International Meeting for Autism Research (London, May 15-17, 2008): GENOMIC COPY NUMBER AND PHENOTYPIC VARIABILITY OF THE AUTISM SPECTRUM DISORDERS

GENOMIC COPY NUMBER AND PHENOTYPIC VARIABILITY OF THE AUTISM SPECTRUM DISORDERS

Thursday, May 15, 2008
Champagne Terrace/Bordeaux (Novotel London West)
10:30 AM
Y. Qiao , Medical genetics and Pathology, University of British Columbia, Vancouver, BC, Canada
X. Liu , Psychiatry, Queen's Univerity, Kingston, ON, Canada
M. Koochek , Medical Genetics, University of British Columbia, Vancouver, BC, Canada
N. Riendeau , Medical Genetics, University of British Columbia, Vancouver, BC, Canada
P. Malenfant , Psychiatry, Queen's Univerity, Kingston, ON, Canada
C. Harvard , Pathology, University of British Columbia, Vancouver, BC, Canada
J. M. Hildebrand , Department of Medical Genetics, University of British Columbia, Vancouver, BC, Canada
J. J. Holden , Psychiatry and Physiology, Queen's Univerity, Kingston, ON, Canada
E. Rajcan-Separovic , Pathology, University of British Columbia, Vancouver, BC, Canada
S. M. Lewis , Medical Genetics, University of British Columbia, Vancouver, BC, Canada
Background:

Autism Spectrum Disorders (ASDs) are heritable and complex neurodevelopmental conditions with unknown aetiology because of their significant clinical and genetic heterogeneity. Clinically relevant genomic copy number variants (CNVs) have been detected in 10~27% of ASD subjects studied in 3 different whole genome screening studies.

Objectives: To investigate the correlation between clinically relevant CNVs and the phenotypes of ASD subjects.

Methods:

We used commercial whole genome 1Mb BAC array comparative genomic hybridization (array-CGH) (Spectral Genomics) to screen for CNVs in 99 subjects with idiopathic autism (78 males and 21 females). FISH and real-time quantitative PCRs were used to validate the array-CGH results.

Results: Nine pathogenic CNVs (7 de novo, 1 inherited and 1 unknown origin) were detected in 9 individuals studied (9%). They include 2 unrelated cases with overlapping regions of de novo del(2)(p15-16.1) (4.5 and 5.7 Mb); an aunt and niece with de novo del(14)(q14.2) (0.7 Mb) and de novo dup(15)(q11) (10 Mb); a boy with de novo del(5)(p15.2-15.31) (2.4 Mb) and de novo del(3)(p24.3) (0.1 Mb); a boy with a de novo dup(18)(p11.3); a girl with unknown familial origin of  dup(7)(q11); and two brothers with a maternally inherited del(X)(p11.22) (470 Kb). The 9 pathogenic CNVs were  found in 9% of simplex, 6.9% of multiplex-immediate (1st degree co-relationship) and 13% of multiplex-extended (≥ 2nd degree co-relationship) families. No significant difference was noticed between parental age at conception and CNV frequency. The severity of clinical phenotype appears to be directly related to the gene content instead of the size of the pathogenic CNV.

Conclusions:

CNVs found in individuals with ASDs signal the locations of ASD-related culprit genes, and whole genome screening coupled with extensive phenotyping including medical and morphological assessments, is an effective and cost-efficient approach to improve prediction of candidate genes and detect those of mild to moderate effect.