Bacterial motility, and in particular repulsion or attraction towards specific chemicals, has been a subject of investigation for over 100 years, resulting in detailed understanding of bacterial chemotaxis and the corresponding sensory network in many bacterial species including Escherichia coli. E. Coli swims by rotating a bundle of flagellar filaments, each powered by an individual rotary motor located in the cell membrane. When all motors rotate counter-clockwise (CCW), a stable bundle forms and propels the cell forward. When one or more motors switch to clock-wise (CW) rotation, their respective filaments fall out of the bundle, leading to the cell changing orientation. Upon switching back to CCW, the bundle reforms and propels the cell in a new direction. Chemotaxis is performed by the bacterium through prolonging runs by suppressing CW rotation when moving towards nutrients and facilitating reorientation by increasing CW bias when close to a source of a harmful substance. Chemicals are sensed through interaction with membrane bound chemosensors. These proteins can interact with a very specific set of chemicals and the concentrations they are able to sense are in the range between 10-⁶ and 10-² M. However, experiments have shown that the osmotic pressure exerted by large (> 10-¹ M) concentrations of solutes, which have no specificity for binding to chemosensors (e.g. sucrose), is able to send a signal down the chemotactic network. Additionally, clearing of bacterial density away from sources of high osmolarity has been previously observed in experiments with agar plates. This behaviour has been termed osmotaxis. The aim of this doctoral thesis work is to understand how different environmental cues influence the tactic response and ultimately, combine at the network output to direct bacterial swimming. As tactic responses to chemical stimuli have been extensively studied, I focus purely on the response to non-specific osmotic stimuli, using sucrose to elevate osmolarity. I monitor the chemotactic network output, the rotation of a single bacterial flagellar motor, using Back Focal Plane Interferometry over a variety of osmotic conditions. Additionally, in collaboration with Vincent Martinez, I studied the effect of elevated osmolality on swimming speed of large (104) bacterial populations, using differential dynamic microscopy (DDM). I have found that sudden increases in media osmolarity lead to changes of both motor speed and motor clockwise bias, which is the fraction of time it spends rotating clockwise. Changes in CW Bias proceed in two phases. Initially, after elevating the osmolarity, CW Bias drops to zero, indicating that the motor is exclusively in the ‘cell run’ mode. This phase lasts from 2-5 minutes depending on the magnitude of the change in solute concentration. What follows then is a distinct second phase where the CW Bias is elevated with respect to the initial levels and this phase lasts longer than 15-20 minutes. In comparison, for defined chemical stimuli, the motor output resets after several seconds, a behaviour termed perfect adaptation. For changes of 100 mOsm/kg and 200 mOsm/kg in magnitude the motors speed up, often by as much as a factor of two, before experiencing a gradual slow down. Despite the slow down, motors still rotate faster 15-20 minutes after the change in osmolarity, than they did before. For changes of 400 mOsm/Kg in magnitude the motors decrease sharply in speed, coming to a near halt, recovering after 5 minutes and eventually, on average, speeding up. DDM studies of free swimming bacteria have shown that elevated osmolality leads to higher swimming speeds, in agreement with single motor data. Using theoretical models of bacterial swimming from the literature, it is discussed how this motor output, although different to what is expected for chemotaxis, is able to drive bacteria away from regions of space with high osmolalities. Additionally, I have started extending the work done with sucrose, to another solute often used to elevate osmolality, sodium chloride. While sucrose is outer membrane impermeable, NaCl can cross the outer membrane into the periplasmic space. Another layer of complexity is that NaCl has some specificty for the chemoreceptors. The preliminary results are shown and qualitatively agree with those obtain with sucrose.

Eukaryotic organisms rely on the coordinated beating of motile cilia for a multitude of fundamental reasons. In smaller organisms, such as Paramecium and the single cell alga Chlamydomonas reinhardtii, it is a matter of propulsion, to swim towards a higher concentration of nutrients or away from damaging environments. Larger organisms use instead the coordinated motion of cilia to push fluid along an epithelium: examples common to mammals are the circulation of cerebrospinal fluid in the brain, the transport of ovules in the fallopian tubes, and breaking the left/right symmetry in the embryo. Another notable example, and one that is central to this thesis, is mucociliary clearance in human airways: A carpet of motile cilia helps keeping the cell surface free from pathogens and foreign particles by constantly evacuating from lungs, bronchi, and trachea a barrier of mucus. The question of how motile cilia interact with one another to beat in a coordinated fashion is an open and pressing one, with immediate implications for the medical community. In order for the fluid propulsion to be effective, the motion of cilia needs to be phase-locked across significant distances, in the form of travelling waves (``metachronal waves''). It is still not known how this long-range coordination emerges from local rules, as there is no central node regulating the coordination among cilia. In the first part of this thesis I will focus on studying the coordination in carpets of cilia with a top-down approach, by proposing, implementing, and applying a new method of analysing microscope videos of ciliated epithelia. Chapter 1 provides the reader with an introduction on motile cilia and flagella, treating their structure and motion and reporting the different open questions currently tackled by the scientific community, with particular interest in the coordination mechanisms of cilia and the mucociliary clearance apparatus. Chapter 2 introduces Differential Dynamic Microscopy (DDM), a powerful and versatile image analysis tool that bridges the gap between spectroscopy and microscopy by allowing to perform scattering experiments on a microscope. The most interesting aspects of DDM for this work are that it can be applied to microscope videos where it is not possible to resolve individual objects in the field of view, and it requires no user input. These two characteristics make DDM a perfect candidate for analysing several hundred microscope videos of weakly scattering filaments such as cilia. In Chapter 3 I will present how it is possible to employ DDM to extract a wealth of often-overlooked information from videos of ciliated epithelia: DDM can successfully probe the ciliary beat frequency (CBF) in a sample, measure the direction of beating of the cilia, and detect metachronal waves and read their direction and wavelength. In vitro ciliated epithelia however often do not show perfect coordination or alignment among cilia. For the analysis of these samples, where the metachronal coordination might not be evident, we developed a new approach, called multiscale DDM (multiDDM), to measure a coordination length scale, a characteristic length of the system over which the coordination between cilia is lost. The new technique of multiDDM is employed in Chapter 4 to study how the coordination among cilia changes as a response to changes in the rheology of the mucous layer. In particular, we show that cilia beating under a thick, gel-like mucus layer show a larger coordination length scale, as if the mucus acted as an elastic raft effectively coupling cilia over long distances. This is corroborated by the coordination length scale being larger in samples from patients affected by Cystic Fibrosis than in healthy samples, and much shorter when the mucus layer is washed and cilia therefore beat in a near-Newtonian fluid. We then show how it is possible to employ multiDDM to measure the effectiveness of drugs in recovering, in CF samples, a coordination length scale typical of a healthy phenotype. In the second part I will focus instead on the single cilium scale, showing how we can attempt to link the beating pattern of cilia to numerical simulations studying synchronisation in a model system. In particular in Chapter 5 I will describe our approach to quantitatively describe the beating pattern of single cilia obtained from human airway cells of either healthy individuals or patients affected by Primary Ciliary Dyskinesia. Our description of the beating pattern, and the selection of a few meaningful, summary parameters, are then shown to be accurate enough to discriminate between different mutations within Primary Ciliary Dyskinesia. In Chapter 6 instead I report the results obtained by coarse-graining the ciliary beat pattern into a model system consisting of two ``rotors''. The rotors are simulated colloidal particles driven along closed trajectories while leaving their phase free. In my study, the trajectories followed by the rotors are analytical fits of experimental trajectories of the centre of drag of real cilia. The rotors, that are coupled only via hydrodynamics interactions, are seen to phase-lock, and the shape of the trajectory they are driven along is seen to influence the steady state of the system.

L'eau est un liquide omniprésent, son omniprésence n'a d'équivalence que la multitude de ses secrets. En effet, dans le cas de l'eau, le comportement de nombreuses grandeurs thermodynamiques et dynamiques se différencie de celui des liquides « standard ». Cette différence est d'autant plus importante dans l'état dit surfondue. Dans cette thèse, on s'intéresse à la viscosité de l'eau légère et lourde dans l'état surfondue à pression atmosphérique. Pour ce faire, nous avons utilisé la méthode de microscopie dynamique différentielle. Nous avons ainsi pu mesurer la viscosité jusqu'à -34°C pour l'eau légère et -25°C pour l'eau lourde. Ces mesures de viscosité corrélées à des mesures de coefficient d'auto-diffusion nous ont permis de confirmer l'existence d'une anomalie dite violation de Stokes-Einstein / Water is an omnipresent liquid, indeed secrets of water are uncountable. In the water case, the behaviour of many thermodynamical and dynamicalquantities is very different from other standard liquids. This difference is more important in the supercooled state. In this thesis we focus on the viscosity of the heavy and light water into the supercooled state at atmospheric pressure. To do this we use the differential dynamic microscopy method. We were able to measure the viscosity to -34 C for light water and -25°C for heavy water. These viscosity measurements correlated to measurements of self-diffusion coefficient allowed us to confirm the existence of a so-called anomaly violation Stokes-Einstein

Nanomedicine is based primarily on the concept of drug formulation through nanotechnology. The main idea is based on the encapsulation of an active ingredient by a nanoparticle (NP) to allow it to accumulate in tumors, to penetrate a biological barrier or to target a biological component. However, the performance of these formulations is disappointing, and, in recent years, it has been noticed that their effectiveness has not improved in the last decade. Some recent hypotheses highlight our lack of knowledge about the interactions of nanotechnologies with living organism and more particularly the lack of techniques to quantify these interactions. We therefore explore in this thesis the development and adaptation of a new microscopy technique, dynamic differential microscopy (DDM), to study the interactions of nanotechnologies with biological matrices. Two subjects are discussed, the first on the interactions of NPs with the proteins of biological fluids and, the second one, on the capacity of NPs to diffuse in interstitial tissues. First, we reviewed quantification techniques that were allowing the measurement of protein adsorption at the surface of NPs. We then identified fundamental questions of this adsorption, namely, if it was generally structured in monolayers or in multilayers and if it was reversible or irreversible. A meta-analysis, based on these questions, could therefore guide the development of the DDM technique to measure protein adsorption and therefore answer these questions. The methodology proposed for the quantification of protein adsorption is based on the measurement of the fluorescence signal which comes from fluorescently tagged proteins adsorbed on non-fluorescent NPs. This methodology was successfully applied for the quantification of the adsorption of lysozyme, albumin and serum proteins. The technique demonstrated that all the proteins studied adsorbed in monolayers and that their adsorption was reversible. An atypical adsorption mechanism which was also hypothesized in our meta-analysis was evidenced by DDM as well. Next, we applied DDM to study the diffusion of NPs in extracellular matrices. The contribution of deformability has been a parameter studied in terms of its relation to improve their diffusion within these confined environments. The diffusion of "soft" NPs was compared to that of "hard" NPs in an agarose gel, mimicking the extracellular matrix. Soft NPs have been observed to diffuse up to 100 times faster than hard NPs of the same size. Evaluation of the hydrodynamic and electrostatic contributions determined that the soft NPs shrinks in the gel, boosting their diffusion in comparison to hard NPs. In summary, this work highlights the important contribution of analytical techniques to the field of nanotechnologies applied to pharmacy and to our understanding of their interactions with living organisms. It is clear that the contribution of these techniques to our detailed understanding of nanomedicine properties has a direct relation with their clinical translation potential. / La nanomédecine repose essentiellement sur le développement de nouvelles formulations pour délivrer les médicaments à partir de nanotechnologies. L’idée principale est que l’encapsulation d’un principe actif par une nanoparticule (NP) pourrait lui permettre de s’accumuler dans des tumeurs, de pénétrer une barrière biologique ou bien pour cibler une composante biologique. Or, les performances de ces « nano-formulations » sont décevantes et, depuis quelques années, il a été remarqué que leur efficacité ne semble pas avoir évoluée dans le temps. De récentes hypothèses mettent de l’avant notre manque de connaissances vis-à-vis les interactions des nanotechnologies avec les éléments du vivant, et plus particulièrement, le manque de techniques robustes permettant de quantifier ces interactions. Nous proposons donc dans cette thèse le développement et l’adaptation d’une nouvelle technique de microscopie, la microscopie différentielle dynamique (DDM), pour étudier les interactions entre les nanotechnologies et les matrices biologiques. Deux thématiques seront abordées, la première, les interactions des NPs avec les protéines des fluides biologiques et, la seconde, la capacité des NPs à diffuser dans des tissus interstitiels. D’abord, nous avons revus les techniques de quantification permettant la mesure de l’adsorption de protéines à la surface des NPs. Nous avons ensuite identifié les questions fondamentales en lien avec cette adsorption. Deux phénomènes sont largement débattus dans la littérature, il s’agit de la formation de multicouches et de la réversibilité de l’adsorption. Une méta-analyse a donc permis d’orienter le développement de la technique par DDM pour mesurer l’adsorption de protéines, dans le but de répondre à ces interrogations. La méthodologie proposée pour la quantification de l’adsorption de protéines à la surface des NPs repose sur la mesure du signal de fluorescence de protéines fluorescentes adsorbées à la surface des NPs non fluorescentes. Cette méthodologie a été appliqué avec succès pour la quantification de l’adsorption des protéines du sérum, du lysozyme et de l’albumine. La technique a d’ailleurs permis de montrer que toutes les protéines étudiées s’adsorbaient en monocouches et que leur adsorption était réversible. Un mécanisme d’adsorption atypique a été mis en évidence dans le cadre de nos expériences et un parallèle a pu être fait avec certaines hypothèses émises avec notre méta-analyse. Ensuite, nous avons appliqué la DDM pour l’étude de la diffusion des NPs dans des matrices extracellulaires. La déformabilité des NPs a été étudiée afin de définir plus précisément sa contribution dans la diffusion à l’intérieur de milieux confinés. La diffusion des NPs « molles » a été comparée à celle des NPs « dures » dans un gel d’agarose, mimant la matrice extracellulaire. Les NPs molles ont été en mesure de diffuser jusqu’à 100 fois plus rapidement que les NPs dures de même taille. L’évaluation des contributions hydrodynamiques et électrostatiques a permis de déterminer que la taille des NPs molles, réduisant dans le gel, leur accordant un avantage diffusif par rapport aux NPs dures. En sommes, ces travaux ont permis de mettre en évidence l’importance des techniques analytiques pour l’étude des nanotechnologies appliquées à la médecine et pour affiner notre compréhension de leurs interactions avec le vivant. Il est clair que la contribution de ces techniques à l’avancement de nos connaissances théoriques relatives aux nanotechnologies aura un impact direct sur leurs chances d’effectuer une transition vers la clinique.