Smith-Magenis Syndrome

A number sign (#) is used with this entry because Smith-Magenis syndrome (SMS) is caused in most cases (90%) by a 3.7-Mb interstitial deletion in chromosome 17p11.2. The disorder can also be caused by mutations in the RAI1 gene (607642), which is within the Smith-Magenis chromosome region.

See also Potocki-Lupski syndrome (PTLS; 610883), which shows overlapping clinical features and is associated with duplication of the same region of 17p11.2.

Clinical Features

Patil and Bartley (1984) reported a 4-year-old girl with an interstitial deletion of chromosome 17p11.2. She had mental retardation, hypotonia, speech delay, small ears, conductive hearing loss, esotropia, dental enamel dysplasia, and prominent premaxilla. Cardiac examination was normal.

Smith et al. (1986) described in detail the phenotype associated with an interstitial deletion of 17p11.2 in 9 unrelated patients. Clinical features included brachycephaly, midface hypoplasia, prognathism, hoarse voice, and speech delay with or without hearing loss, psychomotor and growth retardation, and behavioral problems. A partial deletion was present in 8 patients; 1 patient with complete deletion of band 17p11.2 was more severely affected with facial malformations, cleft palate, and major anomalies of the cardiac, skeletal, and genitourinary systems. Stratton et al. (1986) described 6 additional patients with this chromosome deletion syndrome.

Greenberg et al. (1991) suggested that some SMS patients may have peripheral neuropathy since the disorder is due to deletion of an area of chromosome 17 where a form of Charcot-Marie-Tooth disease (CMT1A; 118220) maps. However, they noted that no patients with CMT1A had shown signs of SMS. Among 32 SMS cases, all unrelated and all with an interstitial deletion of 17p11.2, Greenberg et al. (1991) observed broad flat midface with brachycephaly, broad nasal bridge, brachydactyly, speech delay, and a hoarse deep voice. Fifty-five percent of patients had signs suggestive of peripheral neuropathy, including decreased or absent deep tendon reflexes, pes planus or pes cavus, decreased sensitivity to pain, and decreased leg muscle mass. However, unlike patients with CMT1A, patients with SMS had normal nerve conduction velocities. Two-thirds of the patients demonstrated self-destructive behavior, including head-banging, onychotillomania (pulling out fingernails and toenails), and polyembolokoilamania (insertion of foreign bodies into their ears). Sixty-two percent of patients showed significant symptoms of sleep disturbance such as difficulty falling asleep, difficulty staying asleep, and frequent awakening during the night. Polysonographic studies in 2 patients showed absence of REM sleep. Absence of REM sleep had previously been reported in association with CMT1A; the possibility that a gene associated with REM sleep is in the proximity of the CMT1A locus had been suggested by Tandan et al. (1990). The deletion was determined to be paternal in origin in 9 patients and maternal in 6. The apparent random parental origin suggested that genomic imprinting does not play a role in expression of the SMS clinical phenotype. Greenberg et al. (1991) concluded that SMS is a contiguous gene deletion syndrome.

Moncla et al. (1991) added 3 patients to the 21 previously reported. Hearing loss had been described in 10 of 20 cases. Most had speech delay, hyperactivity, and behavioral problems. Craniofacial changes were described and pictured.

Zori et al. (1993) described an infant with del(17)(p12p11.2) and manifestations consistent with SMS. The mother, who was mosaic for the same deletion, had minor craniofacial changes as well as brachydactyly, consistent with partial manifestation.

Greenberg et al. (1996) reported on a multidisciplinary clinical study of 27 SMS patients. Significant findings included otolaryngologic abnormalities in 94%, eye abnormalities in 85%, sleep abnormalities (especially reduced REM sleep) in 75%, hearing impairment in 68% (approximately 65% conductive and 35% sensorineural), scoliosis in 65%, brain abnormalities (predominantly ventriculomegaly) in 52%, cardiac abnormalities in at least 37%, renal abnormalities (especially duplication of the collecting system) in 35%, low thyroxine levels in 29%, low immunoglobulin levels in 23%, and forearm abnormalities in 16%. The measured IQ ranged between 20 and 78, most patients falling in the moderate range of mental retardation at 40-54, although several patients scored in the mild or borderline range. Greenberg et al. (1996) noted that the diagnosis of SMS is usually secured by cytogenetic analysis during the evaluation of developmental delay and/or congenital anomalies. However, in older individuals the phenotype is distinctive enough that a diagnosis can be made by an experienced clinician before cytogenetic confirmation.

Other Features

Kondo et al. (1991) commented on the presence of fingertip pads in 4 patients with SMS. Fryns (2001) commented on the characteristic clasping of the hands or arms in patients with this disorder.

Barnicoat et al. (1996) reported a 5-year-old boy with deletion of 17p11.2 who, in addition to manifestations of the Smith-Magenis syndrome, had iris dysgenesis characterized by atrophy of iridal stroma, ridges of iridal tissue stretching over the pupils, 'doubled' pupils, and microcornea. Barnicoat et al. (1996) suggested that a gene for anterior chamber development may be found in 17p11.2.

SMS patients are reported to have fewer sleep disturbances when given a nighttime dose of melatonin. Potocki et al. (2000) measured urinary excretion of 6-sulfatoxymelatonin (aMT6s), the major hepatic metabolite of melatonin, in 19 SMS patients in conjunction with 24-hour sleep studies in 28 SMS patients. Of the 28 patients studied, 5 did not have the common SMS deletion. All patients showed significant sleep disturbance. Abnormalities in the circadian rhythm of aMT6s were observed in all but 1 patient, who did not have the deletion. All patients studied were haploinsufficient for COPS3 (604665). De Leersnyder et al. (2001) documented sleep disturbances in 20 children with SMS and an inverted circadian rhythm of plasma melatonin, urinary melatonin, and urinary 6-sulfatoxymelatonin in 8 of them. The authors suggested that haploinsufficiency for a circadian system gene may be responsible. They stated that the administration of melatonin to SMS patients is not necessarily warranted because the amount of secreted hormone is largely normal but its kinetic is erratic.

Smith et al. (2002) studied fasting lipid profiles in 49 children between the ages of 0.6 years to 17.6 years (mean 6.9 years) with Smith-Magenis syndrome. Observed values for serum total cholesterol, triglycerides, LDL cholesterol, and HDL cholesterol were compared with published norms. Mean total cholesterol was significantly higher than pediatric norms (P less than 0.0008). Overall, 57% of Smith-Magenis syndrome subjects had lipid values greater than the 95th percentile for age and gender for at least 1 or more of total cholesterol, triglycerides, and/or LDL. Only 16 subjects (32%) were within normal limits for all 3 variables. These values were not correlated with body mass index.

Clinical Management

De Leersnyder et al. (2001) treated 9 children with SMS with acebutolol, a selective beta-1-adrenergic antagonist, which was given early in the morning. A significant improvement of inappropriate behavior with increased concentration, delayed sleep onset, increased hours of sleep, and delayed waking were noted.

Cytogenetics

Using probes from the juxtacentromeric region of chromosome 17p, Moncla et al. (1993) mapped 3 microdeletions in patients with Smith-Magenis syndrome. Using Southern blot analysis, they demonstrated that all patients had deletion of markers D17S29 and D17S71. One breakpoint was located between D17S58 and D17S29 and the other breakpoint was distal to D17S71. Imprinting did not appear to be involved; one deletion was of paternal origin, the other maternal. Moncla et al. (1993) suggested that an unstable region, located between SMS and CMT1A, could be a hotspot for rearrangements, leading proximally to SMS microdeletions and distally to CMT1A duplications.

Juyal et al. (1995) used fluorescence in situ hybridization to demonstrate a deletion in 17p11.2 that was not detectable by flow cytometry and was equivocal by cytogenetics.

By cytogenetic analysis, Park et al. (1998) detected a de novo intrachromosomal insertional rearrangement by which a segment of the proximal long arm of a chromosome 17 (q11.2-q21.3) was inserted into the short arm at p11.2, resulting in an apparent deletion of the SMS critical region, demonstrating the instability of the SMS region of chromosome 17p. Fluorescence in situ hybridization demonstrated that the inserted segment included both the ERBB2 (164870) and RARA (180240) loci, and dual color hybridizations defined the insertion as direct, with ERBB2 located more proximally on the short arm of the derivative chromosome. The FLI1 locus (600362) and several markers were deleted from 17p, whereas CMT1A remained in its immediate location on the short arm of the metacentric der(17) chromosome.

Molecular Genetics

Chevillard et al. (1993) described a 5-Mb YAC contig spanning the CMT1A duplicated segment and the distal part of 4 SMS microdeletions. They identified the first expressed sequence located in the SMS critical region: the gene coding for small nuclear RNA U3 (180710).

Koyama et al. (1996) localized the human homolog of the murine Llglh gene (LLGL1; 600966) to chromosome 17p11.2. In SMS patients, a probe representing LLGL1 failed to hybridize with 1 of the 2 chromosome 17 homologs, suggesting that this gene may play a role in the pathogenesis of SMS.

Elsea et al. (1999) assessed the potential effect of haploinsufficiency of SGN3 (COPS3; 604665), which encodes subunit 3 of the COP9 signalosome and maps to the distal portion of the SMS critical region, in SMS patient lymphoblastoid cell lines. Although SMS patients were haploinsufficient for SGN3, analyses showed that the SGN3 protein was present at equivalent levels in patient and parental control cells, and that the COP9 signalosome complex was assembled and in normal quantities in transformed lymphoblastoid cell lines from patients. The authors concluded that SGN3 probably does not play a significant role with respect to SMS, although its involvement could not be ruled out since the importance of the COP9 signalosome in embryogenesis or differentiation was not well understood.

Lucas et al. (2001) created a contiguous physical and transcription map of the 1.5-Mb SMS critical interval. Within this interval, they identified 13 known genes, 14 ESTs, and 6 genomic markers. To identify possible candidate genes, they performed sequence analysis and determined the tissue expression pattern of 10 novel ESTs mapping to the SMS critical interval. Lucas et al. (2001) also presented a detailed review of 6 SMS candidate genes, of which NT5M (605292) was considered especially intriguing because of the possible role of the NT5M protein on the modulation of dTTP substrate pools in the mitochondria. Lucas et al. (2001) speculated that some of the characteristics of SMS, including hypotonia, mental retardation, and behavioral abnormalities, may be an effect of excess dTTP and defective mitochondria.

Slager et al. (2003) studied 3 individuals who had clinical features consistent with SMS but did not have 17p11.2 deletions detectable by standard fluorescence in situ hybridization techniques. They found that 2 patients had deletion of a single cytosine in a run of Cs in the RAI1 gene (607642.0001, 607642.0002). They compared clinical findings in these 2 patients with those with the typical deletion, with a small deletion, and with a 29-bp deletion (607642.0003). This led to the conclusion that SMS may be similar to previously described microdeletion syndromes in which a single gene is implicated in most of the features but other deleted genes may modify the overall phenotype, e.g., Williams syndrome (194050) and Angelman syndrome (105830). Haploinsufficiency of RAI1 is probably responsible for the behavioral, neurologic, otolaryngologic, and craniofacial aspects of this syndrome, but more variable features such as heart and renal defects are probably due to hemizygosity of other genes in the 17p11.2 region.

Kaminsky et al. (2011) presented the largest copy number variant case-control study to that time, comprising 15,749 International Standards for Cytogenomic Arrays cases and 10,118 published controls, focusing on recurrent deletions and duplications involving 14 copy number variant regions. Compared with controls, 14 deletions and 7 duplications were significantly overrepresented in cases, providing a clinical diagnosis as pathogenic. The 17p11.2 deletion was identified in 16 cases and no controls for a p value of 0.00045 and a frequency of 1 in 984 cases.

Molecular Mechanisms of Deletion and Duplication

Chen et al. (1997) identified 3 copies of a low-copy-number repeat located within and flanking the SMS common deletion region. They showed that this repeat, called SMS-REP by them, represents a repeated gene cluster. They isolated a corresponding cDNA clone that identified a novel junction fragment from 29 unrelated SMS patients and a different-sized junction fragment from a patient with duplication of 17p11.2. Their results suggested that homologous recombination of a flanking repeat gene cluster is a mechanism for this common microdeletion syndrome.

Recombination between repeated sequences at various regions of the human genome are known to give rise to DNA rearrangements associated with many genetic disorders (Lupski, 1998). Perhaps the most extensively characterized genomic region prone to rearrangement is 17p12, which is associated with peripheral neuropathies, hereditary neuropathy with liability to pressure palsies (HNPP; 162500), and CMT1A. Homologous recombination between 24-kb flanking repeats, termed CMT1A-REPs, results in a 1.5-Mb deletion that is associated with HNPP, and the reciprocal duplication product is associated with CMT1A. In the case of Smith-Magenis syndrome, more than 90% of patients carry deletions of the same genetic markers in 17p11.2, defining a common deletion (Potocki et al., 2000).

Shaw et al. (2002) analyzed the haplotypes of 14 families of patients with SMS and 6 families of patients with duplication of the same region using microsatellite markers directly flanking the SMS common deletion breakpoints. The data indicated that the deletion and its reciprocal duplication of chromosome 17p11.2 result from unequal meiotic crossovers mediated through nonallelic homologous recombination (NAHR) that occurs via both interchromosomal and intrachromosomal exchange events between the proximal and distal SMS repeats. There appeared to be no parental-origin bias associated with common SMS deletions and the reciprocal duplications.

Bi et al. (2003) reported a recombination hotspot associated with both the common SMS deletion and the reciprocal duplication, dup(17)(p11.2p11.2), demonstrating the reciprocity of the crossover events as had been demonstrated for HNPP and CMT1A.

Shaw et al. (2004) reported an additional recombination hotspot within 2 large low-copy repeats (LCRs), which serve as alternative substrates for nonallelic homologous recombination that results in large (approximately 5 Mb) deletions of 17p11.2, which include the SMS region.

Liu et al. (2011) assembled 2 patient cohorts with reciprocal genomic disorders, deletion-associated SMS and duplication-associated Potocki-Lupski syndrome (610883). By assessing the full spectrum of rearrangement types from the 2 cohorts, Liu et al. (2011) found that complex rearrangements (those with more than 1 breakpoint) are more prevalent in copy-number gains (17.7%) than in copy-number losses (2.3%), an observation that supports a role for replicative mechanisms in complex rearrangement formation. Interestingly, for nonallelic homologous recombination-mediated recurrent rearrangements, Liu et al. (2011) showed that crossover frequency is positively associated with the flanking low-copy repeat (LCR) length and inversely influenced by the inter-LCR distance. To explain this, they proposed that the probability of ectopic chromosome synapsis increases with increased LCR length, and that ectopic synapsis is a necessary precursor to ectopic crossing-over.

In an 11-year-old girl, born of nonconsanguineous parents of Ashkenazi Jewish ancestry, who had signs and symptoms consistent with Smith-Magenis syndrome, Adams et al. (2014) analyzed the RAI1 gene and identified heterozygosity for a de novo nonsense mutation (W758X), which they noted had previously been reported in SMS by Vilboux et al. (2011). In addition, the patient and her younger sister, who both experienced multiple episodes of hypoglycemia and lactic acidosis associated with illness or fasting (PCKDC; 261680), were found to be homozygous for a missense mutation in the PCK1 gene (I45T; 614168.0001). Heterozygosity for a de novo missense mutation (E413G) in the GRIN2B gene (138252) was also detected in the younger sister, who exhibited severe neurologic and developmental problems (MRD6; 613970) that were distinct from those of her older sib. Adams et al. (2014) concluded that this family demonstrated that complex medical disorders can represent the cooccurrence of multiple diseases.

Genotype/Phenotype Correlations

Natacci et al. (2000) reported a 22-year-old woman with a deletion in the short arm of chromosome 17 who presented with the clinical manifestations of both Smith-Magenis syndrome and Joubert syndrome (JBTS; 213300). Facial anomalies, brachydactyly, severe mental retardation, and self-injuring behavior were attributed to SMS, whereas the cerebellar vermis hypoplasia, hypotonia, ataxic gait, developmental delay, and abnormal respiratory pattern suggested JBTS. By fluorescence in situ hybridization analyses with YAC mapping to the 17p11.2 region, as well as locus-specific probes generated through a novel procedure, they established that the deletion encompasses a 4-Mb interval. The deletion differed from that commonly found in SMS in its telomeric boundary, and was more distal than usually observed. The presence of the JBTS phenotype in this patient and the detection of an unusual SMS deletion suggested the presence of a JBTS gene in close proximity to the SMS locus. Although Joubert syndrome has been linked to 9q34.3 in some families, no linkage to this area had been demonstrated in other families.

Girirajan et al. (2005) reported 4 individuals in which SMS was caused by mutation in the RAI1 gene. The authors noted that the clinical features of the patients differed from those found in patients with the 17q11.2 deletion by general absence of short stature and lack of visceral anomalies. All 4 patients had developmental delay, reduced motor and cognitive skills, craniofacial and behavioral anomalies, and sleep disturbances. Seizures, not previously thought to be associated with RAI1 mutations, were present in 1 individual. A patient with an S1808N mutation (607642.0004) had had neonatal jaundice, sleep disturbance, and mildly delayed motor and cognitive milestones as major features during early childhood. He had high myopia, a loud and hoarse voice, a waddling gait, pes planus, and dry skin. Abnormal behavior included sleep disturbances (hypersomnolence as an infant, moving to frequent and early awakenings and daytime napping), reported bipolar episodes, head banging, tantrums, and aggressive and intrusive behavior. He also had deep scarring from obsessively picking his skin. At age 14, he had the developmental age of a 9-year-old child, with an IQ of 89. A boy with a 1-bp deletion (607642.0006) had severely disturbed sleep, head banging, and occasionally self mutilation. He destroyed his toys and the furniture in his bedroom. His intelligence at the age of 9.5 years was evaluated as full scale IQ of 73, verbal IQ of 85, and performance IQ of 65. Craniofacial features included brachycephaly, midface hypoplasia, tented upper lip, and a broad, square face. He also had a hoarse voice. A 19-year-old woman with a 19-bp deletion (607642.0007) was noted as a neonate to have floppy muscle tone, upslanting palpebral fissures, and midface hypoplasia. Down syndrome was initially diagnosed. At 15 years of age she had a developmental age of 8 to 10 years, with an IQ of 67. Her facial and behavioral features were considered consistent with SMS. She had a waddling gait, loud and hoarse voice, decreased sensitivity to pain, and short fingers and hands. She also shared significant sleep disturbance and skin picking.

Edelman et al. (2007) retrospectively analyzed the clinical features of 105 patients with SMS, including 95 (90.5%) with 17p11.2 deletions and 10 (9.5%) with RAI1 mutations. Patients with RAI1 mutations were more likely to exhibit overeating, obesity, polyembolokoilamania, self-hugging, muscle cramping, and dry skin. Those with 17p11.2 deletions were more likely to have short stature, hearing loss, ear infections and heart defects. Female SMS patients were significantly more likely to have myopia, eating problems, cold hands and feet, and frustration with communication compared to male patients, regardless of genotype.

Girirajan et al. (2006) reported the molecular and genotype-phenotype analyses of 31 patients with SMS who carry 17p11.2 deletions or intergenic mutations, respectively, and were compared for 30 characteristic features of the disorder by the Fisher exact test. Eight of the 31 individuals carried a common 3.5-Mb deletion, whereas 10 of 31 individuals carried smaller deletions, 2 individuals carried larger deletions, and 1 individual carried an atypical 17p11.2 deletion. Ten patients with nondeletion harbored a heterozygous mutation in RAI1. Phenotype comparison between patients with deletions and patients with RAI1 mutations showed that 21 of 30 SMS features are the result of haploinsufficiency of RAI1, whereas cardiac anomalies, speech and motor delay, hypotonia, short stature, and hearing loss are associated with 17p11.2 deletions rather than RAI1 mutations (P less than 0.05). Further, patients with smaller deletions show features similar to those with RAI1 mutations. Girirajan et al. (2006) concluded that although RAI1 is the primary gene responsible for most features of SMS, other genes within 17p11.2 contribute to the variable features and overall severity of the syndrome.

As indicated, mutations of the RAI1 gene seem to be responsible for the main features found with heterozygous 17p11.2 deletions. Andrieux et al. (2007) studied DNA from 30 patients with SMS using comparative genome hybridization. Three patients had large deletions. Two of the 3 had cleft palate, which was not found in any of the other patients with SMS. The smallest extra-deleted region associated with cleft palate in SMS was 1.4 Mb, containing less than 16 genes and located at 17p12-p11.2. Gene expression array data showed that the ubiquitin B precursor gene (UBB; 191339) is significantly expressed in the first branchial arch in the fourth and fifth weeks of human development. Together, the data supported UBB as a candidate gene for isolated cleft palate.

Population Genetics

Smith-Magenis syndrome occurs in approximately 1 in 25,000 births (Juyal et al., 1996).

Animal Model

Walz et al. (2003) constructed mouse models of Smith-Magenis syndrome and dup(17)(p11.2p11.2) by engineering rearranged chromosomes carrying a deletion, Df(11)17, or a duplication, Dp(11)17, of the syntenic region on mouse chromosome 11. Mice heterozygous for the deletion showed craniofacial anomalies, seizures, obesity, and male-specific reduced fertility. Mice heterozygous for the duplication were underweight and did not show craniofacial anomalies, seizures, or reduced fertility. Walz et al. (2003) concluded that the phenotypic differences were due to gene dosage effects. Walz et al. (2004) reported that male mice heterozygous for the deletion or the duplication were hypoactive or hyperactive, respectively. In addition, male Dup mutant mice, but not Del mutant mice, had impaired contextual fear conditioning. Circadian rhythm studies revealed period length differences in Del mutant mice, but not Dup mutant mice. Walz et al. (2004) concluded that some behavioral abnormalities were gene dosage-sensitive, whereas other behavioral abnormalities were specific to mice carrying the deletion or the duplication and could be observed in a sex-preferential manner.

Yan et al. (2004) constructed 3 lines of mice with 590- or 595-kb deletions, Df(11)17-1, Df(11)17-2, and Df(11)17-3. Both craniofacial abnormalities and obesity were observed, but the penetrance of the craniofacial phenotype was markedly reduced when compared with Df(11)17 mice. The authors proposed that loss of Rai1 gene may be responsible for the craniofacial abnormalities and obesity.

Bi et al. (2005) generated a null Rai1 allele in mice. Obesity and craniofacial abnormalities were observed in Rai1 +/- mice, but the penetrance of craniofacial anomalies was further reduced in Rai1 +/- mice compared to Df(11)17-1 and Df(11)17 mice. Most Rai -/- mice died during gastrulation and organogenesis, and survivors were growth retarded and displayed malformations in both the craniofacial and the axial skeleton. Using Rai1-fusion constructs, the authors showed that Rai1 is translocated to the nucleus and has transactivation activity. Bi et al. (2005) concluded that Rai1 functions as a transcriptional regulator and may be important for embryonic and postnatal development.

Walz et al. (2006) generated compound heterozygous mice with a Dp(11)17 allele and a null Rai1 allele, thus resulting in normal disomic gene dosage of Rai1. Normal Rai1 dosage rescued many of the phenotypes observed in heterozygous Dp(11)17 mice, including normalization of body weight and partial normalization of behavior. The phenotype was rescued despite altered trisomic copy number of the other 18 or so genes in the region. Walz et al. (2006) concluded that duplication of Rai1 is responsible for decreased body weight in Dp(11)17 mice and that Rai1 is a dosage-sensitive gene involved in body weight control and complex behavioral responses.

Yan et al. (2007) generated a mouse model strain, Df(11)17-4, with a deletion of 1 Mb, intermediate in size between the 2-Mb deletion of Df(11)17 mice and the 590-kb deletion of Df(11)17-1 mice. Remarkably, the penetrance of its craniofacial anomalies in a mixed genetic background was between that of the 2 previous models. They further analyzed the deletion mutations and the Rai1 -/+ allele on a pure C57BL/6 background, to control for nonlinked modifier loci. The penetrance of the craniofacial anomalies was markedly increased for all strains in comparison with the mixed background. Mice with the Df(11)17 and Df(11)17-1 deletions had a similar penetrance of the craniofacial phenotype, suggesting that penetrance may be less influenced by deletion size, whereas that of Rai1 -/+ mice was significantly lower than that of the deletion strains. Yan et al. (2007) hypothesized that potential trans-regulatory sequences(s) or gene(s) that reside within the 590-kb genomic interval surrounding Rai1 are the major modifying genetic element(s) affecting the craniofacial penetrance. Moreover, they confirmed the influence of genetic background and different deletion sizes on the phenotype. The complicated control of the penetrance of one phenotype in Smith-Magenis syndrome mouse models provides tools to elucidate molecular mechanisms for penetrance and clearly shows that a null allele caused by a chromosomal deletion can have different phenotypic consequences than one caused by gene inactivation.