Le rayonnement synchrotron est l'un des facteurs clés du grand succès de la cristallographie macromoléculaire au cours des dernières décennies. Plus de 90% de toutes les structures de protéines de la base de données PDB a été résolu par cristallographie en utilisant des sources de rayonnement synchrotron et environ 95% d'entre elles a été déterminé à partir de cristaux congelés.Cependant, les structures déterminées par des techniques de congélation sont limitées par la nature statique des cristaux congelés. Avec le développement récent des sources de RX produites par lasers à électrons libres (XFEL), qui sont en mesure de produire des impulsions femtosecondes très intenses de rayons X, l'ère de la « diffraction avant destruction » et de la cristallographie sérielle femtoseconde utilisant des micro- ou nano- cristaux a commencé (SFX).Au cours du procédé SFX un cristal de protéine n'est exposé qu'une fois au faisceau de rayons X pendant quelques dizaines de femtosecondes avant qu'il ne soit complètement détruit. Les données sont collectées à partir de cristaux orientés de façon aléatoire par rapport au faisceau de rayons X; une seule exposition par cristal est possible. Afin de recevoir un ensemble de données plus complet, de nouvelles techniques d'analyse de données de diffraction ont été développées.La première expérimentation réussie du procédé (SFX) a été réalisée au LCLS à Stanford en décembre 2009 sur des cristaux du photosystème I et de lysozyme. Les experts en cristallographie des installations XFEL peuvent déterminer des structures de protéines à température ambiante presque exemptes de dégâts d'irradiation, en raison des pulses de FEL femtosecondes si brèves, qu'ils passent à travers l'échantillon avant que des dommages de rayonnement importants ne se produisent.Après la présentation des premières expériences SFX réussies, des efforts pour effectuer une cristallographie en série de cristaux de taille de l’ordre des micromètres à température ambiante ont commencé au sein des synchrotrons. Une première tentative de cristallographie synchrotron en série et à température ambiante a été tentée à PETRA III à DESY à Hambourg. Cette méthode a été nommée SMX (synchrotron serial millisecond crystallography), où des milliers d'échantillons sont collectés à partir de cristaux individuels passant par le faisceau à rayons X. Avec le développement des techniques de cristallographie série à température ambiante au sein des synchrotrons, la répartition de la dose sur un grand nombre de cristaux compense l’augmentation des dommages liés à l’irradiation à température ambiante.Bien que les sources de rayonnement synchrotron n'atteindront probablement jamais la même luminosité que les impulsions de rayons X comme dans les XFELs, elles ont un certains nombre d'avantages. L'un d'eux est la flexibilité de configuration grâce au paramétrage de lignes de faisceaux microfocus. Un autre est que plusieurs expositions par cristal sont possibles. En outre, les synchrotrons sont plus répandus dans le monde: la probabilité d'obtenir un temps d'expérimentation dans un synchrotron est plus élevée que pour une installation XFEL. En effet, maintenant deux installations XFEL sont ouvertes pour les utilisateurs .L'objectif de cette thèse est de proposer et de mettre en œuvre des méthodes qui permettront de recueillir des données en utilisant l'approche de la cristallographie série à l'installation synchrotron européen (ESRF, Grenoble, France). Cette thèse présente différentes techniques pour réaliser la cristallographie sérielle sur la ligne ID13 à l’ESRF. L'objectif était de développer la cristallographie sérielle sur synchrotron basée sur la numérisation micro-diffraction pour démontrer que les sources de rayonnement synchrotron peuvent être utilisées comme un instrument de routine pour cette technique avec des protéines globulaires et membranaires. Les aspects de la collecte de données et leur traitement seront également discutés. / Synchrotron radiation is one of the key factors for the tremendous success of macromolecular crystallography during the past decades. More than 90 % of all protein structures in PDB database were solved by crystallography using synchrotron radiation sources and around 95 % of them were determined from cryocooled crystals1,2. A whole data set can be collected from one flash-cooled crystal. Data-collection at cryogenic temperatures drastically reduces radiation damage effects. However, structures determined using cryo freezing techniques are limited by static nature of frozen crystals.With the recent development of X-ray free-electron laser facilities (XFELs), which are able to produce extremely intense femtosecond X-ray pulses, the era of “Diffraction before destruction” and serial femtosecond crystallography (SFX) for micro-/nano-sized crystals has begun3. In the SFX technique a protein crystal is only exposed once to the X-ray beam for tens of femtoseconds before it is completely destroyed. The data is collected from randomly oriented crystals that are exposed to the X-ray beam; only one shot per crystal is possible. In order to receive a complete data set, new data analysis techniques that are capable of dealing with large quantities of diffraction data have been developed. First experiment where serial femtosecond crystallography (SFX) approach was first carried out was performed at the Linac Coherent Light Source (LCLS, Stanford, USA) in December 2009 on photosystem I and lysozyme crystals4,5,6. At XFELs facilities crystallographers can perform room temperature structure determination of proteins almost free of radiation damage, due to the fact that femtosecond flashes of FEL is so brief, that it passes through the sample before the significant radiation damage occurs.After presenting first successful experiments with SFX technique, efforts to perform serial crystallography of micron-sized crystals at room-temperature started at synchrotron sources. First attempt to perform synchrotron room-temperature serial crystallography has been done at PETRA III at DESY in Hamburg using glass capillary based microfluidics7. This method was named synchrotron serial millisecond crystallography (SMX), where thousands of patterns are collected from individual crystals passing through the X-ray beam8. With development of room-temperature serial crystallography techniques at the synchrotrons, the exposure distributed on a large number of crystals in the sample, which helps compensating the effect of increased radiation damage at ambient temperature.Although synchrotron sources most certainly will never reach the same brightness of X-ray pulses like XFELs, they have some advantages. One of them is possibility to perform any set up due to the flexibility of the parameters of microfocus beamlines. Another advantage of SMX is that several shots per crystals are possible. Also should be mentioned that synchrotrons are more widespread all over the world, so possibility to get a beamtime at the synchrotron is much higher than at XFELs, because currently only two XFEL facilities are open for users (LCLS, USA and SACLA, Japan).The aim of this dissertation is to propose and to implement methods that will allow to collect data using the serial crystallography approach at the European synchrotron radiation facility (ESRF, Grenoble, France). This dissertation presents different techniques to perform synchrotron serial crystallography at ID13 beamline. The goal was to develop synchrotron serial crystallography based on scanning micro-diffraction to demonstrate that synchrotron sources can be used as a routine instrument to perform serial crystallography with soluble and membrane proteins. The aspects of the data collection and data processing also will be discussed.

abstract: Sample delivery is an essential component in biological imaging using serial diffraction from X-ray Free Electron Lasers (XFEL) and synchrotrons. Recent developments have made possible the near-atomic resolution structure determination of several important proteins, including one G protein-coupled receptor (GPCR) drug target, whose structure could not easily have been determined otherwise (Appendix A). In this thesis I describe new sample delivery developments that are paramount to advancing this field beyond what has been accomplished to date. Soft Lithography was used to implement sample conservation in the Gas Dynamic Virtual Nozzle (GDVN). A PDMS/glass composite microfluidic injector was created and given the capability of millisecond fluidic switching of a GDVN liquid jet within the divergent section of a 2D Laval-like GDVN nozzle, providing a means of collecting sample between the pulses of current XFELs. An oil/water droplet immersion jet was prototyped that suspends small sample droplets within an oil jet such that the sample droplet frequency may match the XFEL pulse repetition rate. A similar device was designed to use gas bubbles for synchronized “on/off” jet behavior and for active micromixing. 3D printing based on 2-Photon Polymerization (2PP) was used to directly fabricate reproducible GDVN injectors at high resolution, introducing the possibility of systematic nozzle research and highly complex GDVN injectors. Viscous sample delivery using the “LCP injector” was improved with a method for dealing with poorly extruding sample mediums when using full beam transmission from the Linac Coherent Light Source (LCLS), and a new viscous crystal-carrying medium was characterized for use in both vacuum and atmospheric environments: high molecular weight Polyethylene Glycol. / Dissertation/Thesis / Doctoral Dissertation Physics 2015

abstract: The superior brightness and ultra short pulse duration of X-ray free electron laser (XFEL) allows it to outrun radiation damage in coherent diffractive imaging since elastic scattering terminates before photoelectron cascades commences. This “diffract-before-destroy” feature of XFEL opened up new opportunities for biological macromolecule imaging and structure studies by breaking the limit to spatial resolution imposed by the maximum dose that is allowed before radiation damage. However, data collection in serial femto-second crystallography (SFX) using XFEL is affected by a bunch of stochastic factors, which pose great challenges to the data analysis in SFX. These stochastic factors include crystal size, shape, random orientation, X-ray photon flux, position and energy spectrum. Monte-Carlo integration proves effective and successful in extracting the structure factors by merging all diffraction patterns given that the data set is sufficiently large to average out all stochastic factors. However, this approach typically requires hundreds of thousands of patterns collected from experiments. This dissertation explores both experimental and algorithmic methods to eliminate or reduce the effect of stochastic factors in data acquisition and analysis. Coherent convergent X-ray beam diffraction (CCB) is discussed for possibilities of obtaining single-shot angular-integrated rocking curves. It is also shown the interference between Bragg disks helps ab-initio phasing. Two-color diffraction scheme is proposed for time-resolved studies and general data collection strategies are discussed based on error metrics. A new auto-indexing algorithm for sparse patterns is developed and demonstrated for both simulated and experimental data. Statistics show that indexing rate is increased by 3 times for I3C data set collected from beam time LJ69 at Linac coherent light source (LCLS). Finally, dynamical inversion from electron diffraction is explored as an alternative approach for structure determination. / Dissertation/Thesis / Doctoral Dissertation Physics 2016

abstract: Serial crystallography (SX) is a relatively new structural biology technique that collects X-ray diffraction data from microcrystals via femtosecond pulses produced by an X-ray free electron laser (X-FEL) or by synchrotron radiation, allowing for challenging protein structures to be solved from microcrystals at room temperature. Because of the youth of this technique, method development is necessary for it to achieve its full potential. Most serial crystallography experiments have relied on delivering sample in the mother liquor focused into a stream by compressed gas. This liquid stream moves at a fast rate, meaning that most of the valuable sample is wasted. For this reason, the liquid jet can require 10-100 milligrams of sample for a complete data set. Agarose has been developed as a slow moving microcrystal carrier to decrease sample consumption and waste. The agarose jet provides low background, no Debye-Sherrer rings, is compatible for sample delivery in vacuum environments, and is compatible with a wide variety of crystal systems. Additionally, poly(ethylene oxide) which is amenable for data collection in atmosphere has been developed for synchrotron experiments. Thus this work allows sample limited proteins of difficult to crystallize systems to be investigated by serial crystallography. Time-resolved serial X-ray crystallography (TR-SX) studies have only been employed to study light-triggered reactions in photoactive systems. While these systems are very important, most proteins in Nature are not light-driven. However, fast mixing of two liquids, such as those containing enzyme protein crystals and substrates, immediately before being exposed to an X-ray beam would allow conformational changes and /or intermediates to be seen by diffraction. As a model, 3-deoxy-D-manno-2-octulosonate-8-phosphate synthase (KDO8PS), has been developed for TR-SX. This enzyme initializes the first step of lipopolysaccharide synthesis by a net aldol condensation between arabinose-5-phosphate, phosphoenol pyruvate, and water. During this reaction, a short lived intermediate is formed and has been observed on a millisecond timescale using other methods. Thus KDO8PS is an ideal model protein for studying diffusion times into a crystal and short mixing times (<10 ms). For these experiments, microcrystals diffracting to high resolution have been developed and characterized. / Dissertation/Thesis / Doctoral Dissertation Chemistry 2016

This project arose from the need to filter the sampled data and eliminate non-useful information in Serial Crystallography in Microfluidic Device (MFD)by using a portable optical detector placed around the channel. By testing sixteen different configurations, always using an LED as a source and a photodiode as a light sensor, changes in the channel due to the passage of air bubbles were detected. These changes corresponded to a 13,25% in relation to the changes due to light switching, with a gain factor of 10,11V/V. However, it was not sensitive enough to detect when a microcrystal passed through it, although it can detect bubbles and opens the door to design such sensors for these applications in the future.

Cette thèse porte sur deux aspects complémentaires, le développement et l’implémentation de nouvelles approches de cristallisation et de cristallographie sérielle ainsi que leur mise en œuvre dans l’étude structurale de complexes enzymes : ARNt. La cristallographie est la méthode la plus employée en biologie structurale, mais elle présente encore des points délicats. Plusieurs méthodes avancées ont été déployées dans ce travail pour y pallier qui ont conduit à la résolution de la structure de l’ARNt nucléotidyltransférase du psychrophile Planococcus halocryophilus et à l’étude de son adaptation structurale au froid ; des puces microfluidiques de cristallisation qui ont servi à la résolution de plusieurs structures à température ambiante par cristallographie sérielle ; enfin le Xtal Controller utilisé pour l’étude d’évènements de nucléation et de croissance cristalline dans un but de préparation d’échantillons pour analyse sous rayonnement XFEL. Entre autres systèmes biologiques, cette thèse présente la caractérisation de deux familles d’inhibiteurs visant les aspartyl-ARNt synthétases, notamment du pathogène Pseudomonas aeruginosa. / This thesis focuses on two complementary aspects, the development and implementation of new approaches of crystallization and of serial crystallography as well as their use in the structural study of enzymes/tRNA complexes. Crystallography is the most used method in structural biology, but it presents delicate points. Different methods were implemented in this work to overcome these points, which led to the resolution of the structure of the CCA-adding enzyme of the psychrophilic organism Planococcus halocryophilus and to the study of its structural adaptation to the cold; novel microfluidic crystallization chips that have been used for the resolution of several structures by serial crystallography at room-temperature; finally the Xtal Controller used for the study of nucleation and crystal growth events with the purpose of preparing samples for analysis under XFEL radiation. Among other biological systems, this thesis presents the study and characterization of two families of inhibitors targeting aspartyl-tRNA synthetases, including the one of the pathogenic organism Pseudomonas aeruginosa.