Spatial Principles of Chromatin Architecture Associated With Organ-Specific Gene Regulation

Packaging of the genome in the nucleus is a non-random process that is thought to directly contribute to cell type-specific transcriptomes, although this hypothesis remains untested. Epigenome architecture, as assayed by chromatin conformation capture techniques, such as Hi-C, has recently been described in the mammalian cardiac myocyte and found to be remodeled in the setting of heart failure. In the present study, we sought to determine whether the structural features of the epigenome are conserved between different cell types by investigating Hi-C and RNA-seq data from heart and liver. Investigation of genes with enriched expression in heart or liver revealed nuanced interaction paradigms between organs: first, the log(2) ratios of heart:liver (or liver:heart) intrachromosomal interactions are higher in organ-specific gene sets (p = 0.009), suggesting that organ-specific genes have specialized chromatin structural features. Despite similar number of total interactions between cell types, intrachromosomal interaction profiles in heart but not liver demonstrate that genes forming promoter-to-transcription-end-site loops in the cardiac nucleus tend to be involved in cardiac-related pathways. The same analysis revealed an analogous organ-specific interaction profile for liver-specific loop genes. Investigation of A/B compartmentalization (marker of chromatin accessibility) revealed that in the heart, 66.7% of cardiac-specific genes are in compartment A, while 66.1% of liver-specific genes are found in compartment B, suggesting that there exists a cardiac chromatin topology that allows for expression of cardiac genes. Analyses of interchromosomal interactions revealed a relationship between interchromosomal interaction count and organ-specific gene localization (p = 2.2 × 10(−16)) and that, for both organs, regions of active or inactive chromatin tend to segregate in 3D space (i.e., active with active, inactive with inactive). 3D models of topologically associating domains (TADs) suggest that TADs tend to interact with regions of similar compartmentalization across chromosomes, revealing trans structural interactions contributing to genomic compartmentalization at distinct structural scales. These models reveal discordant nuclear compaction strategies, with heart packaging compartment A genes preferentially toward the center of the nucleus and liver exhibiting preferential arrangement toward the periphery. Taken together, our data suggest that intra- and interchromosomal chromatin architecture plays a role in orchestrating tissue-specific gene expression.