Measuring VHH Kinetics and Specificity at the Single-Molecule Level

Antibodies are a paradigm for high-affinity, protein-based binding reagents and are extremely important in biotechnological, diagnostic, and therapeutic applications. Of special interest are VHHs, recombinant variable domains from heavy-chain-only antibodies. VHHs have several advantages: their small molecular weight, superior solubility, and stability and clearance rate. One particular use of VHHs is their use in imaging, which requires tailoring affinity and specificity for their targets. Despite their benefits, VHHs have only recently been used in single-molecule assays, e.g. PAINT imaging, which requires careful tuning of their kinetic properties. In this work, we demonstrate the detection of VHH-antigen binding at the single-molecule level. We first demonstrate the measurement of interaction kinetics between an immobilised GFP target and an anti-GFP VHH (LaG-16) in solution, with similar results to a bulk-derived kinetics constant. We also demonstrate specific binding of different VHHs against their respective targets (GFP and Lysozyme) in a single experiment, as well as against different epitopes of the same target protein (Lysozyme), where we observe two different VHHs binding simultaneously.

Evolution and Epidemiology of Enterovirus D68

Enterovirus D68 (EV-D68), typically causing mild respiratory illness in children, gained prominence due to its association with acute flaccid myelitis (AFM), a severe paralytic condition. Historically, EV-D68 outbreaks followed a biennial pattern in Europe and North America, with expected outbreaks in 2020 halted by pandemic restrictions. Prior to the pandemic, we analyzed public data to study geographic mixing, antigenic evolution, and age-specific clade associations, finding that subclade A2 disproportionately affected older individuals. Post-pandemic, EV-D68 has resurfaced, but the impact of reduced circulation on its evolution, age demographics, and biennial cycle remains unclear. Drawing on both pre- and post-pandemic case data, I explore shifts in viral epitopes, diversification, and key open questions around immunity and reinfection.

The onset of thrombosis in deep veins (DVT): the coupling effect of stretch and shear stresses on endothelial cells

Deep vein thrombosis (DVT) is the formation of a thrombus in the valvular sinuses of the veins in the lower limbs. It is often associated with blood stasis during prolonged immobilization, however, the triggers for DVT are not well understood. Venous valvular sinuses experience unique blood flow patterns due to the cyclic opening and closing of the valve. We hypothesize that stretching helps maintain vein antithrombotic properties, and its absence could contribute to the onset of DVT. To test this idea, confluent human endothelial cells are cultured on hydrogel coated elastic membranes subjected to uniaxial cyclic stretching at rates of 60 cycles/minute. We study how different levels and duration of stretching influence the expression of two important proteins of the haemostatic system, Thrombomodulin (TM) and the von Willebrand factor (vWf ). Our results show that the cells elongate and align orthogonal to the stretch direction. Stretch amplitude of 10% induces a thromboresistant phenotype up-regulating TM levels by 75% within 6 hours and remaining high up to 24 hours. Interestingly, no significant change occurred at the 5% stretch even after 24 hours. When 10% stretching is interrupted after 24 hours, vWf levels measured 6 hours post interruption increase significantly, simulating a procoagulant state. Additionally, we show that stretching flattens the nuclei and aligns them orthogonal to the stretching direction. Notably, the epigenetic acetylation mark H3K27ac, which regulates TM gene expression, increases by 1.6-fold. Our findings suggest that cyclic stretch contributes to the regulation of endothelium thromboresistance and might prevent thrombosis. In parallel to stretch exploration, two fluidic devices were developed to explore the combined effect of flow and stretch. A microfluidic channel with a flexible membrane that deforms with flow, inducing stretch on cells and a thin-walled PDMS tube that can increase in diameter with pressure. Taken together, these findings empower the study of ECs in complex mechanical situations, highlighting the molecular link between stretch, shear and Thrombosis.

Optimal strategies in navigation and learning: statistical physics meets control theory

To survive, animals must acquire a diverse set of skills and integrate them to respond effectively to complex environments. For instance, to move in a straight line in rough terrains, dung beetles alternate between egocentric strategies, maintaining an internal estimate of position, and geocentric strategies, using landmarks for trajectory correction. In the first part of this talk, I will consider this behavior within a minimal model of navigation and derive the switching strategy that maximizes speed, accounting for environmental, execution, and sensory noise. In the second part of the talk, I will consider the complementary problem of learning multiple skills/tasks, a.k.a., continual learning. While animals typically excel at continual learning, artificial neural networks often forget older tasks when new ones are introduced. By combining exact training dynamics from statistical physics with optimal control methods, I will derive strategies that optimize performance while avoiding forgetting. This flexible framework reveals fundamental principles in multi-task learning and can be adapted across scales to optimize high-dimensional learning processes.

Brownian dynamic simulations of biological condensates forming around DNA

DNA double strand breaks are extremely harmful for the cells. DNA damage signaling, as well as the repair process, involves RNA strands and a large range of proteins. They are recruited at the damage site seconds after the break, in a very precisely orchestrated series of steps[1]. During this process, some of these proteins assemble into phase separated liquid droplets, also called biological condensates, around the damage site. The physical properties of these condensates determine their functions and are a key feature for their rapid formation and dissolution[2]. We aim here to study the physics at play behind the formation of biological condensates. In particular, the role played by the relative strength of protein-protein and protein-chromatin interactions and how this can be related to macroscopic physics of wetting. We then study how the chromatin geometry also impacts the wetting and the dynamic of the individual proteins. For this we study a coarse grained system composed of Lennard-Jones (LJ) spheres representing the proteins and a series of static LJ spheres to represent nuclesomes that form a fiber-like substrate. Brownian dynamic simulations are performed study both the meso-scale wetting behaviour of the proteins liquid droplet on the chromatin fiber and the microscopic dynamic of the proteins. References: [1] J. Miné-Hattab et al. eLife,2021 [2] A. A. Hyman et al. Annual review of cell and developmental biology, 2014

Microbial interactions, multistability, and ecological function in a multispecies context

Understanding microbial community dynamics and functions is key to predicting responses to environmental shifts, including those relevant to host health. First, I will discuss our work on a microbial community that exhibits multistability, where cooperative growth among species drives alternative stable states, an effect modulated by external factors such as glutamate availability. Next, I will present our research on how increasing community complexity influences stability, showing experimental transitions between distinct phases of community dynamics--as predicted by ecological theory. Lastly, I will outline ongoing efforts to predict community functions that emerge from interactions among diverse community members, with the potential to inform the rational design of microbial communities for targeted ecological outcomes, such as minimizing antibiotic resistance.

Optimal sequencing depth for measuring the concentrations of molecular barcodes

In combinatorial genetic engineering experiments, next-generation sequencing (NGS) allows for measuring the concentrations of barcoded or mutated genes within highly diverse libraries. When designing and interpreting these experiments, sequencing depths are thus important parameters to take into account. Service providers follow established guidelines to determine NGS depth depending on the type of experiment, such as RNA sequencing or whole genome sequencing. However, guidelines specifically tailored for measuring barcode concentrations have not yet reached an accepted consensus. To address this issue, we combine the analysis of NGS datasets from barcoded libraries with a mathematical model taking into account the PCR amplification in library preparation. We demonstrate on several datasets that noise in the NGS counts increases with the sequencing depth; consequently, beyond certain limits, deeper sequencing does not improve the precision of measuring barcode concentrations. We propose, as rule of thumb, that the optimal sequencing depth should be about ten times the initial amount of barcoded DNA before any amplification step.