#	Pagina
attuale pagina	/open-h2020/projects/204948/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - 4DGenomeReg (Predictive modelling of 3D genome topology during progressive stages of embryonic development)

Teaser

Summary

Transcriptional regulation is an extremely complex process that allows for appropriate readout of genetic information in every living organism. This process is particularly important during development and disease, and controlled by intricate mechanisms of DNA regulatory elements. These regulatory elements, including gene promoters and distal enhancers, compile regulatory signals received through transcription factor binding. The three-dimensional architecture of the genome brings enhancers, often located at large linear distance along the genome, into spatial proximity to the target genes.

On a larger scale, the genome is partitioned into Topologically Associated Domains (TADs), characterized by having internal contacts more frequent than contacts to the outside regions. TAD boundaries are thought to insulate regulatory environments between adjacent TADs, preventing gene promoters from being inadvertently activated by enhancers located in other TADs. Several studies showed that disruption of a TAD boundary could affect expression of genes on the other side of the boundary through enhancer hijacking. This is also the case for naturally occurring structural variants that disrupt TAD boundaries, and this mechanism has been shown to have a role in driving certain developmental disorders and cancers.

This project used chromosome conformation capture techniques, Hi-C and Capture-C, to measure chromatin topology at a genome-wide scale in the context of embryonic development of Drosophila melanogaster. We performed Hi-C experiments during multiple stages of development spanning the entire embryonic timespan to determine if the chromatin interaction landscape observed during the early stages of development is maintained to the end of embryogenesis. For this purpose, we identified topologically associated domains (TADs) and annotated long-range looping interactions.

To investigate the relationship between genome topology and gene expression, we utilized highly rearranged balancer chromosomes, which contained eight large nested inversions, thousands of smaller structural variants, and hundreds of thousands of single nucleotide variants. We then assessed the impact of these genomic rearrangements on genome topology and gene expression in cis, using a heterozygous cross (balancer over wild-type), which minimizes the contribution of trans effects. For this purpose, we performed allele-specific Hi-C and RNA-seq experiments on F1 embryos.

Work performed

We first aimed to comprehensively characterize chromatin interactions across embryonic development. We identified topologically associated domains (TADs) and long-range interactions in each timepoint, and obtained insights across embryogenesis by comparing data from the early and late stages of development. We also identified chromatin interactions within a single timepoint, separately in wild-type and balancer chromosomes.

Furthermore, we identified differential chromatin interactions between wild-type and balancer haplotypes, in the context of embryonic development. We opted for a comparison between haplotypes in a heterozygous cross, as opposed to comparing two different samples, to minimize confounding experimental and trans effects. To identify changes in chromatin interactions, we employed a quantitative method to compare Hi-C contacts maps obtained at different conditions. We also compared the TAD boundaries identified across haplotypes.

Finally, we attempted to derive a predictive model relating genome topology and gene expression. For this purpose, we correlated allele-specific changes in chromatin contacts to the changes in gene expression. We found that the extensive rearrangements of balancer chromosomes caused many changes to chromatin topology at the level of long-range compartmental loops, TADs and intra-TAD chromatin interactions, yet these changes were not predictive of changes in expression. Our results suggest that there must be other properties, in addition to genome topology, that determine the specificity and productivity of enhancer-promoter interactions.

Final results

We showed that major genomic rearrangements at multiple scales can have a limited impact on transcription â€“ only a few hundred genes had a significant change in their expression. In particular, large nested inversions that disrupted and fused TADs changed the expression of only a fraction (~10%) of the genes in the affected TADs. In the cases where large nested inversions fused TADs together, we found limited evidence for enhancer hijacking â€“ only a small subset of genes had altered expression. Also long-range looping interactions that are broken by genomic rearrangements do not necessarily induce changes in gene expression. While the changes in genome topology influence gene expression at multiple individual loci, the expression of the majority of genes remained unchanged despite the changes caused by genomic rearrangements.

In general, our results are suggestive of uncoupling between gene expression and genome topology. Only a subset of genes responded to changes in their chromatin context. This was the case when TADs were disrupted and fused, but also more locally, around differential contacts constrained within TADs.

While the project used fruit fly Drosophila melanogaster as a model organism, the insights on uncoupling genome topology and gene expression are generalizable to other species, in particular to human health and disease. Our findings will be communicated in open access format to the scientific community, and also to the general public.