Antimicrobial resistance is a major threat worldwide that requires a strong investment in fundamental studies. Resistance, as well as transient tolerance (persistence) to antibiotics, involve a network of intracellular stress responses: e.g., the stringent response, the SOS response, and the RpoS-regulated general stress response. We and others have shown that these stress responses are induced by antibiotic (AB) doses below the minimum inhibitory concentration (sub-MIC) and that they can accelerate acquisition of heritable AB resistance through increased mutagenesis and horizontal gene transfer (HGT). Although low concentrations of antibiotics do not kill bacteria, they can have a major impact on bacterial populations. In particular, it was shown that AB concentrations as low as hundred-fold below the MIC can lead to mutations and the selection of AB resistant cells. Most of the studies describing how bacteria acquire resistance or become persisters are based on experiments dealing with populations of cells. Such measurements yield average quantities for the whole population but they cannot provide a distribution of responses, nor can they follow the temporal evolution of individuals within the population. By contrast, there is mounting evidence that cells within a given population can display widely heterogeneous responses to an AB stress. This project aims at describing precisely individual cell fate during stress responses to low doses of antibiotics, and understanding the emergence of antibiotic resistance on the level of a single cell. We propose to address the profile of induction of four stress responses at the single-cell level: SOS, stringent response, RpoS general stress response and oxidative stress response, in response to three ABs from different families (fluoroquinolones, aminoglycosides, ß-lactams). To this end, we will use a microfluidic platform to culture bacteria, while submitting them to controlled AB stresses to assess heterogeneity and growth on chip. We will develop the theoretical description of bacterial growth dynamics taking into account the AB stress through mathematical modelling relating large-scale heterogeneity to the variability on the scale of individual cells. We will then isolate and extract cells that show phenotypic diversity. The large statistics will allow us to get access to rare events. The extracted cells will be subjected to analysis (NGS, dPCR) in order to detect horizontal gene transfers, mutations or changes of protein expression that can explain the behavior of these cells. This will first require the development of technological tools to genotype the small number of bacterial cells that can be recovered from the microchannel. The second step will be to explore different conditions that lead to the emergence of antibiotic resistance in order to gain insight into the underlying mechanisms and devise strategies to counter them. The impact of this project will be threefold: (i) Concerning the fundamental biological knowledge it will bring, (ii) the technological and quantitative developments that accompany it, and (iii) in understanding the emergence of resistance mechanisms and their implications for the development of new therapeutic strategies.

SCALP considers complex systems being governed by non-linear multiscale dynamics with non-local interactions between scales. These systems present extreme events in their dynamics as well as long-range dependences. Some examples of this kind of processes are fluid turbulence, ocean and atmosphere dynamics, stock markets, avalanche processes. They usually resist traditional analytical/simulation approaches, lacking a reliable and tractable physical model needed to perform any forecasting tasks. High dimensional chaotic systems may display a dynamics so intricated that no analytical theory is available, and the building of an explainable minimal model should provide insights. ML approaches are appealing in this context but require the design of new ML architectures exhibiting specific physics informed properties such as scales symmetries and conservation laws. Furthermore, observations are often scarce, eliminating de-facto too large architectures. Our aim is to develop deep but interpretable models with relevant scale symmetries and respecting physical conservations laws. As study cases we will be focusing on various multiscale dynamical models of turbulence (shell models, wave turbulence, fluid turbulence), chaotic models of atmospheric dynamics, and avalanche processes. Then, real data from observations, experiments and direct numerical simulations will be studied. The long-term impacts of SCALP concern contexts not limited to physics, where bigger non-interpretable networks is not the answer.

How do tissues and organs grow to reach well-defined shapes? For instance, thin tissues, such as animal epithelia and plant leaves, are typically flat, whereas the default state of a growing thin sheet is curved. How is flat shape achieved? Here we address this question in Arabidopsis leaves, a system amenable to live-imaging of growth, biophysical experiments, and genetic manipulation. We hypothesize that cell-to-cell growth heterogeneity enables cells to sense variations in leaf curvature and maintain flatness. We aim to test this hypothesis using experimental and theoretical biophysics with cell biology. We will (i) characterize the relationship between cell-to-cell heterogeneity, cell mechanics, and leaf flatness, (ii) build a theoretical framework to model a thin active growing sheet in 3D space, and (iii) characterize the combinatorial regulation of flatness through model predictions and experimental tests. Altogether, we expect to shed light on the robustness of morphogenesis.

Plants continuously develop throughout their lifetime through the apical meristems located at the tip of growing axes. The genetic regulation of the shoot apical meristems (SAMs), which produces all plant aerial parts, has been extensively studied in model plants, and various key molecular actors have been identified and mapped in space and time. Based on these results, recent works have investigated how these molecular actors induce physical deformation of tissues by modifying cell wall mechanical properties, and in turn result in leaf or flower primordia outgrowth. In most of these studies, cell turgor pressure, that drives wall deformation, is assumed to be constant and homogeneous. Here, with a pluri-disciplinary team, we want to map mechanical and hydraulic properties in tissues and develop computational models to better understand how water pressure builds up in meristem parts. We will challenge this view by using mutants and treatments modifying cell conductivity and mechanics.

Engineering strategically uses compliant components in the design of structures exposed to flows, as those can change their shape and adapt to their surrounding fluid environment. While a flexible structure is more adaptable and versatile, it is also more difficult to control. We thus have to understand the way it deforms and find levers to control it. In this project, we will explore an unconventional route to tailor the deformation of surfaces in a flow, making use of the unique properties of origami (folded sheet) and kirigami (sheet with a network of cuts). Previous literature showed that the meso-structure of folds or cuts allows for surfaces to morph into sophisticated shapes, and produces programmable non-linear mechanical properties. The objective of this project is to study how those original features impact the way the structure interact with a flow, and how it can be harnessed to produce novel mechanical behaviors. Origami and kirigami provide opportunities to revisit fluid/structure interaction out of its regular framework, paving the way for controlled deployment pathways in flows.