The genome in space and time comes of age

ABSTRACT

DNA sequencing is not enough to grasp the complexity of genome organization and function. The four-dimensional (three in space, one in time) configuration of the eukaryotic nucleus varies with cell types, during development and in diseased tissues, and has to be taken into account to decipher genome function. To study, discuss, and advance in such direction, the International Nucleome Consortium COST Action, funded by the European Union, held its concluding symposium ‘The Genome in Space and Time’ at the Ionian University in Corfu, Greece, on September 10–13, 2023.

KEYWORDS:

• International consortium
• nucleome
• 4D genome organization
• chromatin
• nucleus

Introduction

It has been a long road since the discovery of the nucleus in plant cells [1] by Robert Brown, the first suggestion of genome organization following microscopic observations of Salamandra maculata cells by Carl Rabl in 1885 [2], the observation of DNA thanks to in situ Feulgen staining a hundred years ago [3], and of nucleosome 50 years ago [4]. Nowadays, with major advancements in microscopy, sequencing, and computational modeling, the complexity of the nucleus as a dynamic biological system (hereafter referred to as the nucleome) can be apprehended. To decipher the genome’s mysteries, the relationship between structure and function, a multiscale understanding of the nucleus in the three spatial dimensions and its dynamics through time is essential.

Figure 1 To help structure, promote, and organize discussions in nucleome research, following the ‘4D nucleome workshop’ in Mainz 10 years ago, the International Nucleome Consortium was founded. Major goals for the field were defined [5] notably regarding the development of new technologies to assess genome organization at different spatial and temporal scales, including databases and standardization of the data generated [6]. A focus was set on fostering collaborations between biologists, physicists, and mathematicians to explore the complexity of the nucleome, and linking its alterations to disease phenotypes. Fortunately enough, this consortium was financed during the past 4.5 years thanks to the European Union COST Action scheme. And while this funding is coming to an end, the last meeting of the consortium in Corfu, Greece, gives us the opportunity to see the accomplished path and the importance of a lasting international network and collaborations between fields regarding complex scientific questions like nucleus organization and functions, and future perspectives.

Figure 1. The cell nucleus as a city: left) a superresolutive image of H2B-GFP in a human cancer cell nucleus (adapted with permission from Lea Costes); right) an aerial view of Corfu city center from google map (images ©2023 google, image ©2023 Airbus, CNES/Airbus, Maxar Technologies, données cartographiques ©).

A question of organization

Our meeting venue, the Ionian University, the first university of modern Greece, facing east and the sea, opened its doors when the sun rose to the keynote by Thomas Cremer (Munich, Germany). A central question when we think about the nucleome is how two given players, or should we call them ‘actors’ (a TF and its target gene for instance) can find each other? How can genome organization and dynamics favor or prevent such interaction? Based on the latest research on chromatin density and accessibility using super-resolution optical microscopy, Thomas Cremer presented us ‘the city in the cell’ model to address these questions [7,8]. This model is based on the existence of avenues, streets, and alleys, i.e. interchromatin compartments, lined with accessible active chromatin compartments, whereas compact, repressed chromatin compartments are not served by these paths. Protein and molecule movements are therefore constrained, facilitating their search for their targets in a crowded city. Lively discussions addressed the question of how chemical and physical properties of chromatin associated factors contribute or impede this search.

The metaphor of nuclear organization as a city also implies robustness of the ‘city block’ i.e. structural unit of chromatin topology. By studying changes in chromatin organization during sex differentiation in mammalian gonads, Marc A. Marti-Renom (Barcelona, Spain) and his team and collaborators indeed observed moderate changes in A/B chromatin compartment and topological associated domains (TADs) between bipotential and differentiated stages. Nonetheless, by introducing METALoci, a computational approach to integrate Hi-C and ‘omics data for an unbiased quantification of the regulatory environment around each gene in the genome, they identified prominent rewiring of 3D enhancer hubs specifically during male sex differentiation [9]. Such rewiring suggests that even if nuclear compartmentalization is largely stable, individual genes can move from a repressive to a permissive environment in response to cellular cues.

In line with the idea of a global robust chromatin organization with finely tuned changes during differentiation, Philippe Collas (Oslo, Norway) highlighted the heterogeneity of Lamin Associated Domains (LADs), a repressive chromatin compartment associated with the nuclear envelope. While the majority of the LADs show little variability among cell types and during differentiation and remain a repressive compartment, smaller LAD regions (or LAD edges) known as variable LADs are a feature of differentiation and can detach from the lamina to join a permissive compartment, allowing gene regulation. He and his team observed a repositioning of variable LADs harboring adipogenic genes during adipocyte differentiation, whereas the remaining LADs sequester genes irrelevant for the adipose lineage [10].

Heart development relies on specific changes in genome organization. Studied using an elegant cardioid system (a chamber-specific multilineage cardiac organoid), Silvia Becca (Torino, Italy) and her team observed a switch from inactive (B) to active (A) compartments of cardiomyocyte genes during cardiomyocyte differentiation of pluripotent cells. This switch is regulated by progressive CTCF downregulation in association with an upregulation of GATA4 expression (a cardiac specific pioneer transcription factor) at TAD frontiers facilitating their B to A transition. These processes lead to a fine-tuned equilibrium between development and maturation of cardiac cells. To extend the city metaphor, during development sometimes a new street needs to be pierced, some streets will be renovated and other are blocked.

To gain information on chromatin compartmentalization, Francesco Ferrari (Pavia, Italy) presented us a novel experimental method to assess chromatin accessibility based on isolation and sequencing of multiple chromatin fractions, enriched for differences in accessibility: SAMMY-seq [11]. This method, without relying on crosslinking or antibodies, can detect changes in chromatin distribution across nuclear compartments characterized by distinct accessibility (LAD vs non-LAD heterochromatin compartments for instance). Successfully applied on prostate cancer biopsies, this technique led to identification of recurrent alterations of chromatin architecture in patient samples displaying aberrant epigenetic alterations.

Cancer cells can indeed be highly heterogeneous. Argyris Papantonis (Gottingen, Germany) and his team investigate the effect of genomic structural variations (SV) in patient’s glioblastoma. These variations are heterogeneous, uneven, and non-recurrent events in tumors but cluster with location biases. Importantly, they demonstrate that, without a priori information of the 3D structure of the genome, it is difficult to predict the effect of the identified SV. However, using Hi-C, they revealed SV promoting formation of neo-loops that sustained tumor-specific program via new enhancer–promoter interaction [12]. Such neo-loops could help inferring patient-specific vulnerabilities and provide a better diagnostic for glioblastoma patients.

These works highlight the importance of specific organization of chromosomes. While it has been accepted that chromatin loops are set up and maintained through interaction between chromatin, CTCF and Cohesin, Frank Fackelmayer (Ioannina, Greece), and his team are investigating whether chromatin organization can be sustained by scaffolding structures. Frank presented his latest results on SAF-A (hnRNP U), a protein originally identified as a factor that binds to architectural DNA elements [13]. Depletion of SAF-A, preferentially located at the interphase between eu- and heterochromatin, perturbs chromosome territories and chromatin structures. These new elements shed light on a forgotten aspect of chromatin organization, i.e. a non-chromatin skeletal structure.

Looping of interest

Nuclear architecture in mammalian cells is organized at different scales (chromosome territories, chromatin compartment, and topological domains) in a robust manner to ensure regulation of gene expression via management of chromatin loops between distant regulatory elements and the gene. Loops appear as an essential (and primordial) feature of eukaryote chromatin. Indeed, in Caenorhabditis elegans, although autosomes are not organized into TADs, chromatin looping participates in gene regulation. Jennifer Semple (Bern, Switzerland) shows, using Hi-C, that around active enhancer, loose hairpin-like structures are formed, identified as ‘fountains’ in the Hi-C maps, shaped by cohesin to regulate enhancer–promoter interaction and adequate gene expression [14], highlighting the importance of loops.

Biola Javierre (Barcelona, Spain), our second keynote speaker, indeed insisted on the relevance of chromatin loops in gene regulation and reminded us that only 19.5% of the genes are associated with their closest enhancer, urging the field to be careful not to make proximity association assumptions in the absence of 3D data. To this end, she and her team develop low input Capture Hi-C (liCHi-C) to map and compare promoter interactomes at high resolution and suitable for low abundance cell populations. Using this approach, it is possible to define trajectories of the human myeloid versus the lymphoid leucocyte lineages during differentiation. In leukemia, specific topologies are seen to correlate with prominent driver translocations (CEBP-IGH, HoxA, Myc) [15]. LiCHi-C further enabled deciphering how p53 control gene transcription to prevent malignant transformation and how p53 drives direct and indirect changes in chromatin architecture which escort its transcriptional program. Via the establishment of newly formed or preexisting chromatin loops, p53 controls transcription of distal genes, in a cohesin-dependent manner [16].

To investigate chromatin loop formationin particular, small-scale enhancer-promoter contacts, Abrar Aljahani (Gottingen, Germany) develops a MNase-based chromosome conformation approach, Tiled-MCC, reaching 20 bp resolution for interaction data [17]. By depleting various chromatin components, they were able to study the implication of loop extrusion regulators, chromatin remodelers, and mediator complexes on this fine-scale interaction between enhancer-promoter and their effect on transcription [18].

In another aspect of chromatin organizers, Jan Palecek (Brno, Czechia) presented his work on the SMC5/6 complex in fission yeast. Interestingly, he observed that the acetyltransferase complex SAGA interacts with the SMC5/6 complex to target it to gene-rich regions and facilitate its binding by increasing chromatin opening and accessibility. Of note, SAGA deletion does not impair the known binding of SMC5/6 at repair foci or heterochromatin repeated regions [19]. This work highlights the link between molecular organizers and epigenetic modifications.

Epigenetics at play

While we zoom in the levels of genome organization, epigenetic modifications take center stage. Because one can regulate gene expression without modifying the DNA sequence, epigenetic editing (i.e. ‘a locus-specific targeting of epigenetic enzymes to rewrite the local epigenetic landscape of an endogenous genomic site’) triggers the interest of scientists, investors, and clinicians. Marianne Rots (Groningen, the Netherlands) gave us a review of the advancement of epigenetic editing and its use in the clinic. She takes a census of four main concerns regarding the modification of epigenetic landscapes in medicine; first, the instructivity, whether editing an epigenetic mark is indeed causative in gene expression modification cannot always be predicted; second, the accessibility of the target site to modification, notably in the case of heterochromatin landscape targeting; third, the locus specificity of the editing, while it seems possible to achieve, more work and techniques need to be developed to assess off-target effects of the epigenetic editing tools (CRISPR dCas9 platform, notably), and fourth, the sustainability of the modification in time, while it is possible to achieve, it is also context/locus dependent. Overall, epigenetic editing seems to be a promising direction for clinical use but is in its infancy [20]. Having a holistic approach and considering the nucleome in its entirety will help to decipher the remaining obstacles.

The use of epigenetic editing in a clinical point of view was nicely illustrated by the presentation of Melita Vidakovic (Belgrade, Serbia) where pancreatic alpha cells were transiently reprogrammed into insulin-producing cells by enhancing repression via methylation of the Arx gene locus [21]. A similar approach is also developed by her team to repress BRCA1 in Triple Negative Breast Cancer. Taking advantage of the synthetic lethality between BRCA1 and PARP1, repressing BRCA1 potentiates the effect of existing PARP1 inhibitor treatments.

Non-alcoholic fatty liver disease (NAFLD) is also a disease associated with epigenetic alteration. Wim Vanden Berghe (Antwerp, Belgium) and his team show that genetic loss of PPARα in liver or its silencing via hypermethylation triggers compensatory mechanisms promoting an epigenetic transition from a lipid metabolic stress toward ferroptosis and pyroptosis lipid hepatoxicity pathways associated with advance NAFLD. Interestingly, Van den Berghe also observed that ferroptosis is associated with significant changes in histone modification and variants. On top of that, iron can also bind histones [22]. The interdependency between these two observations is still unanswered. Nevertheless, this interesting opening droves us to question the relationship between metabolism and the nucleome and therefore the more profound impact a pollutant on our biology.

To study the dynamics of DNA methylation, the group of Saulius Klimasauskas (Vilnius, Lithuania) redesigned the catalytic domain of DNMT1. This modification enables the enzyme to transfer a synthetically extended analog of S-adenosyl-L-methionine (SAM, the ubiquitous cofactor for methylation reaction). By CRISPR-engineering the Dnmt1 locus in mouse embryonic stem cells and following the pulse internalization of SAM analog, they were able to map the tagged methylation site selectively from Dnmt1 at a high temporal resolution [23]. This new epigenetic tool will provide a way to dissect the dynamic of individual methyltransferases.

And everything is moving

Indeed, as mentioned in the introduction, to improve our comprehension of the nucleome, it is important to add a temporal aspect to it and study the dynamics of chromatin within the nucleus. Extending the city metaphor, we should imagine a phantasmagoric city where not only the inhabitant but also the buildings are moving in complex choreography. This was highlighted in several talks during the conference. Pernette Verschure (Amsterdam, the Netherlands) presented her work on transcription dynamics. Transcription occurs in bursts and it is a substantial source of gene expression variability contributing to phenotypic heterogeneity. Her team adapted genome-wide nascent RNA sequencing techniques to study these transcription bursts. They investigate the relationship between the transcription burst size and frequency of genes and their epigenetic regulation in MCF7 breast cancer cells. Simultaneously, they assess the functional impact of transcription bursting on cell behavior [24].

Transcription and notably RNA Polymerase 2 also influence the motion of the chromatin itself. Everything in the ‘city’ is in motion. This aspect was nicely demonstrated during the talk of Kerstin Bystricky (Toulouse, France). She introduced two real-time imaging techniques allowing the nano-resolution tracking of single loci (ANCHOR technology) or the whole nuclear chromatin (Hi-D). Using these, her team observed that hormone-induced transcription initiation confines chromatin motion locally due to RNA pol II induced forces and enhancer-bound estrogen receptor accumulation [25]. At the scale of the whole nucleus, transcription activation (quiescence exit) globally decreases the diffusion properties of the chromatin [26]. Local confinement may be due to preexisting domain folding patterns brought about by clustering of factors bound to enhancers.

The impact of RNA polymerases on chromatin dynamic and 3D gene folding was further underlined by the work presented by Daniel Jost (Lyon, France). Using micro-C data and polymer modeling, they observed a non-monotonic correlation between intragenic contact density and Pol II occupancy independent of cohesin loop extrusion. Pol 2 occupancy appears as a key determinant of gene folding through phase separation. Interestingly, the speaker suggested that Pol 2-mediated condensation, coupled with transcriptional bursting, may slow down gene mobility, aligning with Bystricky’s observations [27].

Spyros Georgatos (Ioannina, Greece) and his team studies the dynamics of heterochromatin (by assessing the mobility of HP1) during Embryonic Stem Cell differentiation. In contrast to previous findings, they observed that chromatin dynamics are not globally suppressed during differentiation but evolve via subtle state transitions. Importantly, they show that heterochromatin dynamics are cell lineage dependent, suggesting that chromatin remodeling involves a partially stochastic de-stabilization of regional steady states and re-stabilization of local chromatin structure by cell-specific factors [28].

During cell division, DNA sequence, epigenetic marks, and 3D conformation must be transmitted to both daughter cells. Duplication of the genome takes place in a chronological order, in interconnexion with 3D genome organization, with compartment A and B replicating early and late in S phase, respectively. Whether the replication timing program function is important for chromatin organization preservation is an appealing hypothesis. Sara Buonomo (Edinburgh, UK) and her team explore the molecular mechanism coordinating both the temporal organization of replication and of nuclear architecture. Focusing on RIF1, a key regulator of both processes, they were able to demonstrate that RIF1 and PP1 (protein phosphatase 1) are required in early G1 both for the reestablishment epigenetic identity, positioning of genomic loci and the replication timing [29]. However, the functions of RIF1 in these two processes are independent, decoupling replication timing and 3D genome organization, leaving the function of the replication timing program for now unanswered.

On another dynamical aspect of the nucleome, Jerzy Dobrucki (Krakow, Poland) and his team are studying protein dynamics during DNA repair. Using a laser beam, they were able to induce localized double-strand breaks and studied the formation of repair foci by following the movement of 53BP1 in live cell by microscopy. 53BP1 dynamically exchanges between the repair foci and the surrounding environment; however, at a lower rate than between intact chromatin loci and the mobile nucleoplasmic fraction. Additionally, the mobility of the 53BP1 increases in the nucleus following a double-strand break but is less mobile within the repair foci. This observation could be explained in part by the abundance of 53BP1 within the repair foci causing steric hindrance (there is two time more 53BP1 in the foci compared to the rest of the nucleus). They hypothesize that this microenvironment with high protein concentration may stabilize the DNA repair complex.

Concluding remarks

As was emphasized at the creation of the International Nucleome Consortium, to functionally characterize the nucleus, it is essential to visualize the genome in space and time. This meeting gave us the opportunity to appreciate the effort made toward this goal in many aspects in recent years: from a technological point of view, with the development of new methodologies to decipher chromatin organization and motion, from a collaborative point view, with fostering of interdisciplinary partnerships, and from a comprehensive point of view, with the demonstration of the relevance of this 4D way of thinking to tackle scientific questions regarding cancer, development, and nuclear mechanisms.

Author contribution statement

M.R.-M. and KB drafted, conceptualized and wrote the manuscript. K.B. revised the manuscript. M.R.-M. design the figure. All authors have read and approved its submission and agreed to be accountable for all aspect of the work.

Acknowledgments

We would like to thank the organizers of this COST meeting in Corfu, Franck Fackelmayer, Spyridon Doukakis, Panayiotis Valmos and Marios Krokidis and the Ionian University. Our thanks go as well to COST Action International Nucleome Consortium (INC) CA18127 and its organizing committee. Finally, we thank all the participants of the Genome in Space and Time meeting (https://inc-cost.eu/corfu-2023) for sharing their latest results and their scientific generosity, including those we were not able to mention in this synthetic report.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Additional information

Funding

This work is supported by COST Action CA18127 and the Institut Universitaire de France.