This Title All WIREs
How to cite this WIREs title:
WIREs Comput Mol Sci
Impact Factor: 14.016

3D modeling of chromatin structure: is there a way to integrate and reconcile single cell and population experimental data?

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

The genome is organized in a hierarchical fashion within the nucleus in interphase. This nonrandom folding of the chromatin fiber is thought to play important roles in the processing of the genetic information. Therefore, a better knowledge of the mechanisms underlying the three‐dimensional structure of the genome appears essential to fully understand the nuclear processes including transcription and replication. Fluorescent in situ hybridization (FISH) and molecular biology methods deriving from the Chromosome Conformation Capture technique are the methods of choice to study genome 3D organization at different levels. Although these single cell and population methods allowed to highlight similar chromatin structures, they also show frequent discrepancies which might be better understood by improving the capacity to generate actual 3D models of organization based on the different types of data available. This review aims at giving an overview of the principles, advantages, and limits of microscopy and molecular biology methods of analysis of genome structure and at discussing the different approaches of modeling of chromatin classically used and the improvements that are necessary to reach a better understanding on the links between chromatin structure and its spatial organization. WIREs Comput Mol Sci 2017, 7:e1308. doi: 10.1002/wcms.1308

Use of FISH to study the spatial organization of the genome in single cell. Fluorescent in situ hybridization (FISH) can be performed on fixed cells to analyze the different parameters of the spatial organization of chromosomes or genomic domains. For example, whole chromosome painting can be used to study the position of chromosome territories relative to nuclear structures as well as their radial positioning from nuclear center to periphery (dashed lines—in this example, human chromosome 2 was labeled in red). Multiple specific genomic regions can be labeled in parallel to measure the distances that separate them from each other (double arrows—in this example, two different loci of human chromosome 2 were labeled in green).
[ Normal View | Magnified View ]
Chromatin model representation. (a) Different ways of representing modeled chromatin loci: [1] as dots, with no physical consistency, [2] as spheres, [3] as spheres with complex physical properties, for example, a soft external surface and a hard core, and [4] as sphere of different sizes (either representing different content in nucleotides, or different levels of compaction). (b) Schematic representation of the application of the three main physical restraints, the excluded volume (U Excl), the bond restraint (U Bond) here represented either as a spring that could be modeled by a Lennard–Jones potential or as a harmonic restraint, the bending (U Bend) that could in this example be modeled explicitly imposing an angle between three particles or by limiting the distance between two nonconsequtive particles (i, i + 2) using a right‐truncated harmonic. (c) Different kind of potentials to be applied between particles i and j. Dij being the Euclidian distance between them, rij the optimal distance inferred from, for example, the experimental data.
[ Normal View | Magnified View ]
Modeling strategies. Six main different modeling strategies are currently available for modeling genomes and genomic domains. Those methods diverse in their representation, scoring and sampling of the data. [1], [2], [3], [4], [5], and [6]. Refer to the text for detailed explanation of each strategy.
[ Normal View | Magnified View ]
Interpretation of 3C‐based results as compared to single cell FISH. Classical 3C‐based data do not allow to determine whether co‐localization of genomic element occur within a single cell or if they represent the sum of distinct organization co‐existing in the population. (a) i, ii, and iii correspond to different models of organization of two loci (green and red) which would give relatively similar results in 3C‐derived datasets: in i, the red and green loci are co‐localized in every cell of the population whereas the two other loci are located further away from each other. In ii, two subpopulations of cells are depicted, one presenting the co‐localization of the two loci (green and red), the other without any colocalization. In iii, the green and red loci do not colocalize but their movement are restrained in a limited area which favor their random collision and increase the probability of cross‐link. (b) Classically 3C‐derived data give pair‐wise information which do not permit to decipher whether three loci (blue, red, and green) colocalize in a same cell at the same time (population i) or whether two of them are interacting in different cells or at different time (population ii).
[ Normal View | Magnified View ]
Hi‐C allows to study genome conformation at different scales and demonstrates a hierarchical organization. Binning the results of Hi‐C experiment at different scale (e.g., whole chromosome, 1 Megabase, 100 kb windows or lower) permits to identify various levels of organization of the genome: chromosomes are organized as territories which occupy preferential relative positions. Chromatin is segregated into two compartments depending on its active or inactive state and chromosomes are segmented in topologically associating domains (TADs).
[ Normal View | Magnified View ]
Principle of Chromosome Conformation Capture and derivative methods. (a) 3C‐based methods are based on the cross‐link of chromatin with formaldehyde. The probability of cross‐link between two regions is dependent on the spatial distance that separates them. In the case of a randomly organized molecule (top panel), the frequency of cross‐linking will decrease as a function of the genomic distance that separate two loci whereas this relationship is broken in the case two region are establishing preferential contacts (bottom panel). (b) 3C‐based methods use the frequency of ligation of cross‐linked DNA after digestion by restriction enzyme (RE) to recover this information in a quantitative output. If the original 3C methods uses specific primers to detect one‐to‐one those potential interactions, further derivative allows to analyze the interactions at higher levels, either one‐to‐many/all (4C), many‐to‐many (5C), or all‐to‐all (Hi‐C) depending on the detection method used.
[ Normal View | Magnified View ]

Browse by Topic

Software > Molecular Modeling
Structure and Mechanism > Molecular Structures

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts