Periodic Reporting for period 4 - 3DCellPhase- (In situ Structural Analysis of Molecular Crowding and Phase Separation)
Reporting period: 2022-08-01 to 2023-12-31
Summary of the context and overall objectives of the project
Cells are not just salty bags of freely diffusing macromolecules; eukaryotic cells in particular achieved a high level of spatial organization by restricting specific biochemical processes to membrane-bound organelles. A different strategy for compartmentalization, somewhat rediscovered over the past decade, involves the selective partitioning of molecules into cellular condensates. These membrane-less dynamic compartments are found throughout the cytoplasm and nucleoplasm and are involved in a wide range of fundamental cellular processes – from gene expression regulation to metabolism, and stress adaptation. Many of these assemblies behave like condensed liquids; they are highly dynamic and the components within them are in constant exchange with the surrounding cytoplasm or nucleoplasm. Examples are RNA-protein bodies such as stress granules. Others, such as centrosomes have more the appearance of gels. Significant progress on reconstituting such assemblies in vitro has been made over the last decade, and we have contributed to landmark collaborative studies to elucidate their pleomorphic fine structures. However, we still largely lacked information as to whether these compartments in vivo are amorphous liquids, whether they rely on structural scaffolds, or whether both types can coexist. The project 3DCellPhase- aimed at elucidating the molecular organization of phase-separated compartments and of their local cellular environment. Technological advances in sample preparation of cells for cryo-electron tomography (cryo-ET), including cryo-focused ion beam (cryo-FIB) thinning guided by 3D cryo-correlative fluorescence microscopy, enable the preparation of site-specific ‘electron-transparent windows’ in suitable eukaryotic systems, which allow direct examination of fine structural features of cellular compartments in their cellular context. We utilized these techniques to study the structural organization of two very different phase-separated compartments: Stress granules, which are RNA bodies that form rapidly in the cytoplasm upon cellular stress, and centrosomes, which are sites of microtubule nucleation central to cell division. We thereby aimed to derive whether common molecular architectures underly cellular condensates in vivo. We combine these targeted structural studies with computational methods to locally quantify the crowded nature of the cytoplasm and to understand the impact of excluded volume on phase separation in living cells. Our studies provide fundamental insights into the structural basis by which cells dynamically form biochemically distinct compartments.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
3DCellPhase- aimed at developing state-of-the-art cryo-electron tomography (cryo-ET) techniques to unravel, directly inside intact cells, the structural basis of membrane-less compartments formed by liquid-liquid phase separation (LLPS) and to provide a quantitative description of the local organization of the cytoplasm. Three major objectives were defined in the Description of Action: 1: Structural characterization of stress granules, which are RNA bodies that form rapidly in the cytoplasm upon cellular stress, in HeLa cells; 2: Supramolecular and functional architecture of centrosomes, which are sites of microtubule nucleation central to cell division, in C. elegans embryos; 3: Quantitative and structural description of molecular crowding and cytoplasmic reorganization of yeast cells in response to stress. Each eukaryotic model was selected such that it provided the highest level of control and a priori knowledge at the cell-biology, proteomics, genetics and biochemistry levels.
Each of the ambitious research objectives demanded significant technological advance. The need to use three different cellular and organismal models to tackle three distinctly different phase-separation systems in the context of the ERC project allowed us to broadly tackle both experimental and computational bottlenecks in our workflows. We pushed the boundaries of possible applications, achievable resolution and information depth, and improved their robustness to make our methods more widely available for the Life Science research community at large. Specifically, together with outstanding collaborators, we have contributed to the development of tailored grids by photo-micropatterning for in-cell structural studies (Toro-Nahuelpan et al. 2020), software-controlled automation of cryo- focused ion beam (FIB) preparations (Klumpe et al. 2021), and deep-learning-based mining of cryo-ET data (de Teresa et al. 2023). We have provided the first successful demonstration of the possibility to image macromolecular complexes inside cells at close to atomic resolution (Tegunov et al. 2021). These advanced were shared with the community early on through lectures in conferences and courses, preprints, open access publications, open source software and public datasets where applicable.
For all objectives, our unique cryo-ET data revealed for the first time an outstanding structural diversity underlying biomolecular condensates in cells. Complementary experimental approaches provided valuable insights that linked the main structural work to functional cell biology:
Objective 1: Stress granules (SGs) are well-established models for liquid condensates in human cells. Using HeLa lines stably expressing fluorescently-tagged SG proteins, we established the required cryo-ET workflows, from cell vitrification after controlled induction of stress, 3D cryo-confocal fluorescence-based targeted thinning using focused ion beam (FIB), and cryo-ET on correlated locations. Cryo-ET under a variety of stress conditions showed the presence of SG-associated helical filaments. We however had a surprising finding that the helical assemblies were viral nucleocapsids, which prompted an exciting new research direction: negative-stranded RNA viruses, including mumps virus that we identified in our HeLa cells, can establish long-term persistent infection in the form of large intracellular inclusions in the human host and cause chronic diseases. Here we uncovered how cellular stress disrupts the delicate host-virus equilibrium in persistent infection and induces viral replication. Using a combination of cell biology, whole-cell proteomics and cryo-ET, we showed that persistent viral replication factories are dynamic condensates and identified the largely disordered viral phosphoprotein as a driver of their assembly. Upon stress, increased phosphorylation of the phosphoprotein at its interaction interface with the viral polymerase coincides with the formation of a stable replication complex. Atomic models for the authentic mumps virus nucleocapsid elucidated a concomitant conformational change that exposes the viral genome to its replication machinery. These events constitute a stress-mediated switch within viral condensates that establish an environment to support up-regulation of viral replication (Zhang et al. Cell 2023).
Objective 2: Centrosomes consist of barrel-shaped centrioles surrounded by a mass of Pericentriolar Material (PCM). Central to this project, the PCM is known to evolve during the cell cycle from a fast growing structure potentially driven by LLPS at the onset of mitosis, to a gel-like scaffold that withstands microtubule pulling forces during metaphase. Utilizing disassociation of C. elegans embryonic cells from a line fluorescently tagged on the main structural scaffold of the C. elegans PCM, and on histone 2B, facilitated cryo-ET to visualize the native architecture of centrosomes throughout the cell cycle; we describe a pseudo-timeline of centriole assembly and identify novel structural features. We find that centriole and PCM microtubules differ in protofilament numbers (13 versus 11) indicating distinct nucleation mechanisms, that could be explained by atypical γ-tubulin ring complexes with 11-fold symmetry identified in the PCM. We further provide a first geometrical characterization of the porous and disordered network that forms the interconnected PCM. Thus, our work builds a 3D structural atlas that helps explain how centrosomes assemble, grow, and achieve function (Tollervey et al.).
Objective 3: Phase separation is extremely sensitive to changing environmental conditions, and the degree of cytoplasmic crowding leading to excluded volume effects. We aimed to obtain a structural and quantitative description of molecular crowding from cryo-ET. Due to their small size, yeast allow efficient acquisition of large and statistically significant data. Furthermore, under stress conditions, the entirety of the yeast cytoplasm undergoes a global transition from a liquid to a solid state. Our technology developments described in section 1.2 enabled acquisition and analysis of over 500 high-quality label-free cryo-ET datasets, from two evolutionary distant yeast species, under varying stress conditions. We thus visualized mesoscale assembly of macromolecular complexes into ordered structures and condensates under stress, and quantified ribosome and polysomes concentrations as a proxy for molecular crowding under the different conditions. Combined with live cell confocal microscopy, we pinpoint supramolecular assembly driven by cytoplasmic acidification, rather than a uniform increase in crowding, as a key mechanism in the yeast solidification response to environmental stress (Goetz, Spindler et al.).
We have shared our learned perspectives on the study of biomolecular condensates at molecular resolution inside cells in two well-received review articles (Zhang & Mahamid, Current Opinions in Structural Biology 2020; Goetz & Mahamid, Developmental Cell 2020).
Each of the ambitious research objectives demanded significant technological advance. The need to use three different cellular and organismal models to tackle three distinctly different phase-separation systems in the context of the ERC project allowed us to broadly tackle both experimental and computational bottlenecks in our workflows. We pushed the boundaries of possible applications, achievable resolution and information depth, and improved their robustness to make our methods more widely available for the Life Science research community at large. Specifically, together with outstanding collaborators, we have contributed to the development of tailored grids by photo-micropatterning for in-cell structural studies (Toro-Nahuelpan et al. 2020), software-controlled automation of cryo- focused ion beam (FIB) preparations (Klumpe et al. 2021), and deep-learning-based mining of cryo-ET data (de Teresa et al. 2023). We have provided the first successful demonstration of the possibility to image macromolecular complexes inside cells at close to atomic resolution (Tegunov et al. 2021). These advanced were shared with the community early on through lectures in conferences and courses, preprints, open access publications, open source software and public datasets where applicable.
For all objectives, our unique cryo-ET data revealed for the first time an outstanding structural diversity underlying biomolecular condensates in cells. Complementary experimental approaches provided valuable insights that linked the main structural work to functional cell biology:
Objective 1: Stress granules (SGs) are well-established models for liquid condensates in human cells. Using HeLa lines stably expressing fluorescently-tagged SG proteins, we established the required cryo-ET workflows, from cell vitrification after controlled induction of stress, 3D cryo-confocal fluorescence-based targeted thinning using focused ion beam (FIB), and cryo-ET on correlated locations. Cryo-ET under a variety of stress conditions showed the presence of SG-associated helical filaments. We however had a surprising finding that the helical assemblies were viral nucleocapsids, which prompted an exciting new research direction: negative-stranded RNA viruses, including mumps virus that we identified in our HeLa cells, can establish long-term persistent infection in the form of large intracellular inclusions in the human host and cause chronic diseases. Here we uncovered how cellular stress disrupts the delicate host-virus equilibrium in persistent infection and induces viral replication. Using a combination of cell biology, whole-cell proteomics and cryo-ET, we showed that persistent viral replication factories are dynamic condensates and identified the largely disordered viral phosphoprotein as a driver of their assembly. Upon stress, increased phosphorylation of the phosphoprotein at its interaction interface with the viral polymerase coincides with the formation of a stable replication complex. Atomic models for the authentic mumps virus nucleocapsid elucidated a concomitant conformational change that exposes the viral genome to its replication machinery. These events constitute a stress-mediated switch within viral condensates that establish an environment to support up-regulation of viral replication (Zhang et al. Cell 2023).
Objective 2: Centrosomes consist of barrel-shaped centrioles surrounded by a mass of Pericentriolar Material (PCM). Central to this project, the PCM is known to evolve during the cell cycle from a fast growing structure potentially driven by LLPS at the onset of mitosis, to a gel-like scaffold that withstands microtubule pulling forces during metaphase. Utilizing disassociation of C. elegans embryonic cells from a line fluorescently tagged on the main structural scaffold of the C. elegans PCM, and on histone 2B, facilitated cryo-ET to visualize the native architecture of centrosomes throughout the cell cycle; we describe a pseudo-timeline of centriole assembly and identify novel structural features. We find that centriole and PCM microtubules differ in protofilament numbers (13 versus 11) indicating distinct nucleation mechanisms, that could be explained by atypical γ-tubulin ring complexes with 11-fold symmetry identified in the PCM. We further provide a first geometrical characterization of the porous and disordered network that forms the interconnected PCM. Thus, our work builds a 3D structural atlas that helps explain how centrosomes assemble, grow, and achieve function (Tollervey et al.).
Objective 3: Phase separation is extremely sensitive to changing environmental conditions, and the degree of cytoplasmic crowding leading to excluded volume effects. We aimed to obtain a structural and quantitative description of molecular crowding from cryo-ET. Due to their small size, yeast allow efficient acquisition of large and statistically significant data. Furthermore, under stress conditions, the entirety of the yeast cytoplasm undergoes a global transition from a liquid to a solid state. Our technology developments described in section 1.2 enabled acquisition and analysis of over 500 high-quality label-free cryo-ET datasets, from two evolutionary distant yeast species, under varying stress conditions. We thus visualized mesoscale assembly of macromolecular complexes into ordered structures and condensates under stress, and quantified ribosome and polysomes concentrations as a proxy for molecular crowding under the different conditions. Combined with live cell confocal microscopy, we pinpoint supramolecular assembly driven by cytoplasmic acidification, rather than a uniform increase in crowding, as a key mechanism in the yeast solidification response to environmental stress (Goetz, Spindler et al.).
We have shared our learned perspectives on the study of biomolecular condensates at molecular resolution inside cells in two well-received review articles (Zhang & Mahamid, Current Opinions in Structural Biology 2020; Goetz & Mahamid, Developmental Cell 2020).
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
Through 3DCellPhase-, we have significantly expanded the scope of cellular cryo-ET from examining singular macromolecular species in situ towards the realm of a comprehensive description of the organization of the cytoplasm in different eukaryotic systems and at high resolution, thus empowering the emergence of Structural Cell Biology as an interdisciplinary field in the Life Sciences. This was enabled by our development of methods that propelled cryo-ET beyond the state of the art, tackling bottlenecks from specimen preparation to data analysis (Toro-Nahuelpan et al. 2020, Klumpe et al. 2021, de Teresa et al. 2023).
The project provided unique opportunities for integration with different ‘omics’ approaches, most significantly proteomics, thus building towards Systems Structural Biology (Zhang et al. Cell 2023). We have thus been able to contribute to the development of a powerful combination of in-cell methods (including live-cell and quantitative fluorescence light microscopy imaging, cellular fractionations and quantitative mass-spectrometry, structure prediction and modeling) to bridge across different length scales, that have allowed us to provide unprecedented insights into the structural and architectural complexity of biomolecular condensates across different models (detailed below). I expect these technologies and concepts to contribute to the emergence of novel understanding in both fundamental and translational research.
Our research objectives touch on a number of different fields in biology (cellular stress response, infection biology, and cell division, among others) beyond that of biomolecular condensation and functional phase separation, and the unique data and analysis produced in the ERC project provides new insights into all these different topics. As an example, while completely unexpected, our initially unfortunate discovery of a mumps virus infection of the HeLa cell cultures provided the first evidence that cellular stress provokes activation of viral replication in an RNA virus model. The combination of cutting-edge methods allowed a detailed mechanistic understanding that viral replication factories are liquid-like condensates that change behavior under stress, that stress-triggered phosphorylation of an intrinsically disordered viral protein domain by host factors stabilizes the viral replication machinery, and that in response, the viral nucleocapsid structures change to expose the genomic RNA and support increased viral replication. Similar outcomes have resulted from our work on C. elegans centrosomes, that beyond their ability to illuminate the PCM architecture for the first time, also contribute new understanding on the highly conserved centriole structures, their assembly intermediates, and molecular mechanisms of regulation of microtubule nucleation. Our work on the structural reorganization of the yeast cytoplasm in response to different environmental stressors similarly sparked unplanned and exciting exploration of metabolic complexes, their functional states and interaction partners, in relation to their regulation by assembly in microscopic condensed structures. In summary, while the project delivered on its initial objectives to reveal molecular architectures and structural basis of different types of biomolecular condensates, the label-free, minimally perturbing and high-resolution nature of cryo-ET has enabled novel understanding in different fields of biology.
The project provided unique opportunities for integration with different ‘omics’ approaches, most significantly proteomics, thus building towards Systems Structural Biology (Zhang et al. Cell 2023). We have thus been able to contribute to the development of a powerful combination of in-cell methods (including live-cell and quantitative fluorescence light microscopy imaging, cellular fractionations and quantitative mass-spectrometry, structure prediction and modeling) to bridge across different length scales, that have allowed us to provide unprecedented insights into the structural and architectural complexity of biomolecular condensates across different models (detailed below). I expect these technologies and concepts to contribute to the emergence of novel understanding in both fundamental and translational research.
Our research objectives touch on a number of different fields in biology (cellular stress response, infection biology, and cell division, among others) beyond that of biomolecular condensation and functional phase separation, and the unique data and analysis produced in the ERC project provides new insights into all these different topics. As an example, while completely unexpected, our initially unfortunate discovery of a mumps virus infection of the HeLa cell cultures provided the first evidence that cellular stress provokes activation of viral replication in an RNA virus model. The combination of cutting-edge methods allowed a detailed mechanistic understanding that viral replication factories are liquid-like condensates that change behavior under stress, that stress-triggered phosphorylation of an intrinsically disordered viral protein domain by host factors stabilizes the viral replication machinery, and that in response, the viral nucleocapsid structures change to expose the genomic RNA and support increased viral replication. Similar outcomes have resulted from our work on C. elegans centrosomes, that beyond their ability to illuminate the PCM architecture for the first time, also contribute new understanding on the highly conserved centriole structures, their assembly intermediates, and molecular mechanisms of regulation of microtubule nucleation. Our work on the structural reorganization of the yeast cytoplasm in response to different environmental stressors similarly sparked unplanned and exciting exploration of metabolic complexes, their functional states and interaction partners, in relation to their regulation by assembly in microscopic condensed structures. In summary, while the project delivered on its initial objectives to reveal molecular architectures and structural basis of different types of biomolecular condensates, the label-free, minimally perturbing and high-resolution nature of cryo-ET has enabled novel understanding in different fields of biology.