Angelman Syndrome

A number sign (#) is used with this entry because 4 known genetic mechanisms can cause Angelman syndrome (AS). Approximately 70% of AS cases result from de novo maternal deletions involving chromosome 15q11.2-q13; approximately 2% result from paternal uniparental disomy of 15q11.2-q13; and 2 to 3% result from imprinting defects. A subset of the remaining 25% are caused by mutations in the gene encoding the ubiquitin-protein ligase E3A gene (UBE3A; 601623) (Kishino et al., 1997).

See also X-linked mental retardation, Christianson type (300243), which shows phenotypic overlap with Angelman syndrome.

Description

Angelman syndrome is a neurodevelopmental disorder characterized by mental retardation, movement or balance disorder, typical abnormal behaviors, and severe limitations in speech and language. Most cases are caused by absence of a maternal contribution to the imprinted region on chromosome 15q11-q13. Prader-Willi syndrome (PWS; 176270) is a clinically distinct disorder resulting from paternal deletion of the same 15q11-q13 region. In addition, the chromosome 15q11-q13 duplication syndrome (608636) shows overlapping clinical features.

Clayton-Smith and Pembrey (1992) provided a review of Angelman syndrome. Cassidy and Schwartz (1998) reviewed the molecular and clinical aspects of both Prader-Willi syndrome and Angelman syndrome. Horsthemke and Wagstaff (2008) provided a detailed review of the mechanisms of imprinting of the Prader-Willi/Angelman syndrome region.

Van Buggenhout and Fryns (2009) provided a review of Angelman syndrome and discussed genetic counseling of the disorder, which can show a recurrence risk of up to 50%, depending on the underlying genetic mechanism.

Clinical Features

Angelman (1965) reported 3 'puppet children,' as he called them. Angelman (1965) emphasized the abnormal cranial shape and suggested that the depressed occiput may reflect a cerebellar abnormality. (Harry Angelman pronounces his name as though it means 'male angel;' in other words, he uses a 'long a' and a 'soft g.') Bower and Jeavons (1967) coined the name 'happy puppet' syndrome for the condition that they observed in 2 patients. Clinical features included severe motor and intellectual retardation, ataxia, hypotonia, epilepsy, absence of speech, and unusual facies characterized by a large mandible and open-mouthed expression revealing the tongue. The French refer to the syndrome as that of the 'marionette joyeuse' (Halal and Chagnon, 1976) or 'pantin hilare' (Pelc et al., 1976). Williams and Frias (1982) suggested use of the eponym Angelman syndrome because the term 'happy puppet' may appear derisive and even derogatory to the patient's family.

Berg and Pakula (1972) reported a case and reviewed those reported by Angelman (1965) and Bower and Jeavons (1967). All of the patients demonstrated excessive laughter, an occipital groove, a great facility for protruding the tongue, abnormal choroidal pigmentation, and characteristic electroencephalogram (EEG) discharges. Of the 3 patients reported by Angelman (1965), at least 1 developed optic atrophy. Two patients showed jerky movements and had trouble walking, which was believed to result from poor balance. One, a 9-year-old boy who was noticed as an infant to be 'floppy,' could take only a few steps without support. Both patients had major convulsions and showed periods of flapping their arms up and down with the elbows flexed. The EEG pattern seen in these 2 cases and in the cases of Bower and Jeavons (1967) consisted of high amplitude bilateral spike-and-wave activity which was symmetrical, synchronous, and most often monorhythmic, having a slow wave component at 2 cycles per sec. The patient reported by Berg and Pakula (1972) had an unaffected sib who also showed abnormal EEG patterns. Normal karyotype was found in the 5 patients studied.

Williams and Frias (1982) demonstrated unilateral cerebellar atrophy by CT imaging in 1 patient with AS.

In 6 of 8 children with AS, aged 3 to 10 years, Dickinson et al. (1988) found an association of striking deficiency of choroidal pigment with normal foveal reflexes. All 6 had light blue irides with normal iris architecture. All were isolated cases born to healthy, unrelated parents. The presence or absence of 15q microdeletions did not correlate with the ocular findings.

In a review of clinical features in 36 children with Angelman syndrome, Robb et al. (1989) reported global developmental delay, seizures, episodes of paroxysmal laughter, and tongue thrusting. The movement disorder consisted of a wide-based, ataxic gait with frequent jerky limb movements and flapping of the hands.

Fryburg et al. (1991) described the clinical features in 4 patients diagnosed at less than 2 years of age. One of their patients had oculocutaneous albinism, and all were hypopigmented compared to their first-degree relatives. All 4 had choroidal pigment hypoplasia, severe to profound global developmental delay and microcephaly of postnatal onset, seizures, hypotonia, hyperreflexia, and hyperkinesis. Clayton-Smith (1993) reported on observations concerning 82 affected individuals. All of them had absent speech or spoke less than 6 words. Thirty-nine percent were hypopigmented compared to their family members. Frequent smiling was present in 96%. King et al. (1993) concluded from the study of 6 individuals with AS that hypopigmentation characterized by light skin, reduced retinal pigment, low hairbulb tyrosinase activity, and incomplete melanization of melanosomes is part of the phenotype of AS, and is similar to that found in Prader-Willi syndrome.

Viani et al. (1995) found EEG evidence of transient myoclonic status epilepticus in 9 of 18 Angelman patients, which likely corresponded to recurrent jerky abnormal movements observed in these patients. In addition, 7 patients had partial seizures with eye deviation and vomiting similar to those of childhood occipital epilepsies.

Reish and King (1995) established the diagnosis of Angelman syndrome in a 50-year-old woman. She had been healthy without seizures and had a history of pelvic fracture resulting from her unbalanced gait. She was born to a 40-year-old mother. Her height was 148 cm and her IQ was measured at less than 20. She did not speak and had frequent bursts of laughter. Reish and King (1995) demonstrated a 15q11.2-q12 deletion by karyotypic examination and fluorescence in situ hybridization (FISH).

Buntinx et al. (1995) compared the main manifestations of Angelman syndrome in 47 patients at different ages. Most patients between the ages of 2 and 16 years showed at least 8 of the major characteristics of the syndrome (bursts of laughter, happy disposition, hyperactivity, micro- and brachycephaly, macrostomia, tongue protrusion, prognathism, widely spaced teeth, puppet-like movements, wide-based gait) in addition to mental retardation and absence of speech. Most patients (80.8%) had epileptic seizures, starting after the age of 10 months. In children under the age of 2 years, bursts of laughter was found in 42.8% and macrostomia in only 13.3%, but protruding tongue was a constant feature. In patients over 16 years of age, protruding tongue was found in 38.8%, whereas prognathism and macrostomia were almost constant findings. A cytogenetic deletion was found in 61% and a molecular deletion in 73% of the patients. No case of paternal disomy was found. Buntinx et al. (1995) found no differences between patients with or without deletion on chromosome 15q. The authors noted that the diagnosis of Angelman syndrome may be hampered in young children because of the absence of some typical manifestations and in older patients because of the changing behavioral characteristics.

Smith et al. (1996) reviewed the clinical features of 27 Australian patients with AS, all with a DNA deletion involving 15q11-q13 and spanning markers from D15S9 to D15S12 (approximately 3.5 Mb of DNA). There were 9 males and 18 females, all sporadic cases, ranging in age from 3 to 34 years, and all ataxic, severely retarded, and lacking in recognizable speech. Head circumference at birth was normal in all but skewed in distribution, with 62.5% at the 10th centile. Epilepsy was present in 96% with onset during the third year of life in 20 of 26 patients. Hypopigmentation was present in 19 (73%). One patient had ocular cutaneous albinism. A happy disposition was noted in infancy in 95% and they all had a large, wide mouth.

Among 22 institutionalized adults selected for criteria suggestive of Angelman syndrome, Sandanam et al. (1997) found deletion in the 15q11-q13 region in 11 (9 males and 2 females). The mean age at last review was 31.5 years (range 24 to 36 years). Clinical assessment documented findings of large mouth and jaw with deep set eyes and microcephaly in 9 patients (2 having a large head size for height). No patient was hypopigmented; 1 patient was fair. Outbursts of laughter occurred in all patients, but infrequently in 7 of 11 (64%), and a constant happy demeanor was present in 5 of 11 (46%). All had epilepsy, with improvement in 5 (46%), no change in 4 (36%), and deterioration in 2 (18%). The EEG was abnormal in 10 of 10 patients. Ocular abnormalities were reported in 3 of 8 patients (37.5%), with keratoconus present in 2, and 4 of 11 (36%) developed kyphosis. Two had never walked. All 9 who walked were ataxic with an awkward, clumsy, heavy, and/or lilting gait. No patient had a single word of speech, but 1 patient could use sign language for 2 needs, food and drink. The findings of Sandanam et al. (1997) supported the concept that AS resulting from deletion is a severe neurologic syndrome in adulthood.

Lossie and Driscoll (1999) described a pregnancy in a 15-year-old female with AS who had been reported by Williams et al. (1989). Williams et al. (1989) had raised the possibility that the proband's mother, who had normal intelligence, was mosaic for a submicroscopic deletion of 15q11-q13, because she displayed brachycephaly, hearing loss, an enlarged foramen magnum, and mild ataxia. However, extensive cytogenetic and molecular analyses of peripheral blood and skin fibroblasts failed to reveal any abnormality in 15q11-q13 in the mother. The daughter had classic AS features, with severe mental retardation, AS-specific behavior, complete lack of speech, and a movement disorder characterized by ataxia. She showed microbrachycephaly with a head circumference of less than -2 standard deviations, relative prognathism, a protruding tongue, excessive drooling, and an inappropriately happy affect with excessive laughter. Menarche began at 11.5 years. Head CT and MRI were remarkable only for an enlarged foramen magnum. The pregnancy was terminated at 15 to 16 weeks' gestation. The fetus had inherited large deletions of maternal 15q11-q13 and demonstrated paternal-only DNA methylation imprints along 15q11-q13. UBE3A was paternally expressed in eye tissue from the fetus. These results indicated that females with AS are fully capable of reproduction and that UBE3A is not imprinted in fetal eye.

Valente et al. (2006) reported the features of epilepsy of 19 patients with AS caused by deletion of 15q11-q13. All had generalized seizures, and 10 (53%) also had partial seizures. Types of seizures included atypical absence (84%), myoclonic (68%), generalized tonic-clonic or tonic (63%), simple partial with motor phenomena (32%), complex partial (26%), and myoclonic-astatic (11%). The mean age at seizure onset was 13 months (range 4 months to 2 years and 11 months). In 18 patients, seizure onset preceded diagnosis of AS. Sixteen (84%) patients had status epilepticus, of which 7 cases were recurrent, and 53% of patients had worsening with fever. Although complete seizure control was achieved in only 37% of patients, there was a tendency toward age-related improvement during late childhood and puberty.

Michieletto et al. (2011) detailed ophthalmologic findings in 34 consecutive patients with a confirmed diagnosis of Angelman syndrome admitted to their institution for neurologic examination. The patients represented 3 genetic classes: deletion, uniparental disomy, and mutation. Ametropia (refractive error) greater than 1 diopter (D) was present in 97% of cases: myopia in 9%, hyperopia in 76%, and astigmatism in 94%. Myopia and anisometropia (unequal refractive errors) were found only in the genetic deletion group. Strabismus, most frequently exotropia, was found in 24 patients (75%). Ocular hypopigmentation was observed in 18 subjects (53%), with choroidal involvement in 3 cases and isolated iris involvement in 4. Hypopigmentation was observed in all of the genetic classes. Michieletto et al. (2011) stated that ophthalmic alterations were observed more frequently in this study than had previously been reported, except for ocular hypopigmentation, which was observed less frequently.

By gathering data from standardized phone interviews with caregivers, Larson et al. (2015) ascertained the primary health issues of 110 adolescents and adults with Angelman syndrome. Important features included active seizures (41%), sleep dysfunction (72%), constipation (85%), obesity (32%), scoliosis (50%), and self-injurious behavior (52%). Only 13% of patients could speak 5 or more words, suggesting that impaired communication is a significant feature of this condition.

Diagnosis

Boyd et al. (1988) pointed out the usefulness of the EEG in the early diagnosis of Angelman syndrome. Dorries et al. (1988) described 7 cases and concluded that the diagnosis is difficult in the first years of life.

The American Society of Human Genetics/American College of Medical Genetics Test and Technology Transfer Committee (1996) reviewed diagnostic testing for Prader-Willi syndrome and Angelman syndrome.

Stalker and Williams (1998) addressed the challenges of genetic counseling in this disorder with multiple causes. Most cases result from typical large de novo deletions of 15q11-q13 and are expected to have a low (less than 1%) risk of recurrence. AS due to paternal uniparental disomy, which occurs in the absence of a parental translocation, is likewise expected to have a recurrence risk of less than 1%. Parental transmission of a structurally or functionally unbalanced chromosome complement can lead to 15q11-q13 deletions or to UPD and will result in case-specific recurrence risks. In instances where there is no identifiable large deletion or UPD, the risk of recurrence may be as high as 50% as a result of either a maternally inherited imprinting center mutation or a mutation in the UBE3A gene. Individuals with AS who have none of the above abnormalities comprise a significant proportion of cases, and some may be at a 50% recurrence risk. Misdiagnoses can be represented in this group as well. In light of the many conditions that are clinically similar to AS, it is essential to address the possibility of diagnostic uncertainty and potential misdiagnosis before providing genetic counseling. Stalker and Williams (1998) presented an algorithmic chart summarizing the different causal classes of AS for consideration in determining recurrence risks.

Tekin et al. (2000) described a patient with clinical features of Angelman syndrome in whom FISH analysis revealed mosaicism for a deletion in the AS critical region, but whose methylation study results were normal. The authors recommended that FISH studies for detection of mosaicism be done in patients with clinical findings of AS even if methylation studies are normal.

Hall (2002) reported an apparently unique response by Angelman syndrome individuals to the vibrating tuning fork when it was held up to their ears. The response was a wide smile, often with an outburst of laughter, followed by a tendency to lean toward the vibrating tuning fork. In 6 consecutive Angelman individuals ranging in ages from 18 months to 43 years, they demonstrated a positive 'tuning fork response.' The 2 oldest individuals, aged 17 and 43 years, tended to be somewhat less demonstrative with mostly smiles and a more controlled laugh. Parents had observed their affected children as liking sound. This feature was manifested by their lying down or leaning against appliances that made a noise as if it relaxed them or made them feel good. Hall (2002) raised the possibility of the potential use of sound in intervention strategies for these individuals. Hall and Cadle (2002) described a 12-month-old child, later confirmed to have Angelman syndrome, who had a positive tuning fork response. The authors suggested that this test, if found to be positive in Angelman syndrome children at ages 2 to 12 months, may aid in the often difficult first-year diagnosis.

Williams et al. (2006) provided an updated consensus for diagnostic criteria of Angelman syndrome. The list of associated findings was expanded to include abnormal food related behaviors, obesity, constipation, and scoliosis. In addition, some patients show attraction to or fascination with water and 'crinkly' items, such as papers and plastics. Sleep disturbances include abnormal sleep-wake cycles and diminished need for sleep.

The clinical diagnosis of Angelman syndrome is based on the presence of all 4 major criteria, i.e., developmental delay, speech impairment, movement or balance disorder, and behavioral characteristics, as well as the presence of 3 of 6 minor criteria, including postnatal deceleration of head growth, seizures, abnormal EEG, sleep disturbance, attraction to or fascination with water, and drooling (summary by Tan et al., 2011).

Differential Diagnosis

Scheffer et al. (1990) pointed out the possible confusion with Rett syndrome.

Pointing out that the diagnosis of Angelman syndrome can be confirmed by a genetic laboratory in only about 80% of cases, Williams et al. (2001) reviewed several mimicking conditions, including microdeletions or microduplications. Single gene conditions include methylenetetrahydrofolate reductase deficiency (236250), Rett syndrome, alpha-thalassemia retardation syndrome (ATRX; 301040), and Gurrieri syndrome (601187). There are, in addition, symptom complexes, including cerebral palsy (see 603513), autism spectrum disorder (209850), and pervasive developmental delay (PDD), that can suggest Angelman syndrome.

Inheritance

Angelman syndrome results from a lack of maternal contribution from chromosome 15q11-q13, arising from de novo deletion in most cases or from uniparental disomy in rare cases. Most families are therefore associated with a low recurrence risk. Although Angelman syndrome is not typically mendelian, familial occurrence has been reported.

Pashayan et al. (1982) reported Angelman syndrome in 2 brothers, Hersh et al. (1981) reported affected monozygotic twins, and Kuroki et al. (1980) reported 2 affected sisters. Dijkstra et al. (1986) and Fisher et al. (1987) reported affected brothers and sisters. Baraitser et al. (1987) reported 7 cases of Angelman syndrome from 3 families: 2 brothers in the first family, 3 sisters in the second, and 2 brothers in the third. The EEG changes were striking in all 7 patients. Robb et al. (1989) observed 3 sibships with more than 1 affected sib: 3 affected sisters, 2 affected brothers, and 2 affected sisters. Pashayan et al. (1982) found reports of 27 sporadic cases with a male-to-female ratio of 1:1. Paternal age was not remarkable in the patients of Williams and Frias (1982). Willems et al. (1987) reported what they believed to be the fourth family with affected sibs out of a total of 52 cases in the literature. The findings suggested a low but not negligible recurrence risk.

Clayton-Smith et al. (1992) studied 11 AS patients and their parents from 5 families using high resolution chromosome analysis and molecular probes from the region 15q11-q13. No deletions were detected. All sets of sibs inherited the same maternal chromosome 15, whereas in 3 families sibs inherited a different paternal chromosome 15. Polymorphic DNA markers gave the same conclusion. The findings indicated that autosomal recessive inheritance is very unlikely and suggested maternal transmission of a mutation within 15q11-q13.

Abaied et al. (2010) reported a large highly consanguineous Tunisian kindred with a severe form of Angelman syndrome, with mental retardation, motor impairment, seizures, hyperactivity, and frequent laughing. Genetic analysis identified a heterozygous truncating mutation in the UBE3A gene (601623.0011). There were 14 affected individuals, who were all in the same generation, and all patients inherited the mutation from their carrier mothers, who were 4 sisters. These 4 sisters apparently inherited the mutation from their unaffected father, who was deceased. Abaied et al. (2010) noted that the detection of mutations in large AS families emphasizes the importance of available genetic counseling and meticulous family history investigation.

Cytogenetics

Maternal 15q Deletions and Genomic Imprinting

Approximately 70% of cases of Angelman syndrome result from de novo maternal deletions involving the 15q11.2-q13 critical region (Kishino et al., 1997).

Magenis et al. (1987) reported 2 unrelated girls with a deletion of the proximal part of chromosome 15q similar to that observed in Prader-Willi syndrome. However, the girls showed clinical features consistent with Angelman syndrome, including ataxia-like incoordination, frequent, unprovoked and prolonged bouts of laughter, and a facial appearance compatible with that diagnosis. None of the typical features of Prader-Willi syndrome were present. Kaplan et al. (1987) also described deletion in 15q11-q12 in a child with Angelman syndrome. Magenis et al. (1988) proposed that patients with AS and PWS share an identical deletion on chromosome 15q11. Analysis of 6 AS patients and 6 PWS patients suggested that the deletion in AS was slightly larger and also included band q12. Magenis et al. (1988) proposed that genes in band 15q12 are responsible for the greater severity of mental retardation and speech in AS, and that these genes may also suppress or alter the presumed hypothalamic abnormality that results in the uncontrolled appetite and obesity of PWS.

By molecular analyses, Donlon (1988), Williams et al. (1988), and Knoll et al. (1989) showed that similar deletions of 15q11.2 were present in patients with Prader-Willi syndrome and Angelman syndrome. However, whereas the deleted chromosome was of paternal origin in PWS, the deleted chromosome was of maternal origin in AS. Otherwise, the deletions in the 2 disorders were indistinguishable cytogenetically or by molecular genetic methods. The findings were interpreted as indicating imprinting of chromosomes, i.e., changes in the chromosome according to the parent of origin, with resulting consequences for early development. By high-resolution cytogenetic studies, Magenis et al. (1990) found that the same proximal band, 15q11.2, was deleted in both PWS and AS. In general, the deletion in patients with Angelman syndrome was larger, though variable, and included bands q12 and part of q13. The authors confirmed the maternal origin of the deleted chromosome in AS, contrasting with the predominant paternal origin of the deletion in patients with Prader-Willi syndrome.

After discovering 2 unrelated AS patients with a small deletion of proximal 15q, Pembrey et al. (1987, 1989) reassessed 10 further patients. Four showed a deletion within 15q11-q13, 1 showed an apparent pericentric inversion with breakpoints at 15q11 and q13 inherited from the mother, and 5 showed no discernible abnormality. Of the 5 children without discernible chromosome change, 1 had a definitely affected sib and 1 had a possibly affected sib. Of the 4 sets of parents studied, 3 had normal chromosomes, and in 1 the mother had a deletion of 15q11.2 but not 15q12. Like Pembrey et al. (1989), Fryns et al. (1989) found a visible chromosomal change in half of the patients they studied. No deletion was found in 2 affected sisters.

By flow karyotype analysis on lymphoblastoid cell lines, Cooke et al. (1989) confirmed the presence of a de novo 15q deletion in a child with Angelman syndrome. The deleted segment represented 6.1 to 9.5% of chromosome 15, or approximately 6-9.3 million basepairs. Cytogenetic evidence suggested that the deleted chromosome was derived from the smaller chromosome 15 homolog of the mother.

Knoll et al. (1990) studied DNA of 19 AS patients, including 2 sib pairs, using 4 DNA markers specific to 15q11-q13. They identified 3 classes: in class I, deletion of 2 markers was detected; in class II, deletion of 1 marker; and in class III, including both sib pairs, no deletion was detected. High resolution cytogenetic data were available on 16 of the patients, and complete concordance between the presence of a cytogenetic deletion and a molecular deletion was observed. No submicroscopic deletions were detected by the DNA studies. DNA samples from the parents of 10 patients with either a class I or a class II deletion were available for study. In 7 of the 10 families, RFLPs were informative as to the parental origin of the deletion, and in all, the deleted chromosome was of maternal origin.

Imaizumi et al. (1990) described 6 patients, including 2 sibs, with Angelman syndrome. The 4 sporadic cases showed a microdeletion in the proximal part of 15q, whereas the affected sibs had no visible deletion. No clinical difference between the sporadic cases and the sib cases was discerned. Using 2 DNA probes that detect a molecular deletion in most patients with Prader-Willi syndrome, they found by densitometry that 2 patients had only 1 copy of each probe, whereas the other 4, including the sibs, had 2 copies of each sequence. Imaizumi et al. (1990) concluded that the segment causing AS may be different from that causing PWS.

Williams et al. (1990) studied 6 AS patients with de novo deletions of 15q11-q13. In 4 of the patients, cytogenetic studies were informative of parental origin; in all, the deletion was inherited from the mother, suggesting genomic imprinting. Malcolm et al. (1990) studied 37 typical cases. A 15q11-q13 deletion was observed in 18 of 24 isolated cases. No deletion was observed in 13 cases from 6 families with more than 1 affected child. In 11 cases it was possible to elucidate the parental origin of the deleted chromosome and these were shown to be predominantly maternal. Greenstein (1990) presented a kindred in which both the Prader-Willi and Angelman syndromes were found; the inheritance pattern was consistent with genetic imprinting.

Hulten et al. (1991) reported an extraordinary family showing segregation of a balanced translocation t(15;22)(q13;q11) and 2 cases of Prader-Willi syndrome and 1 of Angelman syndrome. It appeared that the females carrying the balanced translocation had a high risk of having children with AS, while their brothers had a high risk of having children with PWS, again indicating genomic imprinting.

All 4 AS patients described by Fryburg et al. (1991) had deletions in the 15q11.2-q13 region. Parental chromosomes were available for study in 3 of these cases; in all 3 the deleted chromosome 15 was maternally derived. Similarly, Smith et al. (1992) found the deletion of band 15q12 to be of maternal origin in all 25 cases of AS that they examined. The parental origin was determined using cytogenetic markers in 13 of the cases, by the pattern of inheritance of RFLPS in 9, and by both techniques in 3. Tonk et al. (1992) found cytogenetic deletion of 15q12 in 3 cases of AS and by heteromorphism studies showed that the deleted chromosome was maternal in all 3.

Chan et al. (1993) presented a series of 93 Angelman syndrome patients, showing the relative contribution of the various genetic mechanisms. Sporadic cases accounted for 81 AS patients, while 12 cases came from 6 families. Deletions in 15q11-q13 were detected in 60 cases by use of a set of highly polymorphic (CA)n repeat markers and conventional RFLPs. In 10 sporadic cases and in all 12 familial cases, no deletion was detectable. In addition, 2 cases of de novo deletions occurred in a chromosome 15 carrying a pericentric inversion. In one of these the AS child had a cousin with Prader-Willi syndrome arising from a de novo deletion in an inverted chromosome 15 inherited from his father. The other case arose from a maternal balanced t(9;15)(p24;q15) translocation. There were 3 cases of uniparental disomy. In the familial cases, all affected sibs inherited the same maternal chromosome 15 markers for the region 15q11-q13. Cytogenetic analysis detected only 42 of the 60 deletion cases. Chan et al. (1993) stated that cytogenetic analysis was still essential to detect chromosomal abnormalities other than deletions such as inversions and balanced translocations, both of which have an increased risk for deletions.

To elucidate the mechanism underlying the deletions that lead to PWS and AS, Amos-Landgraf et al. (1999) characterized the regions containing 2 proximal breakpoint clusters and a distal cluster. Analysis of rodent-human somatic cell hybrids, YAC contigs, and FISH of normal or rearranged chromosomes 15 identified duplicated sequences, termed 'END' repeats, at or near the breakpoints. END-repeat units are derived from large genomic duplications of the HERC2 gene (605837) (Ji et al., 1999). Many copies of the HERC2 gene are transcriptionally active in germline tissues. Amos-Landgraf et al. (1999) postulated that the END repeats flanking 15q11-q13 mediate homologous recombination resulting in deletion. Furthermore, they proposed that active transcription of these repeats in male and female germ cells may facilitate the homologous recombination process.

In a study of 45 Finnish AS patients, Kokkonen and Leisti (2000) found 2 affected sibs, a 16-year-old boy and a 5-year-old girl, in whom the diagnosis was made at 8 years and at 3 months of age, respectively. Both parents and an 18-year-old brother were healthy. The 2 sibs were found to have del(15)(q11q13); the mother's chromosomes 15 were structurally normal, whereas the patients and their unaffected brother shared an identical maternally derived haplotype outside the deletion region. These findings were suggestive of maternal germline mosaicism of del(15)(q11q13).

Angelman syndrome deletions and rearrangements tend to occur at specific 'hotspots' or breakpoint (BP) clusters in proximal 15q (see Pujana et al., 2002): 2 proximal clusters, referred to as BP1 and BP2, are the breakpoints for class I and class II patients, respectively. The most common distal breakpoint, BP3, is located between markers D15S12 and D15S24. Two other breakpoint regions called BP4 and BP5 have been mapped distal to BP3, between markers D15S24 and D15S144. Gimelli et al. (2003) reported that some mothers of AS patients with deletions of the 15q11-q13 region have a heterozygous inversion involving the region that is deleted in the affected offspring. The inversion was detected in the mothers of 4 of 6 AS cases with the breakpoint 2-3 (BP2/3) 15q11-q13 deletion, but not in 7 mothers of AS cases due to paternal UPD 15. Variable inversion breakpoints were identified within breakpoint segmental duplications in the inverted AS mothers, as well as in AS deleted patients. The BP2-BP3 chromosome 15q11-q13 inversion was detected in 4 of 44 control subjects. Gimelli et al. (2003) hypothesized that the BP2/3 inversion may be an intermediate state that facilitates the occurrence of 15q11-q13 BP2/3 deletions in the offspring.

Approximately one-third of Angelman patients have an imprinting defect (ID) but no imprinting center deletion, suggesting that they may mosaicism of ID cells and normal cells. In 2 patients studied, Nazlican et al. (2004) demonstrated somatic mosaicism by molecular and cellular cloning. X-inactivation studies of cloned fibroblasts from 1 patient suggested that ID occurred before the blastocyst stage. Using a quantitative methylation assay based on real-time PCR, the authors detected from less than 1% to 40% normal cells among 24 Angelman patients tested. Regression analysis suggested that patients with a higher percentage of normally methylated cells tended to have milder clinical symptoms.

Paternal Uniparental Disomy

Approximately 2% of cases of Angelman syndrome result from paternal uniparental disomy (UPD) of 15q11-q13 (Kishino et al., 1997).

Malcolm et al. (1991) found evidence of uniparental paternal disomy in 2 patients with AS. Knoll et al. (1991) examined the DNA from 10 AS patients, at least 7 of whom were familial cases, with no cytogenetic or molecular deletion of chromosome 15q11-q13. In each case, 1 maternal copy and 1 paternal copy of 15q11-q13 was observed. The authors concluded that UPD is not a frequent cause of familial AS. Engel (1991), who introduced the concept of uniparental disomy in 1980 (Engel, 1980), took Knoll et al. (1991) to task for their conclusion that uniparental disomy may be rare in this disorder and urged further studies.

Paternal uniparental disomy was demonstrated by Freeman et al. (1993) in a child with a balanced 15;15 translocation. DNA polymorphisms demonstrated that the patient was homozygous at all loci for which the father was heterozygous, suggesting that the structural rearrangement was an isochromosome 15q and not a Robertsonian translocation.

Engel (1993) reviewed the possible mechanisms for uniparental disomy. One possibility is gamete complementation, i.e., the gamete from one parent containing both chromosomes of the pair and that from the other parent containing neither. When gamete complementation is the mechanism, the centromeres of the resulting pair will be heterodisomic if resulting from a meiosis 1 error, and isodisomic if resulting from a meiosis 2 error. Beyond that, meiosis 1 UPD, depending on crossing-over and segregation, may be wholly heterodisomic (holo-heterodisomy) or partially isodisomic (mero-isodisomy); meiosis 2 UPD should always result in an element of isodisomy embodied in the 2 segments of the nonseparated chromatids left unaffected by crossing-over. This unaffected segment thus tends to be juxtacentromeric. Gametic complementation UPD was reported by Wang et al. (1991), who found paternal heterodisomy for chromosome 14 in a 45,XX,t(13q14q)der pat proposita, whose 2 parents were balanced heterozygotes for a translocation involving chromosome 14. This situation is analogous to the effects of biparental translocation as in the mouse experiments of Cattanach and Kirk (1985). A second mechanism of UPD is so-called trisomy rescue or correction. It is expected that the remaining pair, after loss of the extra homolog, will be biparental in two-thirds of cases and uniparental in one-third of cases. In such instances, as in gamete complementation, isodisomy may or may not be present. Cases of UPD in Prader-Willi syndrome whose chromosomal 15 maternal disomy could be traced to a placental mosaicism for trisomy 15 documented at the time of choriocentesis (chorion villus sampling) performed for advanced maternal age were reported by Cassidy et al. (1992) and Purvis-Smith et al. (1992). A third situation is akin to the second; the abnormal initial zygotic situation is monosomy rather than trisomy and the abnormality is 'corrected' through duplication of the single available homolog. The case of cystic fibrosis with maternal chromosome 7 isodisomy and growth delay reported by Spence et al. (1988) may have been of this type, although there is at least one other explanation. Donnai (1993) pointed out that Robertsonian translocations, occurring with a frequency of about 1 in 10,000 live births, may be an important cause of UPD; such has been demonstrated to be the case for 13/15, 13/14, 14/14, and 22/22 translocations. Dysmorphologic features and/or mental retardation are clinical clues for uniparental disomy in apparently balanced offspring of translocation carriers. Among abortion products of balanced Robertsonian translocation carriers, an excess of 'normal balanced' conceptions has been noted. Robertsonian translocations involving chromosomes 13 and/or 21 are frequently ascertained through a trisomic child. Among those ascertained through a mentally retarded but nontrisomic proband, there appears to be overrepresentation of translocations involving chromosome 14. Since nonmosaic trisomy 14 is nonviable, such a conception would survive a pregnancy only by reducing to disomy.

Fridman et al. (1998) reported a patient with AS and the chromosome constitution 45,XY,t(15q15q). She had some unusual clinical features, including hyperphagia and obesity. Methylation analysis with a probe for small nuclear ribonucleoprotein N (SNRPN; 182279) at 15q12, microsatellite analyses of D15S11, GABRB3 (137192) and D15S113 loci, and FISH using SNRPN and GABRB3 probes indicated paternal isodisomy. This was the fourth reported case of translocation 15q15q with paternal uniparental disomy. Fridman et al. (1998) discussed possible explanations such as homozygosity due to paternal isodisomy for sequence variation (mutation) in one of the genes involved in the pathogenesis of Prader-Willi syndrome. They pointed out that hyperphagia and obesity may occur specifically in association with AS in the context of certain genetic backgrounds, as mice with paternal UPD for the Ube3a region have a postnatal onset of severe obesity (Cattanach et al., 1997).

In studies reported by Robinson et al. (1993), most cases of paternal UPD leading to Angelman syndrome were meiosis II errors or, more likely, mitotic errors. In contrast, in more than 82% of cases of maternal UPD leading to Prader-Willi syndrome, the extra chromosome was due to a meiosis I nondisjunction event. A similar observation has been made for trisomy 21: the majority (78%) of maternal errors leading to trisomy 21 are attributable to meiosis I events, whereas most paternal errors are attributed to either meiosis II or mitotic events (40% and 33%, respectively) (Antonarakis et al., 1993).

Defects in the Imprinting Center

Approximately 2 to 3% of cases of Angelman syndrome result from an imprinting defect (Kishino et al., 1997; Buiting et al., 1998).

Reis et al. (1994) demonstrated defects in methylation in 2 AS sibs, 2 patients with sporadic AS, and 2 sibs from another family with PWS with nondeletion, nonuniparental disomy. In the AS patients, the maternal AS chromosome carried a paternal methylation imprint, and the authors postulated an 'imprinting mutation.' Reis et al. (1994) postulated that in some affected families, a germline mutation in 1 of the grandparents results in failure to reset the imprinting signal in the parental germline, thus resulting in an imprinting defect in parental offspring. Buiting et al. (1995) identified inherited microdeletions of 15q11-q13 between D15S63 and SNRPN (182279) in 2 families with AS and 3 families with PWS. Some of the families had been reported by Reis et al. (1994). In the AS families, the deletions were found on the maternal chromosomes of the patients and on the paternal chromosomes of the phenotypically normal mothers. The authors suggested that the deleted region contains an 'imprinting center' (IC), and that mutations in this region can be transmitted silently through the germline of 1 sex and manifest themselves only after transmission through the germline of the opposite sex. Thus, it is the grandparental legacy of an imprinting mutation that determines the clinical phenotype.

Beuten et al. (1996) reported an extended consanguineous Dutch kindred in which 3 patients with nondeletion AS, 2 males and 1 female, occurred in 3 separate sibships sharing common ancestral couples through all 6 parents. Paternal uniparental disomy of chromosome 15 was detected in 1 case, while the other 2 patients had abnormal methylation of D15S9, D15S63, and SNRPN, consistent with an imprinting mutation. Although the 3 patients were distantly related, the chromosome 15q11-q13 haplotypes were different, suggesting that independent mutations gave rise to AS in this family.

Approximately 6% of AS patients have a paternal imprint on the maternal chromosome. In a few cases, this is due to an inherited microdeletion in the 15q11-q13 imprinting center that blocks the paternal-to-maternal imprint switch in the maternal germline. Burger et al. (1997) determined the segregation of 15q11-q13 haplotypes in 9 families with AS and with an imprinting defect. One family, with 2 affected sibs, had a microdeletion affecting the IC transcript. In the other 8 patients, no mutation was found at that locus. In 2 families, the patient and a healthy sib shared the same maternal alleles. In 1 of these families and in 2 others, grandparental DNA samples were available, and the chromosomes with the imprinting defect were found to be of grandmaternal origin. These findings suggested that germline mosaicism or de novo mutations account for a significant fraction of imprinting defects among patients who have an as-yet-undetected mutation in a cis-acting element. Alternatively, Burger et al. (1997) suggested that these data might indicate that some imprinting defects are caused by a failure to maintain or to reestablish the maternal imprint in the maternal germline or by a failure to replicate the imprint postzygotically. Depending on the underlying cause of the imprinting defect, different recurrence risks need to be considered.

Buiting et al. (1998) described the molecular analysis of 13 PWS patients and 17 AS patients who had an imprinting defect but no IC deletion. Furthermore, heteroduplex and partial sequence analyses did not reveal any point mutations in the known IC elements. All of these patients represented sporadic cases, and some shared the paternal PWS or maternal AS 15q11-q13 haplotype with an unaffected sib. In each of the 5 PWS patients informative for the grandparental origin of the incorrectly imprinted chromosome region and 4 cases described elsewhere, the maternally imprinted paternal chromosome region was inherited from the paternal grandmother. This suggested that the grandmaternal imprint was not erased in the father's germline. In 7 informative AS patients reported by Buiting et al. (1998) and in 3 previously reported patients, the paternally imprinted maternal chromosome region was inherited from either the maternal grandfather or the maternal grandmother. The latter finding was not compatible with an imprint-switch failure, but it suggested that a paternal imprint developed either in the maternal germline or postzygotically. Buiting et al. (1998) concluded that (1) the incorrect imprint in non-IC-deletion cases is the result of a spontaneous prezygotic or postzygotic error; (2) these cases have a low recurrence risk; and (3) the paternal imprint may be the default imprint.

In several patients with Angelman syndrome or Prader-Willi syndrome, microdeletions upstream of the SNRPN gene have been identified, defining an imprinting center that appears to control the imprint switch process in the male and female germlines. Ohta et al. (1999) identified 2 large families segregating an Angelman syndrome imprinting mutation; one of these families was originally described in the first genetic linkage study of Angelman syndrome that mapped the AS gene to 15q11-q13 (Wagstaff et al., 1993). Identification of the imprinting mutation demonstrated that the original linkage was for the imprinting center at 15q11-q13. Affected patients in these 2 Angelman syndrome families had either a 5.5- or a 15-kb microdeletion, one of which narrowed the shortest region of deletion overlap to 1.15 kb in all 8 cases. This small region defined a component of the imprinting center involved in Angelman syndrome, i.e., the paternal-to-maternal switch element. The presence of an inherited imprinting mutation in multiple unaffected members of these 2 families, who are at risk for transmitting the mutation to affected children or children of their daughters, raised important genetic counseling issues.

Imprinting in 15q11-q13 is controlled by a bipartite imprinting center which maps to the SNURF-SNRPN locus. Deletions of the exon 1 region impair the establishment or maintenance of the paternal imprint and can cause Prader-Willi syndrome. Deletions of a region 35 kb upstream of exon 1 impair maternal imprinting and can cause Angelman syndrome. In all sibs affected by Angelman syndrome, an inherited imprinting center deletion had been identified. Buiting et al. (2001) reported 2 sibs with Angelman syndrome who did not have a deletion of the imprinting center but instead had a 1-to-1.5 Mb inversion separating the 2 imprinting center elements. The inversion was transmitted silently through a male germline but impaired maternal imprinting after transmission through the female germline. The findings suggested that the close proximity of the 2 imprinting center elements and their correct orientation, or both, are necessary for the establishment of a maternal imprint.

Imprinting Defects Associated with Infertility Treatment

Cox et al. (2002) reported 2 children conceived by intracytoplasmic sperm injection (ICSI) who developed Angelman syndrome. Molecular studies, including DNA methylation and microsatellite and quantitative Southern blot analysis, revealed a sporadic imprinting defect in both patients. In germ cells and the early embryo, the mammalian genome undergoes widespread epigenetic reprogramming. Animal studies had suggested that this process is vulnerable to external factors. The authors discussed the possibility that ICSI may interfere with the establishment of the maternal imprint in the oocyte or pre-embryo.

Orstavik et al. (2003) described a third case of imprinting defect in a girl with Angelman syndrome who was conceived by ICSI. Biparental origin of normal chromosomes 15 and absence of the common large deletion of 15q11-q13 was found. Methylation-specific Southern blot analysis and methylation-specific PCR for the SNRPN locus showed the presence of a normal unmethylated paternal band and the complete absence of a methylated maternal band, indicating that the patient had an imprinting defect.

Among 16 Angelman syndrome patients born to subfertile couples who conceived with or without infertility treatment, Ludwig et al. (2005) found that 4 had an imprinting defect. The relative risk in untreated couples with time to pregnancy exceeding 2 years was identical to that of those treated by ICSI or by hormonal stimulation alone (RR, 6.25; 95% CI, 0.70 to 22.57), and it was twice as high in couples who had received treatment and also had time to pregnancy greater than 2 years (RR, 12.5; 95% CI, 1.40 to 45.13). Ludwig et al. (2005) suggested that imprinting defects and subfertility might have a common cause, and that superovulation rather than ICSI might further increase the risk of conceiving a child with an imprinting defect.

Mapping