Copy number variation in Chinese patients with autism spectrum disorder

Abstract

Introduction Although autism spectrum disorder (ASD) may have a multifactorial etiology, there is a strong genetic basis of the condition. With use of chromosomal microarray according to recommendation of international standards, copy number variations of uncertain significance (CNV VUS) are frequently found. The difficulty encountered in interpretation of CNV VUS may lead to challenges in genetic counselling and clinical management. In this study, we aim to study the CNV findings in Chinese children with ASD in Hong Kong, and to gather information with an aim to reclassify CNV VUS found in these children.

Methods Patients with ASD of the department of Paediatrics QMH/HKU were recruited if their Array Comparative Genomic Hybridization (aCGH) were performed within the period from Jan 2011 to August 2014. Diagnosis of ASD was performed by developmental paediatricians and clinical psychologists using the Diagnostic and Statistical Manual of Mental Disorders, Fourth or Fifth Edition (DSM IV/V) criteria. The array was performed in Tsan Yuk Hospital Prenatal Diagnostic Laboratory (TYHPDL) using NimbleGen CGX-135k oligonucleotide array and Agilent CGX 60k oligonucleotide array. Array results were analyzed, information was gathered from the literature and existing databases with an aim to re-classify CNV VUS occurring in our ASD cohort.

Results Among 288 patients with ASD in our cohort, 10 patients carried pathogenic or likely pathogenic CNVs (3.47%). Among the CNV VUS, we found that one CNV overlapping DPP10 (hg[19]chr2:116,534,689-116,672,358) was commonly occurring in TYHPDL database. The frequency of the CNV overlapping DPP10 in our ASD cohort was 0.35%; and 0.96% (9 from 935 individuals) in controls. The CNV was not statistically associated with ASD probands (P=0.467). The total frequency of the CNV overlapping DPP10 among all samples in TYHPDL was 1.01%. Similar CNVs were thought to be likely pathogenic in previous genetic study recruiting mainly Caucasians. However, there were individuals with typical development or not having ASD possessing similar CNVs in TYHPDL database as well as control databases focusing on Chinese population (9 in TYHPDL database, 1 in Singapore Genome Variation Project SGVP, 24 in The Singapore Prospective Study Program SP2).

Conclusion Our study supported that the CNV overlapping DPP10 observed in our cohort may be more prevalent among Chinese and the study seemed to suggest that it was a CNV polymorphism among Hong Kong Chinese. This result would reclassify the CNV VUS to be likely benign in our locality. It also re-emphasized the need to account for ancestry and ethnicity for the precise interpretation of clinical microarray data.