1Universitätsklinikum Heidelberg Innere Medizin VIII, Institut für Experimentelle Kardiologie Heidelberg, Deutschland; 2University of Freiburg Department of Computer Science Freiburg, Deutschland; 3Institute Insure, German Heart Center Munich, Technical University of Munich, School of Medicine & Health Department of Cardiovascular Surgery Munich, Deutschland; 4Yale University School of Medicine Division of Cardiac Surgery New Haven, CT, USA; 5Universitätsklinikum Würzburg Institut für Pharmakologie und Toxikologie Würzburg, Deutschland
Introduction:
Enhancers are distal non-coding genetic regulatory elements that influence target gene expression and are associated with heart disease. Our research aims to decode and modulate disease-associated enhancers in human cardiomyocytes (CM).
Methods:
We integrated high-resolution data including deep spatial chromatin interactions, DNA methylation, and histone modifications of FACS-sorted CM nuclei isolated from atrial (NF-LA-CM), left ventricular (NF-LV-CM), and failing left ventricular (F-LV-CM, EF < 20%) human tissues from age-matched patients. We decoded the functional relevance of enhancer elements using CRISPRi in induced pluripotent stem cell-derived cardiac myocytes (iPSC-CM).
Results:
Our deep Hi-C data (n=6-8, 11.2 billion reads) identified in average three interaction domains per gene. We then performed differential interaction analysis and found 4,245 differential interactions between NF-LV-CM and F-LV-CM, including a gain in promoter-enhancer interactions of NPPA, correlated to NPPA upregulation in failing CM. However, the derangement of chromatin architecture in failing CM was not tightly linked to gene expression on a genome-wide scale. In contrast, chamber-specific promoter interactions were tightly linked to gene expression between atrial and ventricular CM. These differential interactions contain cis-regulatory elements (CREs) associated with genetic risk factors of heart diseases.
Genetic variants linked to atrial fibrillation were enriched within atrial-specific promoter-interacting CREs (2.9 fold, P-value = 2.5e -09). Conversely, disease variants predominantly affecting the ventricles are abundant within ventricular-specific promoter-interacting CREs, including variants associated with QT-interval alterations (3.2 fold and P-value = 4.1e -13) . To test the functional relevance of QT-associated enhancers, we perturbed eight KCNJ2 enhancers using CRISPRi and identified two QT-syndrome-associated enhancers mediating 70% of KCNJ2 expression.
Conclusions:
Chamber-specific chromatin organization was linked to chamber-specific gene expression programs. In contrast, heart disease-associated rewiring was generally decoupled from pathological gene expression programs. Furthermore, our data connects disease-associated variants with their target promoters in a chamber-specific manner. This knowledge will facilitate the interpretation of known and novel heart disease-associated variants. CRISPRi perturbation experiments revealed the functional relevance of the identified disease-associated chromatin interactions.