Periodic Reporting for period 3 - ChromatinLEGO (Chromatin readout: Dissecting the protein-chromatin interaction code in living cells)
Reporting period: 2023-05-01 to 2024-10-31
Eukaryotic genomes are packaged into highly ordered structures, consisting of nucleosomes as the basic unit of chromatin. Both, DNA and histones are decorated by various chemical modifications that positively or negatively influence biological processes such as transcription, genome organisation, DNA repair or DNA replication. These modifications are deposited and removed by “writer” and “eraser” proteins in a highly regulated fashion. Furthermore, regulatory proteins can read single or multiple modification, enabling the precise execution of various biological processes at defined genomic regions based on spatiotemporal readout of chromatin marks (Figure 1). How chromatin modifications mediate specific recruitment of regulatory factors to the genome based on specific readout is not fully understood, remaining an important and central challenge for the field of chromatin biology and gene regulation.
Detailed knowledge about the underlying interaction logic between regulatory proteins and chromatin is essential to understand how chromatin modifications influence important regulatory events that occur in the nucleus, such as gene transcription, DNA replication, DNA repair etc. Furthermore, alterations in protein-chromatin interactions, either caused by aberrant genomic distribution of chromatin marks, or mutations in chromatin reader proteins are frequently observed in numerous human diseases. Therefore, a detailed understanding about the mechanisms at work would enable to uncover novel disease-related mechanisms and reveal potential therapeutic targets.
We propose a systematic approach where protein-chromatin interactions are dissected into the individual building blocks that mediate specificity: the chromatin reader domains of proteins and their corresponding chromatin modifications. This allows us to reduce the interaction complexity obtained from studying entire proteins and complexes, which so far has convoluted the underlying principles of protein-chromatin interactions. We want to measure specificities of these individual building blocks in a quantitative and comprehensive manner and reveal their individual contribution to more complex interactions observed in nature. In particular we will 1) systematically explore how chromatin reader domains and their synthetic combinations guide protein localisation along the genome by using engineered Chromatin Readers (eCRs), and 2) make use of their selectivity to generate synthetic proteins that enable us to study the local protein composition at chromatin states in living cells based on proximity-biotin ligation using ChromID.
Detailed knowledge about the underlying interaction logic between regulatory proteins and chromatin is essential to understand how chromatin modifications influence important regulatory events that occur in the nucleus, such as gene transcription, DNA replication, DNA repair etc. Furthermore, alterations in protein-chromatin interactions, either caused by aberrant genomic distribution of chromatin marks, or mutations in chromatin reader proteins are frequently observed in numerous human diseases. Therefore, a detailed understanding about the mechanisms at work would enable to uncover novel disease-related mechanisms and reveal potential therapeutic targets.
We propose a systematic approach where protein-chromatin interactions are dissected into the individual building blocks that mediate specificity: the chromatin reader domains of proteins and their corresponding chromatin modifications. This allows us to reduce the interaction complexity obtained from studying entire proteins and complexes, which so far has convoluted the underlying principles of protein-chromatin interactions. We want to measure specificities of these individual building blocks in a quantitative and comprehensive manner and reveal their individual contribution to more complex interactions observed in nature. In particular we will 1) systematically explore how chromatin reader domains and their synthetic combinations guide protein localisation along the genome by using engineered Chromatin Readers (eCRs), and 2) make use of their selectivity to generate synthetic proteins that enable us to study the local protein composition at chromatin states in living cells based on proximity-biotin ligation using ChromID.
We have established workflows for streamlined generation and validation of plasmids and their targeted integration to a defined position in the mouse stem cell genome. These processes allowed us to generate so far over >250 different constructs containing eCRs based on individual and combinatorial reader domains from numerous proteins covering over thirty different chromatin reader domain families. More than 220 mouse stem cell lines were generated and validated to stably express eCRs, either as GFP fusion for live imaging and ChIP-seq, or with TurboID fusions for biotin-based proximity proteomics via ChromID (Villasenor, Nat Biotech 2020 - Figure 2). This setup already allowed us to identify and characterise interesting chromatin readers that dynamically localise in the nucleus in a cell type specific manner or upon specific conditions (i.e. DNA damage, cell cycle phase, DNA methylation inhibition), generating interesting hypotheses that we are currently following up. So far, we have generated over 100 genome-wide maps for different eCRs, reflecting the genomic localisation of the isolated chromatin reader domains in mouse stem cells to identify their genomic binding locations. Furthermore, we have generated more than forty different ChromID datasets based on domains specific for histone-methylation, -phosphorylation, -acetylation, and DNA-methylation, indicating the suitability of this proximity biotinylation technique to detect proteins associated with these chemical modifications (Figure 2).
We have also successfully applied eCRs and ChromID to investigate binding of ZFP57, a known DNA methylation binder at imprinted genes, to identify its genomic localisation in wild type ES cells and ES cells lacking DNA methylation. Finally, we performed ChromID experiments to identify regulatory proteins associated with methylated regulatory sites of imprinted genes using a ZFP57-TurboID fusion. This revealed known and novel factors involved in regulation of genomic imprinting, which we further identified using CRISPR screens as an orthogonal approach, and characterised for their molecular function (Butz, Nat. Genetics 2022).
We furthermore collaborated with the laboratory of Sachdev Sidhu to enhance the affinities of Cbx chromodomains for methylated histone marks by mutating key residues and using phage display. We tested a novel "super-binder" that has a higher affinity for H3K27me3 and can be used in combination with CRISPRi to silence target genes (Veggiani, Nat Communications 2022).
In more recent work, we used ChromID to profile the proteome associated with elongating RNA Pol2 in presence and absence of the H3K36me3 writer SETD2 (Ambrosi et al, biorxiv 2024) and we have developed a DNA damage reader eCR that can be used for live tracking of DNA damage and for the identification of proteins associated with DNA damage sites (Cardoso da Silva et alum biorxiv 2024).
We have also successfully applied eCRs and ChromID to investigate binding of ZFP57, a known DNA methylation binder at imprinted genes, to identify its genomic localisation in wild type ES cells and ES cells lacking DNA methylation. Finally, we performed ChromID experiments to identify regulatory proteins associated with methylated regulatory sites of imprinted genes using a ZFP57-TurboID fusion. This revealed known and novel factors involved in regulation of genomic imprinting, which we further identified using CRISPR screens as an orthogonal approach, and characterised for their molecular function (Butz, Nat. Genetics 2022).
We furthermore collaborated with the laboratory of Sachdev Sidhu to enhance the affinities of Cbx chromodomains for methylated histone marks by mutating key residues and using phage display. We tested a novel "super-binder" that has a higher affinity for H3K27me3 and can be used in combination with CRISPRi to silence target genes (Veggiani, Nat Communications 2022).
In more recent work, we used ChromID to profile the proteome associated with elongating RNA Pol2 in presence and absence of the H3K36me3 writer SETD2 (Ambrosi et al, biorxiv 2024) and we have developed a DNA damage reader eCR that can be used for live tracking of DNA damage and for the identification of proteins associated with DNA damage sites (Cardoso da Silva et alum biorxiv 2024).
We expect to provide a comprehensive and quantitative dataset on chromatin-reader domain binding preferences obtained from using a systematic setup and covering over 300 readers. This unique framework will not only reveal the affinities and site-site-specific localisation of these domains in vivo but will also provide a framework for computational studies aimed to compare specificities between and within domain families and even explore the evolution of specific chromatin binders. These results will enable us to understand how each of these reader domains function on their own, and then, based on this information, to understand how they cooperate to specify genome-wide targeting of multi-domain proteins.
Furthermore, the specific binding properties of these chromatin readers will enable us to address a current challenge in chromatin biology: How do chromatin marks influence the local proteome at defined chromatin states along the genome. By using specific readers we will guide promiscuous biotin ligases to genomic regions and nuclear compartments defined by specific chromatin states. This method, which we termed ChromID (Figure 2), will help us to obtain a detailed and complete view on the proteins associated with defined chromatin states in stem cells. We expect to identify numerous interesting protein candidates, which we will further characterise using functional assays.
Furthermore, the specific binding properties of these chromatin readers will enable us to address a current challenge in chromatin biology: How do chromatin marks influence the local proteome at defined chromatin states along the genome. By using specific readers we will guide promiscuous biotin ligases to genomic regions and nuclear compartments defined by specific chromatin states. This method, which we termed ChromID (Figure 2), will help us to obtain a detailed and complete view on the proteins associated with defined chromatin states in stem cells. We expect to identify numerous interesting protein candidates, which we will further characterise using functional assays.