Matter to Life Fall Days

Studies have shown that liquid-liquid phase separation facilitates the formation of membraneless organelles by compartmentalising biochemical reactions. Complex coacervate droplets regulated by a fuel-driven chemical reaction cycle serve as a model for these membraneless organelles. These reaction cycles provide a tool to drive the system out of equilibrium and generate dissipative structures. Bergmann et al. 2023 reported that continuous fueling of the cycle yields in a morphological change of the dissipative structures; the active droplets transition into a liquid, spherical shell. This work attempts to understand; (1) the pathway of this morphological transition using particle based simulations and soft-coarse grained model for polymer solutions, (2) and the physical and chemical parameters affecting this transition.

Synthetic antimicrobial peptides Api137 and Api88 are promising lead compounds in antimicrobial research. They share the same sequence, apart from the C-terminus, which is amidated in Api88. Although they both suppress protein synthesis by binding to the ribosome's peptide exit tunnel (PET), recent cryo-EM data suggest that they act by different mechanisms. Api137 traps release factors and prevents dissociation of the peptide chain. Api88 binds to the same sites as Api137 but is more deeply positioned into the tunnel, and the cryo-EM density is less defined. The less defined cryo-EM density of Api88 can be tentatively modelled by three different conformations. Our results from all-atom MD simulations, starting from the three modelled conformations of Api88 in the PET, suggest that Api88 adopts metastable conformational states. To estimate how much the metastable states contribute to the overall conformational ensemble, we computed cryo-EM density maps from the MD ensembles of each state. After that, we combined the computed maps and optimised the weights of the states to maximise the correlation with the experimental map. Interestingly, the optimally weighted MD ensemble displayed an improved correlation to the experimental map compared to the optimally weighted initial conformations. Additionally, we identified the minimal number of structures sufficient to describe the ensemble. Our analysis further shows that one of the three initially modelled conformations contributes most to the cryo-EM density. It shows a slight shift toward the peptidyl transferase centre at the tunnel’s beginning compared to the other two structures.

Cellular resource allocation and efficient utilization are paramount for the survival and functionality of living organisms. Central to this process is the regulation of gene expression in response to key metabolites, a fundamental strategy employed by the living cells. However, the current bottom-up assembled in vitro metabolic networks lacks such abilities to self-regulate and regenerate. In this study, we embark on the construction of an in vitro self-regulated metabolic and genetic linked network (MGLN), a synthetic system with the remarkable capability of autonomous "decision-making." Specifically, our MGLN possesses the capacity to activate its genetic layer in response to the production of a target metabolite, which in turn activates the anaplerotic module to regenerate pathway intermediates, effectively mimicking the adaptive behaviors seen in living organisms. We demonstrate this concept by regulating gene expression with glycolate produced from CO2, and glycolate-inducible transcription factor GlcR from Paracoccus denitrificans. We are currently further optimizing such MGLN by using a machine-learning based algorithm (METIS), recently developed in our lab.

Ina Braun, Alexander Ecker, Yunxiao Zhang, Eberhard Bodenschatz, Yong Wong Heart attacks are one of the leading causes of death in developed countries. They damage the heart muscle, leading to scarring and heart failure. Currently, the only routine treatment for heart failure is a heart transplant. However, due to the shortage of donor hearts, other methods of therapy are being developed. One such method is engineered myocardial muscle (EHM) patches, which are implanted onto the damaged areas of the heart to regenerate a failing heart. In order to routinely use EHM patches in a clinical setting, the optimal param- eters for each patient must be calculated. This can be achieved using a fast and computationally inexpensive simulation, such as physics informed neural networks (PINN). The long-term goal of this project is to develop a PINN that predicts the deformation of the damaged heart with an EHM patch throughout the cardiac cycle. This will allow the determination of the optimal EHM patch parameters, thereby guiding bioprinting and implantation of the patch. In the master thesis presented here, first steps are taken towards this goal by simu- lating the deformation of the left ventricle of a healthy heart using a PINN. In addition, the PINN is further modified to simulate the deformation of a dam- aged left ventricle.

The controlled recruitment of modified DNA strands onto lipid membranes, mediated by chemical reaction cycles, offers a promising approach for developing dynamic nanoscale systems. In this study, a DNA oligonucleotide was chemically modified to introduce a dicarboxylic acid motif, forming a complex construct that enabled tracking of the oligonucleotide recruitment throughout the reaction cycle. Optimization of the construct allowed for the observation of DNA strand recruitment to the surface of giant unilamellar vesicle (GUV) lipid membranes, visualized by the formation of fluorescent rings around the membrane. The quantification of the modified oligonucleotide binding revealed an increase in its recruitment to the surface of the lipid membrane only after initiating the reaction cycle with the addition of fuel. This highlights the role of the chemical reaction cycle in driving the controlled recruitment process and underscores the importance of precise molecular interactions for dynamic assembly on lipid membranes.

Pole-to-pole oscillations of Min proteins ensure cell division into two equally sized daughter cells in Escherichia coli. However, during the division event, both daughter cells need to inherit approximately equal amounts of the Min proteins to retain the ability to undergo centralized division. We use a high-throughput microfluidic approach to visualize changes in the MinD protein dynamics during cell growth and division. Employing mean-field simulations and modeling we then elucidate possible mechanisms underlying the changes in the dynamics. We find changes in the MinD dynamics due to longitudinal cell growth throughout the cell cycle. Simulations using a reduced reaction scheme reproduce key aspects of the change in the MinD dynamics in growing and dividing cells. Finally, we propose a mechanism for protein partitioning during cell division based on the increasingly restricted diffusion through the septal opening.

Mechanical stress during cellular processes such as migration or cell division can compromise the integrity of the nuclear envelope (NE), which can, in turn, make genomic DNA vulnerable to nicking and double-strand breaks. In cases of NE rupture, Barrier to Autointegration Factor (BAF) binds to the genomic DNA that is prone to damage and helps recruit downstream proteins required for nuclear envelope reformation. However, whether BAF prevents DNA damage on its own accord remains poorly understood. Using an optical tweezer-based assay coupled with confocal microscopy, we investigate the influence of BAF on structural changes in DNA under mechanical stress. Herein, we report a function of BAF in mechanically protecting DNA with a strand-looping mechanism that arrests pre-existing nicks from peeling off under mechanical stress. Our results demonstrate the ability of BAF dimers to contain DNA damage and to stabilize DNA against breakage.

One of the main goals of bottom-up synthetic biology is the creation of life-like systems that can faithfully reproduce cellular functions. To allow cellular processes to happen in synthetic systems, a constant flux of energy and matter must be provided. Previous attempts to fuel synthetic cells have used light - or chemically driven energy production by the ATP synthase or ATP regeneration enabled by the breakdown of L-arginine . In addition to regenerating ATP itself, these systems have also been used to fuel other chemical reactions inside synthetic cells, for example for the production of proteins . However, without regeneration of the materials required for the production of biomolecules, these systems are limited by the initial concentration of nutrients and the building blocks present inside, whose eventual depletion will cause the protocells to reach equilibrium and stop their activity. Here, we present a minimal synthetic cell, which can dynamically produce RNA at long timescales by continuously recycling all involved nucleotides. A sustained energy flux is achieved by the incorporation of membrane proteins into liposomes for the direct provision and selective transport of the energy carriers. Inside each proteoliposome, we install a simple metabolic reaction network to control the production of RNA, which is synthesized dynamically in dependence of the production and degradation rates inside the active vesicles. With this system, we aim to couple minimal metabolism with the sustained production of RNA in order to obtain higher levels of complexity and autonomous control in self-sustaining life-like systems.

Lubricin, an intrinsically disordered glycoprotein, plays a pivotal role in facilitating the low-friction in boundary lubrication of synovial joints. Consisting of two globular end domains and a mucin-like, disordered central domain, lubricins tripartite structure is known to be essential for its lubricating function. Notably the 11 different O-glycans that glycosylate the serine, threonine and proline rich central domain are necessary for lubricins low-friction behaviour. However, a comprehensive understanding of the contributions of all three domains is lacking. With approximately 1400 amino acids and 200 O-glycans, modeling complete lubricin proteins at an all-atom scale poses significant challenges. To address this, we parameterize the O-glycans for the Martini 3 coarse-grained force field, enabling a computational exploration of lubricins low-friction properties.

An improved method of nuclear magnetisation transfer between long-lived quantum coherences excited on glycine amino acid residues of the studied protein will be described. The magnetisation transfer efficiency for different values of the system’s rotational correlation time were estimated using magnetisation dynamics simulations, and the phenomenon was experimentally tested for the case of the protein Lysozyme.

The flagellar apparatus is one of the costliest investments a bacterium can make during its lifetime. E. coli expresses flagellar genes in a pulsatile manner, allowing for a tight temporal control of the duration and frequency of motile states. Using microfluidics, we track expression profiles of single E. coli cells over multiple days and show their ability to tune the frequency of expressing states in respect to their growth conditions. Using numerical simulations, we assess possible fitness benefits related to pulsatile investment in motility

Blue Copper Proteins (BCPs), predominantly found in bacteria and archaea, are copper-containing redox proteins. They play a crucial role in electron transfer in biological systems. This study focuses on grafting the BCP binding motif onto a WW domain to enable electron transfer reactions. By incorporating the BCP loop, in which cysteine, histidine and methionine residues coordinate copper in a trigonal planar arrangement, we have successfully synthesized a copper-binding WW domain. In addition, we aim to achieve electron transfer between our fusion WW domain and a zinc finger. By re-engineering both peptides we want to enhance their metal binding properties, peptide-peptide interactions and the redox potential of each peptide.

The reproduction-mobility trade-off is frequently observed among cells, as allocating resources to one function reduces the availability of resources for another. This trade-off has been shown to foster biodiversity in natural systems, as demonstrated by the rock-paper-scissors model, where fast-moving cells evade predators and avoid local extinction. Here, we propose a model in which two non-interacting species, one characterized by faster movement and the other by higher reproduction rates, compete for a shared food resource. When food is replenished at regular intervals, these cell types become spatially segregated, forming a traveling wave pattern. This model offers a potential paradigm for achieving both coexistence and spatial segregation. We explore how this trade-off influences the evolutionary outcomes of the cells.

The purpose of our work is the quantitative investigation of the changes in mass density, dry mass, and volume during cell detachment using optical diffraction tomography (ODT). This technique enables the measurement of the 3D refractive index (RI) distribution within biological samples, and the RI value can be converted into the mass density of the samples. ODT provides quantitative characterization of dry mass, volume, and mass density of an object, and has been emerged as a valuable technique in biological research. We conducted time-lapse ODT measurements to observe the process of cell detachment, initiated within a microfluidic device. Furthermore, we also investigated the effects of cytoskeletal perturbation on the dynamics of cell detachment by treating cells with ROCK inhibitor Y27632 or silencing ERM kinases LOK/SLK through RNA transfection.

The cell's response to dsDNA breaks is a vital and complex process. Untreated DNA damage could lead to tumorgenesis or cell death. So it's important for this repair process to happen reliably. The goal of my project is to investigate the role of the chemically active biomolecular condensates that form in response to DNA damage and to understand how the properties of these condensates could help reliably organise the repair process in space and time. In particular, we are interested in developing a theoretical model of a condensate that can transform its composition and function in an ordered, and chemically-controlled manner.

RNA origami enables the creation of a wide variety of structural and functional motifs, paving the way for innovative synthetic biological architectures. In particular, our group recently demonstrated the co-transcriptional folding of an RNA origami cytoskeleton mimic inside synthetic cells as well as membrane-inserting RNA origami nanopores. However, the precise molecular architecture of these RNA origami structures is currently unknown. In this study, we optimized cryo-electron microscopy (cryo-EM) conditions to investigate the structural complexities of RNA origami at Ångström resolution. This high-resolution imaging has provided valuable insights into the structural parameters and heterogeneity of these constructs, revealing discrepancies between theoretical models and the actual folded structures. These findings enhance our understanding of the structural complexities and variations in RNA origami folding allowing us to refine our design strategies to achieve more accurate and functional constructs. To conclude, this work not only advances the field of RNA origami but also contributes to the broader application of cryo-EM in studying complex biomolecular structures.

In response to the COVID-19 pandemic, the World Health Organization issued an Emergency Use Authorization (EUA) for vaccines targeting the SARS-CoV-2 Spike (S) protein to elicit neutralizing antibodies. However, longitudinal studies have shown a decline in antibody-mediated immunity over time and reduced efficacy against emerging Variants of Concern (VOCs). This has shifted focus towards the cytotoxic immune response of CD8+ T-cells, known for their long-lasting memory responses from previous coronavirus outbreaks such as SARS-CoV-1 and MERS-CoV. During natural infection, CD8+ T cells recognize virus-infected cells through T-cell receptors (TCRs) binding to peptides presented by Major Histocompatibility Complex class I (pMHC I) molecules, leading to effective interferon signaling and cytotoxicity. The presentation of peptide epitopes within liposomal membranes is hypothesized to enhance TCR recognition and activation by increasing antigen clustering and reducing epitope competition. This study investigated the presentation of the SARS-CoV-2 Spike protein on liposomal membranes of varying sizes. Results showed that MiniVs (approximately 100 nm) with different ratios of DGS-NTA(Ni2+) to spike protein (1:0.5, 1:1, 1:4) exhibited low responses across all donors, with a slightly better response at the 1:4 ratio. In contrast, SGUVs (15-20 µm) demonstrated increasing responses with higher DGS-NTA(Ni2+) to spike protein ratios, particularly at 1:4. These findings suggest that larger vesicles with lower antigen density more effectively stimulate T-cell responses. Future objectives include designing and synthesizing liposomes incorporating SARS-CoV-2 specific CD8+ peptide epitopes, quantifying in vitro T-cell responses elicited by these immunopeptidoliposomes, and comparing these responses to those elicited by solubilized peptide epitopes. This research aims to provide insights into T-cell activation mechanisms and improve current in vitro T-cell screening methods, ultimately contributing to the development of more effective vaccine strategies that leverage natural antigen presentation for enhanced T-cell priming and immune response evaluation.

Spinal cord injury (SCI) affects millions of people worldwide, oftentimes with little to no hope of recovery. The ReWire project works on the development of innovative techniques for the treatment of SCI. This project is a part of ReWire with the goal of creating an aligned, injectable, macroporous microgel scaffold to support directed nerve growth after SCI. For this, PEG-based microgels are produced using microfluidics or particle replication in non-wetting templates (PRINT). Adding superparamagnetic iron oxide nanoparticles (SPIONs) inside the microgels enables their alignment with an external magnetic field before interlinking them using different chemistries. With this, a porous network can be formed allowing enough space for nerve regrowth and thereby creating a biomimetic anisotropic environment for directed growth preventing less efficient random spreading. For the interlinking during this project, the microgels are post-functionalised with suitable peptides that are designed based on fibrin and can be interlinked by addition of the enzyme Factor XIIIa. An investigative study was carried out to find the optimal peptide concentration for construct preparation. Subsequently, microgels were modified with cell adhesive peptides like fibronectin-derived RGD or laminin-derived IKVAV for cell interaction studies using L929 Fibroblasts and neurites from embryonic chicken dorsal root ganglions

Focal adhesions (FAs) are complex multi-protein structures crucial for cell adhesion, signaling, and mechanotransduction. FAs are dynamic structures that connect the cell to the extracellular matrix (ECM) through transmembrane integrin receptors, which link the actin cytoskeleton via adaptor proteins. While it is known that FAs can be reconstituted from a minimal set of components, the influence of actin-binding proteins such as crosslinkers and motors remains an open question. To investigate this, we introduce $\alpha$-actinin, a crosslinker, and myosin II motors to the polymerized actin from minimally reconstituted FAs on supported lipid bilayers. Following actin polymerization, we observe the dynamic behavior of reconstituted FAs and actin filaments using TIRF microscopy. This project provides insights into the role of crosslinkers and motors in the dynamic behavior and wetting properties of FAs.

Enzymes are fascinatingly complex nanomachines with exceptional catalytic capabilities. To both decipher and harness their mechanism of action, we must try to reconstruct enzyme function in minimal systems by mimicking their essential elements. In this pursuit, we identified a natural trimeric domain called ‘foldon’, which we have functionalized by strategically introducing metal-binding sites to facilitate binding with various metals. Through this approach, we have developed a minimal metallo-enzyme mimic that exhibits hydrolase activity when bound to zinc and reductase activity when bound to copper. This novel minimal metallo-enzyme offers an exciting new platform to study the principles of enzyme function.

RNA plays a crucial role in regulating gene expression across various biological processes in all domains of life. Leveraging this function, synthetic RNA-based regulatory systems have been developed for applications such as metabolic engineering, synthetic gene circuits, and biosensors. However, a major challenge in these systems is RNA's susceptibility to degradation. Circular RNAs present an attractive strategy to address this issue due to their higher resistance to degradation compared to linear RNAs and their demonstrated gene regulatory capabilities. Our project aims to develop an "in vitro" RNA regulation system using toehold regulators and circular RNA triggers. Here, we assessed the stability of circularized RNA molecules in both aqueous and cell extract environments using an aptamer-based assay. Additionally, we examined the potential of these circularized RNAs as triggers by measuring their ability to participate in strand displacement reactions. Our findings contribute to the design of more robust RNA-based regulatory systems.

TBD

Motility-Induced Phase Separation (MIPS) is a notable phenomenon in which self-propelled particles undergo phase separation solely due to their intrinsic motility. This behavior starkly contrasts with passive systems, e.g., active systems constantly form bubbles in the liquid phase. Here, we introduce a stochastic bubble model to elucidate the changes in bubble area within a system of Active Brownian Particle. We demonstrate that the bubble-area evolution can be described by a Langevin equation. Notably, this equation characterizes a unique category of stochastic systems: while it possesses an absorbing state, it concurrently maintains a stable nonequilibrium steady state distribution of areas.

Periodic and sequential formation of the growing vertebrae from pre-somitic mesoderm (PSM) is orchestrated by a genetic oscillatory network termed the segmentation clock. Coupling between cells results in a tissue-wide synchronous clock activity that can be visualized as kinematic waves that travel through the PSM and arrest at the position of each newly forming segment. The coupling rules that guide this synchronization and emergence of phase-waves are not clear with cells showing diverse “winner-takes-it-all” or phase averaging synchronization dynamics in vitro. Here we attempt to quantitatively understand the coupling between segmentation clock oscillators using in vitro setup for mouse embryonic PSM cells, theory and agent-based simulations.

The highly bright, photostable and colour-diverse rhodamine dyes have fueled advancements in live-cell fluorescence microscopy in the recent years. In combination with genetically-encoded self-labeling protein tags, like HaloTag7, they are commonly used to stain target proteins in cultured cells. Yet, in vivo imaging of multicellular organisms with HaloTag remains challenging and requires high concentrations and long labeling times due to the poor pharmacokinetic characteristics of the fluorophore substrates. Here, we present covalent rhodamine-binding protein-tags (cRhoTag) as a tool to label rhodamine chromophores bearing a minimal reactive linker to a small protein tag. cRhoTag1.0 binds tetramethyl rhodamine (TMR) covalently by means of a reactive residue in the binding pocket and a short electrophilic moiety on the dye. Due to their small size, the substrates present ideal pharmacokinetic properties and are therefore ideal candidates for in vivo staining.

Experiments often probe observables that correspond to low-dimensional projections of high-dimensional dynamics. In such situations distinct microscopic configurations become lumped into the same observable state. It is well known that correlations between the observable and the hidden degrees of freedom give rise to memory effects. However, how and under which conditions these correlations emerge remain poorly understood. Here we shed light on two fundamentally different scenarios of the emergence of memory in minimal stationary systems, where observed and hidden degrees of freedom either evolve cooperatively or are coupled by a hidden nonequilibrium current. In the reversible setting the strongest memory manifests when the timescales of hidden and observed dynamics overlap, whereas, strikingly, in the driven setting maximal memory emerges under a clear timescale separation. Our results hint at the possibility of fundamental differences in the way memory emerges in equilibrium versus driven systems that may be utilized as a ``diagnostic'' of the underlying hidden transport mechanism. [1] [1] X. Zhao, D. Hartich, \& A. Godec, Phys. Rev. Lett. 132, 147101 (2024)