Spatial Genome Organization: From Development to Disease

Abstract

Every living organism, from bacteria to humans, contains DNA encoding anything from a few hundred genes in intracellular parasites such as Mycoplasma, up to several tens of thousands in many higher organisms. The first observations indicating that the nucleus had some kind of organization were made over a hundred years ago. Understanding of its significance is both limited and aided by the development of techniques, in particular electron microscopy, fluorescence in situ hybridization, DamID and most recently HiC. As our knowledge about genome organization grows, it becomes apparent that the mechanisms are conserved in evolution, even if the individual players may vary. These mechanisms involve DNA binding proteins such as histones, and a number of architectural proteins, some of which are very much conserved, with some others having diversified and multiplied, acquiring specific regulatory functions. In this review we will look at the principles of genome organization in a hierarchical manner, from DNA packaging to higher order genome associations such as TADs, and the significance of radial positioning of genomic loci. We will then elaborate on the dynamics of genome organization during development, and how genome architecture plays an important role in cell fate determination. Finally, we will discuss how misregulation can be a factor in human disease.

Keywords: CTCF; LAD; TAD; cohesin; development; genome organization.

FIGURE 1

Overview of genome organization in the eukaryotic nucleus. Chromosomes are organized in discrete territories within the nucleus. Specific genomic loci called Lamina Associated Domains (LADs), detected by DamID, interact with the nuclear envelope and are typically repressed upon direct tethering. Topologically Associated Domains (TADs), defined by HiC, are units of the genome that frequently associate with each other. Higher order associations of TADs form A and B compartments which are typically enriched in transcriptionally active and inactive chromatin, respectively. Local chromatin loops are stabilized by architectural proteins such as CTCF and cohesin complex.

FIGURE 2

Functional basis for relocation of genes during differentiation. (A) Pluripotency factor Zfp42 is found in the nuclear interior in mouse embryonic stem cells (mESCs) and is later sequestered to the periphery upon neuronal differentiation. Pcdh9, an important component of neuronal junctions in the brain, is tethered and therefore repressed during pluripotency and is later released, concurrent with its activation in astrocytes. Illustration is based on data from Peric-Hupkes et al., 2010. (B) Fluorescence in situ hybridization images showing the relocation of Ttn locus (green) encoding Titin, an important protein for muscle differentiation, from the nuclear periphery to the interior upon differentiation of myoblasts (MBs) to myotubes (MTs). Data reproduced from Robson et al., 2016 according to the Creative Commons Attribution License (CC BY) (https://creativecommons.org/licenses/by/4.0/).

FIGURE 3

Structural variation can lead to disruption in chromatin conformation and disease. Pitx1 specifies hindlimb identity in mouse. (A) 3D chromatin conformation in mouse forelimb prevents interaction of Pitx1 with its enhancer Pen, leading to preferential expression of Pitx1 in the hindlimb. (B) Introducing an inversion in the locus containing Pen results in a structural variation that leads to a hindlimb-like 3D chromatin conformation in the forelimb allowing Pitx1-Pen interaction and mimicking arm-to-leg transformation observed in Liebenberg syndrome. Illustration is based on the study conducted by Kragesteen et al., 2018.

FIGURE 4

Evolutionary conservation of architectural proteins. (A) Vertebrate cohesin complex is a multi-subunit protein complex made up of a dimer of SMC proteins (SMC3-SMC1α/β) which is the core structural component forming a closed ring along with a kleisin (RAD21/RAD21L/REC8) and STAG1/STAG2/STAG3. The cohesin complex acts in conjunction with CTCF, an architectural protein found at TAD boundaries, to facilitate DNA looping (B) Human genome architectural proteins were queried against NCBI’s non-redundant protein sequence database, using BLASTP with default parameters, for 140 organisms covering all major taxonomic divisions. A representative for each was then selected: Mus musculus (mammals), Gallus gallus (birds), Anolis carolinensis (reptiles), Xenopus laevis (amphibians), Danio rerio (fish), Ciona intestinalis (tunicates), Drosophila melanogaster (insects), Caenorhabditis elegans (worms), Saccharomyces cerevisiae (yeast), Paramecium tetraurelia (protozoans), Arabidopsis thaliana (plants), and Escherichia coli (prokaryotes). When matches were obtained, the percent of the query (human protein) that has significant homology to the target protein were displayed as a heatmap in blue shades (dark = high homology, pale = low homology). Gray with a white dot indicates no matches were found.