| Human Molecular Genetics | Pages |
The spectrum of mutations in UBE3A causing Angelman syndrome
Introduction
Results
Truncating mutations
Missense mutations, amino acid deletion and predicted elongated protein
Benign variants
Analysis for genetic counseling
Discussion
Materials And Methods
Patient population
Mutation analysis
Acknowledgements
References
The spectrum of mutations in UBE3A causing Angelman syndrome
INTRODUCTION
Angelman syndrome (AS) is characterized by moderate to severe mental retardation, absence of speech, tremor, ataxia, abnormal gait, inappropriate laughter, sleep disturbance and seizures. The incidence of AS is estimated to be 1 in 20 000, with most cases being sporadic, although familial occurrence is not rare. Detailed reviews of the clinical phenotype (1,2) and recent molecular studies (3,4) as well as an extensive bibliography (5) are available. AS is caused by deficiency of gene expression from maternal chromosome 15q11-q13, while paternal deficiency for the same region causes Prader-Willi syndrome. This chromosomal region contains numerous genes and transcripts that are expressed from the paternal chromosome and silenced on the maternal chromosome, including the loci encoding small nuclear ribonucleoprotein-associated polypeptide N (SNRPN) and necdin (NDN). There is extensive differential methylation across this region most prominently at a CpG island at the 5[prime] end of the SNRPN locus where the paternal chromosome is unmethylated with SNRPN expressed while the maternal chromosome is methylated with SNRPN silenced. The gene encoding E6-AP ubiquitin-protein ligase (gene symbol UBE3A) maps within the region and is implicated as the AS gene based on the discovery of loss-of-function mutations in patients (6,7). After the discovery of the mutations causing AS, it was recognized that the locus was imprinted with maternal expression in human brain (8,9) and in mouse hippocampus and cerebellum (10).
Cytogenetic and molecular data permit the identification of several molecular mechanisms causing AS as follows. (i) Approximately 70% of patients have a de novo interstitial deletion of ~4 Mb on the maternal chromosome 15q11-q13, and a few patients have analogous but slightly different deletions arising de novo or as unbalanced translocations. (ii) Approximately 3-5% of patients have paternal uniparental disomy (UPD) resulting in maternal deficiency. (iii) Approximately 7-9% of patients have `imprinting mutations' defined as biparental inheritance but with both chromosomes having a paternal methylation and expression pattern (11). Some but not all of these patients have deletions of a putative imprinting center (IC) (12). (iv) Approximately 4-8% of patients have loss-of-function mutations in UBE3A (6,7). (v) Finally, an uncertain fraction of patients (~10-15%) have a clinically typical AS phenotype, but none of the molecular defects of the first four groups are identifiable despite sequencing most or all of the exons for UBE3A. All of the known molecular abnormalities are consistent with the interpretation that AS is caused by deficiency of maternal expression for UBE3A.
Figure 1. Mutation analysis for selected members of a large AS family. A 4 bp deletion (1694del4 in family H-179) was found in a large published pedigree with eight affected individuals (22). Genomic DNA was amplified as described in Materials and Methods and analyzed on a 15% polyacrylamide gel. The migrations of normal and mutant PCR products as well as heteroduplexes are indicated. Half-solid symbols indicate affected with AS and carrying the mutation, half-hatched symbols indicate known or presumed mutation carriers with a normal phenotype, and divided open symbols indicate a normal phenotype with absence of the mutation. Asterisks identify individuals not tested but predicted to carry the mutation. The pathogenesis of AS is unknown at present. E6-AP was discovered on the basis of its ability to inactivate the p53 oncoprotein in the presence of the E6 protein of human papilloma virus (13,14). Subsequently, E6-AP was shown to be an E3 ubiquitin-protein ligase and to be capable of ubiquitinating p53 and, in an E6-independent manner, a human homolog of the yeast RAD23 protein designated HHRAD23A (15). Ubiquitination targets proteins for degradation through the proteosomal pathway (16,18), and it can be hypothesized that the AS phenotype might be caused by failure to ubiquitinate a variety of target proteins in those tissues where the paternal allele for UBE3A is silenced, and expression is dependent on the maternal allele. The lack of ubiquitination could lead to failure to degrade these proteins or to other functional alterations of target proteins (19). Recent studies of a mouse model of AS based on a null mutation generated by gene targeting demonstrated phenotypic features with many similarities to human AS, a contextual learning defect, impaired long-term potentiation in hippocampal slices and increased abundance of p53 in Purkinje cells and hippocampal neurons (20).
Table 1.
| Exon | Primer sequence | Primer location |
Annealing temperature (°C) |
Product size (bp) |
|
| 7 | For | R-ATGGCCACCTGATCTGACCACTTTCa | intron 6 | 58 | 351 |
| Rev | U-AGGTATGTTCCTATCTCCCATT | intron 7 | |||
| 8 | For | U-AGGAGTTGTGGTAAATAGTGCA | intron 7 | 58 | 666 |
| Rev | R-TTTGGCATATGATCTGCTTCTA | intron 8 | |||
| 9a | For | R-TTGCAACAGAGTAAACATACATATT | intron 8 | 55 | 717 |
| Rev | U-CTCCATCATTCTCCGAATCTGGT | exon 9 | |||
| 9b | For | R-TTGCAAAGCGATGAGCAAGCTACCC | exon 9 | 55 | 750 |
| Rev | U-CACTGAACTGTATCATGATATC | intron 9 | |||
| 10 | For | R-ATACTAGCAATCATCTTCTT | intron 9 | 58 | 227 |
| Rev | U-GATACGACACCATAATCACATT | intron 10 | |||
| 11 | For | CGATGCCACCAAATTACTTACTA | intron 10 | 58 | 577 |
| Rev | R-GATAAGAGTATCAACAAAGATTCTA | intron 11 | |||
| 12 | For | U-AGAAGAGTGATATAAATTATTTG | intron 11 | 58 | 465 |
| Rev | CTGCTTCATGTCCTCTTTCTCT | intron 12 | |||
| 13+14 | For | U-TGTTAAGAAGTAGGTGTAAAATTGA | intron 12 | 58 | 558 |
| Rev | R-CCTAGAGATAAAGGTCTGAAGCA | intron 14 | |||
| 15 | For | U-ATGAATGCCAAACTGAAACCAG | intron 14 | 58 | 574 |
| Rev | GCTGGCAATATGACTAAGAAAATGA | intron 15 | |||
| 16 | For | U-ACTGATGTCCTCTCTGTGGTTTTGT | intron 15 | 58 | 566 |
| Rev | R-TTGTACTGGGACACTATCACCACCA | 3[prime] flanking |
Table 2.
| Identifier | Nucleotide | Protein | Family |
| Truncating mutations | |||
| H-144 sibs | 856delG | frameshift | present in mother |
| H-157 | 904del5 | frameshift | de novo |
| H-151 sibs | 980delAG | frameshift | present in mother |
| H-137 | 1500G->A | W305X | present in mother and maternal grandfather |
| H-123 | 1552delA | frameshift | de novo |
| H-118 | 1559del7 | frameshift | de novo |
| H-179 family | 1694del4 | frameshift | large pedigree |
| H-115a | 1835C->T | R417X | de novo |
| H-104a | 1930delAG | frameshift | de novo |
| H-150 sibs | 2185T->G | Y533X | present in mother |
| H-168 | 2567ins4 | frameshift | de novo |
| H-145 sibsb | 2890G->A | W768X | present in mother |
| H-121 | 3093del4 | frameshift | de novo |
| Missense mutations/amino acid deletion/elongated protein | |||
| H-101a | 648G->A | C21Y | present in mother |
| H-173 sibs | 2929del3 | F782[Delta] | present in mother and maternal grandfather |
| H-143 | 2997T->A | I804K | de novo |
| H-124 | 3142del15 | elongated protein | de novo |
| Benign variants | |||
| H-111, H-163, H-172a | 1118G->A | A178T | present in father |
| Multiple | GATGAT->GAT | none | - |
We have now studied 56 index patients with a clinical diagnosis of AS in whom large deletions, UPD and imprinting mutations were not present. We have identified 13 truncating mutations (three reported previously) (6,21), two missense mutations, one single amino acid deletion and one stop codon mutation predicting an elongated protein. We have identified a 4 bp deletion in a large published pedigree, identified mutations in 75% of multiplex families and 23% of sporadic cases, and identified a length polymorphism within an intron and a missense polymorphism. All of the exons encoding the major open reading frame for E6-AP were sequenced in 56 subjects considered to have a highly likely clinical diagnosis of AS and normal methylation analysis. Genomic DNA was amplified using the primers shown in Table 1 followed by direct sequencing. Thirteen truncating mutations were identified in this group (Table 2). Five of these cases involved more than one affected individual in the family, and in these five families, mothers of affected children carried the mutation. In families H-137 and H-179 where analysis was informative, the mutation was on the paternal chromosome in the phenotypically normal mothers and in other normal mutation carriers in families. This was documented most extensively in a large published pedigree (23) with at least eight individuals affected with AS. This family was found to carry a four base deletion causing a frameshift mutation (Fig. RESULTS
Truncating mutations
Missense mutations, amino acid deletion and predicted elongated protein
Two missense mutations, one single amino acid deletion, and one mutation affecting the stop codon and predicting an elongated protein were found in individuals with an AS phenotype (Table 2). It is more complex than in the case of truncating mutations to determine if these mutations are disease-causing or benign variants. Earlier, we reported (23) a substitution of tyrosine for cysteine at codon 21 (C21Y), but it was uncertain if this represented a disease-causing mutation or a benign variant. To determine further if this amino acid substitution is a disease-causing mutation, we analyzed the unaffected sibling and eight maternal relatives (Fig. Figure 2. Analysis of family H-101 for the C21Y missense mutation. The mutation was found only in the affected boy and his mother. Analysis for the GATGAT/GAT polymorphism in the adjacent intron indicated that the C21Y mutation was de novo in the mother (see text for discussion). Pedigree symbols are as for Figure 1.
Benign variants
A 3 bp length polymorphism in the intron immediately upstream of exon 8 was identified as shown above in Figure
An amino acid substitution of threonine for alanine at codon 178 (A178T) was found earlier in an AS subject and was present in the father (H-111) (6). Subsequently, we identified two additional AS cases with the A178T mutation present in the affected child and in the father (H-163 and H-172). Alanine (A178) was found with an allele frequency of 0.97, and threonine (T178) with a frequency of 0.03 in 106 chromosomes from 53 normal subjects.
Analysis for genetic counseling
Numerous families were referred for mutation analysis related to prenatal diagnosis and genetic counseling for family members. In family H-144, the maternal aunt of the affected siblings was pregnant and requested prenatal diagnosis. Earlier analysis for genetic markers in the UBE3A region indicated that the maternal aunt carried the same grandpaternal haplotype as the affected children in the family. Mutation analysis identified a single base deletion (856delG) in one of the affected children and the mother, but this mutation was not present in the maternal grandparents or the maternal aunt.
In the case of the H-137 family, the sample was referred for risk evaluation and possible prenatal diagnosis. The W305X mutation was identified in the affected child and her mother. The mother declined prenatal diagnosis, and her infant has not been tested. In one of the families with a de novo mutation, a family terminated an earlier pregnancy because of concern regarding risk for AS with no possibility for definitive prenatal diagnosis prior to the performance of mutation analysis. This family would now have a relatively low risk of having an affected child, and mutation analysis could be performed for prenatal diagnosis to address concern regarding maternal mosaicism for the mutation.
DISCUSSION
We have identified mutations in UBE3A in 17 of 56 index patients (30%) with a clinical diagnosis of AS and normal DNA methylation analysis. Mutations were found in six of eight families (75%) with more than one individual affected with clinically typical AS, and in 11 of 47 isolated cases (23%). No mutation was found in a family with one sib with clinically typical AS and one sib with atypical features. These findings are similar to the report of Malzac et al. (23) in which mutations were found in 80% of familial cases and 14% of sporadic cases. There are now 29 different mutations reported to cause AS, including 17 frameshift, five nonsense, three missense and one each of single amino acid insertion, single amino acid deletion, splicing and stop codon (6,7,21,23,24) (Fig. Figure 3. Mutations reported to cause Angelman syndrome. All reported mutations (6,7,21,24) are shown. Exons are depicted in proportion to size, but upstream exons are not shown, and introns are not in proportion. The ATG initiation codon designated as amino acid 1 is shown bridging exons 7 and 8. Two mutations (*) have recurred independently. The number for the 3093ins5 mutation is changed from the original publication based on recommended nomenclature (32). Although there is considerable biochemical information regarding the ubiquitin-protein ligase activity of E6-AP, including its ability to ubiquitinate p53 and HHRAD23A (15,25), the reason for the predominance of truncating mutations in UBE3A causing AS is not obvious. Several functional domains have been identified for E6-AP. An 18 amino acid segment (amino acids 391-408) is essential to direct the association of E6 with p53 and promote E6-dependent degradation of p53 (26). The hect (homologous to the E6-AP carboxyl-terminus) domain includes ~350 amino acids in the C-terminal portion of E6-AP, is highly conserved among proteins from yeast, Drosophila, Caenorhabditis elegans and mammals (27) and can interact with certain E2 enzymes to form complexes for ubiquitination of substrates (15). The hect domain contains a cysteine residue at position 820 that serves as the active site for thioester formation with ubiquitin. One possibility is that missense mutations occur with reasonable frequency in UBE3A, but they have no phenotypic effect or cause a milder phenotype that is not recognizable as being related to AS. Of particular interest would be the possibility that patients with milder phenotypes involving mental retardation, seizures or other features might have missense mutations in UBE3A. Although less likely, it is possible that the primary sequence of the locus predisposes to a higher frequency of truncating mutations than missense mutations. Although mutations are identified in 75-80% of families with multiple individuals affected with AS, mutations are not found in some families, as in the case of two families that we studied. Other families with affected sibs and no identifiable mutation are known (7,23), and in one case a recombination suggests that the mutation is toward the 5[prime] end of the locus if it is in cis with UBE3A on the maternal chromosome (7). It is likely that some of the remaining cases of AS have mutations on the maternal chromosome that are yet to be identified. The demonstration of deletions causing imprinting mutations ~0.8-1 Mb centromeric to UBE3A emphasizes that these mutations could affect cis regulatory elements quite distant from the locus. The identification of mutations in a smaller fraction of isolated cases suggests either that there is another unidentified molecular basis for AS that is not associated with a high recurrence risk, as is the case for interstitial deletions or UPD, or that a significant fraction of patients clinically diagnosed as AS have an unrelated condition. Thus possible additional molecular mechanisms involving UBE3A include mutations of 15q11-q13 not yet identified, mutations in genes on other chromosomes regulating the maternal expression of UBE3A and epigenetic events leading to the silencing of the maternal allele for UBE3A without a primary nucleotide change. Possible explanations for failure to identify mutations not implicating UBE3A would include inaccurate clinical assessment or legitimate genocopies or phenocopies meeting the standard diagnostic criteria without alteration of expression of UBE3A. Laboratory diagnosis and genetic counseling for AS are unusually complicated. Although mutation studies do not identify the defect in all cases, they are valuable and indicated in patients with normal DNA methylation analysis.
MATERIALS AND METHODS
Patient population
Many of the patients were seen personally by one of the authors, and the remainder were referred from distant cities in several countries. Medical records, and in many cases video tapes, of patients were reviewed by the authors and primarily by one investigator (C.B.), and patients were characterized as having a phenotype consistent with AS based on standard diagnostic criteria (28). Some patients referred were felt to be unlikely to have AS based on review of medical records and video tapes, and were not included for molecular studies. Family history information was available in all cases.
Mutation analysis
Genomic DNA was isolated from peripheral blood or from cultured lymphoblasts using SDS/proteinase K digestion and phenol-chloroform extraction (29). Exons 7-16 of UBE3A were amplified using primers flanking the intron-exon boundaries (Table 1); more extensive sequence information is available (GenBank accession nos AF016703-AF016708). PCR conditions were 10 mM Tris-HCl, pH 8.3, 1.5 mM MgCl2, 50 mM KCl, 0.2 mM dNTPs, 1 µM primers for exons 9 and 10 and 0.5 µM primers for other exons. PCR conditions were 95°C for 1 min, 55 or 58°C for 1 min and 72°C for 1 min for 30-35 cycles. PCR products were purified using PCR purification or gel extraction kits (Qiagen, Los Angeles, CA or Promega, Milwaukee, WI). Most PCR primers included universal or reverse sequencing tails as shown in Table 1. PCR products were sequenced directly using the ABI PRISM dye primer or dye terminator cycle sequencing ready reaction kits (Perkin-Elmer, Chicago, IL). Automated sequencing was performed using an ABI 377 DNA sequencer (Applied Biosystems, Foster City, CA). All putative mutations were confirmed by cloning and sequencing the mutant allele for each subject.
For many mutations, PCR analysis was established to analyze family members and further confirm the mutation utilizing spontaneously occurring differences in restriction enzyme sites, artificially created differences in restriction enzyme sites or length variation of the product. For analysis of the 1694del4 mutation, genomic DNA was amplified using the primers for exon 9b from Table 1 and nested PCR was performed with additional primers 5[prime]-CAAATGTAGTGGGAGGGGAAGTGG-3[prime] and 5[prime]-CAGCTCGCTGGACTCAGGGATGGG-3[prime]. For the C21Y mutation studies, the primers for amplification were 5[prime]-ACTGTGCTTATTGTTTGAATGTTTG-3[prime] and 5[prime]-CTCATTCGTGCAGGCTTCATTTCC-3[prime], and the mutation was analyzed by restriction enzyme digestion with Tsp45I. For analysis of the 3 bp intronic polymorphism, the primers were 5[prime]-CACAGGTTAACTACTTCAGTGC-3[prime] and 5[prime]-TAAGCACAGTGATTAGTACA-3[prime].
Exons and cDNA nucleotides are numbered according to Kishino and Wagstaff (GenBank accession no. U84404) (30), which differs from some earlier cloning and mutation reports (6,7,21,24,31), but is in agreement with the numbering of Malzac et al. (23). The ATG codon bridging exons 7 and 8 is cDNA bp 587-589 and is codon 1 for amino acid numbering. Mutations are designated according to recommended nomenclature (32) with all nucleotide numbers based on cDNA.
ACKNOWLEDGEMENTS
We wish to thank numerous families and clinicians for providing access to medical records and samples. We thank Diane Dicks, Scott Dudek and Xiao-yun Wang for technical assistance. This work was supported by NIH grants for a Mental Retardation Research Center (HD24064) and for a General Clinical Research Center (RR00188).
REFERENCES
$Department of Molecular Physiology and Biophysics, Vanderbilt University School of Medicine, Nashville, TN, USA
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Jan 1999
Copyright©Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Wu, F. V. Bolduc, K. Bell, T. Tully, Y. Fang, A. Sehgal, and J. A. Fischer A Drosophila model for Angelman syndrome PNAS, August 26, 2008; 105(34): 12399 - 12404. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Chen, J. Visootsak, S. Dills, and J. M. Graham JR Prader-Willi Syndrome: An Update and Review for the Primary Pediatrician Clinical Pediatrics, September 1, 2007; 46(7): 580 - 591. [Abstract] [PDF]
|
|
|
|
 |
|
 L. T. Reiter, T. N. Seagroves, M. Bowers, and E. Bier Expression of the Rho-GEF Pbl/ECT2 is regulated by the UBE3A E3 ubiquitin ligase Hum. Mol. Genet., September 15, 2006; 15(18): 2825 - 2835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Mani and E. P. Gelmann The Ubiquitin-Proteasome Pathway and Its Role in Cancer J. Clin. Oncol., July 20, 2005; 23(21): 4776 - 4789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. Cooper, A. W. Hudson, J. Amos, J. Wagstaff, and P. M. Howley Biochemical Analysis of Angelman Syndrome-associated Mutations in the E3 Ubiquitin Ligase E6-associated Protein J. Biol. Chem., September 24, 2004; 279(39): 41208 - 41217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Muhle, S. V. Trentacoste, and I. Rapin The Genetics of Autism Pediatrics, May 1, 2004; 113(5): e472 - e486. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Salvat, G. Wang, A. Dastur, N. Lyon, and J. M. Huibregtse The -4 Phenylalanine Is Required for Substrate Ubiquitination Catalyzed by HECT Ubiquitin Ligases J. Biol. Chem., April 30, 2004; 279(18): 18935 - 18943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J Clayton-Smith and L Laan Angelman syndrome: a review of the clinical and genetic aspects J. Med. Genet., February 1, 2003; 40(2): 87 - 95. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T.-F. Tsai, K.-S. Chen, J. S. Weber, M. J. Justice, and A. L. Beaudet Evidence for translational regulation of the imprinted Snurf-Snrpn locus in mice Hum. Mol. Genet., July 1, 2002; 11(14): 1659 - 1668. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A C Lossie, M M Whitney, D Amidon, H J Dong, P Chen, D Theriaque, A Hutson, R D Nicholls, R T Zori, C A Williams, et al. Distinct phenotypes distinguish the molecular classes of Angelman syndrome J. Med. Genet., December 1, 2001; 38(12): 834 - 845. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nakagawa and J. M. Huibregtse Human Scribble (Vartul) Is Targeted for Ubiquitin-Mediated Degradation by the High-Risk Papillomavirus E6 Proteins and the E6AP Ubiquitin-Protein Ligase Mol. Cell. Biol., November 1, 2000; 20(21): 8244 - 8253. [Abstract] [Full Text]
|
|
|
|
|
|
M. R. W. Mann and MarisaS. Bartolomei Towards a molecular understandingof Prader-Willi and Angelman syndromes Hum. Mol. Genet., September 1, 1999; 8(10): 1867 - 1873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L A E M LAAN, A M W VAN DEN OUWELAND, P L G BAKKER, D J J HALLEY, and C E CATSMAN-BERREVOETS Angelman syndrome: AS phenotype correlated with specific EEG pattern may result in a high detection rate of mutations in the UBE3A gene J. Med. Genet., September 1, 1999; 36(9): 723 - 724. [Full Text]
|
|
|
|
|
|
A. Moncla, P. Malzac, M.-O. Livet, M.-A. Voelckel, J. Mancini, J. C. Delaroziere, N. Philip, and J.-F. Mattei Angelman syndrome resulting from UBE3A mutations in 14 patients from eight families: clinical manifestations and genetic counselling J. Med. Genet., July 1, 1999; 36(7): 554 - 560. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||