Rare coding mutations cause ∼45% of congenital heart disease (CHD). Noncoding mutations that perturb cis-regulatory elements (CREs) likely contribute to the remaining cases, but their identification has been problematic. Using a lentiviral massively parallel reporter assay (lentiMPRA) in human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs), we functionally evaluated 6,590 noncoding de novo variants (ncDNVs) prioritized from the whole-genome sequencing of 750 CHD trios. A total of 403 ncDNVs substantially affected cardiac CRE activity. A majority increased enhancer activity, often at regions with undetectable reference sequence activity. Of ten DNVs tested by introduction into their native genomic context, four altered the expression of neighboring genes and iPSC-CM transcriptional state. To prioritize future DNVs for functional testing, we used the MPRA data to develop a regression model, EpiCard. Analysis of an independent CHD cohort by EpiCard found enrichment of DNVs. Together, we developed a scalable system to measure the effect of ncDNVs on CRE activity and deployed it to systematically assess the contribution of ncDNVs to CHD.

约45% 的先天性心脏病 (CHD)是由罕见的编码突变引起的。扰乱顺式调节元件（CRE）的非编码突变可能导致了其余病例，但它们的识别一直存在问题。我们在人诱导多能干细胞来源的心肌细胞 (iPSC-CM) 中使用慢病毒大规模平行报告基因检测 (lentiMPRA)，对 750 个 CHD 三人组的全基因组测序优先排序的 6,590 个非编码从头变异 (ncDNV) 进行了功能评估。总共 403 个 ncDNV 对心脏 CRE 活性有显着影响。大多数增强子活性增加，通常在参考序列活性不可检测的区域。在通过引入其天然基因组背景进行测试的 10 个 DNV 中，有 4 个改变了邻近基因的表达和 iPSC-CM 转录状态。为了优先考虑未来 DNV 的功能测试，我们使用 MPRA 数据开发了回归模型 EpiCard。EpiCard 对独立 CHD 队列的分析发现 DNV 富集。我们共同开发了一个可扩展的系统来衡量 ncDNV 对 CRE 活动的影响，并部署该系统来系统评估 ncDNV 对 CHD 的影响。