Relating Copy Number Variation to Phenotype –Bridging Phenotype Gaps

Background: Copy-number variation (CNV) is the most prevalent type of structural variation in the human genome. In a proportion of cases which might be considered ‘syndromal autism’, data from the Autism Genome Project (AGP) supports the role of rare de novo CNV in causing Autistic Spectrum Disorders (ASD). We also find evidence for recurrent CNV impacting neurodevelopmental genes that may increase the susceptibility to ASD. It is not entirely clear how CNV influence the development of the clinical features of ASD. Similar CNV studies in intellectual disability are helping to describe syndromal forms of intellectual disability. In that context it is recognized that CNV also influence broader phenotypic manifestations in addition to intellectual ability, including physical and broader developmental characteristics, and medical comorbidities. Furthermore there is also some evidence in the literature that advanced parental age may carry increased risk for carrying a CNV and in turn could confer risk for ASD. 

Objectives: We hypothesized that developmental abnormalities would be more likely to occur in association with rare CNV in ASD and investigated the association between symptoms of atypical development and medical comorbidities in our data with the presence of CNV. We also investigated the relationship with parental age. Symptom profiles were also investigated by CNV type using clustering methods.

Methods: Data was derived from the Autism Genome Project Illumina Infinium 1M and 1M Duo SNP microarray data and clinical phenotypes (ADI-R, ADOS-G, measures of adaptive and cognitive functioning, brief physical measures) (Stage 1 and Stage 2). >2000 individuals with rare CNV were included in the analyses. Associations were carried out in SAS version 9.2 (SAS Institute Inc., Cary, NC, USA), Chi Square and Fisher’s exact tests via proc freq – Logistic regression via proc logistic. Clustering approaches are also being conducted.

Results: Analyses of relationships between rare CNV and selected phenotypes did not suggest the presence of statistically significant associations. Despite suggestions elsewhere, increased maternal and paternal age were not associated with the presence of rare CNVs impacting ASD- or ID- implicated genes. On a case by case basis we found further evidence supporting the role of known CNV, e.g. duplications of 15q11-q13 and also identified known syndromes within our AGP cohort, e.g. Smith Magenis Syndrome.

Conclusions: Even within a large sample such as the AGP, understanding the role of rare CNV in autism and the relationship with ASD is undermined by low power and missing phenotypic information.. Additional phenotypic information and medical histories will be required on a large scale to better describe ASD syndromes which will require better interdisciplinary integration of clinical genetics, psychiatry, neurology and paediatrics.