Poster #47: Pooja Sonawane MS

Characterizing Regulatory Elements and Noncoding Variants Associated with Structural Birth Defects Using Lentiviral Based Massively Parallel Reporter Assays

Pooja Sonawane1,2, Anthony M. Gacita3,4, Mian Umair Ahsan5, Kai Wang5,6, Justin Cotney2,7 1 University of Connecticut Health Center, Farmington, CT 2 Department of Surgery, Children’s Hospital of Philadelphia, Philadelphia, PA 3 Division of Human Genetics, Children's Hospital of Philadelphia, Philadelphia, PA 4 Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 5 Raymond G. Perelman Center for Cellular and Molecular Therapeutics, Children's Hospital of Philadelphia, Philadelphia, PA 6 Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA 7 Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA

Congenital heart defects and craniofacial abnormalities are among the most common structural birth defects. Although genetics contribute substantially to disease risk, more than half of cases remain unexplained. One hypothesis is that noncoding variants within gene regulatory elements contribute to disease risk by disrupting developmental gene expression. While genome sequencing studies have identified numerous noncoding variants associated with structural birth defects, their functional consequences remain unknown. To address this challenge, we previously performed epigenomic profiling of primary human embryonic heart and craniofacial tissues, identifying thousands of putative developmental enhancers. Here, we used lentiviral based massively parallel reporter assays (LentiMPRA) to test candidate regulatory elements (CREs) and variants. We designed two libraries containing thousands of CREs and variants around developmental genes from either craniofacial or cardiac datasets. These libraries were delivered to stem cells and tissue relevant cell types, including cranial neural crest cells and cardiomyocytes. Across five cell types, thousands of elements exhibited regulatory activity, revealing widespread enhancer activity and cell type specificity. Reporter activity was generally consistent with ChromHMM annotations, supporting epigenomic predictions. Multiple disease associated variants significantly altered activity, implicating regulatory disruption as a mechanism underlying congenital disease. Deep learning models trained on these data showed positive correlation between predicted and observed enhancer activity, supporting their utility for deciphering cis-regulatory code. Together, these studies establish LentiMPRA as a scalable framework for functional characterization of developmental CREs and noncoding variants associated with structural birth defects.