Dynamical Nucleus Architecture : a Polymer physics based approach to Chromosome dynamics imaging – DNA-PolyChrom

In the DNA-PolyChrom project, two sets of imaging experiments will be performed. In a first set, we will monitor ~50 individual non-functional foci that are randomly inserted in the Drosophila genome. They are spread over the different chromosomal arms with on average 10 foci per arm. A second set of experiments consists of ten functional foci that probe the transcriptional activity of ten well-characterized genes. Some of the functional loci shall be paired with one or several non-functional loci close by, at distances ranging from 25 kb to 3 Mb, including a gene expressed in larval cells. This set will be used to directly monitor enhancer-promoter interactions and how the physical parameters describing these interactions scale with distance. In addition, we will combine these two sets of tagged loci to explore the dynamics of pair distances in different genomic regions and for different states of genetic activity. In parallel, we will tag these DNA loci with oligopaints and use super-resolution imaging of the same domains in fixed cells to cross-validate our results and our interpretation. This technique allows to measure distances below 10kb.
These measurements will test our hypotheses concerning the relationship between chromatin folding, loci position distributions, inter-loci distance distributions, and the dynamics of chromatin and transcription: First, the dependence of single loci statistics (position distribution) and dynamics (MSD) on the local chromatin folding state will be set, by comparison with theoretical and numerical modeling. Once these single-loci features established, we will be able to predict their consequences on the pairwise inter-distances behavior. The latter will be measured in experiments and simulated numerically, which will allow us to end with a consistent model of the interplay between folding state and inter-loci dynamics.

We already have achieved important improvement of the numerical and theoretical modeling of polymers.
First, we performed extensive Monte Carlo simulations of polymer domains, enabling us to extract radius of gyration and end-to-end distance statistics, and introduced refinement in the Bayesian inference of the model parameters.

In parallel, we have discovered interesting new scaling relationships that shall lead to a completely new formulation of the polymer free energy allowing us to account at a time for the length and energy dependences.

A theoretical derivation of end-to-end distance distribution for polymer domains of different types has also been finalized, and will be compared to simulations and experiments.
Besides, we developed a new approach based on spectral analysis of the polymer which allows to deduce in a direct and very efficient way its 3D conformation from a very limited number of its monomers. By applying this approach we will be able to dispense with oligopaint imaging of domain conformations and deduce them directly from the dynamics of loci.

On the experimental side, we developed a quantitative imaging technology that allows us to measure physical distances in a small gene locus to show that the transcriptionally active enhancer–promoter pair is not in physical contact, questioning well-established rules of eukaryotic transcriptional regulation. The transcription field is currently undergoing a sea change as there is growing evidence about the three-dimensional structure of chromatin and its orderly packing in the cell nucleus seemingly having an effect on the transcriptional status of a gene. However it is unclear how close distal enhancers have to get to a promoter in order for transcription to ensue. We address these issues, using the early Drosophila embryo as a laboratory to measure physical distances between cis-regulatory elements of an 18 kb gene locus under different topological and functional conditions.

The new formulation of the polymer free energy may have a great interest from a theoretical point of view. Indeed, we have empirically determined new scaling variables, allowing for a universal representation of all polymers sharing the same folding properties, e.g. the same size distribution. Such new formalism considerably decreases the complexity of the model, allowing (1) a much easier inference of its parameters from simulation data, (2) a more convenient mapping of real systems to a polymer representation.

The new approach allowing to assess the polymer configuration (coil, globule) by collecting the relative positions of only three discernible, evenly spaced loci potentially yield a new, very sparing experimental way to assess the epigenetic state, easily performed with modern fluorescent imaging techniques in vivo. This has a particular impact on our project as it avoids the setting up of complex super resolution experiments.

Assessing the polymer coil-globule state from the very first spectral modes, Timothy Földes, Antony Lesage, Maria Barbi
doi: doi.org/10.1101/2021.07.17.452647

Transcription-dependent spatial organization of a gene locus. Barinov L. et al arXiv:2012.15819 [q-bio.MN]. 2020 December.

Using RNA Tags for Multicolor Live Imaging of Chromatin Loci and Transcription in Drosophila Embryos. Chen H, Gregor T. Methods Mol Biol. 2020;2166:373-384. doi: 10.1007/978-1-0716-0712-1_22. PubMed PMID: 32710421.

Action at a distance in transcriptional regulation. Bialek W, Gregor T, Tkacik G. arXiv:1912.08579 [q-bio.SC]. 2019 December.

Chromatin, the plastic DNA-protein assembly that fills the nucleus, is organized in multi-scale compartments. By quantifying the number of interactions between genomic loci that are nearby in 3D space, HiC technology has revealed that one important level of organization is, at the 10 kb to Mb scale, the ?Topologically Associating Domain (TAD), supposed to make different genome segments occupy separate regions such that their individual activities are non-interfering. 3D fluorescent in situ hybridization and super-resolution microscopy have indeed revealed more or less separated compartments and characterized their spatial organization by measuring their spatial variance (or gyration radius). The scaling laws describing how gyration radii increase with the genomic length of the domains in Drosophila show a clear difference for the three different activity states: active (transcribing), inactive (non-transcribing) or repressed (actively silenced transcription).
Functional chromatin domains can also be defined based on the presence of biochemical markers called epigenetic marks. Epigenetic coloring is specific to different gene activity states: in Drosophila, TADs, activity domains and epigenetic domains practically coincide, with three epigenetic colors (“red”, “black” and “blue”) corresponding to the three aforementioned activity states, respectively. Epigenetics is responsible for the temporal and spatial control of gene activity during cell differentiation, and we suggest that the 3D arrangement may be the mechanistic way how epigenetics controls gene activity.
Dynamical features can also be investigated. In vivo tracking of chromosome fluctuations allows to quantify single loci diffusion and reveals multiple dynamical, low-frequency interactions. Partner 3 experimentally showed that sustained physical proximity between enhancer and promoter is necessary for productive transcription to occur, with a longer and closer pairing in transcribing regions. However, no clear results have been obtained allowing to connect loci diffusion, inter-loci distance fluctuations or inter-loci contact frequencies with the underlying folding state of chromatin. Our project aims at filling this gap. It will provide a novel perspective on epigenetically driven regulation of genetic networks. We will thus provide a theory that describes domains in terms of a physical state, that, if close to criticality, would explain the origin of phase transitions in the nuclear environment.
More precisely, we will address the question whether sustained enhancer-promoter proximity is the signature of an optimized dynamics where either the rapidity, encounter frequency, or selectivity of any generic inter-loci encounter is maximized. How do these dynamical features depend on the 3D organization of active, inactive, repressed domains? ?To answer, we will identify a set of epigenetic domains and study the distribution and temporal fluctuations of single and inter-loci distances in living Drosophila by multi-color imaging. Independent information on the folding state of an epigenetic domain will be obtained by super-resolution.
Linking epigenetics, domain folding and dynamics will require the development of efficient data analysis methods and of a coherent theoretical model. Partners 1 and 2 have already laid out a foundation for such a model by developing (i) a scheme to describe single loci diffusion in the framework of polymer dynamics, that brought the concept of Rouse with transient internal contacts, and (ii) a new methodology to extract the best information from super-resolution data, showing that the measured gyration radii distributions are compatible with the behavior of a self-attracting polymer, close to the coil-globule crossover. This points towards a crucial role of criticality to enhance the system responsivity, with possible interesting consequences on dynamics that will be explored in this project.