Dynamic structural models of the 4D genome

The 3-dimensional (3D) organization of chromosomes provides a blueprint for the regulation of expression of our genes (the genome) that together make up the various cell types in our body. In this project, we have studied 1) how chromosomes are organized in 3D, 2) how this organization changes during stem cell differentiation, and 3) how cell metabolic status influences 3D chromosome structure and gene expression. Our work has been both computational, producing 3D models of the genome which enable the formulation of hypotheses that can be tested in the laboratory, and experimental, validating (or not) hypotheses from the models and leading to discoveries on how the metabolic status of a cell influences 3D genome organization over time (the 4D genome, where the 4th dimension is time). We have published a user-friendly software (Chrom3D) to model chromosomes in 3D from experimental data; these data consist of large maps of interactions between chromosomes in the 3D nucleus space, and of interactions between chromosomes and the nuclear envelope, at the periphery of the nucleus. This program allows us to make testable predictions on conformation changes of chromosomes during stem cell differentiation, under various metabolic conditions and in metabolic disease contexts. We have used our 3D modeling tool to predict, and experimentally demonstrate, common and cell type-specific changes in 3D chromosome conformation during stem cell differentiation. This has led to the identification of key features of chromatin, such as open and closed regions, which orchestrate active and repressive domains of the genome. Since, new polymer models of chromosomes give insights on how physical properties of chromosomes affect their interaction patterns and dynamics with the nuclear envelope, and hence spatial genome conformation. In parallel, we have shown that periodic changes in chromatin organization in the nucleus in liver, follow a daily (24 h) rhythm, suggesting a cell-autonomous regulation of genome organization, in addition to what could be metabolically regulated. In human fat stem cells, metabolic stimulation further triggers switches in the levels and form of a specific RNA molecule, which we found works as regulator of protein synthesis and as a determinant of the lipid storage capacity of adipocytes. This RNA could constitute a potent new regulator of fat distribution between upper and lower body, and thereby of metabolic health (lower body fat is healthy, upper body fat is not). Lastly, in international collaborations, we have characterized epigenetic patterns in stem cells from human bone marrow and fat tissue to reveal tissue-specific immune properties of these cells. We have also identified a protein important for epigenetic regulation of genome integrity, which is inactivated in certain types of cancers. Throughout the project, we have disseminated our findings at national and international conferences, and in many publications. In summary, our work has provided new perspectives on how 3D genome organization (with a focus on fat cells) is influenced by cellular metabolic states, how it is important for proper regulation of cell function, and how it is deregulated in metabolic disease contexts.

This was a fundamental science project on the spatial reconfiguration of the genome during stem cell differentiation. This has led to publications which have pioneered new appreciations and debates of the spatial dynamic of chromosomes in the field of nuclear architecture, notably at conferences and on social media. So effects and impact of our results have been very significant at the level of scientific knowledge.

Cell fate decisions are programmed at several levels of regulation of gene expression. One such level involves so-called epigenetic modifications of chromatin, an assembly of DNA and histones. Dynamic interactions between chromosomes in the 3-dimensional (3D) space of the cell nucleus also affect gene expression. This suggests that 3D genome conformation is regulated during development, that is, in a 4D space where the fourth component is time. Nutrients also seem to modulate epigenetic states, adding complexity to the regulation of gene expression. We aim to establish a mechanistic connection between 3D genome conformation, gene expression patterns and cellular metabolic state. We will test three hypotheses. The first is that lineage-specific differentiation shapes developmental transitions in the 3D chromatin landscape. This will be tested by generating novel 3D chromatin models from chromosome-chromosome and chromosome-nuclear envelope interaction data and modeling their transitions during adipogenic and neurogenic differentiation. The second is that 4D chromatin conformation predicts gene expression patterns. We will use Bayesian techniques to learn which aspects of 3D chromatin conformation or changes therein predict gene expression changes during differentiation. The third is that the 4D genome is influenced by cellular metabolism via a nutrient-sensing histone modification. This will be tested by subjecting cells to glucose stress, and monitoring how the 3D genome is affected and whether changes are mediated by H2B GlcNAcylation, a glucose-sensing modification which pre-patterns the association of nuclear lamins with chromatin (and hence the radial positioning of loci). This transdisciplinary project combines stem cell biology, chromatin biology and metabolic studies with high-throughput genomics, statistical modeling and software development. It is designed for 1 postdoc and 2 PhD students with technical assistance.