Analysis of Copy Number Variants on Chromosome 21 in Down Syndrome-Associated Congenital Heart Defects

Abstract

One in five people with Down syndrome (DS) are born with an atrioventricular septal defect (AVSD), an incidence 2000 times higher than in the euploid population. The genetic loci that contribute to this risk are poorly understood. In this study, we tested two hypotheses: (1) individuals with DS carrying chromosome 21 copy number variants (CNVs) that interrupt exons may be protected from AVSD, because these CNVs return AVSD susceptibility loci back to disomy, and (2) individuals with DS carrying chromosome 21 genes spanned by microduplications are at greater risk for AVSD because these microduplications boost the dosage of AVSD susceptibility loci beyond a tolerable threshold. We tested 198 case individuals with DS+AVSD, and 211 control individuals with DS and a normal heart, using a custom microarray with dense probes tiled on chromosome 21 for array CGH (aCGH). We found that neither an individual chromosome 21 CNV nor any individual gene intersected by a CNV was associated with AVSD in DS. Burden analyses revealed that African American controls had more bases covered by rare deletions than did African American cases. Inversely, we found that Caucasian cases had more genes intersected by rare duplications than did Caucasian controls. We also showed that previously DS+AVSD (DS and a complete AVSD)-associated common CNVs on chromosome 21 failed to replicate. This research adds to the swell of evidence indicating that DS-associated AVSD is similarly heterogeneous, as is AVSD in the euploid population.

Keywords: Down syndrome; congenital heart defects; copy number variation.

Figure 1

Summary results from analyses to attempt to replicate previously reported DS+AVSD (Down Syndrome with a complete atrioventricular septal defect)-associated common copy number variants (CNVs) with three technologies. (A and B) shows measures for Sailani-putative CNVs 1 and 2, respectively. Boxplots show inner quartile range of log₂ ratios for tested DS samples with whiskers reaching 1.5 times the interquartile range. Case values are circles, and control values are in outlined boxes. Array comparative genomic hybridization (aCGH) probes are blue, NanoString probes are red, and TaqMan probes are green. Boxplots are ordered by genomic location, and their precise locations are indicated beneath the plots. Neither CGH probes nor TaqMan Copy Number assays detected aberrant copy numbers or differences between cases and controls. Varying results were found across these loci by NanoString probes, with some probes showing differences in log₂ means between cases and controls. Within these same proposed small CNV loci, NanoString probes called all possible combinations of copy gain, loss, and no change within the same small cohort. When compared to adjacent CGH and TaqMan probes, it is clear that the NanoString probes are not reliable predictors of copy number state at this locus.