Skip to main content

Biophysical mechanisms of long-range transcriptional regulation

Periodic Reporting for period 2 - BioMeTRe (Biophysical mechanisms of long-range transcriptional regulation)

Reporting period: 2019-07-01 to 2020-12-31

To establish and maintain their identity, cells require precise control of gene expression. In mammals, the transcriptional control of many genes relies on DNA sequences named enhancers that can be located very far on the genome from the genes that they control. Enhancers are nevertheless fundamental to determine when, where and how much their target genes are expressed. Many lines of evidence point at the fact that distal enhancers control gene expression by looping out intervening DNA and contacting their target genes, so that molecular information can be transmitted between the two genomic elements. This in turn is linked to how chromosomes fold in the three-dimensional space of the cell nucleus. To fully understand transcriptional regulation by enhancers, it is therefore fundamental to quantitatively characterize chromosome structure, including how it varies in time and across different cells in a tissue.

Chromosome conformation capture (3C)-based studies, which measure chromosomal contacts using chemical fixation, have revealed that mammalian chromosomes are partitioned into a complex hierarchy of interaction domains, at the heart of which lie topologically associating domains (TADs), sub-megabase regions of the chromatin fiber that form preferential interactions, and their substructures. Genetic evidence has shown that these specific chromosomal structures are able to restrict the genomic range of enhancer-promoter communication, as well as fine-tune the three-dimensional interactions between regulatory sequences.

However, the mechanistic details of how physical interactions within chromosomes translate into transcriptional outputs are totally unknown and many fundamental questions remain open, such as: What are the mechanisms by which physical interactions determine the activity of enhancer-promoter pairs? Does the manipulation of chromosome interactions perturb transcriptional activities, and how? How dynamic is chromosome conformation at the level of TADs and their sub-structures, and is this linked to variability in gene expression in time and in single cells? These questions overarch molecular biology and biophysics. We are addressing them using an integrated approach combining molecular biology, genome engineering, live-cell microscopy and physical modeling. Our results will enable a quantitative understanding of how chromosome structure contributes to regulate transcription, and how it can be engineered to manipulate and correct aberrant gene expression.
The current understanding of chromosome folding largely relies on methods based on chromosome conformation capture techniques (3C), such as Hi-C and 4C. These methods quantify the number of interactions between genomic loci that are in close physical proximity in the three-dimensional space of the cell nucleus. In 3C methods, chromosomal interactions are detected as ligation products after chromatin crosslinking with formaldehyde, which ‘glues’ proteins and DNA together in an irreversible manner. However, crosslinking and ligation have been criticized as sources of potential experimental bias. This raises the question of whether structures such as TADs and chromatin loops, which are intensively studied because of the fundamental implications for a range of research areas from epigenetic regulation to genetic disorders, actually exist in living cells.

To address this question, we have developed a technique called DamC, which combines DNA-methylation based detection of chromosomal interactions with next-generation sequencing and biophysical modelling of methylation kinetics. Using DamC in mouse embryonic stem cells, we could confirm the existence of key structural features of chromosomes, notably TADs and chromatin loops. This not only validates 3C as an experimental method, but also provides a definitive proof of the fundamentals of chromosomal folding. Our results have been recently published in Nature Structural and Molecular Biology (DOI: 10.1038/s41594-019-0231-0).

We have also made considerable progress in understanding the mechanisms that allow chromosomal interactions and TADs in particular to mediate functional communication between enhancers and promoters. To this aim, we have established a new method to quantitatively measure transcription from a promoter as a function of its distance from a cognate enhancer located at hundreds of different genomic positions within a TAD. Using this method in mouse embryonic stem cells, we have discovered that there is a quantitative, nonlinear relationship between transcription levels at the promoter and its contact probabilities with an enhancer. This explains why enhancers are able to communicate with promoters across large genomic distances within TADs, but are insulated by TAD boundaries. I have presented these results at several international (virtual) meetings and a manuscript is in preparation.

In addition we have collaborated with other laboratories to 1) better characterize the role of chromosome structure at the X inactivation center in mouse embryonic stem cells (van Bemmel et al., Nature Genetics 2019, DOI: 10.1038/ s41588-019-0412-0); and 2) determine the architecture of the regulatory network underlying X chromosome inactivation (Mutzel et al., Nature Structural and Molecular Biology 2019, DOI: 10.1038/s41594-019-0214-1). I have also coordinated and am the corresponding author on a review for Molecular Cell (McCord, Kaplan and Giorgetti 2019, 10.1016/j.molcel.2019.12.021).
Research in our lab, and notably projects funded by the ERC Starting Grant, are fueled by a tight interplay between theory, computational approaches and experiments. This interdisciplinary approach is key to developing new quantitative methodologies allowing to address the big open questions in chromosome structure and gene regulation.

Our recent findings that 3C-based approaches do not distort the detection of chromosomal interactions were enabled by DamC, a new technique that relies on a biophysical model of the underlying molecular processes for the interpretation of experimental data. We anticipate that our research will inspire the development of additional quantitative methods relying on a close interplay between modeling and experiments, which will enable accurate measurements of biological processes in the cell nucleus. DamC also provided the first orthogonal proof of the existence of chromosome structures such as TADs and chromosome loops, which are highly debated because of their potential functional significance but had been detected so far using a single technology (3C). Our results thus provide solid grounds to the current debate around the role of genome organization in different contexts ranging from epigenetic regulation to oncogenesis and genetic disorders.

As witnessed by the overwhelmingly positive feedback at recent conferences, our discovery that chromosome interactions are translated nonlinearly in transcription levels will have a big impact on the field of transcriptional regulation. Our observations provide the first solid, quantitative evidence that chromosome structure is a major control layer in transcriptional regulation, and contribute resolving a long-standing controversy regarding the actual role of CTCF loops and TADs in the control of gene expression. Our results also have far-reaching implications because they reveal quantitative relationships between enhancer strength, interactions with a promoter and chromosome structure that will help fine-tuning enhancer therapy approaches.