Physical Biology Circle Meeting

Targeting a specific cell population expressing a unique receptor combination is defined as phenotypic targeting and can be achieved by multivalent scaffolds where multiple ligands are expressed alongside polymers that prevent non-specific interactions and act as steric modulators [1]. We have recently demonstrated that both steric modulations and multiplexing can be achieved using the polymeric ligand poly(2-(methacryloyloxy)ethyl phosphorylcholine) (PMPC). This synthetic polymer has been employed for several medical devices and demonstrated to avoid unspecific protein adsorption and thrombogenesis [2]. We demonstrated that PMPC binds to CD36, SRB1 and CD81 receptors achieving selective targeting of the cell population that expresses such receptors at specific ratios. We tuned to target mononuclear immune cells, including monocytes, macrophages, dendritic cells, Kupffer cells, and microglia [3]. We used the experimental data to validate our statistical model describing the cell-polymersome multivalent units’ interaction. Our model includes both the free energy of bond formation and the repulsion arising from the cell glycocalyx cushion and the polymer brush. The cell-multivalent unit binding energy is a non-linear function of the number of ligands, and the fraction of bound polymersomes can be controlled by tuning the repulsive and attractive interactions. To further validate our phenotypic targeting strategy and refine our in-silico model, we propose the design of super-selective nanoparticles for specific macrophage phenotypes. Macrophages (MФ), professional phagocytic cells in the frontline defence against potential pathogens, exhibit a marked phenotypic heterogeneity and plasticity. Their differentiation is modulated by environmental stimuli changing their phenotype (known as M1: “pro-inflammatory MФ”, M2: “anti-inflammatory MФ” [4]). The polarisation state of macrophages is a continuum, and the different MФ populations are present in optimal balance promoting tissue homeostasis. Alteration of this equilibrium is translated into disease onset. The manipulation of the macrophage polarisation state is an important pathogenesis mechanism of intracellular bacteria like in Tuberculosis, where a predominant M2 population promotes bacterial survival in granuloma structures [5]. In cancer, TAMs (Tumour Associated Macrophages) display an M2-like phenotype associated with the progression of solid tumours [6]. Other examples of the imbalanced ratio between macrophage subpopulations are chronic inflammation, atherosclerosis [7] and rheumatoid arthritis [8], with evidence of M1 population prevalence associated with disease progression. Phenotypic targeting of a specific macrophage phenotype population to finally modulate macrophage polarisation state represents a promising therapeutic strategy for a wide spectrum of different pathologies with macrophages as the main effector cells. A study of macrophage phenotype is being performed to identify the characteristic receptor expression pattern and the glycans expression profile to determine the receptor expression fingerprint for each macrophage polarisation state. Polymersomes (POs), self-assembled nanoparticles made up of amphiphilic block copolymers, are functionalised with different targeting ligands to identify the most selective combination in receptor-ligand affinity, ligand number and insertion in the polymeric brush. We have selected as targeting ligands the PMPC polymer, the Angiopep2 peptide that binds the LRP1 receptor [1], and the mUNO peptide that binds the MRC1/CD206 receptor [6], which is overexpressed in M2 macrophages. In vitro experiments show that nanoparticle binding to the cell surface has a non-linear behaviour in function of the number of ligands and that the sweet spot to achieve nanoparticle binding occurs only at a specific density of ligands. The optimal multiplexed-multivalent system will be able to discriminate between macrophage states creating on-off association profiles and achieving phenotypic targeting.

Cell progression along the cell cycle relies on the precise and timely completion of DNA replication. In eukaryotic cells replication starts at origins of replication distributed along the genome. Upon activation, origins recruit replication forks, protein complexes that duplicate the DNA as they move along the genome in opposite directions. In our work, we combine theoretical modelling and high-throughput sequencing data to quantitatively characterise the dynamics of DNA replication in budding yeast. Sequencing synchronised cell populations we obtain snapshots of the dynamics from which we estimate replication times along the genome. Furthermore, we identify the position of individual origins and infer the velocity of replication forks from the replication time profiles. Analysing sequencing data from freely cycling populations, we show how replication parameters can be measured without relying on the synchronisation of cells at the start of the S-phase. Our approach illustrates how the combination of recent technical developments and theoretical modelling can be leveraged to quantify the stochastic dynamics of DNA replication in buddying yeast.

Biological systems have a remarkable capacity to react to environmental stresses, quickly adapting over short periods of time and evolving over longer timescales. In this work, we explore how collective behaviour emerges in many-particle systems as a result of fast adaptation to environmental changes induced by the particles themselves. In particular, we study stochastic lattice gases of particles endowed with artificial deep neural networks that we dynamically update using Reinforcement Learning. The system displays a transition between two asymptotic dynamical regimes as a function of the ratio between the timescales of environment and network remodelling. Furthermore, we show that disorder gives rise to dynamic structures formed by a dense core of slow particles and a dilute envelope of fast ones. Our results highlight the wealth of collective effects that can emerge in intelligent interacting systems and pave the way for new numerical and theoretical studies.

Arbuscular Mycorrhizal Fungi (AMF) are widespread symbiotic fungi colonizing more than 60% of plant species. Using plant carbon, they grow away from the root and form a network of hyphae called the Extra Radical Mycellium (ERM). AMF acquire plant inaccessible phosphorous through the ERM and transport it back to the root. ERM growth is the main use of plant carbon, and its final form is reminiscent of the acquisition of phosphorous. Its morphogenesis is particularly informative of the dynamic of trade. ERM characterization has been up to date limited by the difficulty of obtaining time resolved data of its growth. Using an unprecedented, automated imaging set-up and image analysis pipeline, we show that the growth of the ERM colony follow a morphogenetic program that leads to a travelling wave dynamic. This dynamic is essentially described by two parameters: the speed of the wave and the final density reached. These parameters differ for the three AMF strains used and we show that they can directly impact the dynamic of trade in the AM symbiosis. We show that the discussion of these two parameters can prove highly relevant to understand the ecology of AMF. The imaging pipeline is general and can be applied to resolve the morphogenesis of other living branching organisms.

Cis-regulatory elements (CREs), such as enhancers and promoters, control gene expression by binding regulatory proteins called transcription factors (TFs). In contrast to their bacterial counterparts, metazoan CREs typically contain multiple functional TF binding sites of weak specificity. Despite extensive study, the origin and role of such sites remains unclear. Here we use simulations and scaling arguments to study adaptive evolution of CREs under selection for regulatory function. In a novel toy model that recapitulates the essential nonlinearities of metazoan regulation, function requires a gene to be activated by binding of cognate TFs, while remaining inactive in presence of non-cognate TFs that would otherwise cause regulatory crosstalk. Evolutionary outcomes in this model are controlled by two key parameters: selection strength and a newly-identified, dimensionless biophysical parameter. When this parameter is small, multiple strong sites must emerge in a CRE during exceedingly slow adaptation; evolutionary process is lost on the flat plateaus of the model's fitness landscape. When this parameter is large, nearly-random CRE sequences can activate the gene, but adaptation grinds to a halt, unable to select against non-cognate TF binding; evolutionary process is stuck in a frustrated fitness landscape. In the intermediate regime, adaptive evolution of CREs is orders-of-magnitude faster, naturally leading to a diversity of strong and weak binding sites, as reported in empirical studies. Our detailed characterization of this relevant regime suggests that evolution, over a very long timescale, might have optimized metazoan regulatory apparatus to enable rapid adaptation of regulatory sequences.

Living systems process information from their surroundings, exhibiting dynamical adaptation in the response. Here, we propose a paradigmatic chemical model for sensing that encompasses only the necessary ingredients: energy consumption, information storage, and negative feedback. Indeed, equilibrium constraints severely limit the efficiency of information processing, and storage is an unavoidable energy-consuming step to use the information to operate. Additionally, our model architecture is informed by experimental observations that found negative feedback to be a ubiquitous feature. We show that the presence of information storage and negative feedback leads to the emergence of a finite-time memory, essential for dynamical adaptation. Surprisingly, adaptation is associated with both an increase in the mutual information between external and internal variables and a reduction of dissipation in the internal chemical processes. This twofold advantage comes at an energetic cost that can be unambiguously quantified. By simultaneously optimizing energy consumption and information processing features, we find that far-from-equilibrium sensing dominates in the low-noise regime. Finally, we employ our model to shed light on the adaptation of neurons in zebrafish larvae subjected to periodic visual stimuli. We find striking similarities between predicted and observed behaviours, allowing for quantification of dissipation and information-processing performance in this biological system. Our theory provides a stepping stone towards the idea of highlighting the crucial ingredients for information processing at all spatiotemporal scales, starting from the underlying chemical processes.

Two-dimensional packings of cells in developing epithelial tissues are commonly found to be disordered. However, highly organized packings can emerge during development, such as the hexagonal pattern of ommatidia in the eye epithelium of the fruit fly. We analyzed hexatic order parameter $\psi_6$ in the fruit fly pupal wing epithelium and find a sudden increase in the $\psi_6$ value over time, indicating the presence of hexatic and crystalline phases in two-dimensional systems, as described by the classical KTHNY theory. The melting transition scenario with the intermediate hexatic phase has been reproduced in a model of epithelial tissues [Pashupalak et al. Soft Matter, 2020] where the stochastic active forces generated by the cells play the role of an effective temperature. However, both KTHNY theory and recent literature on packings of epithelial tissues assume uniform properties of particles and cells, respectively. In a proliferating tissue, cells grow and divide, which inevitably leads to heterogeneity of cell sizes. We use the vertex model of epithelial tissues to study how the disorder-to-order transition is affected by the heterogeneity of cell sizes. We find that reducing cell heterogeneity as a control parameter drives the system through an ordering transition. We characterize different phases in our model based on the structure factor of cell center positions. We compare our results with the experimental data of the fruit fly wing to identify the role of cell size heterogeneity in the observed disorder-to-order transition.

Increasingly large datasets are acquired for resolving high-resolution molecular structures by means of cryo-ET. Processing of these datasets with established programs requires time-consuming manual intervention. This is particularly the case for the study of small membrane proteins (\textless{}150 kDa) where particles have to be traced via indirect means informed by the context and morphology of membrane systems. Here we describe a workflow that further automates tilt series alignment, 3D reconstruction, segmentation, and particle picking of such membrane proteins by integrating enhanced open-sourced algorithms/softwares with in-house scripts. Automated tilt series alignment, and 3D reconstruction used in this workflow are modified implementations of IMOD while automated segmentation is built on top of TomoSegMemTV, a tensor-voting based membrane detection package. Following automated segmentation, inner and outer lipid leaflet models as well as their sensible normals were generated by means of dipole propagation. Next, particle picking of membrane proteins was realised by sampling sub-tomograms with their centres fixed on membrane point clouds. Their associated normals were used to assign initial particle orientations. Alternatively, subtomogram coordinates and initial orientations could be inferred with membrane models by PySeg, a discrete persistent structure extractor. Subtomogram averaging was then performed with these inputs. Overall, this near-fully automated workflow is capable of efficient handling of large datasets, utilises widely accessible tools, and is comprehensive for configuration. We benchmarked the performance of this workflow in extracting protein particles involved in membrane contact, as an example, the tethering complex of the ER membrane protein VapA dimer (54.8 kDa), and the Golgi-bound cholesterol transporter NPH dimer (75 kDa) in in-vitro reconstituted membrane contact sites.

Active particles have been shown to undergo phase separation based on orientational interactions, where nonreciprocal torques reorient the motion of the particle towards higher density regions [1]. A vectorial coupling between the velocity field and the density gradient leads to this phase separation. The physical properties of the aggregate thus formed, though, markedly differs from those formed due to a particle trapping mechanism. Here, we consider a dense suspension of active particles in the presence of repulsive torques. We will discuss our recent observations of signatures of flocking like behaviour at low activities for such dense systems. Reference [1] J. Zhang, R. Alert, J. Yan, N.S. Wingreen, S. Granick, Nat. Phys. 17, 961-967 (2021)

Non-equilibrium steady states (NESS) pervade nature, from climate dynamics on the planetary scale to the activity of living cells at the molecular level. Key to NESS is the entropy production rate $\sigma$ at which energy is dissipated to the environment. Despite its importance, $\sigma$ remains challenging to measure, especially in nanoscale systems where one has limited access to microscopic variables. We introduce a variance sum rule (VSR) for the displacement and the cumulative force in a NESS that permits us to derive σ from positions and forces trajectories. When all degrees of freedom are (experimentally) accessible, this estimation can be performed from short time measurements. We apply this method both from an analytical and experimental point of view, showing its robustness and usefulness in many practical situations. If all degrees of freedom are not experimentally accessible, then one can resort to a model dependent fitting procedure based on the VSR, that is more reliable and robust than the usual fits based on correlation functions. In particular, this methodology is applied to mechanically stretched red blood cells for the estimation of $\sigma$ associated to membrane flickering. The variance sum rule sets a new resource for exploiting fluctuations to measure physical quantities in non-equilibrium stochastic systems.

The surface of a cell defines a two dimensional space in which biochemical reactions occur. When reactions induce cell shape changes, feedback arises between biochemistry and geometry. In particular, when two cells form a physical contact, surface molecules can interact and transmit biochemical signals that drive the cells into different states. Additionally, such signals can affect the expression of proteins that regulate cellular mechanics and induce changes of the cell-cell contact geometry. We ask how the interplay of signaling, mechanics and cell shape affects the dynamics of cell pairs, focusing on the conditions for symmetry breaking. To this end, we model the cells as a pair of liquid droplets and consider reaction-diffusion processes on their interface. The analysis focuses in particular on the role of the signaling and mechanical time scales. As a possible application of our theory I will discuss symmetry breaking between cells of mouse inner ear explants that engage in Notch signaling.

Living cells maintain and change shape, adhere, move and divide. At the basis of these processes there is the cytoskeleton. Of particular interest is a thin actin layer beneath the plasma membrane, which goes under the name of actin cortex. Indeed, it plays an important role in the cellular morphogenesis and determines the mechanical properties of the cells. Also, instability of the actin cortex might generate protrusions and the contractile ring that cleaves animal cells into two daughter cells during division. We study the dynamics of the actin filaments in this layer and the potential impacts of actin cortex instabilities. For this purpose, we use a hydrodynamic description of active gels. We obtain the formation of an actin cortex with both nematic and polar phase transitions, and the formation of dynamic protrusions above a critical activity. In the absence of polymerization active patterns emerge and determine local changes in density.

During early embryo development, cells need to make several complex decisions: they have to divide, but also change their positions, so that they can differentiate to form organs at the right place and at the right time. Compaction and polarization at the 8-cell stage of mammalian embryogenesis triggers the generation of inside and outside cells, which marks the first cell fate decision. It was shown that the adhesive properties of cells play an important role, especially the surface tensions of cell-cell and cell-medium contacts. Previous biophysical studies have successfully described cells as foams or soap bubbles in this context, but focused either on cell doublets to quartets with simple topology or on large tissues. However, the biophysical mechanisms underlying cellular packings remain largely unknown. Here, we develop a theoretical description of cellular assemblies at the mesoscale, by investigating the evolution of cellular packings up to the 8-cell stage in mouse embryos in collaboration with experimentalists. Our simulation framework consists of spherical soft spheres that pairwise interact with a short-ranged harmonic Morse potential, active fluctuations, and cell divisions. In particular, we investigate a hypothetical path dependence of cellular packings. The growing nature of the embryo with volume and tension heterogeneities as well as asynchronous cell divisions from the 4-cell stage could play an important role in biasing certain 8-cell configurations via initial conditions. Furthermore, we analyze specific biases arising from the adhesion and deformable nature of spheres. Altogether, this research contributes to a better understanding of cellular assemblies at early stages of mammalian embryogenesis.

During development, tissue patterning and shape emergence arise from a complex interplay between cellular mechanics and pathways that control cell fate, leading to changes in physical properties of cells such as cell-cell adhesion, tensions, motility, etc. In vertebrate embryos, somitogenesis, a process giving rise to the formation of musculoskeletal structures in later developmental stages, is a great example of this complex interplay. During this process, mesenchymal cells of the posterior mesodermal tissue differentiate along the anterior-posterior axis into epithelial cells, the mesoderm then breaks up periodically to form the somites. Spatio-temporality of this mechanism is finely controlled under the action of opposing gradients of morphogens along the axis. However, the physical mechanisms involved are not yet fully understood. We aim at deciphering the interplay between mechanics and signaling during somitogenesis to understand how these signaling pathways give rise to mechanical responses leading to the segmentation and somite generation. For that, we have developed an ex vivo approach where we study the mechanical behavior of mesodermal explants as they transform into epithelial segments by assessing their spreading dynamics on adhesive substrates. We observe a spatio-temporal dependence of the spreading dynamics as somitogenesis progresses, and show the role of cell contractility and motility in this process. By combining experimental approaches with theoretical modeling, we aim to reveal mechanical parameters involved in somites generation.

When interacting in large ensembles, cells can undergo collective cell migration. Depending on the type of motions and patterns that are observed, various physical models can be proposed to explain the observations. Ideally, some of the parameters of the model can be estimated through experiments, leading to a better understanding of the system. Here, we plate bronchial epithelial cells (HBECs) into racetracks patterns. The intrinsic behaviour of the cells, combined with the geometrical constraints of the racetracks, leads to a polar flocking, meaning that all the cells migrate in the same direction. Studying the hydrodynamics of this migration motion, we measure the local velocity fields in the direction of the polar flow, and in the transverse direction. They should allow us to access some of the material constants from the equations describing the cell monolayer.

The structure of cells is realized via their cytoskeleton - a polymer network inside cells. It is the main component of structural integrity and it shapes cells. The cytoskeleton is involved in many active cellular activities that includes cell division, migration, and phagocytosis. In macrophages, we know that actin – one protein of the cytoskeleton – plays an important role during phagocytosis. In my project, I work on understanding how actin is involved in sensing the object which needs to be phagocytosed. Particularly, I study the effect of the physical properties such as form, size and stiffness of the objects. Therefore, I am varying the stiffness of objects, e.g. I use gelatine beads, latex beads and polystyrene beads and investigate how these are phagocytosed with the help of actin. Through different microscopy methods (epifluorescence, confocal) I will be able to image the process of phagocytosis. Till now, we have established phagocytosis protocols with macrophages derived from the cell line HL60. For a better biological fit I am now shifting to THP1 derived macrophages. This will help me to understand questions such as how macrophages distinguish between dead cells (ei. Dead RBCs), cell debris, and healthy cells. In addition, we collaborate with theoreticians be build general mathematical model for phagocytosis.

Authors: Gomez Juan Manuel, Bevilacqua Carlo and Robert Prevedel Abstract: Morphogenesis results from coordinated cell behaviours. The cell populations along the Dorso-Ventral (DV) axis of Drosophila embryos at the time of gastrulation are a well-studied example of such coordination. Gastrulation starts when a ventral fold forms, and progresses by the coordinated action of fate-specific cell shape behaviours and additional folding events, which may result from differential cell mechanics. Thus, Drosophila gastrulation is an ideal in vivo model to study the connection between cell fate, cell shape behavior and mechanics. To understand how cell fate determines mechanical properties during development, we combined non-contact Brillouin microscopy to visualise the mechanical properties of tissue volumes at high spatial and temporal resolution with fly genetics to identify effectors of cell mechanics. Here, we describe the time-evolution of mechanical properties of each DV cell population and folding event occurring during gastrulation. The analysis of Brillouin modulus maps along the embryonic DV axis indicated a transient increase in the stiffness of tissues undergoing myosin-dependent folding events, such as Ventral Furrow Formation and Posterior Midgut Invagination. However, we found apical actomyosin did not colocalise with the regions of the tissue undergoing changes in their mechanical properties, suggesting other cellular components are responsible for determining tissue stiffness during gastrulation. We are currently evaluating the role of candidates emerging from the proteomic and phosphoproteomic analyses of DV cell populations (unpublished) in determining the measured changes in Brillouin modulus during gastrulation.

Throughout adult life, epithelial tissues are constantly turned over. At homeostasis, the stochastic loss of cells must be perfectly compensated by the division and differentiation of stem cells. The mechanisms that regulate stem cell fate choice remains in debate. In squamous tissues, such as the skin interfollicular epidermis and oesophagus, emphasis is becoming placed on how the interplay of mechanical and chemical feedback mechanisms regulate of stable density homeostasis. To address this question, mathematical and computational models will be critical in interpreting in vivo clonal lineage tracing and in vitro live imaging experiments in homeostatic and perturbed conditions. To this end, we developed a vertex model with cell loss and replacement, with stochastic cellular decisions either controlled by mechanical density-dependent feedback or a biochemical competition for niche factors. We use our model to explore qualitative and quantitative differences in patterning, clonal dynamics and response to perturbations in tissues where stem cell decisions are controlled by these different feedback rules. Further, we show this model can quantitatively capture the clonal dynamics from in vivo lineage tracing of mouse back skin and tail skin. Our vertex model may prove to be a valuable tool in understanding how homeostasis emerges from collective cellular dynamics in two-dimensional epithelial tissues. Additionally, we can address more physical questions, such as how glass-like or fluid-like phases of epithelial tissues are affected by cell loss and replacement.

Excessive deposition of fibrillar collagen in the interstitial extracellular matrix (ECM) of tissues causes fibrosis, leading to disruption of tissue architecture, with ultimate organ failure. To this end, an FDA approved drug library (767 drugs) screen in primary normal human lung fibroblasts (NHLF) measuring collagen deposition in ECM revealed that Dextromethorphan (DXM), a cough expectorant, significantly reduces the amount of the fibrillar collagen deposited Extracellular Matrix (ECM) (upon TGFb11 treatment). Immunofluorescence, electron microscopy and proteomic analysis of primary human lung fibroblasts showed that the treatment with DXM or TGFb1+DXM causes a trafficking block Collagen 1 (COL1) in the endoplasmic reticulum of the cells. Data showed that observed collagen accumulations are positive for trafficking regulatory proteins specific for collagen transport such TANGO1 and HSP47. Furthermore, we observed that the transport of vesicular stomatitis virus G protein was unaffected upon DXM, hinting towards an ECM cargo specific blocking role. Subsequent, transcriptome analysis (TGFb1 vs TGFb1+DXM) in NHLF showed that this accumulation of pro-fibrotic ECM cargoes causes transcriptional anti-fibrotic regulation of pathways such as those of Semaphorins, MMP-ADAMTS axis, WNT, EGF-ERBB3-axis, ACTA2 and VIM. Next, we observed using Second Harmonic Generation microscopy that DXM also reduced the deposition of fibrillar collagen in the ECM of ex-vivo cultured human precision-cut lung slices (hPCLS) (upon pro-fibrotic stimulation). Transcriptome analysis of DXM treated hPCLS showed anti-fibrotic regulation of ECM related signalling. Taken together the data obtained in both, in-vitro and ex-vivo models of fibrogenesis shows Dextromethorphan has potent anti-fibrotic activity and further yields membrane trafficking of ECM cargoes as a novel target against pulmonary fibrosis and fibrosis in general.

Phase-separated multi-component systems with chemical reactions have become of wide interest in recent years, as they show interesting behaviors of non-equilibrium systems. Much work has been done on the particle concentration dynamics at constant temperature. We extend the existing framework to study the heat dynamics that arises. We furthermore include and study fluctuations in such systems via a kinetic Monte Carlo model. Thereby, we obtain the full stochastic dynamics and entropy production when driving the system out of equilibrium by fueling reactions or via baths at the boundaries. Specifically, we consider stochastic continuity equations for the chemical composition and energy. In systems with multiple coexisting phases, the coupling of these equations can depend strongly on the phase. Using a lattice model, we gain an understanding of how heat and particle flow couple. For this, we write an entropy-driven Metropolis scheme including local reactions and spatial flows. This scheme can also be implemented as a kinetic Monte Carlo simulation, where we explicitely include and track the exchange with heat and particle baths at the boundary. We then go to the isothermal limit and keep the explicit treatment of baths, allowing us to still obtain the heat release by the system. The established framework allows insights into heat flows and reaction heat in fluctuating phase-separated systems. Particular interest for future work is in applying this to understand the energetics of biological cells and organisms.

The three germ layers endoderm, mesoderm, and ectoderm are specified during early embryonic development and give rise to all tissues of the body. The biochemical drivers of germ layer specification are well known, but the concomitant changes in mechanical forces and material properties are less understood. Recently, the Diz-Muñoz lab has found that plasma membrane tension and particularly the attachment of the plasma membrane to the cytoskeletal cortex can control cell fate in mouse embryonic stem cells. These findings raise the question whether cell surface mechanics could also influence germ layer fate specification. I will address this question using 2D in vitro differentiation and 3D organoids (gastruloids), and develop a new optogenetic tool to perturb membrane-to-cortex attachment. Understanding how cell surface mechanics could control the fate of cells in a developing tissue could pave the way to exploit these concepts for stem cell-based tissue engineering.

Morphogenesis consists of a series of complex events, including symmetry breaking, cell differentiation and shape changes. The intrinsic capacity of stem cells to self-organize into functional tissues has enabled the development of in vitro biomimetic structures such as organoids and gastruloids. These approaches have improved the understanding of the mechanisms by which signaling pathways pattern tissues, but the role of tissue mechanics in cell fate patterning remains poorly understood. Differentiation patterns require the apparition of locations with strong expression of genetic markers previously homogeneously expressed, a process called symmetry breaking. Furthermore, fate is coupled to shape, as highly curved areas contain specific cell types. During my PhD I will explore the mechanical cues regulating stem cells morphogenesis in 3D. We aim at understanding how symmetry breaking and coupling between curvature and cell fate are controlled by mechanics of the growing tissue.

M. Marmol$^{1,2}$ , C.Cottin-Bizonne$^{2}$, C.Ybert$^{2}$, D. Faivre$^{1}$ $^{1}$MEM, BIAM, CEA Cadarache, Saint-Paul-Lez-Durance, France $^{2}$Light and Matter Institute, CNRS UMR 5306, Claude Bernard University, Lyon, France Magnetotactic bacteria (MTB) orient in magnetic fields with the help of an intracellular ferrimagnetic nanocrystals chain, and can be seen as self-propelled compass needles, at the frontier of magnetic colloids and micro-swimmers [1]. This magnetically guided motion, combined with an active oxygen concentration sensing (aerotaxis), is a navigation strategy to reach their preferred physiological region, where they accumulate and form a dense aerotactic band. While the dynamic of the aerotactic band formation have been studied in terms of magneto-aerotactic motion [2], the magneto-hydrodynamic aspect remain elusive. In this work, we studied the emergence of magnetically driven shear flows of the aerotactic band, using a custom-designed microscope equipped by a 3D Helmholtz coils set-up. By applying an oblique magnetic field with respect to the band direction, a spontaneous shear flow appears on both sides of the band. The flow speed depends on the bacterial concentration, the field amplitude and the angle between the band and the field. This effect is attributed to gradients of active and magnetic stresses in the suspension and present an analogy with previous theoretical works on the rheological behavior of magneto-active suspensions [2] [3]. [1] Klumpp et al. (2019). Swimming with magnets: From biological organisms to synthetic devices. In Physics Reports (Vol. 789, pp. 1–54). Elsevier B.V. https://doi.org/10.1016/j.physrep.2018.10.007 [2] Lefèvre, C. T., Bennet, M., Landau, L., Vach, P., Pignol, D., Bazylinski, D. A., Frankel, R. B., Klumpp, S., & Faivre, D. (2014). Diversity of magneto-aerotactic behaviors and oxygen sensing mechanisms in cultured magnetotactic bacteria. Biophysical Journal, 107(2), 527–538. https://doi.org/10.1016/j.bpj.2014.05.043 [3] Alonso-Matilla, R., & Saintillan, D. (2018). Microfluidic flow actuation using magnetoactive suspensions. EPL, 121(2). https://doi.org/10.1209/0295-5075/121/24002 [4] Vincenti, B., Douarche, C., & Clement, E. (2018). Actuated rheology of magnetic micro-swimmers suspensions: Emergence of motor and brake states. Physical Review Fluids, 3(3). https://doi.org/10.1103/PhysRevFluids.3.033302

Symmetry breaking in living systems is a fundamental process that remains incompletely understood. In Caenorhabditis elegans, symmetry breaking in the zygote results in anterior-posterior (A-P) polarization defined by the PAR proteins. The established view in the field has been that there are two partially redundant polarization pathways. The main pathway relies on cortical actomyosin flows that are primed by centrosomes, and which induce polarization of the PARs through advective transport, while a secondary pathway leads to A-P polarization without flows through a microtubule-dependent mechanism. Furthermore, previous experiments suggest that the Aurora A kinase AIR-1 somehow acts as a symmetry breaking cue upon its recruitment to centrosomes, since zygotes depleted of AIR-1 undergo spontaneous, albeit aberrant, polarization in a cell shape-dependent manner. Symmetry breaking in the C. elegans zygote has also been investigated through theoretical descriptions in one spatial dimension, by coupling a reaction-diffusion system for the PARs to an active-gel description for the cortex. Given the likely importance of spatial regulation and cell geometry in symmetry breaking, as observed upon AIR-1 depletion, I have started to extend the theoretical description to two dimensions. Furthermore, I have conducted experiments with a mutant allele of non-muscle myosin II, which indicate that actomyosin contractility and flows are, in fact, essential for polarization. Overall, by perturbing key components of the system theoretically, using numerical simulations, as well as experimentally, using select mutant and RNAi conditions, we aim to decipher the mechanisms of centrosome-guided and spontaneous symmetry breaking, i.e. with or without centrosomes and AIR-1.

Plasma membrane tension and its attachment to the underlying acto-myosin cortex (membrane-to-cortex attachment or MCA) are emerging as key mechanical properties regulating animal cell shape and differentiation. Membrane skeletal proteins also provide support and organization to the plasma membrane, but their contribution to cell surface mechanics in vivo is unknown. Spectrin is a membrane skeletal protein which maintains cell shape, interacts with the actin cytoskeleton, and organizes domains of membrane proteins. However, its contribution to cell surface mechanics remains poorly understood, as well as the possible role of these interactions in cell spreading or lateral membrane assembly. We are addressing these questions by assessing the impact of spectrins on the surface mechanics of mouse embryonic fibroblasts (MEFs) and Madin-Darby Canine Kidney (MDCK) epithelial cells. To that end, we have obtainined CRISPR BII-Spectrin knockout MEF cell lines and we will characterize their surface mechanics through atomic force microscopy (AFM) and magnetic pinching. Next, actin and membrane-binding mutants of spectrin will be employed to dissect how spectrin regulates cell surface mechanics. In parallel, we are developing various assays to assess how cortical mechanics and MCA regulate spectrin dynamics and activity. Towards that goal, we have developed an optogenetic LOV2-SSPB actin heterodimerization system to understand how actin crosslinking modulates spectrin dynamics. Furthermore, we have implemented 3D STORM super-resolution imaging of BII-Spectrin in MEFs and MDCKs to identify possible spectrin lattices that can be affected by cell surface mechanics. In the future, we will assess how the mechanical interplay between spectrin and cell surface mechanics regulates cell spreading in MEFs or morphogenesis in epithelial cells.

Hydra has since long served as a model organism for the study of morphologic regeneration from fragments of tissue. Hydra can give us insight into the development of embryonic tissue, the foundation of a body axis, the regeneration of tissue, and the evolution of multicellular organisms. There is an outstanding resemblance to the development of early embryos of other animal species regarding genes expressed and patterns formed. Here, we investigate the mechanisms of axis formation. At the axis-defining moment, the early hydra embryo exhibits a strong sensitivity to external mechanical perturbations [1]. In this study, we develop the hypothesis that these mechanical fluctuations induce the orientation of microtubules that may possibly contribute to β-catenin nuclear translocation, triggering cell differentiation. We are going to investigate this idea by applying mechanical forces on hydra spheres with magnetic tweezers.

Cell morphogenesis is fundamental in living systems since functional anatomical features form through growth and differentiation of primitive cell masses. Cytoskeletal filaments, particularly the cortical microtubule (CMT) network is known to influence cell wall architecture in plant cells through maintenance of cellulose synthesis complex (CSC) movement trajectories, thereby regulating cellulose deposition. The CMT network therefore has an effect on microfibril arrangement within cell walls, thus impacting anisotropic wall expansion patterns, and ultimately cell growth. However, the precise biophysical mechanisms via which the CMT network affects cell morphogenesis, and especially how cytoskeletal defects relate to growth-abnormalities isn’t well known. In this project, I aim to forge a quantitative link between CMT-CSC dynamics and cell wall mechanics within a single numerical framework, thereby going beyond the state-of-the-art modelling approaches, and simulate cell growth as a competition between turgor pressure and cell wall resistance. I will systematically investigate CMT network organization by combining quantitative live-cell imaging with simulations in Cytosim - a C++ based cytoskeletal simulation suite, to evaluate morphometric variations arising from changes within the CMT network. I will focus on unicellular Arabidopsis trichomes, chosen because of their simple morphology, which is functional, reproducible, and easily accessible. However, the underlying principles derived from this study will be general, and applicable to more complex multi-cellular systems. Establishing the cytoskeletal basis behind cell shape and growth, will also further our understanding of the cytoskeleton's response to extrinsic cues. This way, scientists can inversely explore the biophysical pathways of cell shape alterations due to adverse effects of climate change, such as severe droughts and extreme temperatures, that can impair trichome functions vital for the plant.

Deformations of tissues during development often involve cell intercalations including cell neighbour exchanges in T1 transitions, but a general continuum description of such rearrangements is still lacking. Here, we combine morphoelasticity and plasticity to develop a continuum framework for studying tissue dynamics which includes such cell intercalations, intrinsic deformations such as cell shape changes, and elastic deformations of the tissue. We apply our theory to serosa closure in the beetle Tribolium: during this process a layer of serosa cells expands and closes over the embryo. During the closure cells deintercalate from the serosa boundary hence the number of cells at the boundary decreases. These cell intercalations are associated with contraction of the actomyosin cable at the serosa rim. We show that this continuum framework allows us to describe serosa closure with a suggested toy problem. We compare the numerical solution with cell intercalations and exact solution of the toy problem without intercalation and the results show that intercalation reduces the stresses required for the serosa closure.

The intestinal epithelium comprises a single layer of cells that provide a barrier between the intestinal lumen and the underlying tissue. Cells proliferate in the crypt region and mature as they migrate up the crypt-villus axis, and finally extrude from the epithelium and shed into the lumen. Every day, over 10 billion cells extrude from the intestinal epithelium. While some cells extrude after onset of apoptosis and nuclear fragmentation (apoptotic cell extrusion), others undergo live-cell extrusion followed by anoikis. Due to the toxic environment and high pathogenic load of the intestinal lumen, the intestinal epithelium must maintain its integrity during the extrusion process. Although cell extrusion is a key event in intestinal tissue homeostasis, the cellular mechanisms that regulate it and prevent discontinuities in the epithelium remain largely unknown. To investigate the contribution of tissue-scale mechanical forces in the regulation of cell extrusion, we test the response of intestinal organoids to local ablation of the cytoskeleton with a mild infrared laser pulse (laser microsurgery). We find that such laser-stimulation of the basal cytoskeleton – and thus an acute release of local tissue tension – induces live cell extrusion without compromising epithelial continuity. Additionally, we show that laser-induced cell extrusion does not affect the overall cell death rate of stimulated organoids, which is an advantage compared to mechanical, pharmacological and optogenetic methods for triggering cell extrusion. Finally, we present a multiparticle model that quantitatively describes the dynamics of the maintenance of epithelial continuity after cell extrusion with high accuracy. Our findings reveal novel aspects of a mechanical feedback-controlled mechanism that eliminates poorly adhering or damaged cells and suggest a critical role for active tissue mechanics in the regulation and maintenance of epithelial integrity.

Biological polymeric materials form liquid droplets through liquid-liquid phase separation, referred to as biological condensates. Although the material properties of biological condensates are deemed to play essential roles in cellular functions, quantitative studies of condensates' rheology became available very recently~\cite{jawerth2020protein,alshareedah2021programmable}. Particularly the experiments found the glass-like material property of condensates, showing slow relaxation, termed "aging" in the glass field. The aging rheology of biological condensates may have important biological implications. Thus, understanding the phenomenon, not only from experiments but also from the theoretical perspective, is highly anticipated. In this study, we develop a rheological model of biological condensates from the physical pictures: stochastic binding and unbinding of proteins inside condensates. We obtain the constitutive equation for the material property of protein condensates showing aging behavior. We elucidate how aging manifests in the experimental observations in microrheology, both in active and passive rheology. To understand the condensates' aging behavior, we also develop a novel method to characterize the time-dependent rheological properties of aging materials.

Macrophages play a vital role in the immune system by detecting and eliminating bacterial organisms through phagocytosis. Upon activation, macrophages expose vimentin cytoskeletal protein to the extracellular environment. Such extracellular vimentin can either remain bound to the cell surface or it can be released into extracellular space. This phenomenon similarly occurs under circumstances like injury, senescence, and stress. However, the characteristics of the extracellular form of vimentin and its implications on macrophage functionality remain unclear. In this study, we demonstrate that vimentin is released from the back end of macrophages. Activation of macrophages further enhances this polarized secretion of vimentin. Our findings from migration and phagocytosis assays show that extracellular vimentin enhances macrophage functionality in terms of migration and phagocytosis. Through high resolution fluorescence microscopy and scanning electron microscopy techniques, we show that extracellular vimentin is released into extracellular space in the form of small fragments. Interestingly, the extracellular vimentin forms agglomerates on the cell surface prior to its secretion. Taken together, we propose a mechanism of vimentin secretion and its implications on macrophage functionality.

The cytoskeleton is the basic framework which provides structural and mechanical integrity of the cell. Every cell has its own cytoskeletal arrangement and helps in co-ordinating the general functions from providing shape to cell signalling, cell division and actively contributes to cell apoptosis. All these activities are aided by actin, tubulin and intermediate filaments, which are common cytoskeletal proteins of the cells. However, red blood cells (RBCs) have uncommon cytoskeletal framework. The major component of the cytoskeleton in RBCs is spectrin, forming a quasi-hexagonal network under the bi-layer lipid membrane (Flormann et al. 2021a). Spectrin is attached to actin and myosin. Actin and myosin presumably help in giving the red blood cells its characteristic reversible deformability (Gokhin et al. 2015). This association in turn, helps in maintaining the F-actin polymerisation via the Rac GTPases and Hem-1 organization (Chan MM, et al 2013). So far, literature has postulated the absence of cytoskeletal proteins causing diseases in RBCs and consequent systemic effects. However, in the past decade there has been another school of thought which probes the presence of tubulin in RBCs and is being investigated using various proteomic experiments (Nigra A, et al. 2017, 2020). In the presented study, we investigate the presence of tubulin in RBCs using immunofluorescence experiments. We also study how post translational mechanisms of tubulin – if present in the cells- alter mechanical properties of RBCs.

While the importance of studying cell migration in a 3D context to mimic in vivo settings is now well established, the fundamental mechanobiological principles regulating transition between migration modes in response to substrate topography remain poorly understood. In particular, endothelial cells migration on 2D substrate is exclusively characterized by the formation of lamellipodia which is in strong contrast with cells cultured in hydrogels that are able to generate filopodia, characteristics of “leader” tip cells observed in vivo, trailing stalk cells behind them to form the new vessel in an angiogenic context. Using two-photon photopolymerization, we generated 3D microstructures triggering the formation of endothelial filopodia, evocative of a tip cell phenotype. Filopodia were formed with or without Vascular Endothelial Growth Factor, suggesting that geometry alone leads to a transition in the migration mode. Filopodia could give rise to longer branched protrusions reminiscent of structures described in vivo as dactylopodia in the literature. We performed pharmacological experiments in our system and characterized filopodia dynamics under various inhibitions or activations. This led us to identify MLCK as a key player for filopodia induction by microtopography and to postulate a model in which MLCK activation of actomyosin contractility around the microscaffold’s pillars locally increases membrane tension and promotes ionic channel opening and calcium influx regulating filopodia formation.

Recent Cryo-EM experiments have revealed that the protein network controlling the motility of Escherichia coli bacteria — one of the best characterized cell signaling systems at a quantitative level — is spatially organized as a large protein assembly, consisting of thousands of transmembrane receptors, kinases and scaffolding proteins arranged in a highly ordered array [Breigel et al., 2012 PNAS]. Although many lines of evidence suggest that allosteric cooperativity between subunit proteins allow the integration and amplification of input ligand signals, the physical mechanism by which conformational interactions propagate through such a large protein array remains unknown. We recently developed an in vivo single-cell FRET measurement system for probing the signaling dynamics in individual bacteria, and discovered large-amplitude spontaneous fluctuations in signaling activity that were not observable in previous ensemble-averaged experiments [Keegstra et al., 2017 eLife]. Here, we present novel single-cell FRET experiments which reveal, under certain physiological conditions, that these spontaneous fluctuations become oscillatory. We discuss how such an oscillatory phase for the unstimulated dynamics can be explained by a model of this adaptive feedback network that treats the allosteric chemoreceptor array as an Ising lattice tuned close to criticality [Mello et al., 2004], as well as possible functional implications of such spontaneous oscillations. Briegel, A. et al. (2012) “Bacterial chemoreceptor arrays are hexagonally packed trimers of receptor dimers networked by rings of kinase and coupling proteins,” Proceedings of the National Academy of Sciences, 109(10), pp. 3766–3771. Keegstra, J.M. et al. (2017) “Phenotypic diversity and temporal variability in a bacterial signaling network revealed by single-cell fret,” eLife, 6. Mello, B.A., Shaw, L. and Tu, Y. (2004) “Effects of receptor interaction in bacterial chemotaxis,” Biophysical Journal, 87(3), pp. 1578–1595.

Suspensions of motile microswimmers such as bacteria and active colloids often encounter porous environments both in nature and in industrial applications, but much remains unknown about their mechanistic behaviour and transport properties in the medium. Here, we explore microswimmer dynamics in a saturated 2D porous medium, at the scale of the constituent microscopic solid inclusions. To this end, we model the microswimmers as point-sized active Brownian particles in a large doubly-periodic domain containing randomly distributed polydisperse solid inclusions. The Brownian dynamics simulations reveal the influence of the randomness of the medium on the evolution of statistical quantities of the suspension such as the number density and the polarization near the surface of the inclusions and in the fluid channels. In particular, we study how the 'activity' of the microswimmers, together with the spatial constrictions and local heterogeneities of the surrounding medium, affect these statistics. We also examine the effects of an externally-imposed pressure-driven Stokes flow through the porous matrix, where advection and local shear-induced reorientation modify the dynamics and transport of the microswimmers.

Vertex models have enjoyed a high degree of success in modeling the dynamics of confluent, 2D planar epithelia. They predict a transition from a solid phase to a fluid one when the shape index, a circularity measure, exceeds a certain threshold. However, the usual approach treats cells as polygons in the 2D plane, tracking only their apical side, for instance. For curved tissues with notable differences between the apical and basal sides, it is more appropriate to treat cells as 3D polyhedra. It is not understood how to do so in a way that starts from the theory of fluidity in 2D vertex models. The majority of work on curved, epithelial monolayers treating cells as 3D polyhedra is focused on steady states and uses energy functions that cannot support a fluid phase. In intestinal organoids, however, there is evidence for such a fluid phase: cells persistently rearrange in the crypt, and while this might be driven by proliferation, preliminary experiments suggest that it persists in the presence of cell cycle inhibitors. Therefore, we are interested in creating a three-dimensional model of curved monolayers, capitalizing on the theory of fluidity in the well-studied 2D vertex model. We demonstrate that the naive generalization of the 2D energy function, with a target area and perimeter, to 3D, with a target volume and total surface area, cannot support a fluid phase for a simple, columnar epithelial monolayer. We derive an energy function for curved epithelial monolayers that can support a solid-fluid transition starting from the 3D energy function, and by demanding that it reduces to the 2D energy function for a simple, planar epithelium with columnar cells. Our computational framework was designed with intestinal organoids in mind, but it can be applied to generic curved epitelial monolayers.

Sounds are detected and amplified in the cochlea by active mechano-sensory hair cells. The hair cells are each tuned to a characteristic frequency of the sound input and spatially distributed in the cochlea according to a frequency map. Based on the measured essential nonlinearity and the properties of otoacoustic emissions associated with cochlear amplification, it has been proposed that each hair cell within the cochlea may operate as a critical mechanical oscillator poised near the onset of an oscillatory instability—a Hopf bifurcation. However, the relevance of this general physical concept for hearing remains debated. In cochlear models based on critical oscillators that are only coupled hydrodynamically through the surrounding fluid, the response properties to pure tones are dictated by the generic properties of the local oscillators, resulting in peak responses along the cochlea that are way too sharp at low sound levels. Indeed, a single critical oscillator must obey the constraint that the product between the gain and bandwidth of frequency selective amplification be constant, resulting in an active resonant behavior with a “tall and thin peak”, whereas the resonant behavior of the cochlea in vivo is associated with “tall and broad peaks”. Here we develop existing cochlear models to include (i) mechanical coupling between the oscillators and (ii) noise. Our working hypothesis is that noise limits the sensitivity and frequency selectivity of the oscillators at low stimulus levels and that the lateral coupling between hair cells tuned at different frequencies results in synchronization and clustering with a nonlinear tradeoff between sensitivity and frequency selectivity. In such models, we are asking whether sound stimuli, which ought to elicit a nonlocal interaction between coupled critical oscillators, may evoke “tall and broad peaks” of vibration.

Synthetic morphogenesis allows the study of tissue folding and 3D tissue shape generation, which provides us the possibility to also understand the tissue response mechanism during complex shape formation. A study of shape programmability of elastic materials can be useful for modeling morphogenesis of epithelia because of the elastic nature of cells. In this project, we study topology and geometry in deformations of soft matter and apply them to tissue folding problems. By mapping certain desired 3D final shapes onto a 2D sheet and patterning the curvature inducing cells in epithelial monolayers accordingly, tissue shape reconstitution can be achieved. We aim to eventually understand the role of topological defects in elastic materials and in epithelia, which will be helpful to understand the morphogenesis mechanism and to predict the final shapes by looking at the topology in a cell sheet.