Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting over 33 million individuals throughout the world. AF is highly heritable and recent genome-wide association studies (GWAS) have cumulatively identified over 100 loci associated with AF risk. Genome-wide association studies (GWAS) often identify transcription factor (TF) loci in association with complex human diseases, implying that a significant transcriptional component underlies human disease risk and etiology. The transcription factors ZFPM2 (FOG2), GATA4, and TBX5 have all been implicated in human AF risk by genetic studies. We hypothesized that FOG2, GATA4, and TBX5 functionally interact to regulate a gene regulatory network essential for atrial rhythm control. We generated a novel mouse model of spontaneous AF based on FOG2 overexpression. FOG2 ChIP-seq identified FOG2 genomic localization at loci co-occupied by GATA4, a known FOG2 binding partner. However, we found that FOG2 OE caused gene expression alterations that correlated more highly with TBX5-dependent rather than GATA4-dependent gene expression, including a module of calcium handling genes required for atrial rhythm homeostasis. We applied TF-dependent non-coding transcriptional profiling to examine the FOG2 dependent atrial GRN, which identified 805 candidate regulatory regions with accessible chromatin and FOG2 dependent ncRNAs. TBX5 removal and FOG2 OE caused highly correlated dysregulation of ncRNA expression at open chromatin regions genome-wide, suggesting a functional interaction between TBX5 and FOG2. Furthermore, FOG2 OE only affected enhancer activity by altered ncRNA abundance at locations of TBX5 co-binding. The shared TBX5/FOG2 genomic interaction predicted a potential genetic interaction, and we found that cardiac rhythm abnormalities caused by
Tbx5
haploinsufficiency were rescued by
Fog2
haploinsufficiency. Taken together, TF-dependent ncRNA-profiling revealed an interconnected cardiac rhythm gene regulatory network (GRN) between FOG2, TBX5 and GATA4. These data nominate a specific model in which FOG2 is recruited by GATA4 to modulate a co-bound TBX5-dependent atrial gene regulatory network for calcium handling and atrial rhythm homeostasis.
Is Subcellular Localization Informative for Modeling Protein-Protein Interaction Signal?
