There is a growing evidence of the potential health benefits of a diet rich in fruits and vegetables. National nutritional guidelines advise the consumption of at least five portions (400 g ) of these foods per day with the goal being a reduction in the levels of coronary heart disease (CHD) strokes and certain cancers. The beneficial properties of fruit and vegetables may be ascribed, in part to the presence of antioxidants and recent attention in this regard has focused on phenolic and polyphenolic compounds. These compounds are present in a wide variety of commonly consumed foods and beverages. Onions are a rich source of the polyphenolic flavonol quercetin-4′-glucoside. For this compound to have some health effects it must be absorbed and reach target organs in a concentration and form where they can exert an effect. To-date interest has focussed on the levels of the intact quercetin aglycone circulating in plasma and excreted in urine. However, it is now known that quercetin does not circulate in the plasma as the parent compound or the aglycone. However, at the outset of this project, the exact form(s) and concentration of metabolites circulating in plasma were unknown. The need to know what compounds are actually circulating, and at what concentration, is important if in vitro studies are to be made into the mechanisms by which quercetin could, potentially, exert a health benefit. The reasons why these issues have not been addressed are due to a number of factors. The main methodology used in studies into absorption, distribution, metabolism and excretion or (A.D.M.E) as it is know in the drug industry, is by use of chromatography coupled to various detection systems. This can range from a simple isocratic single pump linked to a single wavelength absorbance detector, to a gradient pumping system with an autoinjector linked in series to a diode array absorbance detector and mass spectrometer. The latter instruments, although initially expensive are now becoming more affordable. The original methodology used to determine the level of quercetin in plasma involved hydrolysis of the quercetin conjugates back to the aglycone. The information, which is lost by using this hydrolysis method, is vital if we are to gain a better understanding of the A.D.M.E process. There have been a large number of feeding studies carried out using onions or the flavonol contained in them. However, very little additional information was gained after the initial investigations. The objectives of the studies presented in this thesis were to develop methodology to identify and quantify the major metabolites of quercetin in man after ingestion of onions. This would initially require the use of radiolabelled [2-14C]quercetin-4´-glucoside fed to rats to facilitate the development of the method. Having successfully developed methods that would work both in rats and in man, it was of great interest to establish the fate of the complete dose of [2-14C]quercetin-4´-glucoside in rats. In Chapter 2 radiolabelled quercetin-4´-glucoside was used as a tracer to follow the metabolism of the compound as it was acted on by the digestive system of the rat. After 1 h 93% of the ingested dose was recovered in the gastrointestinal tract (GIT). Analysis using HPLC with a photodiode array (PDA) detector in series with a radioactivity monitor connected to an electrospray ion trap mass spectrometer facilitated the separation, quantification and partial identification of 18 out of 19 metabolites. The 1 h sample was part of a larger study that investigated the fate of the radiolabelled compounds up to 5 h after dosing. The latter samples formed part of another study not reported on in this thesis. Having developed the methodology, using the radiolabelled compound, it was then applied to a feed of onions to healthy human volunteers to determine if metabolite detection, identification and quantification could be carried out without the use of the radioactive tracer. In Chapter 3 plasma samples collected 1 h after a feed of onions and urine collection from 0-4 h post feed were used to test if the method could be transferred to a non labelled assay. A total of 22 metabolites plus the parent compound were identified. The metabolic profile of the plasma and urine showed marked differences, again pointing to major post absorption metabolism. The successful transfer of the method from the initial radiolabelled study to the onion feed allowed pharmacokinetic data to be obtained from all plasma samples taken over a 24 h period, along with the 0-24 h urine samples. In Chapter 4 it was seen that the metabolites are both rapidly absorbed and excreted, with plasma levels returning almost back to baseline by 6 h. The total excretion in urine accounted for 4.5% of the ingested dose. These results were controversial, as the pioneer of this field had published that the elimination half life of quercetin was of the order of 18 h. The differences between the two methods employed are discussed in Chapters 3 and 4. The fact that only 4.5% could be accounted for in this study, which was in agreement with other studies, leaves the question of what happens to the other 95.5%. It is possible that the potential health benefit attributed to this compound may have nothing to do with the parent compound but could be coming from something in the other 95.5%. Studies using patients who have undergone an ileostomy have been used to provide further information into what happens to the majority of the dose. By collecting the ileal fluid after a flavonol feed the amount of intact compound can be measured in ileal fluid (Hollman et al., 1995b; Walle et al., 2000). This work and some results from a similar trial study are discussed in Chapter 4, with regard to the process of metabolite absorption and formation. The only way to follow the parent compound throughout its passage through the body is by use of a labelled compound. In Chapter 5 a second feed of [2-14C]quercetin-4´-glucoside, which focuses on the overall fate of the compound, has samples collected for up to 72 h. The fate of the dose was monitored both in terms of the level of radioactivity excreted and found in the tissues and also the identity of the radioactive compounds detected in these samples. In Chapter 5 the results from this study and what impact they could have on quercetin’s potential ability to be the compound responsible for the health benefits are discussed.

- / -

En forêt tropicale humide Amazonienne, les facteurs gouvernant l'évolution des espèces d'arbres restent peu connus et continuellement débattus. En particulier, les micro-variations environnementales attirent beaucoup d'attention car elles induisent de profondes modifications de structure et composition des communautés. Les variations micro-environnementales associées à la topographie ont couramment été évoquées comme facteur de radiations adaptatives chez les espèces d'arbres. Cependant, l'hypothèse de l'adaptation locale n'a jamais été testée au niveau intra-spécifique chez les arbres de forêt amazonienne alors que l'on sait que la diversité génétique des arbres tropicaux est couramment structurée à faibles échelles spatiales par des processus neutres (en particulier du fait de restrictions de flux de gènes). Dans cette étude, j'ai étudié le processus de différentiation génétique d'une espèce d'arbre (Eperua falcata, Fabaceae) dans les paysages forestiers de Guyane française grâce à la combinaison d'une approche phénotypique (génétique quantitative) et d'une approche moléculaire (génétique des populations). Je me suis attachée à répondre à trois questions principales : 1) Comment se distribue la diversité génétique dans les paysages forestiers de Guyane française ? 2) Quelles forces évolutives sont impliquées dans le processus de différentiation génétique à faible échelle spatiale ? 3) Est-ce que le processus d'adaptation locale contribue à structurer la diversité génétique à faible échelle spatiale ? / In the tropical rainforest of Amazonia, the factors driving the evolution of tree species remain poorly understood, and the relative influence of neutral and adaptive processes is continuously debated. In particular, local habitat patchiness draws much attention, as profound changes in the structure and composition of forest communities occur among micro-habitats. Thus, micro-environmental variations related to topography have frequently been invoked as drivers of adaptive radiation leading to sympatric speciation in Neotropical trees. On one hand, the hypothesis of local adaptation has never been investigated at the intra-specific level, i.e. within species currently undergoing population differentiation; on the other hand, many tree species are genetically structured over local scales due to neutral processes, mainly limited gene flow (caused by restricted pollen and seed dispersal). In this study, I used populations of a common tree species of the Guiana Shield - Eperua falcata (Fabaceae) - to study how neutral and adaptive processes shape the distribution of genetic diversity across forest landscapes characterized by local micro-habitat patchiness. I asked three main questions by combining both phenotypic (quantitative genetics) and molecular (population genetics) approaches: 1) How is the genetic diversity structured in forest landscapes of French Guiana? 2) Which evolutionary drivers are relevant to explain the structure of genetic diversity at local scale? 3) Does local adaptation contribute to structure genetic diversity within continuous populations?

Les polysaccharides sont des composants majeurs des parois cellulaires ayant une structure dynamique et jouant un rôle essentiel dans la croissance des plantes. Les cellules épidermiques du tégument des graines d’Arabidopsis libèrent un halo de mucilage polysaccharidique lors de leur imbibition. Le mucilage séminal s'est avéré être un excellent système modèle pour l’étude de la production, des propriétés des polysaccharides et de leurs interactions. Le premier objectif de ma thèse était de valoriser la variabilité naturelle existant entre accessions d’Arabidopsis pour identifier de nouveaux gènes contrôlant la production de mucilage. Une analyse de génétique d’association a permis l’identification d’une amine oxidase et d’une glycosyltransferase putatives dont j’ai confirmé l’implication dans la biosynthèse des pectines du mucilage.J’ai également associé une famille de protéines aux interactions entre polysaccharides. De part ses propriétés d’hydrogel, le mucilage joue un rôle adaptatif et influence la physiologie de la graine. Il permet la rétention d’eau autour de celle-ci et pourrait ainsi influencer sa longévité. Le deuxième objectif de ma thèse était d’utiliser des mutants impactés dans la production de mucilage pour déterminer si ce dernier influence la longévité des graines après un traitement d’hydratation contrôlée appelé ‘priming’. Les graines ne libérant pas de halo de mucilage à l’imbibition ont une meilleure longévité en lien avec une diminution réduite d’acide salicylique. Mes résultats participent à une compréhension intégrée de la production de mucilage à plusieurs niveaux : écologiques, génétiques et physicochimiques. / Polysaccharides are the major component of cell walls that are dynamic structures playing a fundamental role in plant growth. On imbibition, the epidermal cells of the Arabidopsis seed coat release a mucilage hydrogel formed of polysaccharides. This has proved to be an excellent model system for the study of cell wall polysaccharide production, properties and interactions. The first objective of my thesis was to exploit natural variation between Arabidopsis accessions to identify genes controlling mucilage polysaccharide production. A genome wide association study identified genes encoding proteins with putative functions as either an amine oxidase or glycosyltransferase and these were confirmed to contribute to the synthesis of mucilage pectin. I also found that a family of small proteins, whose function is undetermined, are likely to modulate the interaction of mucilage polymers. Mucilage is also an adaptive trait that may influence various aspects of seed physiology. Recent results indicate that this hydrogel plays a role in the retention of water around the seed and could influence their lifespan. A second objective of my thesis was to use mutants showing altered mucilage production to determine its contribution to seed lifespan after a controlled hydration treatment called ‘priming’. Seeds that do not release mucilage on imbibition retained longevity better after priming. I highlighted that the steady state levels of salicylic acid in primed seeds were influenced by mucilage and correlated negatively with their longevity. My results contribute to our genetic, physicochemical and ecophysiological understanding of mucilage production by seeds.

La régulation redox est un mécanisme ancien présent chez les organismes biologiques et impliquée dans diverses voies métaboliques. En particulier chez les organismes photosynthétiques elle est responsable des mécanismes d‘adaptation rapide dans un environnement constamment modifié. Dans les chloroplastes le système ferrédoxine/thiorédoxine est la cascade redox principale qui relie l‘activité de plusieurs enzymes plastidiales à la source lumineuse. Le rôle central dans ce système est joué par la ferrédoxine-thiorédoxine réductase (FTR), une protéine hétérodimérique qui récupère des électrons à partir de la ferrédoxine photoréduite et les transfère pour réduire des thiorédoxines plastidiales. Ces protéines peuvent alors réduire des enzymes cibles, requérant l‘accessibilité de paires de cystéines dans un disulfure dont la réduction résulte en une activation/ inactivation de la cible. Jusqu‘à présent des plantes viables n‘ont pu être obtenues en l‘absence de ce système de régulation. Dans cette thèse des secteurs du système redox ont été explorés chez la plante modèle Physcomitrella patens (une mousse). Par manipulation de gènes l‘influence de l‘enzyme FTR sur la croissance et le développement de la plante a été analysée suivant différents paramètres. De manière à impacter la fonction de la réductase des changements nucléotidiques simples ont été introduits au niveau des codons programmant les cystéines catalytiques et dans un deuxième temps le gène complet a été supprimé. De façon inattendue nous n‘avons observé aucun effet significatif sur la viabilité et le développement des plantes mutantes. De plus, nous avons détecté dans P. patens des thiorédoxines additionnelles absentes chez les plantes à graine qui sont fonctionnelles vis à vis des enzymes cibles mais non-réduites par la FTR. Ceci rend possible un scénario de compensation chez les mutants via un système de réduction FTR-indépendant qui reste à caractériser. Deux des cibles photorégulées, la fructose-1,6-bisphosphatase (FBPase) et la sédoheptulose-1,7-bisphosphatase (SBPase), fonctionnent dans la phase de régénération du cycle de Calvin-Benson cycle et elles possèdent plusieurs caractéristiques de catalyse et de régulation similaires. En combinant des approches biochimiques et structurales, une comparaison fonctionnelle et structurale des deux phosphatases de P. patens a été conduite. De plus l‘analyse phylogénétique a révélé une origine procaryotique indépendante des deux séquences en dépit de leurs similitudes structurales et catalytiques. De plus trois articles de revue résument la plasticité et la représentativité du modèle P. patens pour la recherche forestière, les principes généraux de la régulation redox relativement aux aspects évolutifs et fonctionnels chez les plantes ainsi que l‘ état de l‘art de la régulation redox chez les espèces ligneuses en utilisant principalement le peuplier comme modèle / Redox regulation is an ancient mechanism present in biological organisms and is involved in diverse cellular pathways. In particular in photosynthetic organisms it is responsible for fast adaption mechanisms to a constantly changing environment. In chloroplasts the ferredoxin/thioredoxin system represents the main redox regulatory cascade which links the activity of several plastid enzymes to the energy source, light. A central role in this system is played by the heterodimeric ferredoxin-thioredoxin reductase (FTR), which gains electrons from the photo-reduced ferredoxin and transfers those further on via reduction to plastidal thioredoxins. Those proteins in turn reduce their target enzymes and require therefore the availability of redox sensitive cysteine pairs whose reduction results in an inactivation/activation switch of the targets. So far no viable plants could be obtained in complete absence of this redox regulation system. In this thesis single sections of the system were explored in the model plant Physcomitrella patens. Through gene manipulation the influence of the FTR enzyme on plant growth and development was analysed. In order to impact on the function of the reductase, firstly single nucleotide exchange of the catalytic cysteines was performed and later on the gene was completely deleted. Surprisingly, no significant effect could be observed on the viability and development of mutant lines compared to WT plants. Furthermore we found that P. patens possesses in contrast to seed plants additional thioredoxins which are functional for reduction of FTR target enzymes but are most likely not supplied with electrons by this reductase. Thus a possible rescue scenario independent of FTR could be assumed for P. patens and also by other redox regulation systems present in chloroplasts. Two of the FTR target enzymes, fructose-1,6-bisphosphatase and sedoheptulose-1,7-bisphosphatase, are functional in the regeneration phase of the Calvin-Benson cycle and share similar characteristics in regulation and catalysis. By combining biochemical and structural approaches, a functional comparison of both phosphatases was conducted using cDNAs from P. patens. A stricter TRX-dependent regulation and catalytic cleavage ability for both substrates, FBP and SBP, could be observed for PpSBPase, whereas PpFBPase is only capable of cleaving FBP. By obtaining the oxidized X-ray structure of both enzymes these observations can be associated with the distinct positions of regulatory sites and the various sizes of the substrate binding pocket. In addition, the phylogenetic analysis revealed an independent prokaryotic origin for both phosphatases. Furthermore we summarized in three review articles the amenability of P. patens as model plant for forest research, the general principles of redox regulation in respect of evolution and functional mechanisms in plants, and the current state of the art in forest redox regulation using poplar as exemplary model

The aim of this thesis was to develop stable-isotope steady-state metabolic flux analysis (MFA) based on ¹³C labeling to quantify intracellular fluxes of central carbon metabolism in plants. The experiments focus on the analysis of a heterotrophic cell suspension culture of Arabidopsis thaliana (L) Heynh. (ecotype Landsberg erecta). The first objective was to develop a robust methodology based on combining high quality steady-state stable labeling data, metabolic modeling and computational analysis. A comprehensive analysis of the factors that influence the outcome of MFA was undertaken and best practice established. This allowed a critical analysis of the subcellular compartmentation of carbohydrate oxidation in the cell culture. The second objective was to apply the methodology to nutritional perturbations of the cell suspension. A comparison of growth on different nitrogen sources revealed that transfer to an ammonium-free medium: (i) increased flux through the oxidative pentose phosphate pathway (oxPPP) by 10% relative to glucose utilisation; (ii) caused a substantial decrease in entry of carbon into the tricarboxylic acid cycle (TCA); and (iii) increased the carbon conversion efficiency from 55% to 69%. Although growth on nitrate alone might be expected to increase the demand for reductant, the cells responded by decreasing the assimilation of inorganic N. Cells were also grown in media containing different levels of inorganic phosphate (Pi). Comparison of the flux maps showed that decreasing Pi availability: (i) decreased flux through the oxPPP; (ii) increased the proportion of substrate fully oxidised by the TCA cycle; and (iii) decreased carbon conversion efficiency. These changes are consistent with redirection of metabolism away from biosynthesis towards cell maintenance as Pi is depleted. Although published genome-wide transcriptomic and metabolomic studies suggest that Pi starvation leads to the restructuring of carbon and nitrogen metabolism, the current analysis suggests that the impact on metabolic organisation is much less extreme.