domains. One approach is antisense technology to block an exon splicing enhancer sequence and generate a protein with an in-frame deletion that encodes a protein with some residual function (i.e., a BMD rather than a DMD phenotype). That mice with dystrophin-negative muscular dystrophy can show spontaneous muscle repair indicates that there could be a way of switching on an alternative compensatory protein such as utrophin. This is expressed in the fetus rather than dystrophin, with which it shares a large degree of homology. Genetically engineered mice deficient for both dystrophin and utrophin develop a typical DMD dystrophy. If the utrophin gene could be reactivated as in the dystrophin-negative mouse, this may have a therapeutic benefit.

Hemophilia

There are two forms of hemophilia: A and B. Hemophilia A is the most common severe inherited coagulation disorder, with an incidence of 1 : 5000 males. It is caused by a deficiency of factor VIII, which, together with factor IX, plays a critical role in the intrinsic pathway activation of prothrombin to thrombin. Thrombin then converts fibrinogen to fibrin, which forms the structural framework of clotted blood. The existence of hemophilia was recognized in the Talmud, and the tendency for males to be affected much more often than females was acknowledged by the Jewish authorities 2000 years ago when they excused from circumcising the sons of the sisters of a mother who had an affected son. Queen Victoria was a carrier and, as well as having an affected son—Leopold Duke of Albany—she transmitted the disorder through two of her daughters to most of the royal families of Europe (Figure 19.13).

FIGURE 19.13 Pedigree showing the segregation of hemophilia among Queen Victoria’s descendants.

Hemophilia B affects approximately 1 : 40,000 males and is caused by deficiency of factor IX. It is also known as Christmas disease, whereas hemophilia A is sometimes referred to as ‘classic hemophilia’.

Clinical Features

FIGURE 19.14 Lower limbs of a male with hemophilia showing the effect of recurrent hemorrhage into the knees.

(Courtesy Dr. G. Dolan, University Hospital, Nottingham, UK.)

Genetics

Both forms of hemophilia show X-linked recessive inheritance. The loci lie close together near the distal end of Xq.

Hemophilia A

The factor VIII comprises 26 exons and spans 186 kb with a 9-kb mRNA transcript. Deletions account for 5% of all cases and usually cause complete absence of factor VIII expression. In addition, hundreds of frameshift, nonsense, and missense mutations have been described, besides insertions and a ‘flip’ inversion, which represented a new form of mutation when first identified in hemophilia A in 1993. Inversions account for 50% of all severe cases with <1% factor VIII activity. They are caused by recombination between a small gene called A located within intron 22 of the factor VIII gene and other copies of the A gene, which are located upstream near the telomere (Figure 19.15). The inversion disrupts the factor VIII gene, resulting in very low factor VIII activity. The genetic test is straightforward but detection of the numerous other mutations may require direct sequencing.

FIGURE 19.15 How intrachromosomal recombination causes the ‘flip’ inversion, which is the most common mutation found in severe hemophilia A.

(Adapted from Lakich D, Kazazian HH, Antonarakis SE, Gitschier J 1993 Inversions disrupting the factor VIII gene are a common cause of severe hemophilia A. Nat Genet 5:236–241.)

As in DMD, point mutations usually originate in male germ cells whereas deletions arise mainly in the female. The ‘flip’ inversions show a greater than 10-fold higher mutation rate in male compared with female germ cells, probably because Xq does not pair with a homologous chromosome in male meiosis—so that there is much greater opportunity for intrachromosomal recombination to occur via looping of the distal long arm (see Figure 19.15).

Factor VIII levels are about half normal in carrier females and many are predisposed to a bleeding tendency. Carrier detection used to be based on assay of the ratio of factor VIII coagulant activity to the level of factor VIII antigen but, as with CK assay in DMD, this is not always discriminatory, and direct mutation analysis is now routine. Linkage analysis may sometimes be necessary to resolve carrier status.