Epigenetic memory enables the propagation of gene expression patterns following transient stimuli. Although three-dimensional chromatin organization is emerging as a key regulator of genome function, it is unknown whether it contributes to cellular memory. Here we establish that acute perturbation of the epigenome can induce cellular memory of gene expression in mouse embryonic stem cells. We uncover how a pulse of histone deacetylase inhibition translates to changes in transcription, histone modifications and genome folding. While most epigenomic and transcriptional changes are initially reversed once the perturbation is removed, some loci remain transcriptionally deregulated and genome architecture partially maintains its perturbed conformation. Consequently, a second pulse of transient hyperacetylation induces stronger memory of transcriptional deregulation. Using ultradeep Micro-C, we associate memory of gene expression with repressive Polycomb-mediated chromatin topology. These results demonstrate how cells can record transient stresses in their genome architecture, thereby enabling an enhanced response to subsequent perturbations.

Abstract Chromosomes constitute deeply conserved evolutionary units in many metazoan genomes, with chromosomal fusions and fissions, accompanied by sub-chromosomal rearrangements, rewiring three dimensional genome architecture. How chromatin loops and compartments that define distal regulatory interactions within chromosomes impose functional constraints that affect this long-term evolutionary process (and vice-versa) is an emerging research topic. Genome expansions, especially through transposable element (TE) activity, test these constraints by increasing the genomic distances over which regulatory interactions must function and were thus suggested to be the drivers of chromosomal rearrangements. To study dynamics and stability of such distal interactions in the light of genome expansions, we focus on the cnidarian Hydra vulgaris , which, based on its simple and well-understood biology as well as one of the largest genomes among cnidarians, is particularly suited to test how chromatin loop and genome architecture respond to genome expansion. We investigate genome architecture using Micro-C, single-cell Hi-C (Dip-C), and DNA FISH, and perform comparative analysis using available genomic and epigenomic data. Contrary to prior expectations, our analysis of whole-genome data and particular loci (e.g., Wnt) suggests a scenario where genome expansion did not only result in chromatin loops often reaching several megabases in hydra, but also led to regulatory contact mixing and entanglement, introducing additional constraints to maintain ancestral genomic architecture. Generalizing these findings across hundreds of metazoan genomes, we show a new mechanistic role for genome expansion in yielding entangled long-range regulatory configurations that, in turn, decelerate chromosomal rearrangements, thus maintaining (and not breaking) ancestral regulatory states and synteny. Significance statement Some of the largest and most repetitive animal genomes retain a surprisingly high level of deeply conserved metazoan synteny. As repetitive regions are often associated with chromosomal rearrangements, it has been enigmatic why syntenic retention is so frequently observed. In this study, using hydra as a model system for both stem cell biology and an evolutionary history rich in transposable-element driven genome expansion, a multi-level conformational landscape dissection reveals a multitude of long-range regulatory states. We show that mixing of multiple such regulatory links accompanied by genomic expansion is associated with maintained ancestral synteny, thus pointing to the counter-intuitive role of genome expansions as “fossilization” agents across metazoan genomes.

Three-dimensional DNA fluorescent in situ hybridization (3D-FISH) has been extensively used for the study of chromosome organization, from large-scale chromosome territories to finer sub-megabase structures, such as topologically associating domains (TADs) or internal TAD features such as chromatin loops. The development of oligo-based technology has rendered this technique highly versatile and allowed to study chromosomal structures at different scale levels, as well as structural variations such as deletions, duplications, or translocations, as those occurring in cancer and chromothripsis. Here, we present a comprehensive Oligopaint FISH protocol, from probe design and production to FISH, microscopy, and specific analysis that can be used with various cell lines.

Summary Non-genetic information can be inherited across generations by the process of Transgenerational Epigenetic Inheritance (TEI). TEI can be established by various triggering events, including transient genetic perturbations. In Drosophila , hemizygosity of the Fab-7 regulatory element triggers inheritance of the histone mark H3K27me3 at a homologous locus on another chromosome, resulting in heritable epigenetic differences in eye colour. By mutating transcription factor binding sites within the Fab-7 element, we demonstrate the importance of two proteins in the establishment and maintenance of TEI: GAGA-factor and Pleiohomeotic. We show that these proteins function by recruiting the Polycomb Repressive Complex 2 and by mediating interchromosomal chromatin contacts between Fab-7 and its homologous locus. Finally, using an in vivo synthetic biology system to induce them, we show that chromatin contacts alone can establish TEI, providing a mechanism by which hemizygosity of one locus can establish epigenetic memory at another in trans through long-distance chromatin contacts.

Non-genetic information can be inherited across generations in a process known as transgenerational epigenetic inheritance (TEI). In Drosophila, hemizygosity of the Fab-7 regulatory element triggers inheritance of the histone mark H3K27me3 at a homologous locus on another chromosome, resulting in heritable epigenetic differences in eye color. Here, by mutating transcription factor binding sites within the Fab-7 element, we demonstrate the importance of the proteins pleiohomeotic and GAGA factor in the establishment and maintenance of TEI. We show that these proteins function by recruiting the polycomb repressive complex 2 and by mediating interchromosomal chromatin contacts between Fab-7 and its homologous locus, respectively. Using an in vivo synthetic biology system to induce them, we then show that chromatin contacts alone can establish TEI, providing a mechanism by which hemizygosity of one locus can establish epigenetic memory at another distant locus in trans through chromatin contacts.

Upon differentiation chromatin rewires to reflect its new cellular identity and function. While it is widely known that this process involves cooperative changes in transcription, chromatin composition and 3D conformation, it is unclear what exactly drives these changes and how they influence one another. Here we used ESC-to-NPC differentiation to study rewiring at a 3 Mb large neuronal Zfp608 locus. During this process, this large chromatin domain splits in half right at the Zfp608 promoter, local chromatin gets littered with activating marks, compacts in 3D space and Zfp608 abounds in transcription. We investigated the cis and trans elements using capture Hi-C (cHi-C), extensive biophysical modelling, and 3-colour 3D-FISH with technical and analytical breakthroughs and found that transcription abundance modulates the contacts in the region as well as the insulation at the domain split. Furthermore, we found a genetic element we named scaffolding element, with a dual enhancer and architectural function that is essential for chromatin rewiring and loop formation at the NPC stage. The loss of this element disrupts the formation of all local NPC-loops irrespective if they are anchored in this element or not, highlighting the hierarchical relationship between elements that act as loop anchors. Furthermore, we uncovered that the scaffolding function, although driven by multiple mechanisms, can form loops independent of loop-extrusion and that other molecular attractions were necessary to form NPC-specific contacts in the region. Together, these results demonstrate that a hierarchy of genetic elements in cis allows successful rewiring during differentiation and that multiple trans acting elements contribute to make this rewiring efficient.### Competing Interest StatementThe authors have declared no competing interest.

Polycomb (Pc) group proteins are transcriptional regulators with key roles in development, cell identity, and differentiation. Pc-bound chromatin regions form repressive domains that interact in 3D to assemble repressive nuclear compartments. Here, we use multiplexed chromatin imaging to investigate whether Pc compartments involve the clustering of multiple Pc domains during Drosophila development. Notably, 3D proximity between Pc targets is rare and involves predominantly pairwise interactions. These 3D proximities are particularly enhanced in segments where Pc genes are co-repressed. In addition, segment-specific expression of Hox Pc targets leads to their spatial segregation from Pc-repressed genes. Finally, non-Hox Pc targets are more proximal in regions where they are co-expressed. These results indicate that long-range Pc interactions are temporally and spatially regulated during differentiation and development but do not induce frequent clustering of multiple distant Pc genes.

Introduction À l’intérieur du noyau, le génome est fortement compacté et plié. La combinaison des nouvelles techniques de génomique et de microscopie a révélé que notre génome est organisé selon quatre niveaux. Des plus grands, avec les territoires chromosomiques et les compartiments A/B, aux plus petits avec les Topologically Associating Domains (TADs) et les boucles chromatiniennes. Cependant, les mécanismes impliqués dans le repliement de la chromatine au-dessus du nucléosome sont encore mal caractérisés.En combinant le FISH oligopaint et la microscopie à super résolution, notre laboratoire a récemment montré que les TADs sont subdivisés en une nouvelle couche d’organisation, appelée ChromatinNano Domains (CNDs). Alors que la structure et la formation des TADs dépendent largement de deux protéines (CTCF et cohésine), les CNDs ne sont pas affectés par la déplétion de ces facteurs, indiquant que d’autres mécanismes pourraient être impliqués dans leur formation [1]. Objectif Le projet vise à caractériser les mécanismes de formation des CNDs. L’analyse des données de Chip-Seq et de microscopie a révélé une corrélation entre les niveaux d’acétylation des TADs et leur fragmentation en CNDs. Nous avons émis l’hypothèse que les interactions nucléosome-nucléosome, due à l’acétylation des histones, pourraient jouer un rôle central dans la formation des CNDs. Méthodes Afin d’étudier l’impact de l’acétylation sur les CNDs, nous avons induit une hyperacétylation et une hypoacétylation globale au sein du génome de cellules souches embryonnaires de souris. Résultats L’examen des loci à l’aide de la microscopie 3D à super-résolution dans ces conditions a démontré une forte association entre l’organisation des CNDs et le niveau d’acétylation des histones. Ces résultats soulignent que des mécanismes distincts et indépendants régissent l’organisation des TADs et des CNDs. Discussion Les CNDs pourraient servir de chaînon manquant entre les nucléosomes et les TADs et être une structure mettant en contact les enhancers avec les promoters régulant la transcription des gènes.

Gene duplications and transcriptional enhancer emergence/modifications are thought having greatly contributed to phenotypic innovations during animal evolution. Nevertheless, little is known about how enhancers evolve after gene duplication and how regulatory information is rewired between duplicated genes. The Drosophila melanogaster bric-a-brac (bab) complex, comprising the tandem paralogous genes bab1 and bab2, provides a paradigm to address these issues. We previously characterized an intergenic enhancer (named LAE) regulating bab2 expression in the developing legs. We show here that bab2 regulators binding directly the LAE also govern bab1 expression in tarsal cells. LAE excision by CRISPR/Cas9-mediated genome editing reveals that this enhancer appears involved but not strictly required for bab1 and bab2 co-expression in leg tissues. Instead, the LAE enhancer is critical for paralog-specific bab2 expression along the proximo-distal leg axis. Chromatin features and phenotypic rescue experiments indicate that LAE functions partly redundantly with leg-specific regulatory information overlapping the bab1 transcription unit. Phylogenomics analyses indicate that (i) the bab complex originates from duplication of an ancestral singleton gene early on within the Cyclorrhapha dipteran sublineage, and (ii) LAE sequences have been evolutionarily-fixed early on within the Brachycera suborder thus predating the gene duplication event. This work provides new insights on enhancers, particularly about their emergence, maintenance and functional diversification during evolution.

Regulatory landscapes drive complex developmental gene expression, but it remains unclear how their integrity is maintained when incorporating novel genes and functions during evolution. Here, we investigated how a placental mammal-specific gene, Zfp42, emerged in an ancient vertebrate topologically associated domain (TAD) without adopting or disrupting the conserved expression of its gene, Fat1. In ESCs, physical TAD partitioning separates Zfp42 and Fat1 with distinct local enhancers that drive their independent expression. This separation is driven by chromatin activity and not CTCF/cohesin. In contrast, in embryonic limbs, inactive Zfp42 shares Fat1's intact TAD without responding to active Fat1 enhancers. However, neither Fat1 enhancer-incompatibility nor nuclear envelope-attachment account for Zfp42's unresponsiveness. Rather, Zfp42's promoter is rendered inert to enhancers by context-dependent DNA methylation. Thus, diverse mechanisms enabled the integration of independent Zfp42 regulation in the Fat1 locus. Critically, such regulatory complexity appears common in evolution as, genome wide, most TADs contain multiple independently expressed genes.

SUMMARYCohesin loop extrusion facilitates precise gene expression by continuously driving promoters to sample all enhancers located within the same topologically-associated domain (TAD). However, many TADs contain multiple genes with divergent expression patterns, thereby indicating additional forces further refine how enhancer activities are utilised. Here, we unravel the mechanisms enabling a new gene,Rex1,to emerge with divergent expression within the ancientFat1TAD in placental mammals. We show that such divergent expression is not determined by a strict enhancer-promoter compatibility code, intra-TAD position or nuclear envelope-attachment. Instead, TAD-restructuring in embryonic stem cells (ESCs) separatesRex1andFat1with distinct proximal enhancers that independently drive their expression. By contrast, in later embryonic tissues, DNA methylation renders the inactiveRex1promoter profoundly unresponsive toFat1enhancers within the intact TAD. Combined, these features adapted an ancient regulatory landscape during evolution to support two entirely independentRex1andFat1expression programs. Thus, rather than operating only as rigid blocks of co-regulated genes, TAD-regulatory landscapes can orchestrate complex divergent expression patterns in evolution.HIGHLIGHTSNew genes can emerge in evolution without taking on the expression pattern of their surrounding pre-existing TAD.Compartmentalisation can restructure seemingly evolutionarily stable TADs to control a promoter’s access to enhancers.Lamina-associated domains neither prevent transcriptional activation nor enhancer-promoter communication.Repression rather than promoter-specificity refines when genes respond to promiscuous enhancer activities in specific tissues.

During animal evolution, de novo emergence and modifications of pre-existing transcriptional enhancers have contributed to biological innovations, by implementing gene regulatory networks. The Drosophila melanogaster bric-a-brac ( bab ) complex, comprising the tandem paralogous genes bab1 - 2 , provides a paradigm to address how enhancers contribute and co-evolve to regulate jointly or differentially duplicated genes. We previously characterized an intergenic enhancer (named LAE) governing bab2 expression in leg and antennal tissues. We show here that LAE activity also regulates bab1 . CRISPR/Cas9-mediated LAE excision reveals its critical role for bab2 -specific expression along the proximo-distal leg axis, likely through paralog-specific interaction with the bab2 gene promoter. Furthermore, LAE appears involved but not strictly required for bab1 - 2 co-expression in leg tissues. Phenotypic rescue experiments, chromatin features and a gene reporter assay reveal a large “pleiotropic” bab1 enhancer (termed BER) including a series of cis -regulatory elements active in the leg, antennal, wing, haltere and gonadal tissues. Phylogenomics analyses indicate that (i) bab2 originates from bab1 duplication within the Muscomorpha sublineage, (ii) LAE and bab1 promoter sequences have been evolutionarily-fixed early on within the Brachycera lineage, while (iii) BER elements have been conserved more recently among muscomorphans. Lastly, we identified conserved binding sites for transcription factors known or prone to regulate directly the paralogous bab genes in diverse developmental contexts. This work provides new insights on enhancers, particularly about their emergence, maintenance and functional diversification during evolution. Author summary Gene duplications and transcriptional enhancer emergence/modifications are thought having greatly contributed to phenotypic innovations during animal evolution. However, how enhancers regulate distinctly gene duplicates and are evolutionary-fixed remain largely unknown. The Drosophila bric-a-brac locus, comprising the tandemly-duplicated genes bab1 - 2 , provides a good paradigm to address these issues. The twin bab genes are co-expressed in many tissues. In this study, genetic analyses show a partial co-regulation of both genes in the developing legs depending on tissue-specific transcription factors known to bind a single enhancer. Genome editing and gene reporter assays further show that this shared enhancer is also required for bab2 -specific expression. Our results also reveal the existence of partly-redundant regulatory functions of a large pleiotropic enhancer which contributes to co-regulate the bab genes in distal leg tissues. Phylogenomics analyses indicate that the Drosophila bab locus originates from duplication of a dipteran bab1 -related gene, which occurred within the Brachycera (true flies) lineage. bab enhancer and promoter sequences have been differentially-conserved among Diptera suborders. This work illuminates how transcriptional enhancers from tandem gene duplicates (i) differentially interact with distinct cognate promoters and (ii) undergo distinct evolutionary changes to diversifying their respective tissue-specific gene expression pattern.

To understand the role of the extensive senescence-associated 3D genome reorganization, we generated genome-wide chromatin interaction maps, epigenome, replication-timing, whole-genome bisulfite sequencing, and gene expression profiles from cells entering replicative senescence (RS) or upon oncogene-induced senescence (OIS). We identify senescence-associated heterochromatin domains (SAHDs). Differential intraversus inter-SAHD interactions lead to the formation of senescence-associated heterochromatin foci (SAHFs) in OIS but not in RS. This OIS-specific configuration brings active genes located in genomic regions adjacent to SAHDs in close spatial proximity and favors their expression, We also identify DNMT1 as a factor that induces SAHFs by promoting HMGA2 expression. Upon DNMT1 depletion, OIS cells transition to a 3D genome conformation akin to that of cells in replicative senescence. These data show how multi-omics and imaging can identify critical features of RS and OIS and discover determinants of acute senescence and SAHF formation.