Article Text

Download PDFPDF

Microdeletion/duplication at 15q13.2q13.3 among individuals with features of autism and other neuropsychiatric disorders
  1. D T Miller1,2,3,4,
  2. Y Shen1,3,4,5,
  3. L A Weiss3,4,5,6,
  4. J Korn5,6,
  5. I Anselm3,7,
  6. C Bridgemohan3,8,
  7. G F Cox2,3,9,
  8. H Dickinson10,
  9. J Gentile2,3,11,
  10. D J Harris2,3,
  11. V Hegde2,3,
  12. R Hundley3,8,
  13. O Khwaja3,4,7,
  14. S Kothare3,7,12,
  15. C Luedke3,13,
  16. R Nasir3,4,8,
  17. A Poduri3,7,
  18. K Prasad3,7,
  19. P Raffalli3,7,
  20. A Reinhard1,2,
  21. S E Smith2,3,9,
  22. M M Sobeih3,4,7,
  23. J S Soul3,7,
  24. J Stoler2,3,
  25. M Takeoka3,7,12,
  26. W-H Tan2,3,
  27. J Thakuria2,3,
  28. R Wolff3,7,
  29. R Yusupov2,3,
  30. J F Gusella3,4,5,6,
  31. M J Daly3,4,5,6,
  32. B-L Wu1,3,4
  1. 1
    Department of Laboratory Medicine, Children’s Hospital Boston, Boston, Massachusetts, USA
  2. 2
    Division of Genetics, Children’s Hospital Boston, Boston, Massachusetts, USA
  3. 3
    Harvard Medical School, Boston, Massachusetts, USA
  4. 4
    Autism Consortium, Boston, Massachusetts, USA
  5. 5
    Center for Human Genetic Research, Massachusetts General Hospital, Boston, Massachusetts, USA
  6. 6
    The Broad Institute of Harvard and MIT, Cambridge, Massachusetts, USA
  7. 7
    Department of Neurology, Children’s Hospital Boston, Boston, Massachusetts, USA
  8. 8
    Division of Developmental Medicine, Children’s Hospital Boston, Boston, Massachusetts, USA
  9. 9
    Genzyme Corporation, Cambridge, Massachusetts, USA
  10. 10
    Department of Otolaryngology and Communication Enhancement, Children’s Hospital Boston, Boston, Massachusetts, USA
  11. 11
    Department of Psychiatry, Children’s Hospital Boston, Boston, Massachusetts, USA
  12. 12
    Division of Epilepsy and Clinical Neurophysiology, Children’s Hospital Boston, Boston, Massachusetts, USA
  13. 13
    Division of Endocrinology, Children’s Hospital Boston, Boston, Massachusetts, USA
  1. Dr B-L Wu, Department of Laboratory Medicine, Children’s Hospital Boston, 300 Longwood Ave, Boston, Massachusetts 02115, USA; bai-lin.wu{at}


Background: Segmental duplications at breakpoints (BP4–BP5) of chromosome 15q13.2q13.3 mediate a recurrent genomic imbalance syndrome associated with mental retardation, epilepsy, and/or electroencephalogram (EEG) abnormalities.

Patients: DNA samples from 1445 unrelated patients submitted consecutively for clinical array comparative genomic hybridisation (CGH) testing at Children’s Hospital Boston and DNA samples from 1441 individuals with autism from 751 families in the Autism Genetic Resource Exchange (AGRE) repository.

Results: We report the clinical features of five patients with a BP4–BP5 deletion, three with a BP4–BP5 duplication, and two with an overlapping but smaller duplication identified by whole genome high resolution oligonucleotide array CGH. These BP4–BP5 deletion cases exhibit minor dysmorphic features, significant expressive language deficits, and a spectrum of neuropsychiatric impairments that include autism spectrum disorder, attention deficit hyperactivity disorder, anxiety disorder, and mood disorder. Cognitive impairment varied from moderate mental retardation to normal IQ with learning disability. BP4–BP5 covers ∼1.5 Mb (chr15:28.719–30.298 Mb) and includes six reference genes and 1 miRNA gene, while the smaller duplications cover ∼500 kb (chr15:28.902–29.404 Mb) and contain three reference genes and one miRNA gene. The BP4–BP5 deletion and duplication events span CHRNA7, a candidate gene for seizures. However, none of these individuals reported here have epilepsy, although two have an abnormal EEG.

Conclusions: The phenotype of chromosome 15q13.2q13.3 BP4–BP5 microdeletion/duplication syndrome may include features of autism spectrum disorder, a variety of neuropsychiatric disorders, and cognitive impairment. Recognition of this broader phenotype has implications for clinical diagnostic testing and efforts to understand the underlying aetiology of this syndrome.

View Full Text

Statistics from

Recurrent microdeletion or microduplication events are a common cause of developmental delay and mental retardation.1 2 Most of these events are mediated by recombination between segmentally duplicated sequences through an established mechanism of non-allelic homologous recombination (NAHR).3 Clinical genetic testing of individuals with developmental delay, mental retardation, and autism spectrum disorders using whole genome high resolution array comparative genomic hybridisation (CGH) has revealed the clinical importance of ever smaller microdeletions and microduplications such as deletions in the range of 500–650 kb at 17q21.3 associated with developmental delay46 and at 16p11.2 among individuals with autism.79

The proximal portion of chromosome 15q is a well known region of genomic instability that contains many segmental duplications.5 1014 Deletions at 15q11–q13 that result in Prader–Willi syndrome and Angelman syndrome (PWS/AS) are typically ∼4 Mb, and are mediated by repetitive elements with clustered breakpoints (BP) at either of two proximal sites (BP1 and BP2) and one distal site (BP3).15 16 Genomic imbalance on proximal 15q has been associated with developmental delay and autism, most notably maternally inherited duplications of the PWS/AS region at 15q11q1314 1719 and also a large duplication from BP1–BP5 in a simplex case of autism.20

A recurrent microdeletion syndrome mediated by more distal segmental duplication breakpoints on chromosome 15q13.2q13.3, designated BP4–BP514 at chromosome 15q13.2q13.3, was recently described.21 Minor dysmorphic features, mental retardation, epilepsy, and/or electroencephalogram (EEG) abnormalities were common in the reported cases of BP4–BP5 deletion. This deletion is ∼1.5 Mb and includes six reference genes (MTMR15, MTMR10, TRPM1, KLF13, OTUD7A and CHRNA7) and an miRNA gene (hsa-mir-211). The alpha7-nicotinic receptor subunit gene (CHRNA7) is a candidate gene for seizures based on several lines of evidence: the presence of seizures in seven out of nine probands with BP4–BP5 deletions that include CHRNA7,21 genetic linkage to 15q among subjects with epilepsy,22 23 a Chrna7 knockout mouse model that has an abnormal EEG,24 and genetic linkage studies supporting susceptibility to juvenile myoclonic epilepsy.25

We present eight patients with BP4–BP5 deletions and duplications, and two with a smaller duplication within BP4–BP5, identified by whole genome high resolution array CGH on patients undergoing clinical genetics evaluations from Children’s Hospital Boston (CHB), or by genome-wide oligonucleotide genotyping arrays on samples from the Autism Genetic Resource Exchange (AGRE). We report a broader phenotype than originally recognised, including autism spectrum disorder and other neuropsychiatric disorders including anxiety, attention deficits, mood disorder, and cognitive impairment with or without mental retardation and without epilepsy.


CH Boston samples

We performed whole genome high resolution oligonucleotide array CGH on 1445 consecutively submitted clinical samples with referring diagnoses including developmental delay (DD; n = 639), mental retardation (MR) or learning disability (LD; n = 49 for MR/LD), autism spectrum disorder (ASD; n = 177) or pervasive developmental disorder (PDD; n = 85; total for ASD/PDD = 262), multiple congenital anomalies (n = 118), dysmorphic features (n = 224), seizures (n = 49) or undefined/other phenotypes (n = 104). All patients with 15q13 imbalance were examined by a developmental specialist and a clinical geneticist. A team of specialists from clinical genetics, neurology, and developmental medicine at CHB conducted a medical record review approved by the CHB institutional review board (IRB).

Array CGH and confirmatory studies

CGH was performed according to previously published methods of analysis using the Agilent 244K human genome oligonucleotide CGH microarray; all coordinates reflect human genome build 18 (G4411B, Agilent Technologies, Palo Alto, California, USA).26 Independent confirmation of deletion/duplication of the 15q13.2q13.3 region was performed by multiplex ligation dependent probe amplification (MLPA) and fluorescence in situ hybridisation (FISH) according to previously described methods.26

AGRE samples

DNA samples from 751 multiplex families were obtained from the AGRE collection of multiplex families27 using previously described sample selection criteria.8 Our final dataset included 1441 individuals affected with autism spectrum disorders, 1420 parents, and 132 unaffected/unknown siblings. This study was approved by the Massachusetts Institute of Technology (MIT) IRB.

Genotyping and confirmatory studies

AGRE samples were genotyped on Affymetrix 5.0 arrays at the Genetic Analysis Platform of the Broad Institute, and analysed for copy number variants with the COPPER and Birdseye algorithms.8 28 SNP genotype data and raw intensity files have been released to AGRE, and are available to the research community under AGRE guidelines. Independent confirmation of deletion/duplication of the 15q13.2q13.3 region among AGRE samples was performed using Agilent 244k array CGH and MLPA at CHB.


We identified 10 patients with genomic imbalance at chromosome 15q13.2q13.3, including five with BP4–BP5 microdeletions from the CHB cohort (chr15:28.7 Mb to ∼30.3 Mb; hg18). Patient 5 was identified after the manuscript was originally submitted, and was not included among the original 1445 DNA samples. We did not find any cases of BP4–BP5 microdeletion among 1420 parents, and 132 unaffected/unknown siblings in the AGRE samples. We identified three patients with reciprocal BP4–BP5 duplications; and two siblings with a smaller duplication of ∼500 kb within BP4–BP5 (chr15:28.9–29.4 Mb; hg18) (fig 1). BP4 is more than 1 Mb distal to the telomeric breakpoint (BP3) of the 15q11q13 deletion associated with PWS/AS and the reciprocal duplication that has been associated with autism. None of the patients from CHB or AGRE had other clinically significant copy number variants elsewhere in the genome. All CHB patients had normal karyotypes and fragile X testing by Southern blot and polymerase chain reaction (PCR). All deletion and duplication events in these samples were confirmed by dye reversal array CGH (fig 1) and through a customised MLPA assay or FISH (supplemental figs 1 and 2). Genomic coordinates of all deletions and duplications are listed in tables 1 and 2.

Figure 1

In the top panel, an ideogram of proximal chromosome 15q (15q11q14) shows the Prader–Willi syndrome and Angelman syndrome (PWS/AS) region and the more distal 15q13.2q13.3 region between BP4 and BP5. Lower panels show scatter plots of array comparative genomic hybridisation (CGH) data for a deletion of ∼1.5 Mb superimposed with dye-swap scatter plot (note the mirrored distribution of spots). The lower scatter plot represents a duplication of ∼500 Kb within the BP4–BP5 interval. The relative positions of seven genes (six reference genes and one miRNA gene) are shown in the bottom panel (grey bars). The 1.5 Mb deletions (chr15:28.719–30.232 Mb; hg18) include all seven genes, while the 500 kb duplications (chr15:28.902–29.404 Mb; hg18) contain four genes (MTMR15, MTMR10, TRPM1 and hsa-mir-211) within the BP4–BP5 at chromosome 15q13.2q13.3.

Figure 2

Photographs of five patients with 15q13.2q13.3 BP4–BP5 microdeletion, and two patients with BP4–BP5 microduplication, all from the Children’s Hospital Boston cohort (numbering corresponds to patient number in the text and supplemental data). We obtained written consent to publish photographs for each individual included in this figure.

Table 1 15q13.2q13.3 microdeletions
Table 2 15q13.2q13.3 microduplications

In general, cognitive performance of patients with 15q13.2q13.3 microdeletion/microduplication was variable. Test scores ranged from moderate MR to the normal range. Although some patients had full scale IQ in the normal range, they all had some degree of language impairment and/or learning disability. Expressive language was consistently more delayed than receptive language. Many of these individuals showed capacity for ongoing improvement in academic and social skills. Neurobehavioural symptoms were very common in microdeletion/microduplication patients. All had difficulties with social interactions. Overall, dysmorphic features were mild (fig 2), and the neurobehavioural symptoms were the most significant cause of disability among these patients.

15q13.2q13.3 microdeletion patients

Clinical features of the five individuals from the CHB cohort with 15q13.2q13.3 BP4–BP5 deletions are presented in table 1 and supplemental data. All have subtle dysmorphology findings based on examination by a clinical geneticist. Cognitive testing was performed on all individuals. Patient 4 has mental retardation, but the other four patients have variable scores in the range of “below average” to “average”. Patients 1 and 3 showed significant non-verbal learning disability. None has a history of developmental regression. All have impaired language skills ranging from mild to severe, accounting for a diagnosis of developmental delay in all cases. Those with developed language have significant early expressive language impairment but with better receptive abilities. Profiles commonly included developmental or oro-motor dyspraxia with disarticulation. Older children have good language abilities if they have developed expressive language.

Motor delays were not prominent, especially compared to cognitive and behavioural issues, but were observed among patients with BP4–BP5 deletion. Patients 1–4 had hypotonia that resolved over time, and patient 5 was delayed in walking until age 18 months, suggesting the possibility of undiagnosed mild hypotonia. Patients 1 and 5 had problems with fine motor coordination. None of the patients with BP4–BP5 deletion had examination findings consistent with cerebral palsy.

The majority of individuals with BP4–BP5 deletion have a diagnosis of an autism spectrum disorder or autistic features such as variably poor eye contact and other difficulties with social interactions. Beyond concerns about autistic features, many of these patients have other behavioural problems. All have some degree of difficulty with attention, hyperactivity, mood regulation, and impulsive behaviours. Patient 3 has attention deficit hyperactivity disorder (ADHD), bipolar disorder, and anxiety disorder. None have problems sleeping at night.

15q13.2q13.3 microduplication patients

Clinical features of the five individuals from the CHB and AGRE cohorts with 15q13.2q13.3 BP4–BP5 duplications are presented in table 2 and supplemental data. Four of five patients with the duplications have a diagnosis of autism. Patient 7 does not carry an autism diagnosis, but displays some repetitive behaviours and expressive language delay. Patients 6 and 8 also have severe expressive language delay, but language testing results were not available on patients 9 and 10. Patient 6 has a history of anxiety spectrum disorder/obsessive compulsive disorder in addition to autism. Cognitive and behavioural test results are not available for duplication patients from the AGRE cohort, although patient 8’s Vineland score suggests he would fall in the range of mental retardation. Clinical examination of patients 6 and 7 (CHB cohort) did not suggest a consistent pattern of dysmorphology and neither had a history of seizures. Patients 8–10 (AGRE cohort) were not available for examination.

Patient 6 (CHB cohort) and patient 8 (AGRE cohort) have a de novo duplication and an autism diagnosis. Patient 7 (CHB cohort) has a maternally inherited duplication from an apparently unaffected parent and has developmental delay with a cleft lip and palate. The paternal grandfather also carries the duplication and was apparently unaffected. Patients 9 and 10 (AGRE cohort) are siblings with autism and a smaller duplication nested within BP4–BP5 inherited from their apparently unaffected mother.


The recent recognition of genomic imbalance at chromosome 15q13.2q13.3 is certainly due to increasing use of whole genome high resolution array CGH in the evaluation of individuals with developmental delay, mental retardation, and autism spectrum disorder. Our results expand the phenotype of the 15q13.2q13.3 microdeletion/duplication syndrome to include clinically significant developmental delays and features of autism spectrum disorder in a majority of our patients with 15q13.2q13.3 microdeletion/duplication. A larger proportion of patients with duplication had a confirmed diagnosis of autism, while patients with deletion had either pervasive developmental disorder–not otherwise specified (PDD-NOS), autistic features, or other neurobehavioural disabilities.

Among patients 1–7 from the CHB cohort, and patient 8 from AGRE, all had some degree of developmental delay, particularly in expressive language, although not necessarily with mental retardation. These findings are not inconsistent, as verbal cognitive measures have minimal demands for language formulation. For example, they do not consider articulation, pragmatics, and reciprocity. Six of these patients (patients 1, 4, 6, 8–10) have a clinical diagnosis on the autism spectrum, and another is considered to have autistic features (patient 5).

Many of these patients also exhibited other neurobehavioural symptoms such as attentional problems, anxiety, mood instability, and impulsivity. Our patients had minor dysmorphic features previously described in this syndrome, but did not show the high prevalence of epilepsy/EEG abnormalities (7 of 9 patients) reported in the earlier study of Sharp et al.21 In that study, patients were ascertained for mild or moderate mental retardation, but in our study cases were referred for a variety of diagnoses.

We found BP4–BP5 microdeletions in 4/950 (0.4%) and microduplications in 2/950 (0.2%) of CHB cases with developmental delay, mental retardation, or autism spectrum disorder, and 1/751 (0.13%) of AGRE index (proband) cases. Four families had a proband with maternally inherited imbalance; two involved a BP4–BP5 deletion from a mother with learning issues (patients 1 and 2) and two had duplications also present in apparently unaffected mothers (patient 7; patients/siblings 9 and 10). Two BP4–BP5 duplication cases were de novo (patients 6 and 8). BP4–BP5 duplication was also reported in 1/960 (0.1%) European–American controls,21 and an overlapping 2.3 Mb duplication (chr15:28,393,128–30,740,356) was reported in 2/776 (0.3%) controls (506 unrelated healthy individuals from Northern Germany, 270 HapMap individuals).29

Based on this sample, the BP4–BP5 microdeletion appears to be fully penetrant with variable expressivity, while the microduplication may not be fully penetrant. All patients with BP4–BP5 deletion showed clinically significant symptoms, including the parents of patients 1 and 2. Sharp et al21 found no BP4–BP5 deletions among 2962 unaffected control subjects, and the combined experience from both studies indicates that 14/14 subjects with BP4–BP5 deletion exhibited significant developmental disability. Parent-of-origin effects could account for variable penetrance and expressivity of both deletions and duplications, but we were not able to test this in our cohort. There is no evidence of any parent-of-origin effect based on the prior report.21 The importance of parent-of-origin has been well established for the 15q11 locus.30 By comparison, duplications and other rearrangements in the 15q11–q13 region can result in an autism and mental retardation phenotype, but are also observed in phenotypically normal individuals,31 underscoring the phenotypic variability common in many genomic imbalance disorders such as microduplications at 1q41q42,32 3q29,33 16p11.2,8 and 22q11.34

Further research efforts to understand this variable phenotype should focus on additional imbalance events and the genes located in this region. CHRNA7 is a candidate gene for epilepsy, but is also a candidate for the broader phenotype of neuropsychiatric and neurological disease based on genetic association with schizophrenia,3539 and bipolar disorder,40 as well as biological studies that support a role for CHRNA7 in neuropsychiatric disease.41 Negative association studies with schizophrenia have also been reported,42 43 and the literature is likely biased toward more reports of positive association. The 15q13.2q13.3 deletion and duplication events described here show relatively consistent breakpoints at BP4 and BP5, but the smaller duplication suggests that further study may reveal a smaller deletion event that could identify a smallest region of overlap (SRO) that includes CHRNA7.

Awareness of the broad neurobehavioural phenotype will alert clinicians to consider testing for a chromosome 15q13.2q13.3 imbalance on the basis of developmental delay, autistic features, or the other neuropsychiatric issues such as expressive language delay, ADHD, anxiety disorders and/or obsessive compulsive disorder, bipolar or mood disorder, and subclinical EEG or magnetic resonance imaging abnormalities. Many of these conditions are known to be common comorbid psychiatric disorders in individuals with autism,4446 and vice versa.47 The subtlety of physical examination findings underscores the importance of clinical diagnostic tests that include this locus to facilitate an earlier diagnosis and more accurate recurrence risk counselling. In most cases of autism spectrum disorder, some clinical symptoms are apparent before the age of 3 years, but the current average age at clinical diagnosis is 5 years.48 Clinical genetic testing that leads to early behavioural interventions could notably improve the developmental outcome for individuals with a chromosome 15q13.2q13.3 imbalance.


We would like to thank the families and individuals cited in this work, and also the families from AGRE who agreed to share their information and samples for this research. For helpful discussion: Drs Orah Platt and Leonard Rappaport from Children’s Hospital Boston. For assistance with clinical evaluations: Drs Mustafa Sahin and Susan Waisbren from Children’s Hospital Boston and Dr Eileen Bickford from Salud Family Health Center in Longmont, CO. For technical support of aCGH and MLPA: Va Lip, Xiaoming Sheng, Hong Fang, Hong Shao from Children’s Hospital Boston. For FISH confirmation of clinical cases: Drs Christa Lese-Martin and David Ledbetter from Emory University; Dr Arthur R Brothman and Ms Emily Aston from University of Utah. For assistance with IRB documentation: Ms Stephanie Brewster and Dr Ingrid Holm from Children’s Hospital Boston. We gratefully acknowledge the resources provided by the Autism Genetic Resource Exchange (AGRE) Consortium and the participating AGRE families. The Autism Genetic Resource Exchange is a program of Autism Speaks and is supported, in part, by grant 1U24MH081810 from the National Institute of Mental Health to Clara M Lajonchere (PI). We also thank the Autism Consortium for their support and enthusiasm.


View Abstract

Supplementary materials


  • ▸ Additional material is published online only at

  • Funding: We acknowledge the following sources of financial support: Young Investigator Award from the Children’s Tumor Foundation (to YS). Ruth L Kirschstein National Research Service Award (to LAW); National Institutes of Health grant P01-GM061354, Autism Speaks, and NARSAD Distinguished Investigator Award (to JFG), Harvard Scholars in Clinical Science Program K30 grant RRO22292-07 (to JT), grants from the Autism Consortium and the Ellison Foundation (to MJD), and grants from the NIH-CETT Program and the MDA Foundation (to BLW).

  • Competing interests: None declared.

  • Patient consent: Obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.