Résumé Le complexe V mitochondrial (F1FO-ATP synthétase) catalyse la phosphorylation de lADP par le phosphate inorganique en utilisant la force proton-motrice générée par la chaîne de transport délectrons. Ce complexe protéique possède deux domaines : un secteur associé à la membrane, FO, impliqué dans la translocation des protons, et un domaine extrinsèque, F1, qui catalyse la synthèse dATP. Les deux secteurs sont connectés par deux bras : un bras central qui couple la translocation des protons à la région catalytique, et un bras latéral qui est considéré comme faisant partie du stabilisateur (stator) de lenzyme. Au cours de ce travail, nous nous sommes intéressés à lenzyme de deux algues appartenant à la classe des Chlorophyceae, Chlamydomonas reinhardtii et Polytomella sp.. L'enzyme des deux algues présente une composition sous-unitaire atypique, les sous-unités classiquement retrouvées chez les eucaryotes et impliquées dans larchitecture du bras périphérique ou dans la dimérisation du complexe en étant absentes. En contrepartie, 9 sous-unités dorigine évolutive inconnue sont associées à lenzyme. Elles ont été appelées Asa1 à 9 pour ATP Synthase Associated protein. Chez C. reinhardtii et Polytomella sp., lATP synthétase présente une stabilité accrue de sa forme dimérique in vitro, et, in vivo, les cellules de C. reinhardtii sont insensibles à loligomycine, un puissant inhibiteur de la translocation de protons au travers de FO. Nous avons dans un premier temps tenté détablir la composition sous-unitaire du complexe V chez des espèces appartenant aux différentes classes de Chlorophytes (Chlorophyceae, Trebouxiophyceae, Prasinophyceae et Ulvophyceae) en combinant analyses génomiques et protéomiques. Plusieurs sous-unités Asa ont ainsi pu être détectées chez des algues appartenant à divers ordres de Chlorophyceae. Au contraire, les analyses de séquences disponibles chez les autres classes de Chlorophytes (Trebouxiophyceae, Prasinophyceae et Ulvophyceae) indiquent une composition canonique de lenzyme. Lanalyse de la stabilité de la forme dimérique du complexe de différentes espèces d'algues vertes sur BN PAGE (Blue Native PolyAcrylamide Gel Electrophoresis) suggère également que la présence dun dimère stable est caractéristique aux Chlorophyceae. Par ailleurs, leur croissance, respiration, et niveaux d'ATP sont à peine affectés par la présence d'oligomycine à des concentrations inhibitrices chez les représentants des autres classes de Chlorophytes. Les nombreuses particularités communes aux algues appartenant à cette classe suggèrent que la perte d'éléments canoniques du stator est apparue lors de la séparation des Chlorophyceae et a été accompagnée du recrutement de nouvelles sous-unités. Ce réarrangement drastique de la composition de stator et du module de dimérisation pourrait avoir conféré de nouvelles propriétés à lenzyme, notamment une meilleure stabilité et une plus grande résistance à loligomycine. Nous avons également étudié la fonction de la sous-unité atypique Asa7 en inactivant son expression chez C. reinhardtii. Bien que la perte de la sous-unité Asa7 n'aie aucun impact sur la bioénergétique des cellules ou sur la structure mitochondriale, elle déstabilise lenzyme in vitro et rend la croissance, la respiration, et de le niveau d'ATP sensible à oligomycine. L'impact de la perte de l'activité ATP synthétase mitochondriale chez un organisme photosynthétique a été étudié chez C. reinhardtii par linactivation de l'expression du gène ATP2, codant pour la sous-unité catalytique beta. Les résultats démontrent que, en l'absence de beta, l'ATP synthétase ne peut plus être assemblée et les cellules deviennent dépendantes de la photosynthèse. La respiration en présence ou en absence du découpleur CCCP suggère que le passage des protons à travers la membrane interne mitochondriale est bloqué chez la souche mutante. Enfin, la morphologie des mitochondries est affectée, et les chloroplastes montrent un réaménagement massif de l'appareil photosynthétique, suggérant des répercussions importantes sur la synthèse dATP par les chloroplastes. Ces résultats contribuent à la compréhension des interactions entre organites bioénergétiques chez les organismes photosynthétiques.

Karenia brevis is the major harmful algal bloom-forming species in the Gulf of Mexico, and produces neurotoxins, known as brevetoxins, that cause large fish kills, neurotoxic shellfish poisoning, and human respiratory distress. Brevetoxins are polyethers that bind voltage-sensitive sodium channels, opening them for prolonged periods of time. Clonal cultures of K. brevis exhibit unique brevetoxin profiles, which not only differ from one another, but also change when subjected to different environmental conditions. The brevetoxin structures were elucidated 30 years ago without any breakthroughs for the biosynthetic pathway. These unique ladder-like polyethers have 10 (PbTx-1) or 11 (PbTx-2) rings, indicating that they are synthesized as secondary metabolites by polyketide synthases. The extensive size of the genome and the lack of histones and nucleosomes combined with the additional regulatory step of a trans-splicing spliced leader sequence make normal molecular techniques ineffective in determining the genes involved in toxin synthesis. The goal of this project is to identify a potential link between toxin, gene, and function. One objective is to take the next step towards identifying the genes associated with the synthesis and regulation of brevetoxins and to help elucidate the hypothesized gene clusters of multi-protein enzymatic complexes involved in brevetoxin production, one for each backbone. The second objective is to make an effort to determine the in vivo function of the costly brevetoxins by identifying possible ion channels, which could be osmotically regulated by the toxins. Genes for polyketide synthases (PKS) were identified in K. brevis, obtained from Expressed Sequence Tag (EST) libraries. In this work, reverse transcription polymerase chain reactions (RT-PCR) were used to generate pools of complementary DNA (cDNA), which was used in real-time quantitative polymerase chain reactions (qPCR) to give relative amounts of PKS transcripts. K. brevis clones have shown a significant increase in toxin production after a rapid shift from high salinity to low salinity, indicating a regulation of brevetoxin synthesis. To gain a better understanding of regulation of toxin production during algal blooms, we compared the toxin levels under different conditions to the transcript levels of PKS genes, as determined by quantitative RT-PCR. In a separate line of investigation, an in silico analysis of the EST library was performed to identify ion channel genes expressed by K. brevis, which may be the in vivo binding site of brevetoxin. The information generated from this project will help to elucidate the effects of environmental variations on toxin production and the biological function of toxin production -- valuable information for the shellfish industries and public health.

Association mapping (AM), based on linkage disequilibrium, is a complementary strategy to traditional quantitative trait loci (QTL) mapping for describing associations between genotypes and phenotypes in crop plants. Yellow endosperm colour, an important quality trait in durum wheat (<i>Triticum turgidum L. var. durum</i>), was studied to determine the potential of AM to (1) identify previously reported QTL using a genome wide scan and (2) to determine allelic association of the phytoene synthase 1 (Psy1) gene using a candidate gene analysis. At present, a number of QTL for endosperm colour have been identified, and phytoene synthase, the initial enzyme of the carotenoid biosynthetic pathway, has been associated with QTL on the group 7 chromosomes which are considered to play a significant role in expression of yellow pigment concentration. CIE 1976 b*, a light reflectance measurement, and water-saturated butanol extracted pigments were assessed on a collection of 93 elite accessions from a variety of geographic origins, and genotyped with 245 markers. Population structure was assessed using genetic distance and Bayesian model based approaches, identifying five sub-populations consistent with breeding origin and pedigree. Association analysis identified significant associations with yellow endosperm colour on all chromosomes, including several previously identified QTL as well as new regions for genomic dissection. Pairwise LD mapping of Psy1-B1 and Psy1-A1 located the genes to chromosomes 7B and 7A respectively, to regions which have previously been identified for yellow pigment concentration QTL. The results of this study indicate that AM can be used to complement traditional QTL mapping techniques, and identify novel QTL for further study.

MosA is an enzyme from Sinorhizobium meliloti L5-30, a beneficial soil bacterium. Initial investigation into this enzyme categorized it as a methyltransferase. Further investigation revealed that this was incorrect, and that MosA is actually a dihydrodipicolinate synthase, part of the N-acetylneuraminate lyase superfamily. One of the characteristics of enzyme superfamilies is their low sequence identity, but relatively high structural similarity. The structural investigation reported here confirms the high structural similarity between MosA and other superfamily members. <p>Investigation of MosA was carried out by means of x-ray crystallography. It was believed that detailed structural information may shed light into not only the enzymatic mechanism, but also the inhibition of MosA by lysine, the final product of the enzymatic pathway. Insight into enzyme mechanism and inhibition may ultimately prove useful in herbicide or insecticide development, as other dihydrodipicolinate synthases from harmful fungi, bacteria, or plants, make attractive targets for inhibition. Lysine is an essential amino acid for humans, meaning that there is no endogenous lysine production to block the use of these hypothetical inhibitors. Specific inhibitors based on crystal structures have proven to be effective in the past and hopefully, will continue to be useful in the future. <p>Here we report the structure of MosA, solved to 1.95 Å resolution with lysine 161 forming a Schiff-base adduct with pyruvate. This adduct is consistent with the currently accepted dihydrodipicolinate synthase enzyme mechanism.

Adenosine triphosphate (ATP) is the molecular currency of intracellular energy transfer in living organisms. The enzyme ATP synthase is primarily responsible for ATP production in eukaryotes. In archaea and some bacteria, ATP is synthesized by V-ATPase that is related to ATP synthase both in structure and function. Both of these enzymes are reversible rotary motors capable of catalyzing ATP synthesis or hydrolysis. The rotation of the central rotor, which is powered by the flow of proton (or sometimes sodium ion) down the electrochemical gradient through the membrane-bound Fo/Vo region, leads to the chemical synthesis of ATP in F1/V1 region. The F1/V1 region, on the other hand, can catalyze ATP hydrolysis, which in turn leads to proton (or sodium) pumping across the membrane through rotation of the central rotor in the opposite direction. This thesis describes structure determination of both the intact F-type and V-type enzymes using single particle electron cryomicroscopy (cryo-EM), with the aim of better understanding their overall architecture, subunit organization and the mechanism of proton translocation. Our cryo-EM structural analysis on the F-type ATP synthase from Saccharomyces cerevisiae uncovered the arrangement of subunits a, b, c, and the two dimer-specific subunits e and g within the membrane-bound region of Fo. A model of oligomerization of the ATP synthase involving two distinct dimerization interfaces was proposed.The rotor-stator interaction within the membrane-bound region of both enzymes is responsible for proton translocation. Our cryo-EM structures of the V-ATPase from Thermus thermophilus reveal that the interaction between the rotary ring (rotor) and the I-subunit (stator) is surprisingly small, with only two subunits from the ring making contact with the I-subunit near the middle of the membrane. Furthermore, the spatial arrangement of transmembrane helices resolved in subunit I can form two passageways that could provide proton access through the membrane-bound region and is consistent with a two-channel model of proton translocation.

Adenosine triphosphate (ATP) is the molecular currency of intracellular energy transfer in living organisms. The enzyme ATP synthase is primarily responsible for ATP production in eukaryotes. In archaea and some bacteria, ATP is synthesized by V-ATPase that is related to ATP synthase both in structure and function. Both of these enzymes are reversible rotary motors capable of catalyzing ATP synthesis or hydrolysis. The rotation of the central rotor, which is powered by the flow of proton (or sometimes sodium ion) down the electrochemical gradient through the membrane-bound Fo/Vo region, leads to the chemical synthesis of ATP in F1/V1 region. The F1/V1 region, on the other hand, can catalyze ATP hydrolysis, which in turn leads to proton (or sodium) pumping across the membrane through rotation of the central rotor in the opposite direction. This thesis describes structure determination of both the intact F-type and V-type enzymes using single particle electron cryomicroscopy (cryo-EM), with the aim of better understanding their overall architecture, subunit organization and the mechanism of proton translocation. Our cryo-EM structural analysis on the F-type ATP synthase from Saccharomyces cerevisiae uncovered the arrangement of subunits a, b, c, and the two dimer-specific subunits e and g within the membrane-bound region of Fo. A model of oligomerization of the ATP synthase involving two distinct dimerization interfaces was proposed.The rotor-stator interaction within the membrane-bound region of both enzymes is responsible for proton translocation. Our cryo-EM structures of the V-ATPase from Thermus thermophilus reveal that the interaction between the rotary ring (rotor) and the I-subunit (stator) is surprisingly small, with only two subunits from the ring making contact with the I-subunit near the middle of the membrane. Furthermore, the spatial arrangement of transmembrane helices resolved in subunit I can form two passageways that could provide proton access through the membrane-bound region and is consistent with a two-channel model of proton translocation.

Avec le vieillissement, les protéines accumulent des dommages qui affectent leur structure et activité. La protéine L-isoaspartate (D-aspartate) méthyltransférase (PIMT) est impliquée dans la réparation de protéines endommagées au niveau de résidus aspartates. Bien que son mécanisme d'action soit connu, les voies de régulation qui influencent l'expression de la PIMT ne sont pas très bien établies. Dans cette étude, il a été observé que la protéine nommée glycogène synthase kinase 3 (GSK-3) est impliquée dans la régulation de la PIMT. En effet, le traitement de cellules gliales, devenues cancéreuses, par différents inhibiteurs connus de GSK-3, dont le lithium, a démontré une augmentation de l'expression de la PIMT. Dans la littérature, il a été rapporté que le lithium possède deux cibles principales dans la cellule: la voie de l'inositol et la protéine GSK-3. Cette dernière kinase est notamment contrôlée par la voie de Wnt. Or, le traitement des cellules avec le myo-inositol n'influence pas la PIMT, ce qui exclut la voie de l'inositol comme voie de régulation de la PIMT. Les résultats obtenus en présence de lithium suggèrent cependant une régulation de la PIMT par GSK-3, par l'intermédiaire de la voie de Wnt. Par ailleurs, un traitement des cellules avec différents inhibiteurs de protéines kinases, pouvant agir sur GSK-3 ne semble pas affecter l'expression de la PIMT. Cependant, un traitement des cellules avec un inhibiteur de la synthèse protéique, en présence de lithium, empêche l'augmentation de l'expression de la PIMT, ce qui suggère une régulation au niveau de sa synthèse. De plus, une analyse par RT-PCR démontre une modulation du niveau d'ARNm de la PIMT en présence de différents inhibiteurs de GSK-3. L'établissement de nouvelles voies de régulation pourrait permettre d'identifier de nouvelles cibles thérapeutiques dans des maladies impliquant la PIMT, telle que dans l'épilepsie, l'Alzheimer et surtout dans les états maniaco-dépressifs. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : PIMT, GSK-3, Lithium, Cellules gliales cancéreuses.

The increase of calcium in the cell can induce cellular functions such as fertilization, cell division and cell communication. Calcium (Ca2+) carries out these processes through proteins called calcium sensors. An important calcium modulator is calmodulin. Calmodulin has four possible Ca2+ binding sites that have the characteristic helix-loop-helix (EF hand) motif. When the EF hands bind to Ca2+, methionine rich hydrophobic patches are exposed allowing for CaM to interact with target proteins. However, there are proteins that can interact with CaM at low levels of Ca2+ or in the absence of Ca2+. An enzyme that is activated by CaM is nitric oxide synthase (NOS), which converts L-arginine to L-citrulline and nitric oxide (•NO), where •NO is used to carry out important cellular functions. There are three isoforms of the enzyme; endothelial, neuronal and inducible NOS. The first two isoforms are activated by Ca2+-bound CaM when there is an influx of Ca2+ and are therefore Ca2+-dependent whereas inducible NOS (iNOS) is activated and binds tightly to CaM regardless of the Ca2+ concentration and is therefore Ca2+-independent. Of particular interest is the iNOS enzyme, since no three-dimensional structures of the reductase domain or the CaM-binding region have been solved. All three isoforms of NOS exist as homodimers, where each monomer consisting of a reductase domain and an oxygenase domain separated by a CaM-binding region. The reductase domain contains binding sites for NADPH and the flavins, FAD and FMN, which facilitate electron transfer from the NADPH to the catalytic heme in the oxygenase domain of the opposite monomer. The transfer of electrons from the FAD to the heme is carried out by the FMN domain which is proposed to swing between the two docking points since the distance between the two points is too large for electron transfer. This electron transfer point is under the control of CaM, which is essential for NOS activation. This dynamic process and the direct role of CaM have yet to be observed structurally. A method to monitor dynamics structurally is through the use of nuclear magnetic resonance (NMR) spectroscopy. Therefore as the first step to determine the NMR structure of the FMN domain with the CaM-binding region, the structure of the iNOS CaM-binding region bound to CaM will be determined. The structure will allow for further characterization and identification of important interactions between the iNOS CaM-binding region and CaM which contribute to the unique properties of iNOS.

The increase of calcium in the cell can induce cellular functions such as fertilization, cell division and cell communication. Calcium (Ca2+) carries out these processes through proteins called calcium sensors. An important calcium modulator is calmodulin. Calmodulin has four possible Ca2+ binding sites that have the characteristic helix-loop-helix (EF hand) motif. When the EF hands bind to Ca2+, methionine rich hydrophobic patches are exposed allowing for CaM to interact with target proteins. However, there are proteins that can interact with CaM at low levels of Ca2+ or in the absence of Ca2+. An enzyme that is activated by CaM is nitric oxide synthase (NOS), which converts L-arginine to L-citrulline and nitric oxide (•NO), where •NO is used to carry out important cellular functions. There are three isoforms of the enzyme; endothelial, neuronal and inducible NOS. The first two isoforms are activated by Ca2+-bound CaM when there is an influx of Ca2+ and are therefore Ca2+-dependent whereas inducible NOS (iNOS) is activated and binds tightly to CaM regardless of the Ca2+ concentration and is therefore Ca2+-independent. Of particular interest is the iNOS enzyme, since no three-dimensional structures of the reductase domain or the CaM-binding region have been solved. All three isoforms of NOS exist as homodimers, where each monomer consisting of a reductase domain and an oxygenase domain separated by a CaM-binding region. The reductase domain contains binding sites for NADPH and the flavins, FAD and FMN, which facilitate electron transfer from the NADPH to the catalytic heme in the oxygenase domain of the opposite monomer. The transfer of electrons from the FAD to the heme is carried out by the FMN domain which is proposed to swing between the two docking points since the distance between the two points is too large for electron transfer. This electron transfer point is under the control of CaM, which is essential for NOS activation. This dynamic process and the direct role of CaM have yet to be observed structurally. A method to monitor dynamics structurally is through the use of nuclear magnetic resonance (NMR) spectroscopy. Therefore as the first step to determine the NMR structure of the FMN domain with the CaM-binding region, the structure of the iNOS CaM-binding region bound to CaM will be determined. The structure will allow for further characterization and identification of important interactions between the iNOS CaM-binding region and CaM which contribute to the unique properties of iNOS.

MosA is an enzyme from Sinorhizobium meliloti L5-30, a beneficial soil bacterium. Initial investigation into this enzyme categorized it as a methyltransferase. Further investigation revealed that this was incorrect, and that MosA is actually a dihydrodipicolinate synthase, part of the N-acetylneuraminate lyase superfamily. One of the characteristics of enzyme superfamilies is their low sequence identity, but relatively high structural similarity. The structural investigation reported here confirms the high structural similarity between MosA and other superfamily members. <p>Investigation of MosA was carried out by means of x-ray crystallography. It was believed that detailed structural information may shed light into not only the enzymatic mechanism, but also the inhibition of MosA by lysine, the final product of the enzymatic pathway. Insight into enzyme mechanism and inhibition may ultimately prove useful in herbicide or insecticide development, as other dihydrodipicolinate synthases from harmful fungi, bacteria, or plants, make attractive targets for inhibition. Lysine is an essential amino acid for humans, meaning that there is no endogenous lysine production to block the use of these hypothetical inhibitors. Specific inhibitors based on crystal structures have proven to be effective in the past and hopefully, will continue to be useful in the future. <p>Here we report the structure of MosA, solved to 1.95 Å resolution with lysine 161 forming a Schiff-base adduct with pyruvate. This adduct is consistent with the currently accepted dihydrodipicolinate synthase enzyme mechanism.