Identifying genetic modifiers of congenital heart disease and mental illness in 22q11.2 deletion syndrome

Abstract

The 22g11.2 deletion syndrome (22q11DS) is the most common microdeletion syndrome occurring in 1:4000 live births. It is caused by a 3 Mb deletion on chromosome 22g11.2 and presents with a highly variable phenotype expressivity. Mouse and human genetic studies have identified haploinsufficiency of Tbxl as the strongest genetic factor causing congenital heart disease (CHD) and brain and behavior (BB) disorders in 22q11DS; which are observed in 65% and 40% of 22g11DS patients, respectively. However, genetic causes of variable expressivity are not well described. We hypothesize that rare coding and non-coding variants can modify the cardiac or brain and behavior phenotypes in 22q11DS. To test this hypothesis, we performed two integrative genomic studies using high throughput sequencing (HTS) data in human subjects with 22q11DS.;I first performed a personal genomics analysis on two unrelated 22q11DS trios where the probands had contrasting clinical findings and parents were unaffected. Proband P1 had tetralogy of Fallot (TOF), psychotic episodes, and other BB phenotypes; proband P2 had juvenile rheumatoid arthritis and short stature but no other major clinical findings. In P 1, I identified an enrichment of predicted deleterious rare variants in genes related to his cardiac and BB phenotypes including a de novo SNV in ADNP2, a gene previously implicated in schizophrenia. In P2, I identified a predicted deleterious SNV in ZFPM2, a known TOF causative gene, which may act as a protective variant downstream of TBX1.;Next, I performed a candidate gene analysis using HTS data from individuals with 22g11DS (138 cases with CHD, 141 controls) and searched for high impact rare predicted deleterious single nucleotide variants (rdSNVs) in the TBX1 cardiac gene interaction network, consisting of 79 genes. GO enrichment analysis found that case-unique and control-unique genes were enriched for anterior-posterior patterning and cardiac progenitor cell proliferation terms, respectively. This suggests that genes affected by rdSNVs in cases compared to controls affect different biological processes. Overall, 9.3% of the cohort had an rdSNV in a TBX1 gene network gene, suggesting other genes may act as modifiers of CHD in the syndrome.