De novo deletions and duplications of 17q25.3 cause susceptibility to cardiovascular malformations

F. J. Probst, R. A. James, L. C. Burrage, J. A. Rosenfeld, T. P. Bohan, C. H. Ward Melver, P. Magoulas, E. Austin, A. I. A. Franklin, M. Azamian, F. Xia, A. Patel, W. Bi, C. Bacino, J.W. Belmont, S. M. Ware, C. Shaw, S.W. Cheung, S. R. Lalani

Genomic disorders resulting from deletion or duplication of genomic segments are known to be an important cause of cardiovascular malformations (CVMs). In our previous study, we identified a unique individual with a de novo 17q25.3 deletion from a study of 714 cases with CVM. To understand the contribution of this locus to cardiac malformations, we reviewed the data on 60,000 samples submitted for array comparative genomic hybridization (CGH) studies to Medical Genetics Laboratories at Baylor College of Medicine, and ascertained seven individuals with segmental aneusomy of 17q25. We validated our findings by studying another individual with a de novo submicroscopic deletion of this region from Cytogenetics Laboratory at Cincinnati Children's Hospital. Using bioinformatic analyses including protein-protein interaction network, human tissue expression patterns, haploinsufficiency scores, and other annotation systems, including a training set of 251 genes known to be linked to human cardiac disease, we constructed a pathogenicity score for cardiac phenotype for each of the 57 genes within the terminal 2.0 Mb of 17q25.3. We found relatively high penetrance of cardiovascular defects (~60 %) with five deletions and three duplications, observed in eight unrelated individuals. Distinct cardiac phenotypes were present in four of these subjects with non-recurrent de novo deletions (range 0.08 Mb-1.4 Mb) in the subtelomeric region of 17q25.3. These included coarctation of the aorta (CoA), total anomalous pulmonary venous return (TAPVR), ventricular septal defect (VSD) and atrial septal defect (ASD). Amongst the three individuals with variable size duplications of this region, one had patent ductus arteriosus (PDA) at 8 months of age. The distinct cardiac lesions observed in the affected patients and the bioinformatics analyses suggest that multiple genes may be plausible drivers of the cardiac phenotype within this gene-rich critical interval of 17q25.3.