Paris Biological Physics
Community Day 2022

Invited Talk

Neural mechanisms for memory and emotional processing during sleep

The hippocampus and the amygdala are two structures required for emotional memory. While the hippocampus encodes the contextual part of the memory, the amygdala processes its emotional valence. During Non-REM sleep, the hippocampus displays high frequency oscillations called "ripples". Our early work shows that the suppression of ripples during sleep impairs performance on a spatial task, underlying their crucial role in memory consolidation. We more recently showed that the joint amygdala-hippocampus activity linked to aversive learning is reinstated during the following Non-REM sleep epochs, specifically during ripples. This mechanism potentially sustains the consolidation of aversive associative memories during Non REM sleep. On the other hand, REM sleep is associated with regular 8 Hz theta oscillations, and is believed to play a role in the regulation of emotional reactions and the consolidation of emotional memories (emotional processing). Unraveling the fine neuronal dynamics related to REM sleep, Non-REM sleep and the transitions between states in the hippocampus-amygdala network will further our understanding of the implication of these sleep stages and related brain patterns in emotional processing.

Short Talks

Learning and Unlearning: towards the optimal memory retrieval in recurrent neural networks

Recurrent neural networks and their dynamic properties have interested physicists for long time, as an effective tool to study the associative memory performance of real and fictitious systems of neurons. A crucial problem, both in the field of theoretical neuroscience and physics, is the one regarding synaptic plasticity and optimisation, dynamic processes aimed at finding the optimal configurations of the synaptic weights that maximise the memory performance of the network. In this talk I will introduce the audience to the Unlearning algorithm, a surprisingly effective training procedure inspired by the biology of REM sleep in mammals. I will thus present some recent results in the realm of the statistical mechanics of Hopfield-like networks and I will illustrate an application of Unlearning in the context of biologically plausible models of hippocampus where spatial maps are stored and retrieved.

Data-driven discovery of long timescale behavioural strategies during sensory evoked locomotion

Survival requires the animal brain to generate flexible and adaptive behavioral mechanisms on multiple timescales. To identify the internal states that modulate behavior across scales, we require an understanding of how short timescale behaviors are chained to generate long sequences. We leverage the larval zebrafish to characterize such sequences because of the availability of high-resolution, high throughput recordings in various sensory conditions. Larval zebrafish naturally swim in short timescale burst like motions called bouts. We construct a maximally predictive state space by stacking consecutive bouts, and then study the time evolution of state-space densities through transfer operators. Their spectral decomposition reveals slowly decaying modes corresponding to stereotyped long timescale behaviors. We find two long-lived strategies in larval zebrafish locomotion lasting tens to hundreds of seconds which occur naturally during exploratory behavior in light - "Roaming", which causes fast changes in orientation and "Cruising", which is dominated by forward locomotion. Our analysis reveals how stimuli modulate such long timescale behaviors by either triggering or driving the fish into roaming or cruising. We discover a clear structure in larval zebrafish behavior at long timescales and how it is modulated by external stimuli, enabling discovery of the internal states regulating behavior.

Invited Talk

Mechanical and molecular control of cell division dynamics in stem cell

Cell division allows the formation of two new daughter cells. The final step of division, after the chromosomes have segregated, is the separation of the cytoplasm of the two cells. This process starts with the ingression of the cleavage furrow and ends in abscission, a membrane scission event which finally isolates the two daughter cells. Abscission is a stepwise process culminating in the recruitment of the ESCRT-III complex that drives membrane scission. In cultured cells, abscission can be delayed when the so-called abscission checkpoint or NoCut checkpoint is active. This checkpoint relies on the ectopic activity of the kinase Aurora B which inhibits proper localization and/or assembly of ESCRT-III complex. This pathway can be activated as a consequence of mitotic defects, such as lagging DNA between the two sisters cells and impaired nuclear pore complex assembly. A high membrane tension can also delay abscission. We recently showed that pluripotent mouse embryonic stem cells (mESC) undergo a slow abscission, while differentiating cells undergo faster abscission. Yet, nothing is known about the mechanisms for delayed abscission in mammalian pluripotent stem cells. Here, we will show that Aurora B activity controls abscission dynamics in stem cells. We also present data suggesting that a change in tension could be causing the delay in abscission in naive cells. We are currently investigating the possible relation between tension and P-Aurora B activity.

Short Talks

Physical basis of the cell size scaling laws

The dimensions and compositions of cells are the result of the interplay between active biological processes and passive physical constraints. This interplay is embodied in the robust scaling laws relating cell size, dry mass, and nuclear size. Despite accumulating experimental evidence, the physical basis of these laws is still unclear. In this talk, I will show that these laws and their breakdown can be explained quantitatively by three simple, yet generic, physical constraints defining altogether the Pump-Leak model. Based on estimations, I will clearly map the Pump-Leak model coarse-grained parameters with the dominant cellular events they stem from. I will propose that dry mass density homeostasis arises from the scaling between proteins and small osmolytes, mainly amino-acids and ions. Our theory predicts this scaling to naturally fail, both at senescence when DNA and RNAs are saturated by RNA polymerases and ribosomes respectively, and at mitotic entry due to the counterion release following histone tail modifications. Based on the same equations, I will further show that nuclear scaling requires osmotic balance at the nuclear envelope (NE) and a large pool of metabolites, which dilutes chromatin counterions that do not scale during growth. I will finally end my talk by showing that the osmotic balance at the NE is likely the result of a mechanical instability.

A blueprint of the topology and mechanics of the human ovary for next-generation bioengineering and diagnosis

Although the first dissection of the human ovary dates back to the 17th century, the biophysical characteristics of the ovarian cell microenvironment are still poorly understood. However, this information is vital to deciphering cellular processes such as proliferation, morphology and differentiation, as well as pathologies like tumor progression, as demonstrated in other biological tissues. Here, we provide the first readout of human ovarian fiber morphology, interstitial and perifollicular fiber orientation, pore geometry, topography and surface roughness, and elastic and viscoelastic properties. By determining differences between healthy prepubertal, reproductive-age, and menopausal ovarian tissue, we unravel and elucidate a unique biophysical phenotype of reproductive-age tissue, bridging biophysics and female fertility. While these data enable to design of more biomimetic scaffolds for the tissue-engineered ovary, our analysis pipeline is applicable for the characterization of other organs in physiological or pathological states to reveal their biophysical markers or design their bioinspired analogs.

Carbon footprint of our lab: how to do our share?

Invited Talk

Finite Strain Analysis reveals mechanical basis and evolutionary conservation in ATP synthase function

One of the tenets of molecular biology is that the function of proteins and their assemblies is determined by changes in their structure. However, despite a wealth of "structural snapshots" for functional assembly states, their comparative analysis remains predominantly qualitative. We adapted the formalism of Finite Strain Analysis (FSA) to the study of structural changes in protein assemblies. As an application, we studied the ATP synthase functional cycle. Without any a priori knowledge, our method locates strain in relevant regions, e.g. near the site of ATP synthesis. We further compared strain patterns of ATP synthase structures from different species. This showed that strain, unlike e.g. displacement, is conserved in functional transitions across species. Our results support FSA as a robust method to analyze protein structural changes, and suggests that strain is a structural property that is evolutionary conserved in functional transitions.

Short Talks

Quantitative Characterisation and Modelling of Genetic AND-Gates: From Robust Molecular Models to Predictive Design of Synthetic Gene Circuits

In the past few decades, the rapid progression of synthetic biology has brought along the development of synthetic gene circuits, which have grown from simple transcriptional cascades to much more sophisticated architectures often involving boolean logic gates. This in turn has opened the way to new potential therapeutic strategies, with notable examples in gene therapy and cancer treatment. Synthetic gene circuits are commonly engineered using well-described building blocks, promising a highly tunable and reliable behaviour for high-specificity cell targeting. However, implementing circuits in mammalian cells remains challenging and precise response functions are still hard to predict. An important effort is now given to the rationalisation of network design, coupling model-based prediction and systematic measurement of circuits responses, in order to optimise existing circuits or create de novo functional devices. To this aim, we are developing an experimental platform combining multi-colour fluorescence imaging, pixel-correlation analysis and thermodynamic modelling, in order to build a predictive model of AND-gate like circuits based on the dimerisation of two independent proteins into a synthetic transcription factor (sTF). Here, we present the study of two systems. Our first results, extracted from a simplified gene cascade, yield interaction energies between an active or inactive sTF and different synthetic promoters. Through the variation of single molecular parameters, we were able to gain insight on how our system works, thus increasing the robustness of our theoretical model. Using the combinatorial assembly of dimer-affinity mutants, we are currently experimentally characterising variants of AND-gate circuits. This ongoing work will allow the construction of a comprehensive model able to predict the design of circuits with advanced functionalities such as complex cell-state classifiers, multi-output switches or oscillators.

Optogenetic control of antagonist functions with one protein

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 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 activated 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 switch in the phenotype induced by the recruitment of 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 phenotypes.

Mutational paths with sequence-based models of proteins: from sampling to mean-field characterisation

Identifying and characterizing mutational paths is an important issue in evolutionary biology and in bioengineering. We here propose an algorithm to sample mutational paths, which we benchmark on exactly solvable models of proteins in silico, and apply to data-driven models of natural proteins learned from sequence data with Restricted Boltzmann Machines. We then use mean-field theory to characterize the properties of mutational paths for different mutational dynamics of interest, and show how it can be used to extend Kimura's estimate of evolutionary distances to sequence-based epistatic models of selection.

Invited Talk

From topological defects to fruiting bodies in bacterial colonies

The soil bacterium Myxococcus xanthus lives in colonies of millions of cells that migrate on surfaces. When nutrients are scarce, the colony develops droplet-like multicellular aggregates called fruiting bodies, which allow cells to resist starvation. The colony builds these three-dimensional structures by sequentially adding one cell layer on top of each other. But how do bacteria manage to form new cell layers? I will show that the bacterial colony organizes into an active liquid crystal, and that its topological defects promote the formation of new cell layers. Thus, our work shows how bacteria exploit the physics of active matter in their collective response to starvation.

Short Talks

Cells distinguish between concave and convex curvatures while migrating on curved surfaces

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 /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.

Interplay between patterning and mechanics during amniote gastrulation

Large-scale morphogenetic movements and specification of embryonic territories are the hallmarks of gastrulation in avian embryos. Previous work from our lab has shown that mechanics regulates both of these processes. First, active fluidization mediated by cell division has been shown to be a key controller of morphogenetic flows within the epiblast during gastrulation. Second, the redirection of tissue motion by various mechanical perturbations in live embryos reroutes cell fate specification to eventually redefine embryonic territories. To investigate how tissue mechanics coordinate morphogenetic flows and embryonic territories' specification, we developed a piezo-actuated micromanipulator that can interact with developing embryos. Using such a device, we are able to measure the mechanical properties of live embryonic tissues but also to impose controlled and localized deformation. We are now testing the effect of mechanical perturbations on critical gene expression.

Boundary-driven epithelial ordering: from the mouse embryo to topological defects

In physical problems boundaries are typically considered to be simple, static and externally fixed. Biological systems however not only interact with their surroundings, but also alter them in ways that feed back on their own dynamics. We address this complex interaction in the context of epithelial development. Motivated by observations of an interplay between apico-basal polarity and boundary geometry in mouse epiblast morphogenesis, we develop a theory for epithelial ordering based on the Landau-de Gennes approach to surface-induced order in liquid crystals. We introduce a vector order parameter to represent the polarity, and model its interaction with the boundaries by a weak anchoring energy. We calculate the alignment fields arising from different boundary curvatures, and compare our predictions with imaging data of the morphogenetic process. Our work highlights the role of extra-embryonic tissue in embryogenesis, while identifying interesting physical phenomena such as boundary-dependent transitions in the structure of topological defects.