Physical Biology Circle Meeting

The mammalian plasma membrane is composed of over a hundred distinct lipid species. It has been proposed that this complex lipid composition serves to maintain mechanical properties critical to membrane function. In particular, cholesterol and lipid saturation are known to regulate membrane fluidity. As the plasma membrane is mechanically coupled to the active actomyosin cortex through membrane-to-cortex linker proteins, we investigate how the lipid composition-mediated fluidity of the membrane affects the dynamics of the cortex. To address this problem, we combine experiments on cultured mammalian cells with in vitro reconstitution and computational modeling of minimal membrane-tethered actomyosin networks. In cells, we find that lowering plasma membrane fluidity via perturbations to either cholesterol or lipid saturation levels results in increased cortical stiffness and cell rounding. In the reconstituted membrane-cortex system lacking the biochemical signaling pathways present in cells, reducing membrane fluidity causes an arrest in actin network remodeling by myosin motors. These findings suggest that membrane fluidity directly regulates cortical mechanics in a signaling-independent manner. Simulation of a minimal membrane-tethered actin network being acted upon by myosin motors further shows that a decrease in membrane fluidity can cause network arrest by increasing the drag on membrane-to-cortex linkers. As the actomyosin cortex is a central regulator of cell shape and integrity, these findings highlight the importance of the homeostatic regulation of plasma membrane lipid composition, which constrains membrane fluidity to a narrow, favorable range in cells.

Cells can sense external chemical or physical signals, make internal computation within their protein network, and change their behavior accordingly. A fundamental step in our understanding of living system is to decipher how biochemical signals are transmitted within the cell, leading to distinct morphogenetic responses. Far from a simple linear picture where one protein belongs to one pathway, it is now well known that many molecules can transmit very different signals depending on the cellular context in which they are used. However, such clear examples of multifunctionality of one protein are rare, as it is often hard to determine the causal relationships within this network complexity. In this project, using optogenetics to recruit one protein at the cell membrane, we show that we can control two opposite migratory phenotypes with the same protein, in the same cell type. The recruited protein creates either a ‘front’ in the activated region or a ‘rear’, triggering the migration of the cell in two opposite directions, despite being caused by the exact same acute perturbation. How can one protein causally control such antagonist behaviors? We show that basal concentration is the main determining factor responsible for the two opposite phenotypes induced by this protein. We explain these two phenotypes by a balance between two effectors with different affinities, and we build a small model to describe it. Using this model, we show that we can control the two phenotypes in the same cell by fine tuning both the intensities and timing of the protein activation. Overall, the project uncovers how cell use the dynamic complexity of their network in space, time and concentration to control migratory phenotype.

Cells rely on signalling pathways to transduce signals from the environment to the inside of the cell, generating behaviour and facilitating communication with their neighbours. The activity of signalling pathways — and consequently communication— is normally modulated through ligand binding. The phenomenon of communication by electrical signals is a well-known hallmark of neuronal function and has until recently been largely overlooked in non-neuronal cells. However, the growing body of evidence demonstrates that the electrical potential at the plasma membrane (Vmem) plays an important role in the function of the cell. There are intrinsic differences in the Vmem between different cell types, and its changes modulate the activity of the signalling pathways. In this work, we delve into the mechanisms of intra- and inter-cellular communication by means of electrical signals, focusing on the differences between healthy and cancer cells. Using a pair of isogenic cell lines and optogenetic tools, we measure, manipulate and quantify the electrical potential at the plasma membrane. Simultaneously, we track the activity of ERK, one of the most important signalling pathways and a key player in the development of cancer, and quantify the impact of Vmem modulation on its activity. Finally, we simulate the in-vivo setting by performing the measurements on the co-culture of healthy and cancer cells, and gain the first insight on cell-to-cell communication through the coupling of electrical signals and ERK activity in a mixed epithelium.

The segmentation clock is a genetic oscillator that generates the segmented body pattern in vertebrate embryos. The coupling of these cellular oscillators leads to synchronization and to the emergence of traveling waves of gene expression. Synchronization of these oscillators is important for correct development in vivo. Moreover, it has recently become possible to characterize and manipulate the emergence of these traveling wave patterns in vitro. Coupling between these cellular oscillators is mediated by the Delta-Notch pathway, which depends on cell-cell contact. This implies that coupling strength, and thus synchronization, may depend on mechanical properties such as cell shape, contact area and the relative motion of cells with respect to their neighbors. However, it is not clear to what extent these mechanical features are important in this system. Here, I will explain how we are combining experiments and mathematical modeling to investigate the role of mechanics in synchronization and the emergence of wave patterns. We use both nonlinear dynamics and the theory of phase response curves to study how oscillators respond to perturbations, and continuum models that link oscillator phase and density to study under which conditions synchronization and pattern formation occur.

How cell differentiation and morphogenesis in developing tissues during embryonic growth are regulated by mechanical and signaling interactions between cells is a key question in developmental biology. The general physical principles underlying the interplay between molecular signaling processes and mechanical changes in pattern formation are still unclear. Drosophila pupal wing is an ideal system to address these questions, because the development of the wing can be imaged in vivo at single cell resolution and because cells can acquire only two different fates: vein or intervein. A marker for the fate of intervein cells (i.e. cells which are not vein cells), DSRF, allows to follow the formation of the vein pattern. Experimental observations reveal that an initial broad, poorly defined pre-pattern ($\sim$16 hours after pupal formation) is refined into a highly precise and stereotypical venation pattern in the adult fly ($\sim$36 hours after pupal formation). Concurrently with this refinement process the wing as a whole experiences morphogenetic flows causing the tissue to shrink along the anterior-posterior axis and to elongate along the proximal-distal axis, a process known as convergent extension. What is the relative role of mechanical flows and cell fate adjustments in veins refinement? Analysis of Dumpy mutant flies, flies in which the wing is prevented to undergo convergent extension, shows that the pattern is preserved (same vein width as in wild type), suggesting that the vein refinement is mainly driven by cell fate changes. Experiments indicate that the key components of the signaling system involved in fate decision are the EGFR and Dpp pathway acting to promote vein fate in nearest neighbor cell neighbors, and a Delta/Notch inhibition pathway that extends beyond nearest cell neighbors. Therefore, to explore whether signaling interactions alone can lead to vein refinement we developed a cell fate adjustment model in which the variable describing the vein/intervein state evolves according to an effective reaction-diffusion equation with short range activation and long range inhibition of vein fate. We then simulated the model onto a lattice of cells obtained by segmenting images from a wing tissue. These simulations predict a vein refinement that is in good agreement with experiments. However, data analysis of in vivo experiments reveals also a mechanical contraction due to cellular flows in the vein regions. In the talk, I will present results of the simulations of the chemical model based only on cell fate changes, the role of mechanical deformations in vein narrowing, and a possible mechanism to control vein width based on a feedback between morphogenetic deformations and cell area. This work is the result of a collaboration with Marc de Gennes, Sophie Herszterg, Jean-Paul Vincent, and Guillaume Salbreaux.

Understanding how fluid-filled lumens form and grow is key to a broad variety of biological processes in physiological and pathological contexts. During preimplantation development, the position of the first mammalian lumen breaks the radial symmetry of the embryo. This establishes the first axis of symmetry of the mammalian embryo, which determines the implantation site in the uterus. To grow the lumen, epithelial cells at the surface of the embryo build up an osmotic gradient across the epithelium. This draws water from the outside medium that accumulates into microlumens trapped between cell-cell contacts, which fracture under the hydrostatic pressure of the fluid. Then, guided by the cell contractile activity and the cell adhesion, microlumens pour into a common lumen at the embryo periphery. However, how fluid dynamically accumulates into the embryo could influence the lumen growth and positioning. This could not be investigated due to the lack of tools to measure the fluid transport properties of the embryo. In my talk, I will present a microfluidic device that I have developed to control the environment of preimplantation embryos in space and time, and how we can use it to characterize in vivo the biophysical parameters of blastocyst fluid transport. Preliminary results suggest that both inward and outward transport rates scale equally with the osmotic gradient built by the embryo. However, preliminary analysis suggests that elastic constraints provided by the egg shell and the epithelium itself could influence transport rates. PhD Student: Louise DAGHER (2nd year) Thesis supervisors: Jean-Léon Maître and Stéphanie Descroix

The physical properties in living systems are fundamentally different from the ones found in inert materials. In particular, biological mixtures are composed of many distinct components of mesoscopic size that interact in combinatorial ways. Yet, despite this heterogeneity of interactions, ordered structures emerge in these mixtures. For example, the cytoplasm is composed of thousands of components that have been shown to display deterministic phase behavior, from multiphase coexistence of liquid droplets, to the assembly of solid-like protein complexes. A natural question is: what are the necessary characteristics of the interactions that lead to the emergence of such high order structures? To answer this question, we investigate the phase stability of multicomponent fluid mixtures. Inspired by classical work in neural networks, we show how to encode specific phase compositions in the interactions. The results of our mean-field model predict different regimes. A mixed phase is found at high temperatures and as the temperature is lowered, localised states where few components are dense become predominant. Remarkably, at intermediate temperatures a retrieval phase where the assembly of multiple pre-programmed phases is present. Our results highlight the importance of non-linearities and self-interactions in the phase behaviour of multicomponent mixtures.

Compartmentalization into functional units is a key principle of cellular life. In addition to membrane-bound organelles, eukaryotic cells utilize membrane-less biomolecular condensates to locally concentrate proteins and nucleic acids. While we are beginning to understand how membrane-less condensates assemble and disassemble, we know very little about the biological processes that take place at the surface of such condensates. The surface of the largest membrane-less cellular assembly, the mitotic chromosome, is covered by the intrinsically disordered protein Ki-67. Our previous studies have revealed that Ki-67 has dual functionality. In early mitosis, Ki-67 functions as a surfactant to prevent chromosomes from collapsing into a single chromatin mass, whereas it actively promotes chromosome clustering during exit from mitosis. How Ki-67 switches between these two opposing processes – chromosome dispersal and chromosome clustering – has remained unknown. Here, we demonstrate that Ki-67’s biophysical properties radically change during anaphase onset, when all chromosomes merge into a single cluster. Ki-67’s amphiphilic character is lost as its molecular brush structure collapses and the soluble pool of the protein forms condensates. Our study uncovers a cell-cycle-regulated mechanism that controls individualization and coalescence of chromosomes during mitosis.

Interactions among proteins in living cells can lead to molecular assemblies of different sizes and large-scale coexisting phases formed via phase separation. Both are essential for the spatial organization of cells and for regulating biological function and dysfunction. A key challenge is understanding the interplay between molecular assembly and phase separation. However, a corresponding theoretical framework that relies on thermodynamic principles is lacking. Here, we present a non-equilibrium thermodynamic theory for a multi-component mixture that contains assemblies of different sizes, which can form, dissolve, and phase-separate from the solvent. We show that the size distributions of assemblies differ between the phases and that the formation of assemblies affects the volume of the condensed phases. We also find a gelation transition of the dense phase that coexists with a dilute phase mainly composed of small assemblies. Our theory can explain how molecular assembly is intertwined with phase separation, and our results are consistent with recent experimental observations on protein phase separation.

Stress propagation in nonlinear media is crucial in cell biology, where molecular motors exert anisotropic force dipoles on the fibrous cytoskeleton. While the force dipoles can be either contractile or expansile, a medium made of fibers which buckle under compression rectifies these stresses towards a biologically crucial contraction. A general understanding of this rectification phenomenon as a function of the medium's elasticity is however lacking. Here we use theoretical continuum elasticity to show that rectification is actually a very general effect in nonlinear materials subjected to anisotropic internal stresses. We analytically show that both bucklable materials rectify small forces towards contraction. Using simulations, we moreover show that these results extend to larger forces.

Morphogenesis involves the translation of patterns, often defined by the concentration of signalling molecules, into the shape of a cell or tissue. Robustness of such processes can be enhanced by feedbacks where the shape impacts back on the formation of the pattern. Here, we show that such shape sensing may result from the same forces that drive shape changes. We consider sheets of active matter, such as the cell cortex or developing tissue layers, that behave as active fluid surfaces on long time scales. In these systems, forces drive flows within the surface that inevitably depend on the surface geometry. In-plane torque dipoles, for example, drive flows only in the presence of curvature. When molecules are advected by this flow, a pattern arises, reflecting the symmetries of the geometry. In particular, we show that viscous shear forces result in an effective friction force being proportional to the Gaussian curvature, such that patches of contractile stresses are advected towards regions of minimal Gaussian curvature. On a surface with spherical topology, this implies that a contractile ring such as the cytokinetic ring aligns perpendicularly to the long axis of the surface. Hence, the actomyosin cortex can drive alignment of the division axis with the long axis of the cell by a rotation of the entire cell, consistent with recent experiments

Active fluids exhibit spontaneous flow under confinement and above a critical active stress. So far, this phenomenon has been extensively studied in two dimensions. Here. we study the dynamics of a 3D active fluid and characterize the spontaneous flow transition for various boundary conditions. Using perturbation analysis, we find that perpendicular anchoring permits an extensile Fredriks-type transition with in-plane and out-of-plane flows. Parallel anchoring allows for both an in-plane contractile transition, and an out-of-plane extensile wrinkling transition. We confirm the predicted transitions with high-fidelity numerical solutions of the nonlinear hydrodynamic equations describing active polar fluids. We then show numerically that 3D active fluids exhibit spatiotemporal chaos when further increasing active stresses, and we characterize the flow regimes between spontaneous flow and chaos. Our results demonstrate that 3D active fluids exhibit rich behavior that is not possible in two-dimensions.

Biological tissues are often described as viscoelastic fluids on long time-scales. However, on shorter time-scales, tissues can behave as amorphous solids, such as clay, changing shape only when exposed to a shear stress above the material yield stress $\Sigma_c$. Amorphous solids near $\Sigma_c$ display critical behaviour with a diverging correlation length-scale characterising dynamics of plastic activity. Here, we ask how would this critical behaviour be affected by active processes present in biological tissues, such as cell divisions. In order to model yielding of biological tissues we employ the mesoscopic elasto-plastic model, commonly used to describe yielding of amorphous solids. Here, we extend the classical elasto-plastic model by introducing cell divisions as an additional source of plastic activity. We find that cell divisions strongly fluidise the solid phase of the system at stresses lower than $\Sigma_c$, consistent with literature. Furthermore, we find that critical behaviour is strongly suppressed, leading to localised dynamics of plastic activity nucleated by cell divisions. Finally, in our model, we can describe how well is the cell division orientation aligned with local shear stress. We find that low alignment strength leads to less mechanically stable tissues where most of the plastic flow arises from cell rearrangements, and vice versa.

Cells often migrate on curved surfaces, that can be the curvature of extracellular matrix or the cylindrical protrusions by the other cell. Recent experiments provide clear evidence that motile cells are affected by the topography of the substrate on which they migrate. The origin and underlying mechanism that gives rise to this curvature sensitivity are not well understood. Here, we try to understand how a migrating cell sense and respond to such curvature cues using a theoretical framework as well as experiments performed on different cell types. We systematically study the cell migrating on different types of curved surfaces, such as on a sinusoidal substrate, outside or inside of a cylindrical substrate etc. We note that cell alignment and direction of migration are highly determined by the local curvature of the substrate: on the ridges (maxima) of a sinusoidal substrate, the cell prefers to align perpendicular to the axial direction, while on the grooves (minima), it prefers to align along the axis. While migrating from one groove to another, cells often cross the ridges with much higher angles (greater than $\pi/4$). The speed of the migrating cells show oscillatory behaviour as it migrates along the sinusoidal surface. On the outside (inside) of a cylindrical substrate, cells behave in a similar way as on the ridges (grooves) of a sinusoidal substrate, and explains the behaviour of cells migrating on sinusoidal surfaces within a simplified geometry of constant curvature.

Cytotoxic T lymphocytes (CTLs) are the key player in the adaptive immune system to eliminate tumorigenic and infected cells. The killing efficiency of CTLs can be, however, greatly impaired by dense extracellular matrix in solid tumors, because of reduced migratory capacity. The cytoskeleton plays a major role in regulating all types of cell motility. Among cytoskeletal components, microtubules (MTs) contribute to polarity maintenance, directional cell migration, and nuclear squeezing through mechanical barriers. The organization, positioning, and functions of MTs during cell migration greatly differ between slow-moving mesenchymal cells and fast-moving cells like leucocytes. In amoeboid fast-moving lymphocytes like T cells, MTs are enriched behind the nucleus and assembled in an MT organizing center (MTOC) during migration. However, the precise contribution of MTs dynamics in T-cell migration is only partially understood. We aim to characterize MTs as a major cytoskeletal component regulating the migration of CTLs in simplified micro-fabricated environments. The key question we would like to tackle here is whether MTs is a promising target to improve CTL migration and functionality using pharmaceutical approaches. To this end, we expose CTLs to MT disturbing drugs. We characterized the migratory proprieties of CTLs in self-manufactured micro-channels and in collagen matrix and we examined the impact of drug treatment on the killing capacity of CTLs in 3D. We found that both speed and persistence are increased in treated CTLs as well as the killing efficiency, mainly due to improved migration in 3D. Additionally, using a probe of real-time tubulin polymerization, SiR-tubulin, we describe the position of the MTOC during migration of CTLs in two scenarios with different constriction levels. We found that high confinement of the cells forces a polarization status that dictates faster and more persistent migration, linked to less freedom of the MTOC to rotate around the nucleus. Thus, we conclude that MT-network has a promising potential for therapeutic applications to improve CTL migration and killing efficiency, both in one- and three-dimensional scenarios.

Cells and microorganisms like bacteria use chemotaxis to move in a directed manner in response to concentration gradients of nutrients and toxins. Likewise, synthetic delivery systems could take inspiration from nature to mimic this kind of transport. Ultimately, it could guide nanomotors in the body in a specific way, according to concentration gradients occurring physiologically. We present a system with the basic characteristics to achieve a minimal chemotactic cell. It consists of an asymmetric phospholipid vesicle of 100 nm (liposome) with an encapsulated enzyme (glucose oxidase). The asymmetry is given by the presence of pores in the membrane, inserted by the protein alpha-hemolysin. Mass ratios of 0.075 and 0.1 Hly/lipid were used, resulting in ~2 and ~3 pores per liposome. When the liposome is placed in an environment with glucose, the catalysed reaction occurs in its lumen. The products diffuse outwards through the pores, creating a local concentration gradient. The asymmetric distribution of products along its surface generates a slip velocity that moves the vesicle in response to the glucose concentration gradient. The active motion of the liposomes labelled with rhodamine octadecyl ester perchlorate (1%) was investigated in an Ibidi microfluidic device in which a concentration gradient of 0.05 M of glucose was established in the channel. While the pristine liposome and the one with encapsulated glucose oxidase (GOX-L) presented a velocity towards low glucose concentration, the displacement of liposomes with pores (~3) was reverted to the opposite direction. The movement of the pristine liposome and the GOX-L are due to diffusioosmophoresis, as a result of the interaction of glucose and the channel walls. In the absence of a glucose concentration gradient, vesicles present only Brownian motion. The drift velocity is the sum of the diffusioosmophoresis velocity and the chemotactic velocity. With ~2 pores, the chemotactic velocity is on the same order magnitude of the diffusioosmophoresis, cancelling out any drift movement. With ~3 pores, the chemotactic velocity suppresses the diffusioosmophoresis, resulting in a net drift towards high glucose concentrations.

Eukaryotic cells use clathrin-mediated endocytosis to take up a large range of extracellular cargos. During endocytosis, a clathrin coat forms on the plasma membrane, but it remains controversial when and how it is remodeled into a spherical vesicle. Here, we use 3D superresolution microscopy to determine the precise geometry of the clathrin coat at large numbers of endocytic sites. Through pseudo-temporal sorting, we determine the average trajectory of clathrin remodeling during endocytosis. We find that clathrin coats assemble first on flat membranes to 50% of the coat area, before they become rapidly and continuously bent, and confirm this mechanism in three cell lines. We introduce the cooperative curvature model, which is based on positive feedback for curvature generation. It accurately describes the measured shapes and dynamics of the clathrin coat and could represent a general mechanism for clathrin coat remodeling on the plasma membrane.

Lipid bilayers are highly fluid. This unique property determines many of the mechanical properties of cellular membranes, in particular the fast equilibration of membrane tension on every point of their surface. Thus, at the size of single cells, no membrane tension gradient is theoretically expected to arise. Paradoxically, reports have shown that in neurons (Dai and Sheetz, 1995) and migrating cells, a membrane tension gradient exists, and participates to the force dipole required for movement (Hetmanski et al., 2019; Lieber et al., 2015; Mueller et al., 2017). Other works proposed that in non-migratory cells tension fluctuations rapidly dissipated and are only local (Shi et al., 2018). This raises the question on what sets the length scale of force propagation in cells (Cohen and Shi, 2020), which is crucial to understand how membrane tension gradients are formed and maintained. Here, we reconstitute membrane tension gradients in vitro using supported lipid bilayers (SLBs) expanding on a glass surface, as a proof-of-principle experiment for the study of tension dissipation in membranes. We show that in expanding SLBs a tension gradient exists and it can be measured by flipper-TR, a FLIM probe sensitive to lipid packing. The speed of the expansion correlates with the tension gradient: if the tension is higher at the leading edge, the expansion is faster; if the tension homogenizes, the expansion stalls. We compare this to migrating cells and especulate on what parameters are important to maintain a gradient across a lipid bilayer.

The auditory organ of chick, called the basilar papilla, comprises of two terminally differentiated cells: sensory hair cells and non-sensory support cells. In prenatal development, these cells arrange into a highly regular pattern where hair cells stand isolated from each other separated by support cells. This patterning of the hair cells becomes even more pronounced as the apical surface area of the hair cells increases 10-fold. We suggest that non-muscle myosin, found to be located at certain junctions, drives this regular packing by active contraction. Here, we use a mathematical vertex model, that considers mechanical forces at the scale of single cells and cellular junctions, to identify the contributions of forces, that are expressed in precise patterns by cells, to the process of cellular rearrangement. Active forces of cell size regulation and myosin contraction stand in contrast to global shear forces, which are known to be involved in similar processes of tissue reorganisation but are experimentally only observed to a limited extend in the basilar papilla.

Myoblasts are undifferentiated muscle cells with elongated cell shape. Recently, it has been shown that, at defect site, myoblasts can form bilayers in confluent layers [ \,1] \, and three-dimensional ( \,3D) \, cell mound in a small circular confinement [ \,2] \,. However, the role played by topological defects in 3D cell organization remains elusive. Here, we investigate how C2C12 myoblast monolayers form mounds on defect microstructures in the form of concentric circle, mimicking +1 defect. Myoblasts exhibit contact guidance along the micro rails of the concentric pattern. With time increases, cell proliferation causes density to increase. Above a critical density, myoblasts spontaneously extrude from the defect site, forming multicellular mounds. The system experiences a global transition from azimuthal flow ( \,along the rails) \, to radial flow ( \,perpendicular to them) \, toward the defect site, which feeds and sustain the growing 3D cell mounds. Cells then exhibit vortex-type orientation that evolves in other geometries as the structure develops. Subsequent layers exhibit aster-type geometry and vortex/spiral configurations at the top of the 3D cell mound. \textbf{References} [ \,1] \, T. Sarkar, V. Yashunsky, L. Brézin, C. Blanch-Mercader, T. Aryaksama, M. Lacroix, T. Risler, J.-F. Joanny, P. Silberzan, bioRxiv 2021.06.22.449403. (2021) [ \,2] \, P. Guillamat, C. Blanch-Mercader, G. Pernollet, K. Kruse, and A. Roux, Nat. Mater. \textbf{21}, 588 (2022).

Dynamic properties of allosteric complexes are crucial for signal processing by diverse cellular networks. However, direct observations of switches between distinct signaling states have been limited to compact molecular assemblies of fixed size. Here, we report in vivo FRET measurements of spontaneous discrete-level fluctuations in the activity of the Escherichia coli chemosensory array - an extensive membrane-associated assembly comprising thousands of molecules. Finite-size scaling analysis of the temporal statistics by a two-dimensional conformational spread model revealed nearest-neighbor coupling within 3$\%$ of the Ising second-order transition, indicating that chemosensory arrays are poised at criticality. Our analysis yields estimates for the intrinsic timescale of conformational changes (12-35 ms) of allosteric units, and identifies near-critical tuning as a design principle for balancing the inherent tradeoff between response amplitude and response speed in higher-order signaling assemblies. These results demonstrate the power of using fluctuation signatures of protein complexes to understand their function.

In the past decade, new microfluidic devices have been designed to monitor single lineages of cells for many generations with great precision. In classical bulk cultures where the full population is grown, cells with high reproductive success lead to larger populations of offsprings, while no such selection effect is present in single-lineage experiments. In this work, we ask two questions: What is the lineage-population statistical bias for the cell size distribution? Can we learn the laws of cell growth and division from single-lineage steady state size distributions? To do so, we derive analytical steady-state cell size lineage distributions for size-controlled cells. These distributions are compared to population distributions, and the role of stochasticity, both in single-cell growth and in volume partitioning at division, is explored. In addition, in simple cases, analytical distributions are fitted to Escherichia coli data in order to infer the parameters of the model, such as the single-cell growth rate, the strength of the size control or the asymmetry of division.

Extensive regeneration of the body plan is found in a few exceptional vertebrates, including salamanders such as the axolotl (Ambystoma mexicanum) and the Spanish ribbed newt (Pleurodeles waltl). These organisms are able to regenerate complex body parts throughout life, most notably their entire limbs, through a process which relies on extensive modulation of cellular plasticity and which results in the faithful recapitulation of the missing structure. Further to this, salamanders display additional noteworthy traits, namely extraordinary longevity, indefinite regenerative potential, and lack of traditional signs of age-related decay or ‘negligible senescence’. As such, they constitute valuable models for addressing the mechanisms underlying complex regeneration and the interplay between regeneration and ageing. Here, I will give an overview of the research paradigms available and their potential to illuminate biological phenomena, and discuss our lab’s efforts towards leveraging quantitative biology and biophysical approaches to tackle problems pertinent to regeneration and ageing. Website: https://www.theyunlab.com, Twitter: @MaxYunLab

The spatial organization of the genome is essential for its functions, including gene expression, DNA replication and repair, as well as chromosome segregation. Loop extrusion has been proposed to underlie the formation of non-random chromatin structures such as topologically associating domains (TADs). However, whether the 3D-folding of DNA in single cells is consistent with this mechanism has been difficult to address in situ. Here, we developed a fluorescence imaging workflow for high-resolution reconstruction of 3D genome architecture that conserves chromatin structure at the nanoscale and can resolve the 3D-fold of chromosomal DNA with better than 5-kb/25nm precision in single human cells. Our results show that the chromatin fibre behaves as an ideal random coil up to the megabase scale. Onto this rather open architecture, Cohesin dependent loops are superimposed, compacting chromatin. Computational mining of thousands of single cell 3D DNA traces reveals that long-range looping interactions are sparse and that the random coil nature of intra-loop DNA leads to a large heterogeneity of folds, whose non-random feature are loop bases positioned by CTCF. We find single cell 3D folds are consistent with a model of progressive loop extrusion, while averaging extruded chromatin loops from many cells cumulatively explains the overall more compact structure of TADs as an emergent feature.

Single molecule localization microscopy (SMLM) has become the method of choice in several biological investigations. The ability to image individual fluorophores and localize their centers very precisely allow to break the diffraction barrier and observe molecular organization with unpreceded resolution. Super-resolution (SR) methods such as STORM and PALM accumulate the centers of individual molecules appearing stochastically in time in order to recover a pointillist super-resolution image of the structure of interest. A main limitation of SMLM is the time it takes to acquire a SR image which exceeds 10min typically. A chemical fixation is thus required to avoid cellular motion during the acquisition. Recently, reversible cryo-arrest of cells was proposed as a mean to block molecular motion and acquire single-molecule localization images of the same cell at different time steps. Still many gaps limit the use of such method for regular biological investigations. Here we show a new implementation of the method with a new user-friendly design optimized for biological applications. Our experimental set-up includes a dedicated aluminum mount that enables homogeneous temperature control on the sample and efficient replacement of the medium by adapted DMSO concentrations during the cooling process. The design allows direct on-stage cooling of cells and imaging with high NA objective. In addition, 3D super-resolution images are obtained thanks to a multi-focus microscope in order to probe the 3D molecular organization within living cells. Finally, because the photophysics of fluorescent molecules can be affected at low temperatures, several quantification tools are implemented to characterize the behavior of molecules and their emission properties at low temperature, such as signal to background measurements and blinking behavior. Suitable dyes for SMLM at -45°C are identified. We finally discuss cell survival and show some examples of SR acquisitions performed at low temperatures.

The self-organisation of cells into complex tissues relies on a tight regulation of cell behaviour. The interaction of genes in gene regulatory networks is considered to be a main layer of cell fate regulation. The concentration of gene products, mRNA molecules and proteins, has been shown to be subjected to strong fluctuations. Here, by combining a theoretical framework with single-cell sequencing data, we show that cells can reside in a state where gene expression noise exhibits glass-like properties. Specifically, we develop a theory that maps fluctuations in gene regulatory networks to bipartite asymmetric spin glasses. The dynamics on these networks may be characterized by multiple attractors in a rough landscape and the transition between these attractors are of particular interest in the context of cellular decision making. We show that biologically plausible parameters pose cells in the vicinity of a phase transition to a glass-like phase, where fluctuations are strongly correlated in time and between genes. By mapping these findings to single-cell RNA sequencing we find that stem and progenitor cells exhibit signatures of glassy fluctuations in neural tissues. Our work highlights the possibility that long-lived noise could be a carrier of biological information.

Abstract: Obtaining and processing information is of vital importance to all organisms. A central operation in any information processing device is a copy operation, in which a databit is copied into a memory bit. Living cells frequently perform such a copy step, for example in copying the state of one protein into that of another protein. While cellular systems can be mapped onto computational copying devices, they are also markedly different: cellular systems operate fully autonomously. Using methods from stochastic thermodynamics, we derive optimal copy protocols, compare the cost of cellular, autonomous copy protocols to that of canonical, non-autonomous copy protocols, and study how close these protocols come to the fundamental bounds as set by thermodynamics. Surprisingly, we find that cellular copy protocols can be more efficient than canonical copy protocols, highlighting the power of autonomous information processing devices.

We develop numerical and analytical approaches to calculate mutual information between complete paths of two molecular components embedded into a larger reaction network. In particular, we focus on a continuous- time Markov chain formalism, frequently used to describe intracellular processes involving lowly abundant molecular species. Previously, we have shown how the path mutual information can be calculated for such systems when two molecular components interact directly with one another with no intermediate molecular components being present. In this work, we generalize this approach to biochemical networks involving an arbitrary number of molecular components. We present an efficient Monte Carlo method as well as an analytical approximation to calculate the path mutual information and show how it can be decomposed into a pair of transfer entropies that capture the causal flow of information between two network components. We apply our methodology to study information transfer in a simple three-node feedforward network, as well as a more complex positive feedback system that switches stochastically between two metastable modes.

Organoids and other 3D in vitro multicellular systems have frequently been observed to display rotational motion within their matrix. Collective rotational motion can also be witnessed in vivo, for example in the Drosophila egg chamber. We propose that this motion results from cell-matrix interactions. Cells are thought to move similarly to 2D migration - however, a (near-)spherical geometry of cell clusters can lead to the observed rotation in 3D. Here, we present a continuum mechanics description of an organoid rotating within its embedding matrix. We discuss the extreme cases of the matrix being either purely elastic or purely viscous. The organoid is considered to be a non-deformable solid with a surface polarity field, exerting traction forces on the matrix. Importantly, our study is not limited to perfectly spherical organoids but takes small shape deformations into account. This permits to distinguish between purely rotational and deformation-induced cell-matrix interactions. Our work clarifies how matrix material properties and cellular traction forces enable collective organoid rotation. Reciprocally, the rotating organoid can serve as an active rheology probe, revealing key information about the matrix properties.

Long term time-lapse microscopy of organoids provides a unique way to study the dynamics of organ development in space and time at single cell resolution. Unfortunately, manual tracking of individual cells can be prohibitively time consuming, while current automated tracking methods do not provide data quality guarantees, necessitating manual review either way. We have developed OrganoidTracker 2.0 which fully automates 3D cell tracking and crucially enables quality control by providing error rate estimates on its predictions. High fidelity tracking is achieved with the use of neural networks. Error rate estimates are calculated using techniques borrowed from statistical physics. OrganoidTracker 2.0 first identifies the center points of cells. Probabilities are then assigned to potential connections between cell centers at subsequent time-points. Both these tasks are fully automated using convolutional neural networks. The actual tracking problem is then solved using a graph representation of the data, containing all cells at all time-points and their potential connections with associated probabilities. In order to estimate error rates, we interpret connection likelihoods as energies and let possible trajectories on the graph correspond to different microstates of the system. This framework allows us to define a partition function and thereby determine the error probability of a specific connection given the graph structure and predictions made on other connections. Integrating this information improves error rate estimates by orders of magnitude and makes them an usable upper bound on the true error rates. In our datasets more than 99% of the predicted connections are guaranteed to have an error probability under 1%, showcasing both the quality of the tracking algorithm and our error predictions.

Throughout development, homeostasis and disease, physical forces are central to the regulation of tissue function. Despite the crucial role mechanics plays in our tissues, few current experimental systems allow us to control these forces or tissue shape, limiting our research and understanding. In the last decade, organoid models provided us with unprecedented experimental access. Nevertheless, their canonical 3D culturing makes mechanical measurements difficult or impossible. To bridge the gap between physiological advantages of organoids and mechanical measurements, we use a system of '2D' organoid monolayers. This self-organized tissue allows fundamental biomechanical insights into epithelial organization and function, by measurements such as cell-substrate traction forces, laser ablation, cell shape and migration analyses. I am currently using such 2D organoids of the mouse small intestinal epithelium to identify forces that drive live cell extrusion, a process essential for fast intestinal cell turnover and regeneration. I show that the broad neighborhood of the extruding cell mechanically participates in the extrusion process, which is a behavior different from that reported for apoptotic cell extrusion. By taking advantage of accessibility that 2D intestinal monolayers offer for tissue mechanics studies, we will learn how homeostatic cell numbers are maintained to support health of the small intestine.

Dynamic cellular processes such as cell migration and axon pathfinding are regulated by the mechanical properties of the cells’ environment. The mechanosensitive ion channel Piezo1, which is thought to be activated by changes in membrane tension, is a key player in cellular responses to mechanical signals such as tissue stiffness. However, how membrane tension is influenced by substrate stiffness is currently poorly understood. Here, we used optical tweezers to measure effective membrane tension in fibroblasts and neurons as a function of substrate stiffness. To our surprise, global membrane tension was independent of substrate stiffness within the physiological range. However, we found strong differences between membrane tension in cells cultured on compliant substrates and on glass. To explain our observations, we introduce a toy model connecting substrate mechanics, actomyosin contractility and membrane tension. These results provide a deeper understanding of how the complex interplay of membrane and cortical structures influences effective membrane tension and pave the way for a more fundamental understanding of the downstream biological processes.

Hearing is initiated by vibrations of the hair bundle—the mechanosensitive antenna of the inner ear’s sensory hair cells. Mechanosensitivity results from the direct mechanical activation of ion channels by tension changes in tip-links interconnecting the “hairs” (called stereocilia) of the hair bundle. Channel gating in return changes tip-link tension, producing an internal force—the gating force—associated with an effective reduction of the hair-bundle stiffness. The intimate association between channels gating and stiffness provides a mean to probe mechanosensitivity through mechanical measurements of a bundle’s force-displacement relation. Here, using a two-chamber excised preparation of the frog saccule, we studied the effects on hair-bundle mechanics of a static electric potential difference applied across the sensory epithelium. Upon application of a negative potential in the fluid bathing the hair bundles, we found that the magnitude of channel-gating forces could increase by up to twofold. This effect was graded and reversible for negative potential down to -40 mV. Remarkably, larger absolute potential (> 75 mV) resulted in linear stiffness, corresponding to a sudden drop in gating force and this in a nonmonotonic relation between the gating-force magnitude and potential. Our observations suggest that the transepithelial potential can serve as a control parameter of transduction channels’ gating forces and in turn of hair-bundle mechanosensitivity.

Control of microtubule abundance, stability, and length is crucial to regulate intracellular transport as well as cell polarity and division. How microtubule stability depends on tubulin addition or removal at the dynamic ends is well studied. However, microtubule rescue, the event when a microtubule switches from shrinking to growing, occurs at tubulin exchange sites along the shaft. Molecular motors have recently been shown to promote such exchanges. Using a stochastic theoretical description, we study how microtubule stability and length depends on motor-induced tubulin exchange and thus rescue. Our theoretical description matches our in vitro experiments on microtubule dynamics in presence of kinesin-1 molecular motors. Although the overall dynamics of a population of microtubules can be captured by an effective rescue rate, by assigning rescue to exchange sites we reveal that the dynamics of individual microtubules within the population differ dramatically. Furthermore, we study in detail a transition from bounded to unbounded microtubule growth. Our results provide novel insights into how molecular motors imprint information of microtubule stability on the microtubule network.