Deciphering the function of Nuclear Speckles in 3D genome architecture

The nucleus of human cells is compartmentalized into functional DNA and protein assemblies that execute essential biological processes in a tightly controlled manner. Biomolecular condensates are membraneless bodies with essential roles in compartmentalizing biochemical reactions in cells. Nuclear speckles (NS) are amongst the most prominent condensates in the human nucleus and are implicated to have a role in regulating transcription, RNA processing and export. Moreover, the proximity of a gene locus to the NS in three dimensional space correlates with high transcription rates from this locus. However, the majority of these connections between NS and gene expression regulation originate from correlative studies, especially in the absence of direct evidence from an NS-absent cell state. We identified two evolutionarily conserved, large and disordered proteins, SON and SRRM2, as the essential proteins for NS formation. This finding does not only allow us to investigate the function of NS in humans, but also opens up new possibilities for studying the evolution of NS and of its functions. The main goal of the project proposed here is to bring two compartmentalisation processes in the nucleus —chromatin topology and biomolecular condensates— together using nuclear speckles as a model. We will investigate the role of NS in nuclear architecture and establish connections between its role in RNA processing and nuclear architecture. This work will be further expanded into fruit flies in order to find evolutionary parallels between gene expression programs and the nuclear architecture. Since loss of function mutations of SON were identified in human patients with intellectual disability and developmental delay, the proposed project will have the possibility to provide more mechanistic understanding into the disease phenotype.

Visualising nanoscale 3D genome architecture and transcriptional state during cell fate specification in the early mouse embryo

Mammalian development is a highly plastic process that begins with fertilisation of the oocyte bythe sperm to form the zygote, a diploid totipotent cell containing two pro-nuclei, which undergoes several rapid cell divisions to build a blastocyst that is competent for implantation into the uterine wall of the mother. The blastocyst contains the first two lineage-committed cell types in mammalian development, the extraembryonic trophectoderm and the inner cell mass that provides embryonic stem cells. These have different morphologies and differential expression profiles. The field has recently started to understand that regulation of early differentiation steps is associated with major changes in the hierarchical spatial organization of chromatin, in addition to many changes in the transcriptome and the epigenome. However, it remains to be investigated whether or how the spatial restructuring of a genomic locus affects its activity, or vice versa, and how the changes in spatial genome architecture differ between lineages and modulate gene expression during lineage specification. This gap in our knowledge is largely due to the fact that direct combined visualization of the physical 3D structure of the genome and transcriptional activity in single differentiating cells is lacking, which would allow us to reveal when and how changes in the spatial genome architecture are linked to changes in function such as gene expression, in situ inside single embryonic cells. In the proposed project we plan to address this gap in our knowledge and decipher the relation between genomic architecture and transcription in single cells of the early mouse embryo. To achieve this, we will combine our recently developed 3D chromatin tracing technology with imaging of single-allele transcriptional activity and nuclear architecture and relate these to cellular fate. This novel approach will allow us to quantitatively map how genome architecture changes when identical sister cells differentiate into inner cell mass and trophectoderm. Our experiments will thus reveal which structural hallmarks of the genome underlie the first fate specification in mammalian life. In summary, the proposed project will for the first time directly visualise changes in genome architecture associated with transcription and cell fate at the nanoscale in single blastomeres during early mammalian development.In combination with the single-cell transcriptomics and live-cell imaging technologies available within the consortium, this will allow us to create a complete view of the structure-function relationship between genome, transcriptome and fate specification in the developing embryo.

Assessing the functional role of cohesin and transcription in genome topology during Drosophila embryogenesis

The three-dimensional organization of the genome plays an important role in bringing enhancers (E) in close proximity to their cognate promoters (P) via chromatin looping. Such E-P loops occur within topology associated domains (TADs), although how TADs facilitate E-P function is still unresolved. In mammals, TADs are formed by Cohesin mediated loop extrusion. However, it remains unclear how TADs are formed in other species, and how E-P looping is regulated, especially during embryogenesis. In Drosophila, and to a lesser degree also in mouse and humans, many TAD boundaries have features of active transcription such as promoters and the binding of the transcription machinery. Moreover, ectopic insertion of transcriptionally active transposons are able to form a boundary de novo, suggesting that transcription itself can also shape genome topology. We are in a unique position to address these questions thanks to a new optogenetic nuclear depletion system that we developed. Distinguishing cause and consequence between transcription and topology has so far been very difficult using conventional loss-of-function mutants as they often block embryogenesis at very early stages, before TADs are even established. We can now solve these issues using our recently developed iLEXY system, to deplete nuclear proteins in minutes, inside living embryos. We will combine iLEXY with high-resolution genomics to determine the role of cohesin and the basal transcriptional machinery in the establishment and maintenance of TADs and E-P loops during Drosophila embryogenesis. We will target three components each of the cohesin complex (Mau2, Smc3, WapL) and the basal transcriptional machinery (M1bp, Spt4 and Spt5) for depletion. M1bp is essential for the establishment of the pre-initiation complex at a subset of constitutively active genes that are present at TAD boundaries. Spt4 and Spt5 are required for Pol II pausing and its subsequent elongation. By targeting all three, we can thereby distinguish between a role of the PIC assembly, versus Pol II pausing and elongation in TAD boundary formation. Embryos with rapid depletion and repletion of each of these six proteins will be used for low input high resolution genomic assays to quantify (1) chromatin occupancy of proteins via CUT&Tag, (2) changes in TADs and E-P loops via MicroC, and (3) nascent transcription to measure the impact on gene expression. This will reveal if cohesin-mediated loop extrusion is an evolutionarily conserved mechanism in chromatin folding or if alternate mechanisms exist. It will also provide the first functional data on the role of Pol II pausing in embryonic development, as different models have been proposed. Taken together, our proposal has the potential to decouple, for the first time, the interdependency between genome topology and transcription in vivo.

Understanding how changes in spatial chromatin organization cause defective gene expression in acute myeloid leukemia with STAG2 mutations

Cohesin genes are frequently mutated in acute myeloid leukemia (AML), but the exact mechanisms underlying their relevance in hematopoietic stem cell biology are unknown. STAG2 is the most frequently mutated cohesin member in AML and characterized by a specific set of co-mutations, while its paralogue STAG1 is frequently affected in bladder and breast cancers but hardly ever mutated in myeloid malignancies. With this follow-up proposal, we would like to continue unraveling the importance of STAG2 mutations in spatial chromatin organization and gene expression in AML. Based on our preliminary work, we hypothesize that STAG2 mutations specifically affect interactions between regulatory elements. Here, we propose to focus on the effects of STAG2 loss on promoter-enhancer interactions in STAG2 mutant AMLs using Hi-C (already available; see preliminary work) as well as high resolution Micro-C-enhancer-promoter capture for the same patients. In addition, we aim to study the synergistic effects with distinct co-occurring mutations, including SRSF2 and ASXL1.We will further address the mechanisms responsible for different roles of STAG1 or STAG2 in human myeloid progenitors by identifying immediate STAG1 and STAG2 targets. Here we will utilize degron-technology in a human hematopoietic cell model to follow the effects of rapid STAG protein depletion on epigenetic features and nascent transcription, both, on the level of progenitor cells as well as during myeloid differentiation. The results obtained from this project will reveal fundamental insights into overlapping and distinct functions of STAG1 and STAG2 in hematopoietic cells and the effects of mutated STAG2 on 3D chromatin folding and gene regulation in primary AML cells.

Activity-dependent gene expression in male and female neurons: 3D genome architecture, transcription and chromatin mechanisms

Activity-regulated gene expression, the sequential expression of immediate early genes (IEG) and secondary response genes (SRG), facilitates the conversion of external stimuli into a transcriptional response which in turn impacts on e.g. synapses to modulate post-stimulus neuron function. Like other genes, IEG and SRG are regulated by transcription factors that bind to regulatory elements that interact with their respective target genes in the context of the various levels of 3D genome organization (e.g. topologically associating domains or TADs, high frequency looping interactions). How TF binding, chromatin changes and long range interactions feed into enhancer activation, enhancer-promoter interactions and gene activation at these loci remains less clear. Also, how specificity at the level of TF binding to regulatory elements and of enhancer-promoter communication is achieved is not well understood. Here we will delineate the molecular mechanisms underlying changes in genome topology and gene activity at activity-regulated loci. To gain mechanistic insights, we will perturb key transcription factors (e.g. CTCF, c-Fos) and chromatin modifiers (e.g. Polycomb Group proteins (PcG, e.g. Eed), Lsd1, CBP) involved in activity-dependent gene expression, enhancer biology and various aspects of 3D genome organization using targeted protein degradation in in vitro differentiated excitatory neurons. We also aim to explore sex-specific activity-dependent gene expression. Sex bias in neurodevelopmental disorders is known to be linked to the X chromosome and several X-linked genes have been linked to neurodevelopmental disorders (including important chromatin factors such as the histone lysine demethylase Utx and DNA methylation binding protein MeCP2). In female mammals, one of the two X chromosomes is almost completely epigenetically silenced by a process termed X-chromosome inactivation (XCI), but several genes can escape XCI. Due to escape from XCI, the protein products of these genes could thus be present in different doses in female and male neurons. In general, very little is known about sex-specific modulation of gene regulation. We will thus also study how activity-dependent expression impacts on gene expression on the X chromosome and vice versa how dosage differences of X-encoded gene regulators impact on stimulus response. We will also perturb gene dosage of X-encoded factors in female cells to assess the contribution of dosage to neuronal gene regulation in the context of chromatin state and 3D genome organization. This will include important transcriptional regulators and epigenetic factors such as MeCP2 or factors involved in enhancer-promoter communication such as Med14 for which sex-specific disease phenotypes have already been reported.

Endogenous retroviruses in genome organization during early embryogenesis

The 3D organization of the genome plays essential roles in the control of gene expression programs, and many recent studies have focused on the contribution of nuclear proteins to 3D genome organization. Here we propose to investigate the roles of notoriously understudied and abundant repetitive DNA elements that originate from endogenous retroviruses (ERVs) in genome organization in pluripotent stem cells. ERVs are repressed by heterochromatin in stem cells, but are transiently transcribed during embryonic development, and in several disease contexts. Our preliminary data suggest that transcribed ERVs can engage in long-range chromatin contacts with transcribed genes, but the biological roles and importance of these contacts are unknown. In the proposed work, we will investigate the contribution of long-range contacts mediated by heterochromatin-adapter proteins to ERV repression in murine embryonic stem cells; and we will identify the components and molecular mechanisms utilized by transcribed ERVs to mediate long-range chromatin interactions with cellular genes. We will rationally mutagenize the TRIM28 heterochromatin-adapter protein that binds ERVs, and measure genomic interactions and ERV transcription genome-wide in embryonic stem cells. Furthermore, we will systematically perturb transcription factors that bind ERVs, perturb RNA species produced by ERVs, and determine the impact of the perturbations on genome organization and transcription. The results will provide new insights into the contributions of ERVs to 3D genome organization, and ultimately shed new light on the mechanisms transposable elements engage in shaping mammalian genomes on an evolutionary scale.

Characterization and testing of looping and non-looping enhancers

The spatial organization of the genome has been identified as a critical factor in controlling gene expression. How enhancers and promoters interact within the three-dimensional organization of chromatin is just beginning to be understood. Genome-wide assays revealed that mammalian genomes fold into distinct units of ~1Mb called topologically associated domains (TADs), in which enhancers and their target genes are located. However, to what degree 3D chromatin looping between enhancers and promoters within TADs is required for gene regulation is less clear. While some studies show a direct connection between enhancer-promoter looping and gene activity, other studies find less evidence. This discrepancies in part stem from methodological differences in analyzing 3D chromatin structure, in particular differences between sequencing-based (chromosome conformation capture – 4C-seq, HiC) or imaging-based (DNA-FISH) methods. Furthermore, the local influence of other regulatory elements within the TAD can impact the interpretation of enhancer-promoter looping. Despite these specific challenges, in analyzing high-resolution epigenomic and 3D-chromatin folding data of mouse and chicken embryonic hearts we found that indeed distinct enhancer classes exist: looping and non-looping enhancers. This poses the fundamental question how important 3D-chromatin looping is for enhancer-driven gene activity and if some enhancers rely more on 3D looping than others. However, current functional reporter assays are not well suited to systematically test the influence of 3D looping for enhancer function, since they typically position enhancer and reporter gene within a few kilobases at best. In this proposal, we aim to overcome these limitations and elucidate the role of long-range enhancer-promoter looping using a combination of bioinformatic analysis, genome-engineering, and microscopy. Starting from our cardiac enhancer dataset, we will first identify the genomic signatures that distinguish looping from non-looping enhancers. By adapting and expanding genome-engineering techniques, we will develop a novel long-range enhancer-reporter assay that is tailored to investigate long-range enhancer activity and can be combined with sequencing- and imaging-based analyses. We will combine the long-range enhancer assay with 4C-seq, DNA-FISH, and expression analysis, which will deliver complimentary readouts linking 3D-chromatin looping to gene activation. This setup will allow us to characterize the genetic features driving enhancer-promoter looping, establish a novel experimental platform for enhancer analysis, and directly compare the how chromatin looping relates to gene expression.

Understanding the interplay between transcription and cohesin-mediated loop extrusion at the single molecule level

The spatial structure of the genome is tightly connected to its function. While it is known that the genome organization influences gene expression, the impact of transcription on the genome organization remains to be understood. The process of loop extrusion is central to chromosome organization. In eukaryotes, the cohesin complex extrudes large chromatin loops by translocating along arms of the chromosome. During translocation, cohesins encounter the transcription machinery, which could interfere with loop extrusion. In this proposal, we aim to investigate the interplay between these two important active processes and determine whether and how transcription can modulate loop extrusion and vice versa. To this end we will employ in vitro single molecule imaging for the visualization of loop extrusion mediated by cohesin. Of particular interest is whether the transcription machinery acts as a ‘moving barrier’ to loop extruding cohesin and how the directionality of transcription and the topological state of transcribed DNA interfere with the kinetics of loop extrusion. The prospective findings will provide a molecular basis for understanding the functional relation between gene expression and loop extrusion. In the long term this knowledge will be crucial for understanding how loop extrusion actively regulates gene expression and how transcriptional activity shapes functional genomic organization.

Dissecting structural and functional cooperation between cis-regulatory elements in higher-order chromatin structures

The establishment of precise gene expression patterns during the development of multi-cellular organisms is controlled by the cisregulatory elements of the genome, which include gene promoters, enhancers, and boundary elements. In mammals, these elements can be separated by large genomic distances. To activate gene expression, spatial proximity between gene promoters and enhancers is facilitated within 3D genome structures, whose formation is dependent on the boundary elements. Many mammalian genes are controlled by multiple enhancers. However, it is not clear how these elements structurally and functionally cooperate with other cisregulatory elements to regulate gene expression. The aim of this project is to investigate when and how 3D genome structures form during cellular differentiation, how multiple cis-regulatory elements interact within these structures, and how this relates to gene activity levels. To achieve this, we will make use of a Chromosome Conformation Capture-based approach (“Tri-C”), which we have previously developed to enable high-resolution analysis of spatial interactions formed between multiple cis-regulatory elements in genomic regions of interest. Using this approach, we will examine the formation of 3D genome structures during cellular differentiation with high throughput. Furthermore, we will genetically delete enhancers and boundary elements in representative 3D structures and examine their functional interplay in the regulation of gene expression and the formation of genome structures. We anticipate that the proposed large-scale analysis of structural and functional properties of cisregulatory elements at high spatio-temporal resolution will contribute to our understanding of the complex relationship between genome structure and function.

Exploring the contribution of RNA polymerases to mammalian 3D genome architecture

Mammalian chromosomes are three-dimensional entities shaped by converging and opposing forces. Mitotic cell division induces drastic chromosome condensation, but following reentry into the G1 phase of the cell cycle, chromosomes reestablish their interphase organization. During the first funding round of the SPP2202, we tested the role of RNAPII in this transition, as well as in asynchronous G1 cells, by using a system allowing its auxin-mediated degradation. For the mitosis-to-G1 transition, in situ Hi-C coupled to super-resolution 3dSTORM imaging and computer simulations showed that RNAPII is required for both compartment and loop establishment upon mitotic exit. This is due to reduced and aberrant cohesin loading onto chromatin, which we can now show relies on the physical presence of RNAPII at accessible sites. Notably, the positions most affected are those bookmarked during mitosis by polymerase cofactors, which also show differential accessibility upon RNAPII depletion. In contrast, 3D folding of chromosomes in asynchronous G1-cells appeared less affected at the large scale. However, multiple new and larger CTCF /cohesin-anchored loops emerged in the absence of RNAPII. To mechanistically understand these effects, for this second funding round of the SPP2202, we will generate ultra-resolution Micro-C data and identify different scenarios affecting loop formation along chromosomes. We will combine Micro-C and super-resolution 3D-SIM imaging with epigenetic mark mapping and three new cell lines engineered to allow for the acute depletion of different factors in order to address the following questions: (1) How do loop-level changes in the 3D architecture of interphase chromatin arise in proliferating versus post-mitotic cells? (2) How does RNAPII orchestrate cohesin loading onto chromatin after mitosis? (3) Is there a role for bookmarking transcription factors in this process? In the end, we anticipate to obtain new insights into how the transcriptional apparatus acts to organize chromatin directly or indirectly. These rules of engagement would allow us to revisit the concept of transcription-based 3D chromatin organization, and thus reconcile the role of RNAPII in gene expression with that in chromosomal architecture.

Nuclear pore complex proteins-directed 3D nuclear architecture in neural development and disease

3D chromatin organization underlies lineage-specific gene expression and genome instability, which both are affected by nuclear structural proteins such as the nuclear pore complex. Recent studies and our pilot data indicate that nuclear pore complex proteins (Nups) directly or indirectly interact with chromatin and provide a structural scaffold for epigenetic regulators, transcription factors and DNA repair. Furthermore, Nup153, one of Nups was found to interact with the CTCF/cohesion complex to possibility organize topology associated domains (TAD). Our preliminary genomic analyses also indicate prominent roles for Nup153 in topological and directional gene regulation. This accumulating evidence suggests mechanisms of chromatin reorganization around pores to balance gene regulation and genome stability. Still, causal relationships among the accumulated damage on Nups, mechanisms behind Nups-directed 3D genome architecture, and its impact on regulation and genome instability remain largely elusive, especially in neural cells, where we expect high impact on disease development. In the proposed project, by combining interdisciplinary expertise, we will address, 1) how 3D genome organization at the nuclear pores is reorganized during neural differentiation, 2) to what extent disruption of Nups leads to disorganization of 3D chromatin organization, 3) what the spatial relationship of the balance between genome instability and Nups-directed 3D nuclear architecture is. Understanding spatio-temporal mechanisms underlying the organization of 3D genome-architecture at the nuclear pore is vital to unravel how cell type-specific epigenetic programs are maintained, while also preserve stability. Deregulation of these mechanisms can lead to dysfunctional neurodevelopment with cancer and neurodevelopmental diseases as potential consequences. To this end, using neural cells at different developmental stages as a model, we will employ multi-omics approaches to unravel relationships of multilayered epigenetic characteristics around nuclear pores. By integrating HiChIP, ChIP-seq, ATAC-seq, AP-seq, and END-seq, we will characterize changes of nuclear architecture at nuclear pores in 3D-chromatin interactions, binding of epigenetic regulators, chromatin accessibility, and DNA damage, upon differentiation and loss of Nup153. We will also interrogate the spatial relationship between identified changes in chromatin architecture and genomic mutations in cancer and neurodevelopmental disorders. The analyses will uncover the risk of genomic vulnerability associated with nuclear pores.

Deregulation of 3D genome structure in models of memory and learning disability

Chromatin regulators are often mutated in developmental disorders accompanied by neurological impairments. For example, intellectual disability is associated with mutations in chromatin proteins with diverse functions, such as CTCF, ATRX and SATB2. Depletion of either factor in post-mitotic pyramidal neurons, using conditional knockout (cKO) mouse models, results in eletrophysiological dysfunction and impairs hippocampal-related learning and long-term memory in fear-based experiments. However, the direct effects of CTCF, SATB2 and ATRX loss on chromatin regulation and the affected genes remain to be investigated.Here, we propose to dissect the consequences CTCF, ATRX or SATB2 loss in vivo, directly in pyramidal glutamatergic neurons of the mature murine hippocampus. We will use state-of-the-art methodologies, Genome Architecture Mapping (GAM), to map 3D chromatin structure, and 10x Genomics multiome, to profile both gene expression and chromatin accessibility in single cells. We will deliver an atlas of cKO-affected regulatory regions for all hippocampal cells, and the landscape of altered chromatin contacts. We will map the changes in chromatin structure genome-wide at different levels of 3D genome organization, and the transcription factors and chromatin regulatory pathways underlying the altered gene expression and 3D genome folding. We will focus on changes in chromatin structure and gene expression common to the three mutants, to identify candidate pathways involved in the learning and memory dysfunctions. We will validate genes affected in all cKOs, as these provide candidate targets with potential for intervention to ameliorate learning disabilities. Ultimately, we also aim to advance our knowledge of chromatin regulation in specialized post-mitotic neurons. Our hypothesis is that the inception of hippocampal dysfunction in learning disabilities is rooted in chromatin dysfunction mechanisms, and expect to develop a framework that can be adopted to accelerate the evaluation human patient samples.

Functional organization of co-regulated RNA polymerase II nuclear subcompartments activated by TNFα or TGFβ

During development or in response to external signals, the genome organizes into active or silenced chromatin subcompartments to implement cell type/state specific gene expression programs. How this segregation on the µm scale is established and relates to transcriptional activity is a long-standing question. Recent studies propose that the unmixing of soluble factors from the nucleoplasm into protein droplets around chromatin loci by a liquid-liquid phase separation (LLPS) drives the assembly of active “transcriptional condensates”. However, alternative models exist that are based on the chromatin binding of proteins and RNAs together with long- and short-range bridging interaction between these factors and looping of the intervening nucleosome chain. The project proposed here will investigate structure-function relationships underlying the assembly of endogenous promoters and enhancers into co-regulated active RNA polymerase II (Pol II) subcompartments. Transcriptional activation will be induced by stimulation of human cells with TNFα (signaling via the NFκB transcription factor) and TGFβ (signaling via SMAD2/3 transcription factors). These two cytokines cause acute and delayed induction of specific gene expression programs. In this cellular system the activation of co-regulated gene clusters has been established previously. Furthermore, p65/RelA as part of NFκB as well as SMAD3 have intrinsically disordered regions and can form liquid droplets by LLPS when ectopically expressed. Thus, stimulation with TNFα and TGFβ is ideally suited to assess a potential contribution of LLPS to the formation and/or activation of Pol II subcompartments. To dissect the underlying mechanisms, we will combine state-of-the-art single cell sequencing and fluorescence microscopy methods into a spatially resolved transcriptomics approach. By applying an integrative analysis of co-regulation, changes of RNA production, protein mobility and interactions and chromatin features we will reveal organization principles that underly the formation/activation of Pol II subcompartments upon TNFα and TGFβ induction. In this manner, we will assess the contribution of LLPS to this process and advance our understanding on how the dynamic nuclear organization of Pol II subcompartments is coupled to gene activation.

Architectural Rearrangements at the Xist locus during the onset of X-chromosome inactivation

Developmental genes are often regulated by long-range enhancer elements. To activate gene expression, these elements must come into close physical proximity of the gene promoter. Enhancer-promoter contacts mostly occur within topologically-associating domains (TADs), which are submegabase regions, where chromatin preferentially interacts. TAD positions are mostly invariant across cell types, but intra-TAD contacts often accompany changes in gene expression. Several mechanisms have been proposed to drive such intra-TAD rewiring, but how they cooperate to precisely tune gene expression remains poorly understood. In the proposed project, we will systematically test several alternative hypotheses of how intra-TAD rewiring is controlled and how these architectural changes shape transcriptional output. We will use the Xist locus as a model, which encodes the master regulator of X-chromosome inactivation. We will make use of a system in murine embryonic stem cells (TX XXΔXic/XO) we have recently developed, which allows high-resolution profiling of the Xist locus at the inactive X, where Xist is expressed, and at the active X, where Xist is silent. Using this cell model we have discovered structural rearrangements during initial Xist upregulation within the TAD that harbors the Xist promoter. In the proposed project, we will build on our extensive characterization of the TX XXΔXic/XO model, with respect to chromatin modifications, transcription and chromosome conformation. We will complement the existing data with factors that have been proposed to govern 3-dimensional genome structure and with high-resolution quantification of 3D contacts. We will then use computational modelling and experimental perturbations to test several alternative hypotheses of how 3D rewiring at the Xist locus might be controlled. Specifically, we will perform polymer simulations of cohesin-mediated loop extrusion, assuming (1) differential binding of CTCF, which can halt extrusion, (2) preferential cohesin loading at activated enhancers, or (3) lncRNA transcription as extrusion barrier. All three hypotheses will also be tested by (epi)genomic perturbations using the CRISPR/Cas9 system. In addition, we will use the high-resolution contact mapping data for unbiased identification of proteins bound at interacting genomic sites, to potentially develop additional hypothesis of how 3D rewiring might be regulated. Finally, we will assess the functional consequences of 3D rewiring, by quantifying transcriptional output in response to perturbations of 3D contacts. By combining a series of state-of-the-art experimental and computational approaches we will thus dissect how dynamic changes in genome architecture are regulated during an essential developmental process. The results will be relevant beyond the X inactivation field, as we will, for the first time, precisely dissect the relative contributions of different mechanisms to the dynamics of chromatin contacts at a specific locus.

Algorithms for inferring haplotype-specific chromatin contact maps in cancer using Genome Architecture Mapping

DNA and proteins together form chromatin, which is packaged in the nucleus in a highly structured way. This structure plays an important role in tissue development and in maintaining tissue integrity and is frequently disrupted in cancer and other diseases. Despite its importance, identifying changes in chromatin structure in cancer is challenging. Cancer genomes are highly unstable and evolve in the course of disease progression, rearranging their chromosomes and creating changes in chromosome copy-number. These copy-number changes confound the analysis of chromatin structure and to understand their interplay both high-resolution copy-number data and high-resolution chromatin conformation data are required. In addition, since healthy cells contain two parental copies of each chromosome (two haplotypes), it would be highly beneficial to be able to distinguish these haplotypes when analysing copy-number and chromatin structure.Genome Architecture Mapping (GAM) is a new experimental method for determining chromatin structure based on cryosectioning and sequencing of cell nuclei. In our previous work we demonstrated that during the sectioning process, GAM locally mostly captures one of the two haplotypes. This allowed us to devise an algorithm to assign genetic variants to the haplotype they originate from and to generate haplotype-specific chromatin contact maps, albeit only in genomes with a sufficiently large number of variants, such as mouse. Independently, on cancer sequencing data, we have developed phasing algorithms that allowed us to determine haplotype-specific copy-number profiles with very high accuracy.To be able to jointly infer chromatin structure and copy-number variants in cancer in a haplotype-specific manner we propose several algorithmic advances. We first propose to extend our algorithm GAMIBHEAR to be amenable to human sequencing data with lower variant densities. Using a “co-phasing” strategy, we will be able to assign sequencing reads that do not overlap heterozygous variant positions to their parental haplotype-of-origin. This will enable us to derive high-resolution haplotype-specific chromatin contact maps in human. In the next step, we will leverage our experience in developing phasing algorithms for copy-number variants to develop a novel algorithm for detecting haplotype-specific copy-number changes in GAM data. Finally, this algorithm will be extended to also include genomic rearrangements which are copy-number neutral, such as balanced translocations and inversions. This set of algorithms and tools will ultimately enable the routine application of GAM to clinical cancer specimen.

Revealing the function of heterochromatin spatial organization in response to early-life environmental challenges in C. elegans

The spatial distribution of transcriptionally active euchromatin and repressed heterochromatin is not random and enables a functional compartmentalization of the genome. In particular, silent heterochromatin is actively sequestered at the nuclear periphery, while active euchromatin is centrally located. Yet, the function of 3D chromatin distribution per se remains largely unknown, mainly due to the inability to selectively perturb chromatin spatial organization while leaving other nuclear processes unaltered. In this proposal, we will take advantage of the roundworm C. elegans, the only organism where a highly specific perinuclear anchor of heterochromatin, termed CEC-4, was identified to date, to impair heterochromatin spatial distribution globally and unravel the consequences for tissue integrity and organismal health. Despite a genome-wide impairment of spatial genome positioning, ablating cec-4 does not alter transcription globally and cec-4 mutants display no phenotype under optimal growth conditions unless challenged with an ectopic transcriptional stimulus. This raises the fascinating possibility that the spatial compartmentalization of chromatin is functional in coordinating the transcriptional response to non-programmed cues. Intriguingly, animals growing in their natural context are constantly exposed to environmental stresses, which induce profound gene expression changes. However, how environmental cues affects the spatial organization of the genome and whether this critically contributes to the stress response is currently unknown. Our preliminary data suggest that an accurate spatial segregation of euchromatin and heterochromatin is indeed functional when animals are exposed to environmental stress. In this proposal, we plan to characterize the functional interaction between the environment and 3D chromatin organization further. In particular, we will expose cec-4 mutants and wt animals to environmental stress and measure i) genome-nuclear lamina interactions, ii) chromatin accessibility and iii) gene expression, during stress and upon recovery in two different tissues, allowing us to determine how different cell types within an organism respond to the same environmental cue. Interestingly, in several species an early life stress exposure can influence events that occur later in life, with mechanisms that remain largely unknown. In this work, we will identify genes de-regulated during the stress response when heterochromatin spatial distribution is impaired and determine their impact on development and aging later in life, upon restoration of optimal growth conditions. To achieve our goals, my team will employ a combination of microscopy, molecular and genetic approaches. Because all organisms, including humans, are constantly exposed to a changing environment, the implications of this study are likely to be vast.

Spatial organization of transcribed genes in mammalian cells

The current grant proposal is a continuation of my previous research project within SPP2202, which was focused on spatial organization of transcribed eukaryotic genes. In the course of the past period, we studied several long highly expressed mouse genes as models and demonstrated that a transcribed gene expands from its harboring locus and forms an open-ended transcription loop (TL) with polymerases moving along the loop and carrying nascent RNAs. Remarkably, TLs can span across microns arguing against recent propositions that interphase chromatin is a gel or a solid. Extension and shape of TLs suggest their intrinsic stiffness, which we attribute to dense decoration of highly expressed genes with multiple voluminous nascent ribonucleoproteins (nRNPs). The stiffness hypothesis was successfully tested experimentally and by polymer modeling of expressed genes. In summary, our data contradict the popular model of transcription factories and suggest that although microscopically resolvable TLs are specific for long highly expressed genes, the mechanisms underlying their formation could represent a general aspect of eukaryotic transcription. The work is now published in the journal Nature Cell Biology (2022). Whereas the previous work brought a clear message about spatial organization of expressed genes, it raised new intriguing questions concerning eukaryotic transcription, which I am going to address in the new funding period. (1) The canonical view on splicing is that it occurs strictly co-transcriptionally with introns being excised shortly after they are read through. Using the Tg gene as a model, I plan to study rapidity of co-transcriptional splicing in case of massive transcription. (2) The previous observations of TLs were performed utilizing FISH in fixed tissues or cells. By targeting nRNAs of the Ttn and Cald1 genes via the stem-loop/coat protein system, I plan live-cell observations of spatiotemporal dynamics of TLs and transcriptional bursting. (3) Very little is known about structure of nRNPs and mRNPs owing to their small size. Based on our preliminary EM data, I plan to study nRNPs and 3D structure of TLs formed by Ttn in myotubes, using correlative microscopy (cryo-CLEM) and cryo-EM tomography. (4) The textbook knowledge that chromosome territories are the major feature of an interphase nucleus is not supported by our previous works indicating that territoriality is likely a mere consequence of the last mitosis. To address this question, I plan to evaluate territoriality based on oligopainting of mouse chromosomes in cells that vary by duration of their postmitotic period. (5) And finally, I plan to complete the study of the Tg gene as a model for high and robust upregulation of transcription. In particular, I plan to test Tg circadian rhythmicity and intron retention in Tg transcripts as possible mechanisms for separation of exocrine (thyroglobulin secretion) and endocrine (hormone production) activities in thyrocytes.

Position effects in the 3D genome as the cause of neurodevelopmental disorders

The discovery of topologically associating domains (TADs) and our increased understanding of long-range regulation have allowed us to better understand the mechanisms underlying “position effects. Structural variants (SVs) have the potential to disrupt higher order chromatin organization thereby rewiring the complex three-dimensional chromatin organization of the locus. SVs that disrupt TAD boundaries and promoter-enhancer interactions are relatively common in skeletal malformations and certain tumors. However, the role of SVs changing the 3D genome architecture in neurodevelopmental disorders (NDDs) is less understood. In the first funding period we have successfully shown that SVs cause position effects in the 3D genome in NDDs and that HiC can effectively be used to identify SVs in patient cells. In the second funding period we now aim to functionally characterize specific deletions at the Zfp608/Lmnb1 locus and an insertion at the Sox3 locus to decipher the molecular drivers of position effects causing NDDs. In Aim 1 we will decipher the molecular drivers of 3D position effects at the Zfp608/Lmnb1 locus associated with autosomal dominant demyelinating leukodystrophy (ADLD). In Aim 2 we will dissect the molecular mechanism of an insertion at the SOX3 locus associated with spastic paraplegia and reduced SOX3 expression. And finally in Aim 3 we will map the cell type specific activity of enhancer elements located in the regulatory landscapes of Zfp608/Lmnb1 in vivo at single-cell resolution. Our data will contribute to a better understanding of the molecular biology of position effects in patients with NDDs.

Coordination and function of nuclear lamina and nuclear pore compartmentalisation in genome organisation during early mouse development

The packaging of the chromosomes in the cell nucleus while retaining fundamental DNA-based activities such as transcription and replication represents an enormous organisation challenge. Indeed, the way in which the DNA is folder into the 3 dimensions of the cell nucleus can affect gene expression and thus cellular identity. How nuclear organisation is first established after fertilisation in mammals is largely unknown. Following fertilisation, a period of intense chromatin remodeling and epigenetic reprogramming of the two parental genomes occurs. Such remodeling is necessary to start a developmental programme capable of forming a new organism. Several nuclear landmarks can provide positional cues to the genome and thus are key contributors to genome organization. Amongst them, the nucleolus, the nuclear lamina and speckles organise the chromatin into functionally relevant domains and have been extensively studied. In addition, a role for the nuclear pore in nuclear organization has been suggested and studies in somatic cells in culture have revealed a role for components of the nuclear pore in gene regulation. However, how the nuclear pores first assemble at the beginning of development or the components that characterise embryonic nuclear pores are largely unknown. In addition, the genomic regions that interact with the nuclear pore as the embryo develops and whether such association is functionally relevant for key developmental landmarks such as zygotic genome activation and cell fate allocation has not been investigated. This project thus aims to i) characterize the components and cellular architecture of the nuclear pore at the beginning of mammalian development, ii) to map genome interactions with nuclear pore components during reprogramming as development proceeds, and iii) to determine the functional role of nuclear pore components in the establishment of genome topology, de novo in mammalian development. Using a combination of cell and molecular biology as well as perturbations through experimental embryology in mouse embryos, we will provide a comprehensive functional characterization of the nuclear pore and its components during early development. Altogether, our data will shed light into the organization and the molecular determinants of nuclear architecture during development.

Modeling embryonal neuroblastoma tumorigenesis by activation of chromosomal 3D super enhancer interactions and genomic instability

Human embryonal neuroblastoma is a paradigmatic tumor that is thought to develop as a consequence of recurrent structural chromosomal alterations (e.g. translocations, gene amplifications) leading to oncogene activation. Enhancer hijacking events, where strong regulatory regions (super enhancers, SEs) controlling cell identity genes are brought into close proximity of oncogenes (e.g., MYCN, MYC, TERT), are recurrent features defining highly aggressive neuroblastoma subtypes. Recent studies on neuroblastoma evolution indicate that these tumors arise during embryogenesis in a neuroblast population that usually gives rise to adrenal medullary cells. Here, we will test the hypothesis that states of extensive 3D chromosomal changes along the normal developmental trajectory (e.g. changes in cis/trans interaction ratio) in the cell(s) of origin predispose to structural chromosomal alterations in the context of genomic instability. We will define physiological time-resolved epigenetic states and 3D interactions during normal adrenal gland development and functionally recapitulate neuroblastoma tumorigenesis in neuroblasts derived from hiPSCs by experimentally inducing neuroblast-specific 3D interactions and genomic instability. In summary, our studies will develop a new molecular framework for embryonal tumorigenesis that relies on naturally occurring structural rearrangements and that will be the foundation for more accurate embryonal tumor models.