Abstract Body (Do not enter title and authors here):
Introduction/Background: Congenital heart defects (CHD) are the most common form of developmental abnormalities, occurring in ~1% of live births, and can arise due to altered dosage of genes essential for cardiogenesis. Aneuploidy accounts for nearly 15% of CHD and the most frequent form involves trisomy of chromosome 21 (Ch21), resulting in Down Syndrome (DS). CHD is present in ~50% of DS cases, with a 1000-fold enrichment of atrioventricular canal (AVC) defects that disrupt the junction of the atria and ventricles.

Hypothesis: We hypothesize that the presence of a third copy of one or more genes on Ch21 underlies AVC and other cardiac defects; however, the dosage-sensitive CHD gene(s) on Ch21 causing this are not fully defined.

Methods: We used human pluripotent stem cells derived from an individual mosaic for DS, allowing us to compare Ch21 trisomy cells with isogenic control cells disomic for Ch21. We performed single cell RNA-sequencing (scRNA-seq) to define the transcriptional dysregulation associated with DS in atrioventricular canal myocardial cells (AVCM). We then introduced a CRISPR-activation transgene to disomic cells and delivered small guide RNAs (sgRNAs) against 66 candidate Ch21 CHD genes. We developed a new machine learning algorithm to assign a trisomic score to CRISPR perturbed (CROP-seq) cells. Finally, we reduced the copy number of our top candidate gene in a mouse model of Down Syndrome and used microCT to assess incidence of cardiac septal defects.

Results: Using human pluripotent stem cell and mouse models of DS, we identify HMGN1, a nucleosome-binding epigenetic factor on Ch21, as a dosage-sensitive regulator of cardiac defects in DS. Single cell transcriptomics revealed that human AVC cardiomyocytes shifted to a more ventricular myocardial state in trisomy 21 hiPS-derived cells. CRISPR-activation in isogenic disomic cells of 66 candidate Ch21 genes expressed in heart development, followed by single cell RNA-sequencing (CROP-seq) combined with machine learning predictions indicated that increased expression levels of HMGN1 altered the transcriptional state of atrioventricular cardiomyocytes similar to trisomy 21. Finally, reduction of Hmgn1 from three to two alleles in a mouse model of DS attenuated the incidence of cardiac septal defects.

Conclusion: Our results identify HMGN1 as a dosage sensitive regulator of heart defects in DS and represent a generalizable approach for identifying candidate genes within areas of aneuploidy.