Mutation Screening of the UBE3A/E6-AP Gene in Autistic

Home , Ligase

ory elements are also excellent candidates for mimick- ORIGINAL RESEARCH ARTICLE ing the effects of maternal duplication of this region. Currently, UBE3A is the only gene in the 15q11–q13 region found to exhibit maternally imprinted Mutation screening of the expression in regions of interest in autism. Autopsy studies in adult humans have revealed that certain UBE3A/E6-AP gene in mRNA transcripts of UBE3A are solely maternally autistic disorder expressed in the brain but are expressed biparentally elsewhere in the body.14 mRNA expression studies sug- J Veenstra-VanderWeele1, D Gonen1, gest that all transcripts of UBE3A are predominantly BL Leventhal1,2 and EH Cook Jr1,2 maternally derived in total fetal human brain.15 Fur- thermore, evidence from the mouse indicates the hip- 1Laboratory of Developmental Neuroscience, Child and pocampus and the cerebellar Purkinje cells as the pri- Adolescent Psychiatry, Department of Psychiatry, mary areas where expression is solely from the University of Chicago, MC3077, 5841 S Maryland Avenue, maternal chromosome.16 Abnormalities in cerebellar Chicago, Illinois 60637; 2Department of Pediatrics, size and structure have been the most frequent brain University of Chicago, MC3077, 5841 S Maryland Avenue, correlate to autism, both in post mortem17 and MRI Chicago, Illinois 60637, USA studies.18,19 Cytoarchitectural abnormalities in the hippocampus have also been identified in a post mortem study of autistic subjects.20 Keywords: autism; genetics; mutation screen; polymor- Based on these separate lines of evidence, truncating phism; sequencing mutations of UBE3A in AS, imprinting of UBE3A in Previous reports of individuals with autistic disorder brain regions implicated in autism, and maternal with maternal duplications of 15q11–q13,1–11 the Prader- 15q11–q13 duplications in autistic disorder, UBE3A is Willi/Angelman syndrome region, suggest this area as a a strong candidate gene in autistic disorder. We used source of candidate genes in autistic disorder. Maternal sequence ascertained in determining the genomic truncation mutations in UBE3A, which encodes for E6- organization of the gene to design primer sets to screen AP ubiquitin-protein ligase, have been shown to cause the coding region, one portion of a 5Ј UTR exon, and 12,13 Angelman syndrome, which can also result from the a putative promoter sequence for mutations in autis- absence of maternal chromosomal material from this tic subjects. region. Despite showing no evidence for imprinting in Ten autistic subjects were sequenced in both direc- other tissues, this gene was recently discovered to be preferentially maternally expressed in human brain14,15 tions for each of the 12 primer sets shown in Table 1. and expressed solely from the murine maternal chromo- Screening of the coding region in ten autistic subjects some in the hippocampus and cerebellar Purkinje revealed no mutations that would be likely to be cells,16 regions implicated in the neuropathology of responsible for disrupting expression or activity of the autism.17–20 Based on this evidence, the coding region UBE3A gene product. Additionally, no mutations were and a putative promoter region were sequenced in ten found in the area between exons 4 and 5 which has autistic subjects. Several polymorphisms were been predicted to be a promoter region15 (exon naming detected, but no evidence was found for a functional conventions from Kishino and Wagstaff).23 Finally, no mutation. Evidence for likely altered regulation of mutations were detected in exon 6, which when tran- UBE3A expression in maternal 15q11–q13 duplications scribed contains a stop codon that terminates trans- suggests further investigation of the regulatory regions lation that begins at start codons in exons 4 or 5, rather of this gene in autistic disorder. than the usual start codon in exon 7. Duplications of 15q11–q13 of maternal origin are the Three polymorphisms were detected in the mutation most consistent chromosomal abnormalities detected screen, none of which affects amino acid sequence. 1–8,11 in autistic disorder. Recently, it was shown in five Table 2 shows the positions of these three polymor- families that these duplications cause autism or other phisms and four other known polymorphisms in close developmental disorders when maternally but not proximity to the UBE3A gene relative to a PAC gen- 9–11 paternally derived. Studies of autistic populations omic clone (GenBank Accession No. AC004259) that suggest that at least 1% of the cases may be accounted includes almost the entire gene. One of the three poly- 5,11,21 for by chromosome 15 abnormalities. In contrast, morphisms observed, a 3-bp deletion in intron 7 of maternal deletions of this same region cause Angelman bases 408–410 of the exon 8 genomic sequence Syndrome (AS), a disorder typified by mental retar- (GenBank Accession No. AF016704), had been pre- dation, seizures, ataxia, and inappropriate laughter. viously identified in a population of AS subjects Truncating mutations in UBE3A, which encodes the (Beaudet, unpublished data) and tested for linkage dis- E6-AP ubiquitin-protein ligase gene, have also been equilibrium in autistic disorder subjects.21 A 9–10 T found to cause familial AS.12,13 The presence of autism repeat polymorphism was identified in intron 6 from in many patients with AS22 allows for the possibility bases 373–381 of the exon 7 genomic sequence that a maternally transmitted mutation of smaller effect (GenBank Accession No. AF016703). A G to C transver- could lead to a milder phenotype that could resemble sion that fails to alter the amino acid sequence was autistic disorder.23 Mutations in UBE3A or its regulat- identified within the coding sequence at base 687 of Mutation screening of the UBE3A/E6-AP gene J Veenstra-VanderWeele et al 65

Figure 1 Map of UBE3A on chromosome 15. The two new polymorphisms detected in this mutation screen are indicated and underlined. Five previously identiﬁed polymorphisms are also indicated. D15S122, D15S10, the OP2 7-bp deletion, and the intron 7 3-bp deletion, were previously tested for linkage disequilibrium with autistic disorder.21 The repeat within exon 16 in the 3Ј UTR30 has not yet been tested for linkage disequilibrium with autistic disorder. Exons are shown by solid boxes. The GABRB3 155CA-2 marker found to be in linkage disequilibrium with autistic disorder21 is telomeric to UBE3A, as is D15S156, the marker where a suggestive linkage ﬁnding peaked.24

Table 1 Oligonucleotide primers

Exon Sense Antisense TA Source Position

4–5 5Ј-caccctgatgtcaccgaatggccacagc-3Ј 5Ј-ccatcctgcaagccactccttttacctccactg-3Ј 60.0°C X98035, 83–110, X98036 8–40 65Ј-ttttgttttgcaagccagctctg-3Ј 5Ј-taagtaccccaagattgcttccaaac-3Ј 53.2°C X98037 17–173 75Ј-tggccacctgatctgaccactttctttc-3Ј 5Ј-aatttaggcgtgtgcatttatctatctagtagtagaaac-3Ј 54.2°C AF016703 260–696 85Ј-acttcagtgcttgactaacttttgccttgatg-3Ј 5Ј-atgattagacctgtaaaatgtagttattattcctgtccg-3Ј 55.6°C AF016704 377–951 95Ј-agcatgggtgaggtcattggtataggtaggtagttcattg-3Ј 5Ј-gagatttcagttcttccttggtgtgttgtttaactttccg-3Ј 55.7°C AF016705 147–540 95Ј-aacagctatgaccatgcaacagagtaaacatacatatt-3Ј 5Ј-tctccatcattctccgaatctggtc-3Ј 55.0°C AF016705 307–1023 95Ј-ttgcaaagcgatgagcaagctacc-3Ј 5Ј-aatgaattcactgaactgtatcatg-3Ј 55.0°C AF016705 926–1683 95Ј-ccctttatattgaatgctgtcacaaagaatttggg-3Ј 5Ј-aaccgatactttgtagaacacatctataaacttgcacagg-3Ј 55.2°C AF016705 1440–1810 10 5Ј-ttttactatgctagtgtctgttttgtgtctgtc-3Ј 5Ј-tgatacaaatcagttcctaaaacctacaaattc-3Ј 52.7°C AF016706 299–841 11 5Ј-ttatagcagatatatgtggacattgttttctcaagtg-3Ј 5Ј-tttaagtgtttctggtacttcggtcagattaaaac-3Ј 52.3°C AF016707 402–956 12 5Ј-acatcacagtgactgacaatctttcttttataggttgg-3Ј 5Ј-aacaccctgcttcttgctttatactacaactcaacttagc-3Ј 55.0°C AF016707 1372–1870 13–14 5Ј-atggccttttcatagcttaatattg-3Ј 5Ј-tgacaatttgtgagtttgcttatttg-3Ј 51.1°C AF016707 2604–3245 15 5Ј-tacgtattcagagctgttcaggataaatttgc-3Ј 5Ј-ggcaatttcttaaaagtttcctcacacaatg-3Ј 54.2°C AF016708 359–872 16 5Ј-cctaaaatggtatctgttatgtagtaaa-3Ј 5Ј-aaatttatccctcgttatatttttaa-3Ј 48.9°C AF016708 1262–1725

the exon 11 genomic sequence (GenBank Accession The continuing localization of mutations in Angelman No. AF016707). Syndrome25 may assist in this effort by identifying Linkage disequilibrium mapping of the 15q11–q13 areas outside of the known coding region that merit region revealed significant family-based association screening in autism. Some caution must be taken in between autistic disorder and the GABRB3 155CA-2 interpreting these results due to the small sample size. marker, also distal to UBE3A.21 A preliminary finding Additionally, one reported crossover in a familial AS of linkage in autistic disorder began at GABRB3 mark- case has been localized telomeric to D15S122,26 which ers, extended telomerically, and peaked at the D15S156 would implicate a disruption in the untranslated first marker, distal to the UBE3A gene on chromosome 15.24 exon or in 5Ј regulatory elements, neither of which was No significant linkage disequilibrium was detected covered in the present mutation screen. Analysis of the between four markers within the UBE3A gene and promoter region or other regulatory gene elements autism.21 However, given the evidence for maternal imprinting of UBE3A expression in the brain, linkage Table 2 Polymorphism positions disequilibrium findings may be weakened since only maternal transmissions may be informative. The well-documented finding of 15q11–q13 maternal Polymorphism Position within 15q11–13 PAC clone (GenBank duplication events in autism suggests that an alteration Accession No. AC004259) in the UBE3A gene product’s activity could be found in subjects with autistic disorder. In patients without OP2 7-bp deletion 92010–92016 this duplication, an alteration in activity could derive D15S10 (CA repeat) 91633–91653 from either altered gene expression due to a change in D15S122 (CA repeat) 86746–86786 regulatory elements or dysregulation of the protein’s 9/10 T repeat in intron 6 57520–57528 activity due to a change in the coding sequence. The 3-bp deletion in intron 7 27798–27800 absence of a coding region mutation in this current Transversion in exon 11 8774 study suggests the former as a more likely possibility. Mutation screening of the UBE3A/E6-AP gene J Veenstra-VanderWeele et al 66 would be the next logical step in searching for possible Acknowledgements mutations of this gene in autistic disorder. The linkage Tahseen Mohammed, Zhi-Ying Yang, Shuya Yan and and linkage disequilibrium evidence also points to a Kristi Nix provided expert technical assistance. Arthur region distal on the chromosome to the coding region Beaudet and Toshinobu Matsuura generously shared of UBE3A, suggesting upstream regulatory elements as sequence and genomic organization information. This not only functional candidates, but positional candi- study was supported in part by NIMH R01 MH52223 dates as well. (EHC), NIH K02 MH01389 (EHC), the Jean Young and Walden W Shaw Foundation (BLL), the Harris Foun- dation (BLL), and the Brain Research Foundation (EHC). Methods

Subjects References Subjects consented to participate in a study of candi- 1 Gillberg C, Steffenburg S, Wahlstrom J, Gillberg I, Sjostedt A, Mar- date loci for autistic disorder at the University of tinsson T et al. Autism associated with marker chromosome. JAm Chicago Developmental Disorders Clinic. Inclusion cri- Acad Child Adolesc Psychiatry 1991; 30: 489–494. teria included the following: (1) diagnosis of autistic 2 Baker P, Piven J, Schwartz S, Patil S. Duplication of chromosome disorder, by use of the Autism Diagnostic Interview 15q11–13 in two individuals with autistic disorder. J Autism Dev 27 Disord 1994; 24: 529–535. Revised (ADI-R) and age-appropriate versions of the 3 Bundey S, Hardy C, Vickers S, Kilpatrick M, Corbett J. Duplication 28 Autism Diagnostic Observation Schedule or the Pre- of the 15q11–13 region in a patient with autism, epilepsy, and Linguistic Autism Diagnostic Observation Schedule;29 ataxia. Dev Med Child Neurol 1994; 36: 736–742. (2) mental age Ͼ18 months, as assessed by the Vine- 4 Leana-Cox J, Jenkins L, Palmer CG, Plattner R, Sheppard L, Flejter Ͼ WL et al. Molecular cytogenetic analysis of inv dup(15) chromo- land Adaptive Behavior Scales; (3) nonverbal IQ 35; somes, using probes specific for the Prader-Willi/Angelman syn- (4) confirmation of the diagnosis of autistic disorder, drome critical region: clinical implications. Am J Hum Genet 1994; by a child psychologist and a child psychiatrist; and 54: 748–756. (5) exclusion of known etiologies of autistic disorder, 5 Schinzel AA, Brecevic L, Bernasconi F, Binkert F, Berthet F, Wuil- by physical examination, including neurological exam- loud A et al. Intrachromosomal triplication of 15q11–q13. J Med Genet 1994; 31: 798–803. ination and Wood’s lamp examination to exclude 6 Crolla J, Harvey J, Stitch F, Dennis N. Supernumerary marker 15 tuberous sclerosis, and by FRAXA DNA testing. chromosomes: a clinical, molecular and FISH approach to diagnosis and prognosis. Hum Genet 1995; 95: 161–170. 7 Hotopf M, Bolton P. A case of autism associated with partial tetra- somy 15. J Autism Dev Disord 1995; 25: 41–49. Mutation screen 8 Flejter WL, Bennett-Baker PE, Ghaziuddin M, McDonald M, Shel- Oligonucleotide primers were designed from genomic don S, Gorski JL. Cytogenetic and molecular analysis of inv dup(15) sequence flanking the exons of the UBE3A coding chromosomes observed in two patients with autistic disorder and region. The primers, with their annealing temperatures mental retardation. Am J Med Genet 1996; 61: 182–187. and positions, are listed in Table 1. Amplification was 9 Cook EH, Lindgren V, Leventhal BL, Courchesne R, Lincoln A, ␮ ␮ Shulman C et al. Autism or atypical autism in maternally but not carried out in a total volume of 10 l containing 0.5 M paternally derived proximal 15q duplication. Am J Hum Genet of each primer, 200 ␮M of each dNTP (Perkin Elmer, 1997; 60: 928–934. Foster City, CA, USA), 2.5 mM MgCl2, 10 mM Tris-HCI, 10 Browne CE, Dennis NR, Maher E, Long FL, Nicholson JC, Silli- 50 mM KCI, 0.001% gelatin, 0.3 Units of AmpliTaq bourne J et al. Inherited interstitial duplications of proximal 15q: genotype-phenotype correlations. Am J Hum Genet 1997; 61: Gold DNA Polymerase (Perkin Elmer), and 50 ng of 1342–1352. DNA template. DNA was extracted from the blood of 11 Schroer RJ, Phelan MC, Michaelis RC, Crawford EC, Skinner SA, 10 subjects seen at the University of Chicago Develop- Cuccaro M et al. Autism and maternally derived aberrations of mental Disorders Clinic as previously described.21 PCR chromosome 15q. Am J Med Genet 1998; 76: 327–336. was performed on a Perkin Elmer 9600 thermocycler 12 Matsuura T, Sutcliffe JS, Fang P, Galjaard RJ, Jiang YH, Benton CS ° et al. De novo truncating mutations in E6-AP ubiquitin-protein with an initial activating step of 95 C for 12 min, fol- ligase gene (UBE3A) in Angelman syndrome. Nature Genet 1997; lowed by 40 cycles of PCR: denaturing at 94°C for 15 s, 15: 74–77. annealing at optimum temperature (see Table 1) for 13 Kishino T, Lalande M, Wagstaff J. UBE3A/E6-AP mutations cause 30 s, and extension at 72°C for 30 s, and followed by a Angelman syndrome. Nature Genet 1997; 15: 70–73. ° 14 Rougeulle C, Glatt H, Lalande M. The Angelman syndrome candi- final extension step of 72 C for 2 min. PCR products date gene, UBE3A/E6-AP is imprinted in brain. Nature Genet 1997; were sized on 4% agarose gels, and excised fragments 17: 14–15. were extracted using the QIAquick Gel Extraction Kit 15 Vu TH, Hoffman AR. Imprinting of the Angelman syndrome gene, (Qiagen, Valencia, CA, USA). Cycle sequencing was UBE3A, is restricted to brain. Nature Genet 1997; 17: 12–13. performed using the ABI prism dRhodamine Termin- 16 Albrecht U, Sutcliffe JS, Cattanach BM, Beechey CV, Armstrong D, Eichele G et al. Imprinted expression of the murine Angelman ator Cycle Sequencing Ready Reaction Mix (PE syndrome gene, UBE3A, in hippocampal and Purkinje neurons. Applied Biosystems, Foster City, CA, USA) on the Per- Nature Genet 1997; 17: 75–78. kin Elmer 9600 thermocycler. Sequencing fragments 17 Bauman ML, Kemper TL. Neuroanatomic observations of the brain were separated by capillary electrophoresis and in autism. In: Bauman ML, Kemper TL (eds). The Neurobiology detected by laser-induced fluorescence on an ABI of Autism. Johns Hopkins University Press: Baltimore, 1994, pp prism 119–145. 310 Genetic Analyzer using POP6 polymer (PE 18 Courchesne E. Brainstem, cerebellar, and limbic neuroanatomical Applied Biosystems). abnormalities in autism. Curr Opin Neurobiol 1997; 7: 267–278. Mutation screening of the UBE3A/E6-AP gene J Veenstra-VanderWeele et al 67 19 Ritvo ER, Freeman BJ, Scheibel AB, Duong T, Robinson H, Guthrie 27 Lord C, Rutter M, Le Couteur A. Autism Diagnostic Interview— D et al. Lower Purkinje cell counts in the cerebella of four autistic Revised: a revised version of a diagnostic interview for caregivers subjects: initial findings of the UCLA-NSAC autopsy research of individuals with possible pervasive developmental disorders. J report. Am J Psychiatry 1986; 143: 862–866. Autism Dev Disord 1994; 24: 659–685. 20 Raymond GV, Bauman ML, Kemper TL. Hippocampus in autism: 28 Lord C, Rutter M, Goode S, Heemsbergen, J, Jordan H, Mawhood A Golgi analysis. Acta Neuropathologica 1996; 91: 117–119. L et al. Autism Diagnostic Observation Schedule: a standardized 21 Cook EH, Courchesne RY, Cox NJ, Lord C, Pizzo S, Gonen D et al. observation of communicative and social behavior. J Autism Linkage disequilibrium mapping of autistic disorder with 15q11– Develop Disord 1989; 19: 185–212. 13 markers. Am J Hum Genet 1998; 62: 1077–1083. 29 DiLavore P, Lord C, Rutter M. Pre-linguistic Autism Diagnostic 22 Steffenburg S, Gillberg CL, Steffenburg U, Kyllerman M. Autism in Observation Schedule (PL-ADOS). J Autism Dev Disord 1995; 25: Angelman syndrome: a population-based study. Ped Neurol 1996; 355–379. 14: 131–136. 30 Kishino T, Lalande M, Wagstaff J. UBE3A/E6-AP mutations in Ang- 23 Kishino T, Wagstaff J. Genomic organization of the UBE3A/E6-AP elman syndrome indicate a role for the ubiquitin pathway in brain gene and related pseudogenes. Genomics 1998; 47: 101–107. development. Am J Hum Genet 1997; 61: A39. 24 Pericak-Vance M, Wolpert C, Menold M, Bass M, DeLong G, Beaty L et al. Linkage evidence supports the involvement of chromosome 15 in autistic disorder. Am J Hum Genet 1997; 61: A40. Correspondence: EH Cook Jr, MD, University of Chicago Department 25 Malzac P, Webber H, Moncla A, Graham JM, Kukolich M, Williams of Psychiatry, 5841 South Maryland Avenue MC3077, Chicago, IL C et al. Mutation analysis of UBE3A in Angelman syndrome 60637, USA. E-mail: edȰyoda.bsd.uchicago.edu patients. Am J Hum Genet 1998; 62: 1353–1360. Received 3 June 1998; revised 17 and 30 July 1998; accepted 30 26 Greger V, Reis A, Lalande M. The critical region for Angelman syn- July 1998 drome lies between D15S122 and D15S113. Am J Med Genet 1994; 53: 396–398.