The single-scattering properties of six non-spherical ice crystals, droxtals, plates, solid columns, hollow columns, aggregates and 6-branch bullet rosettes are simulated. The anomalous diffraction theory (ADT) is applied to the simulation of the extinction efficiency and the absorption efficiency. Because the first order reflection is considered, the accuracy of the absorption efficiency increases with the increasing of the size parameter. Compared with the reference single-scattering properties from an improved geometric optics method (IGOM), the errors in the extinction and absorption efficiencies are small. In addition, the asymmetry factor is formulated within the framework of diffraction and external reflection. The asymmetry factor based on the ADT matches very well with the IGOM counterpart when the absorption is strong, but needs an improvement in the solar region. The errors in conjunction with the application of the ADT-based optical properties to the computation of atmospheric fluxes and heating rates, based on the Fu-Liou model also are investigated. Two cases, one for tropical cirrus clouds and the other for mid-latitude cirrus clouds, are designed. It is found that the errors of bulk asymmetry factor between ADT-based and IGOM-based result in an overestimation of downward infrared (IR) fluxes and upward solar fluxes, and an underestimation of upward IR fluxes and downward solar fluxes. The errors of the fluxes and heating rates based on two sets of single-scattering properties are caused mainly by the underestimation of the bulk absorption efficiency based on ADT. It is also shown that ADT-based optical properties generate more accurate radiative properties for tropical cirrus clouds than for the mid-latitude cirrus clouds. In conclusion, the ADT-based method can generate reasonably accurate single-scattering properties of ice crystals, and can result in reasonable upward IR and solar fluxes at top of atmosphere (TOA), downward IR fluxes at the surface, and net heating rates.

Conformational changes in human proteins can induce several types of diseases. The nature of the conformational changes is largely unknown, but some lead to amyloid fibril formation. Amyloid fibrils accumulate in the extra-cellular space of tissues resulting in disruption of organ function. Transthyretin (TTR) is a plasma protein involved in three amyloid diseases, familial amyloidotic polyneuropathy, familial amyloidotic cardiomyopathy, and senile systemic amyloidosis. The latter disease involves conformational changes in the wild-type structure of the protein, whereas the others are caused by a gene mutation. Our goal is to increase the knowledge of why and how some proteins aggregate into amyloid fibrils by solving and analyzing structures of different TTR variants of which some can form amyloid fibrils, whereas others cannot. The crystal structures of wild-type TTR and many of its disease-causing mutants have previously been determined, and observed structural discrepancies between mutant and wild type were claimed to be of importance for amyloid formation. We performed a comparative analysis of all, at that point, known structures of TTR. As a reference for our study, we determined a 1.5 Å resolution structure of human wild-type TTR. We found that the previously reported structural differences between wild type and mutant TTR were insignificant and did not provide clues to the mechanism for amyloid formation. We showed the double mutant TTR-Ala108Tyr/Leu110Glu to be less amyloidogenic than wild-type transthyretin. Since the structure of few non-amyloidogenic mutants are known, we solved its structure in two space groups, C2 and P21212, where the latter was consistent with most of the structures of transthyretin. Only the highly amyloidogenic mutant ATTR-Leu55Pro has previously been solved in C2. The packing of molecules in our C2 crystal was close-to-identical to the ATTR-Leu55Pro crystal structure, ruling out the described ATTR-Leu55Pro packing interactions as significant for amyloidosis. The C2 structure displayed a large shift in residues Leu55-Leu58, a structural change previously found only in amyloidogenic TTR variants. Combined with previous data, this suggests that transthyretin in solution contains a mixture of molecules with different conformations. This metastability of transthyretin provides insight to why some proteins aggregate into amyloid fibrils. The natural ligand thyroxine has been shown to stabilize TTR. Small molecules, based on thyroxine, with the potential to serve as inhibitors for amyloid fibril formation are under development. Iodine is a component of thyroxine and we found that TTR also bound free iodide ions. Taking advantage of the anomalous scattering of iodide, we solved the iodide-bound TTR structure using the single-wavelength anomalous dispersion method. In addition, we determined the TTR-chloride structure. Both chloride and iodide stabilized transthyretin where iodide stabilized better. From the thyroxine-TTR structure, three halogen-binding pockets have been identified in each TTR monomer. We found three bound iodides per TTR monomer, two of which were in the thyroxine-binding channel. This indicates that only two of the three halogen-binding pockets in the thyroid-hormone binding channel are optimal for halogen binding. Our results might be useful for the continuing design of small molecule ligands, which in the end can lead to inhibitors for amyloid diseases.

<p>Conformational changes in human proteins can induce several types of diseases. The nature of the conformational changes is largely unknown, but some lead to amyloid fibril formation. Amyloid fibrils accumulate in the extra-cellular space of tissues resulting in disruption of organ function. Transthyretin (TTR) is a plasma protein involved in three amyloid diseases, familial amyloidotic polyneuropathy, familial amyloidotic cardiomyopathy, and senile systemic amyloidosis. The latter disease involves conformational changes in the wild-type structure of the protein, whereas the others are caused by a gene mutation. </p><p>Our goal is to increase the knowledge of why and how some proteins aggregate into amyloid fibrils by solving and analyzing structures of different TTR variants of which some can form amyloid fibrils, whereas others cannot. The crystal structures of wild-type TTR and many of its disease-causing mutants have previously been determined, and observed structural discrepancies between mutant and wild type were claimed to be of importance for amyloid formation. We performed a comparative analysis of all, at that point, known structures of TTR. As a reference for our study, we determined a 1.5 Å resolution structure of human wild-type TTR. We found that the previously reported structural differences between wild type and mutant TTR were insignificant and did not provide clues to the mechanism for amyloid formation.</p><p>We showed the double mutant TTR-Ala108Tyr/Leu110Glu to be less amyloidogenic than wild-type transthyretin. Since the structure of few non-amyloidogenic mutants are known, we solved its structure in two space groups, C2 and P21212, where the latter was consistent with most of the structures of transthyretin. Only the highly amyloidogenic mutant ATTR-Leu55Pro has previously been solved in C2. The packing of molecules in our C2 crystal was close-to-identical to the ATTR-Leu55Pro crystal structure, ruling out the described ATTR-Leu55Pro packing interactions as significant for amyloidosis. The C2 structure displayed a large shift in residues Leu55-Leu58, a structural change previously found only in amyloidogenic TTR variants. Combined with previous data, this suggests that transthyretin in solution contains a mixture of molecules with different conformations. This metastability of transthyretin provides insight to why some proteins aggregate into amyloid fibrils.</p><p>The natural ligand thyroxine has been shown to stabilize TTR. Small molecules, based on thyroxine, with the potential to serve as inhibitors for amyloid fibril formation are under development. Iodine is a component of thyroxine and we found that TTR also bound free iodide ions. Taking advantage of the anomalous scattering of iodide, we solved the iodide-bound TTR structure using the single-wavelength anomalous dispersion method. In addition, we determined the TTR-chloride structure. Both chloride and iodide stabilized transthyretin where iodide stabilized better. From the thyroxine-TTR structure, three halogen-binding pockets have been identified in each TTR monomer. We found three bound iodides per TTR monomer, two of which were in the thyroxine-binding channel. This indicates that only two of the three halogen-binding pockets in the thyroid-hormone binding channel are optimal for halogen binding. Our results might be useful for the continuing design of small molecule ligands, which in the end can lead to inhibitors for amyloid diseases.</p>

Cette thèse concerne le développement de complexes de lanthanide et leur utilisation pour résoudre les structures de macromolécules biologiques par diffraction des rayons X, en particulier celles de protéines membranaires et de complexes protéiques de grande taille. Les complexes de lanthanide sont formés d’un atome de lanthanide et d’un ligand chimique qui assure une liaison non-covalente avec les surfaces protéiques. Introduits dans les cristaux de protéine, ces derniers constituent une sous-structure qui, déterminée par les méthodes de phasage de novo courantes, permet de résoudre la structure de la macromolécule d’intérêt. La première partie de ce travail de thèse est une étude menée sur la grande famille des complexes picolinates de lanthanide, complexes luminescents dont la fixation au sein des cristaux est aisément décelable. En premier lieu, nous avons défini les conditions dans lesquelles le complexe tris-dipicolinate de lanthanide peut être utilisé ainsi que ses éventuelles capacités à promouvoir la cristallisation (effet supramoléculaire). En second lieu, un nouveau complexe, dérivé du précédent, a été développé au cours de cette thèse : le tris-hydroxymethyltriazoledipicolinate de lanthanide (« [Ln(TDPA)3]3- »). Il nous a permis de réaliser un phasage de novo très précis conduisant à la détermination des structures du lysozyme de blanc d’œuf de poule et de la thaumatine de Thaumatococcus daniellii à haute résolution. Par ailleurs, différents ligands pour ce nouveau complexe ont été synthétisés par chimie-click, nous permettant de créer une panoplie de complexes uniques et des complexes hybrides. Nous avons montré que cette nouvelle famille de complexes présente une meilleure affinité pour les protéines permettant leur utilisation à de faibles concentrations. Les essais menés avec ces LnTDPA ont aussi permis d’entrevoir l’importance de la charge globale pour expliquer l’effet supramoléculaire. En troisième lieu, un tripicolinate cagé, le LnTNTPA, a également été évalué. Constitué d’une cage chimique de charge globale nulle, nous avons montré que ce nouveau complexe luminescent est le seul de cette famille picolinate qui puisse être utilisé en présence d’ions divalents. Dans la seconde partie de cette thèse, nous décrivons l’utilisation des complexes de lanthanide pour le phasage de protéines multimériques de grandes tailles par les méthodes de phasage de novo. Premièrement, la structure de l’aminopeptidase dodécamérique PhTET1-12s de 480 kDa a été déterminée à 4 Å de résolution à l’aide du tris-dipicolinate d’europium. Ceci nous a permis de démontrer que les complexes de lanthanide peuvent être utilisés pour obtenir un phasage précis, même à basse résolution. Deuxièmement, les complexes de lanthanide issus de l'imagerie médicale ont aussi permis de déterminer la structure de trois nouvelles enzymes homotétramériques de la famille des malate déshydrogénases. Ces structures permettent d’apporter de nouveaux éclaircissements sur l'adaptation halophile. Enfin, en utilisant ces enzymes en tant que bibliothèque de fonctions chimiques, nous avons pu mettre en place une nouvelle approche méthodologique pour comprendre finement les modes d'interaction des complexes de lanthanide. / This thesis aims to develop lanthanide complexes for solving biological macromolecular structures, especially membrane proteins and large protein complexes. Lanthanide complexes are composed of a lanthanide atom and a chemical ligand, which provides non-covalent binding to protein surfaces. Incorporated in protein crystals, they make up the substructure, determined by the widely-used de novo phasing methods, needed for solving the whole protein structure. The first part of the present work shows a study of the luminescent lanthanide picolinate complexes family, easily detectable in protein crystals. First, we defined the conditions in which the known lanthanide tris-dipicolinate can be used and we examined its possible ability to promote protein crystallization (supramolecular effect). Secondly, a newly developed lanthanide tris hydroxymethyltriazoledipicolinate complex "[Ln(TDPA)3]3-", derived from this previous complex, allowed us to obtain a very precise de novo phasing for solving the high-resolution structure of the hen egg white lysozyme and the thaumatin from Thaumatococcus daniellii. Besides, various ligands for these new complexes were synthetized by click chemistry, enabling us to create both unique complexes outfit and hybrid complexes. We have shown that this new complexes family presents a better affinity for proteins allowing their use at very low concentrations. Tests conducted with those LnTDPA have also evidenced the importance of complex global charge to explain the supramolecular effect. Third, a caged tripicolinate, the LnTNTPA, was evaluated. Characterized by a neutrally charged chemical cage, we have shown that this new luminescent complex is the only one of the picolinate complexes family that can be used in the presence of divalent ions. In the second part of this thesis, we describe the use of lanthanide complexes for phasing large multimeric proteins by de novo phasing methods. First, the structure of the 480 kDa dodecameric aminopeptidase PhTET1-12s was solved at 4 Å resolution by using the europium tris-dipicolinate which demonstrates that the lanthanide complexes can be used to obtain an accurate phasing, even at low resolution. Secondly, the lanthanide complexes from medical imaging also helped to solve the structures of three new homotetrameric enzymes from the malate dehydrogenases family. Those structures provide new insights into halophilic adaptation. Finally, by using these enzymes as a library of chemical functions, we developed a new methodological approach for a better understanding of the precise binding modes of those lanthanide complexes.

<p>The malfunction of the transcriptional regulator RUNX1 is the major cause of several variants of acute human leukemias and its normal function is to regulate the development of the blood system in concert with other transcriptional co-regulators. RUNX1 belongs to a conserved family of heterodimeric transcription factors that share a conserved DNA binding domain, the Runt domain (RD), named after the first member of this group – Runt - found in Drosophila melanogaster. The binding partner CBFβ serves as a regulator of RUNX by enhancing its DNA binding affinity through an allosteric mechanism.</p><p>The main focus ofo my thesis work has been the crystallization and structural analysis of the RUNX1 RD and involved also more technical methodological aspects that can be applied to X-ray crystallography in general.</p><p>The high resolution crystal structure of the free RD shows that this immunoglobulin-like molecule undergoes significant structural changes upon binding to both CBFβ and DNA. This involves a large flip of the L11 loop from a closed conformation in the free protein to an open conformation when CBFβ and/or DNA are bound. We refer to this transition as the “S-switch”. Smaller but significant conformational changes in other parts of the RD accompany the “S-switch”. We suggest that CBFβ triggers and stabilizes the “S-switch” which leads to the conversion of the RD into a conformation enhanced for DNA binding.</p><p>During the structural analysis of the RD we identified two chloride ions that are coordinated by residues otherwise involved in DNA binding. In electrophoretic mobility-shift analyses (EMSA) we demonstrated a chloride ion concentration dependent stimulation of the DNA binding affinity of RUNX1. We further showed by NMR line width broadening experiments that the chloride binding occurred within the physiological range. A comparable DNA binding stimulation of RUNX1 was seen in the presence of negative amino acids. This suggests a regulation of the DNA binding activity of RUNX1 proteins through acidic amino acid residues possibly provided by activation domains of transcriptional co-regulators that interact with RUNX1.</p><p>The use of the anomalous signal from halide ions has become a powerful technique for obtaining phase information. By replacing the sodium chloride with potassium bromide in the crystallisation conditions of the RD, we could demonstrate in a single wavelength anomalous diffraction (SAD) experiment that the anomalous signal from 2 bromide ions were sufficient to phase a 16 kDa protein. Due to lack of completeness in the low-resolution shells caused by overloaded intensities, density modification schemes failed and the resulting electron density maps were not interpretable. By combining the highresolution</p><p>synchrotron data with low-resolution data from a native data set collected on a home X-ray source, the density modified bromide phases gave easily traceable maps.</p>

The malfunction of the transcriptional regulator RUNX1 is the major cause of several variants of acute human leukemias and its normal function is to regulate the development of the blood system in concert with other transcriptional co-regulators. RUNX1 belongs to a conserved family of heterodimeric transcription factors that share a conserved DNA binding domain, the Runt domain (RD), named after the first member of this group – Runt - found in Drosophila melanogaster. The binding partner CBFβ serves as a regulator of RUNX by enhancing its DNA binding affinity through an allosteric mechanism. The main focus ofo my thesis work has been the crystallization and structural analysis of the RUNX1 RD and involved also more technical methodological aspects that can be applied to X-ray crystallography in general. The high resolution crystal structure of the free RD shows that this immunoglobulin-like molecule undergoes significant structural changes upon binding to both CBFβ and DNA. This involves a large flip of the L11 loop from a closed conformation in the free protein to an open conformation when CBFβ and/or DNA are bound. We refer to this transition as the “S-switch”. Smaller but significant conformational changes in other parts of the RD accompany the “S-switch”. We suggest that CBFβ triggers and stabilizes the “S-switch” which leads to the conversion of the RD into a conformation enhanced for DNA binding. During the structural analysis of the RD we identified two chloride ions that are coordinated by residues otherwise involved in DNA binding. In electrophoretic mobility-shift analyses (EMSA) we demonstrated a chloride ion concentration dependent stimulation of the DNA binding affinity of RUNX1. We further showed by NMR line width broadening experiments that the chloride binding occurred within the physiological range. A comparable DNA binding stimulation of RUNX1 was seen in the presence of negative amino acids. This suggests a regulation of the DNA binding activity of RUNX1 proteins through acidic amino acid residues possibly provided by activation domains of transcriptional co-regulators that interact with RUNX1. The use of the anomalous signal from halide ions has become a powerful technique for obtaining phase information. By replacing the sodium chloride with potassium bromide in the crystallisation conditions of the RD, we could demonstrate in a single wavelength anomalous diffraction (SAD) experiment that the anomalous signal from 2 bromide ions were sufficient to phase a 16 kDa protein. Due to lack of completeness in the low-resolution shells caused by overloaded intensities, density modification schemes failed and the resulting electron density maps were not interpretable. By combining the highresolution synchrotron data with low-resolution data from a native data set collected on a home X-ray source, the density modified bromide phases gave easily traceable maps.

Les films de nitrures métalliques nanostructurés sont généralement utilisés comme revêtements protecteurs. Ti1-xAlxN (0 ≤ x ≤ 1) peut être considéré comme un système modèle, où TiN (cubique) et AIN (hexagonal) sont partiellement miscibles. L’élaboration par dépôt physique en phase vapeur donne au film une microstructure colonnaire complexe composée de phase métastable pouvant cohabiter avec des précipités localisés aux joints de grains. Une haute dureté et une grande résistance à l’oxydation sont observées pour un maximum d’atomes de Ti substitué par des atomes de Al en réseau cubique. Les conditions de dépôt et la composition jouent un rôle majeur sur la substitution des éléments métalliques (Ti ,Al). Nous avons préparé deux séries de films déposés par pulvérisation cathodique magnétron réactive à partir de cibles TiAl compartimentées et frittées. La micro- et nanostructure des films ont été analysées par Diffraction, Spectroscopie d’Absorption des rayons X et Microscopie Electronique à Transmission. L’usure des revêtements a été étudiée par microtribologie. Nous observons pour les films riches en Ti (x < 0,5) des directions de croissances [200]c et [111]c, caractéristiques d’un réseau cubique. Tandis que, les films riches en Al (x > 0,7) présentent une croissance de domaines bien cristallisés suivant la direction [002]h du réseau hexagonal. De plus, nous avons mis en évidence l’apparition de la transition cubique / hexagonal à des teneurs en Al plus élevée pour les films issus de cible frittée. Ces films montrent une meilleure résistance à la fissuration et à l’usure que ceux déposés à partir de cible compartimentées. / Ti1-xAlxN (0 ≤ x ≤ 1) is considered as a model system, where TiN (fcc) and AlN (hcp) do not mix over the whole composition range due to their low miscibility. However, the physical vapour deposition (PVD) allows achieving metastable phases of Ti1-xAlxN, where Al atoms are partially substituting for Ti in fcc lattice. Ti1-xAlxN coatings exhibit high hardness and oxidation resistance for the maximum Al substituted to Ti in fcc lattice (about x=0.6). The proportion of grain boundaries and the limit solubility play a major role on the mechanical properties and resistance to wear of the coatings. Several techniques are employed to investigate two sets of Ti1-xAlxN thin films deposited by magnetron reactive sputtering from two types of metallic targets onto Si (100). Lattice symmetry of crystallised domains and columnar growth structure of the films are characterized by X-ray diffraction (XRD) and electron microscopy (TEM, HRTEM). Several local probes such as X-ray absorption fine structure (XAFS), diffraction anomalous fine structure (DAFS) and Electron Energy Loss Spectroscopies (EELS) which are very sensitive to the symmetry of the atomic sites either octahedral for fcc lattice or tetrahedral for hcp one are carried out. For Ti-rich films (x < 0.5), the competitive growth of cubic domains between [200]c and [111]c is observed. For Al-rich films (x > 0.7) have a domain growth well crystallized in the direction [002]h the hexagonal lattice. In addition, the cubic / hexagonal transition in Al contents higher is observed for films from sintered target. These films show better wear resistance than those deposited from target compartmentalized.

Le travail illustré par ce manuscrit de thèse présente l'étude structurale d'hétéro-structures semi-conductrices à base de nitrures d'éléments III avec l'un des outils les plus puissants de la recherche scientifique: le rayonnement synchrotron. La cartographie haute résolution de l'espace réciproque, la diffraction anomale multi-longueur d'onde, la spectroscopie d'absorption X et la spectroscopie en condition de diffraction nous ont permis de caractériser la structure à l'échelle atomique de différentes régions d'un même système. Dans un premier temps, nous montrons que les nanofils GaN sur Si(111) ont une polarité N et proposons un mécanisme de nucléation. Dans un second temps, nous mettons en évidence un phénomène d'inter-diffusion stimulée par la contrainte dans les boîtes quantiques GaN / AlN recuites à haute température. Enfin, nous observons la présence d'un ordre local à courte distance dans les nanofils coeur-coquille InGaN / GaN. Cette organisation atomique pourrait être induite par la présence de contrainte, nous avons initié une étude de l'anisotropie de l'ordre à courte distance pour explorer cette hypothèse. / The work presented in this manuscript deals with the structural investigation of III-nitrides semiconductor heterostructures with one of the most powerful tools of materials science: the synchrotron radiation. We used high resolution reciprocal space mapping, multi-wavelenght anomalous diffraction, x-rays absorption spectroscopy and diffraction anomalous fine structure to characterize the structure at the atomic scale of complex nano-structured systems. First, we show that GaN nanowires on Si(111) are N-polar and we suggest a nucleation mecanism. Then, we highlight a strain assited inter-diffusion phenomenon in the GaN / AlN quantum dots annealed at high temperature. Finally, we observe the presence of short-range order in InGaN / GaN core-shell nanowires. This atomic organization could be induced by the presence of stress and we have begun a study of the anisotropic ordering to explore this hypothesis.