Data-driven genome compartmentalization with nuclear landmarks.

Abstract

In eukaryotic cells, the three-dimensional (3D) organization of genomes plays a crucial role in genome function and has inspired the development of numerous innovative experimental techniques for its characterization. Although researchers have made a great deal of significant progress in deciphering the folding mechanisms of an individual chromosome, the principles of the dynamic large-scale spatial arrangement of the whole genome inside the nucleus are poorly understood. Following the maximum entropy principle pioneered by Zhang and Wolynes, we used polymer simulations to develop a predictive model for the whole genome at one hundred kilo-base resolution (100 KB) with nuclear bodies such as nuclear lamina, nucleoli, and speckles and parameterized a force field to study genome structure and dynamics using genome-wide chromosome conformation capture data. We discovered that self-organization based on a co-phased separation between chromosomes and nuclear bodies process can capture various features of genome organization, including the formation of chromosome territories, and phase separation of A/B compartments. Both sequencing-based genomic mapping and imaging assays that probe chromatin interaction with nuclear bodies are quantitatively reproduced with the simulated 3D structures. Essentially, we captured the heterogeneous distribution of chromosome positioning across cells, while simultaneously producing well-defined distances between active chromatin and nuclear speckles with our 100 KB model. Such heterogeneity and preciseness of genome organization can coexist due to the non-specificity of phase separation and the slow chromosome dynamics. In addition, with our 100 KB model, we analyzed the rheology of the cell nucleus noninvasively using the storage and loss moduli and investigated the different time scales existing in the model.