Chronically elevated brain glucocorticoid (GC) levels impair cognition. Age-related cognitive deficits or "sickness" behaviour is often associated with neuroinflammation. In rodents, raised GC levels prior to lipopolysaccharide (LPS) administration potentiate neuroinflammation although GC suppresses neuroinflammation if administered after LPS. 11β-hydroxysteroid dehydrogenase-1 (11β-HSD1) reductase activity can increase intracellular GC levels, including in the brain, without alteration in circulating levels. Deficiency/pharmacological inhibition of 11β-HSD1 is protective against age related cognitive impairment in both rodent and humans. However, the underlying mechanism remains unclear. 11β-HSD1 reductase activity is coupled to hexose-6-phosphate dehydrogenase activity, itself dependent on cellular energy status. Processes affected by deficiency/inhibition of 11β- HSD1 (e.g. acute inflammation, angiogenesis) are associated with increased glycolysis. Additionally, compared to C57BL/6J controls, adipose tissue of 11β-HSD1 deficient mice shows increased expression of glycolytic and oxidative metabolism genes in a rodent model of obesity, characterised by low-grade chronic inflammation. I hypothesised that 11β-HSD1 has a role in regulation of cellular energetics basally and following inflammation. 11β-HSD1 expression in the brain will be up-regulated during systemic inflammation. Following inflammation, 11β-HSD1 deficiency will attenuate the pro-inflammatory response and subsequently alter energy substrate uptake and/or utilisation in the key areas of brain (i.e. hypothalamus and the hippocampus) that sense and respond to inflammation and energy balance. To test my hypothesis, global 11β-HSD1 KO mice, primary macrophages in vitro and murine models of inflammations were utilised. 11β-HSD1 mRNA and protein expression were confirmed in the hypothalamus and the hippocampus of C57BL/6J mice. In the absence of inflammation, expression of inflammatory markers is low or negligible in the brains of Hsd11b1-/- mice similar to C57BL/6J controls. However, compared to C57BL/6J, Hsd11b1-/- mice show altered mRNA levels of metabolic transporters and enzymes in the hypothalamus and the hippocampus. Overall, the mRNA profiling suggests reduced dependence on glucose in the brains of Hsd11b1-/- mice, either through increased lactate availability (in the whole brain and hippocampus) or through increased glycolysis and mitochondrial number/function (in the hypothalamus). Primary macrophages were utilised to investigate the role of 11β-HSD1 in cellular energetics in vitro. In these cell based assays, glycolysis was found to be the predominant glucose metabolising pathway in C57BL/6J primary macrophages, consistent with the literature. Preliminary data suggested reduced glycolytic activity in Hsd11b1-/- compared to C57BL/6J primary macrophages. However, initial attempts to utilise these cell based assays on primary microglia were unsuccessful. Moreover, Hsd11b1 mRNAs in the brain (down-regulation with inflammation, discussed later) was found to be differentially regulated in comparison to Hsd11b1 mRNA levels in the macrophages (up-regulation with inflammation) hence further investigation was not pursued. To identify a model of peripheral inflammation where 11β-HSD1 is regulated in the brain in vivo, Staph. aureus induced acute lung inflammation and the K/BxN serum transfer induced model of arthritis were utilised. Increased expression of inflammatory markers in the brain was associated with reduced Hsd11b1 mRNA levels in the hippocampus of control mice in these models. Comparison of Hsd11b1-/- and C57BL/6J mice showed increased levels of mRNAs encoding metabolic transporters in the hypothalamus and the hippocampus of Hsd11b1-/- mice following inflammation in the K/BxN serum transfer model of arthritis suggesting increased energy substrate availability. Additionally, increased levels of mRNA encoding metabolic enzymes suggested increased glycolytic capacity and mitochondrial oxidative phosphorylation activity in the hippocampus but not the hypothalamus of Hsd11b1- /-, compared to C57BL/6J mice, following K/BxN serum induced arthritis. Overall, these data suggest that the reduction in expression of 11β-HSD1 could be a potential mechanism to increase energy substrate availability, glycolytic capacity and mitochondrial activity in the hippocampus to provide metabolic support for neuronal metabolism and function following peripheral inflammation. The role of 11β-HSD1 in the pro-inflammatory response and cellular energetics in the hippocampus was further investigated in a well characterised sterile peritonitis model of systemic inflammation in which a low to moderate dose of LPS was used. Mice were administered LPS or vehicle (0.9% saline) by a single i.p. injection and culled 3h, 6h or 9h post injection. Inflammation resulted in significant reduction in burrowing activity both in Hsd11b1-/- and C57BL/6J mice suggesting sickness behaviour.. The number of circulating immune cells, as a measure of peripheral inflammation, did not differ between genotypes. Similarly, plasma corticosterone levels were elevated following inflammation but no genotype difference was observed. However, levels of plasma 11-dehydrocorticosterone, the inert substrate for 11β- HSD1, were significantly elevated in the Hsd11b1-/-, compared to C57BL/6J mice, following inflammation. Levels of mRNA encoding inflammatory markers were lower in the hippocampus of Hsd11b1-/-, compared to C57BL/6J mice, following inflammation. Also, Hsd11b1 mRNA levels were reduced in the hippocampus of C57BL/6J mice following inflammation, consistent with the finding above. Principal component analysis on levels of mRNA encoding metabolite transporters and enzymes revealed a distinct metabolic response in the hippocampus of Hsd11b1-/-, compared to C57BL/6J mice, 6h post LPS. At the same time point in the hippocampus, levels of mRNAs encoding metabolite transporters and enzymes suggested an attenuated switch to aerobic glycolysis with maintenance of mitochondrial function/activity. Quantification of hippocampal energy metabolites using targeted metabolomics in the Hsd11b1-/- compared to C57BL/6J mice 6h post LPS showed correspondence with the mRNA results. Overall, these results suggest that reduced expression of 11β-HSD1 could be a potential mechanism to reduce the pro-inflammatory response and provide better metabolic support for neuronal function and metabolism in the hippocampus, following systemic inflammation. In summary, the current work provides evidence for neuroprotection with 11β-HSD1 deficiency, following systemic inflammation. The suggestive neuroprotection is at least in part mediated via an attenuated pro-inflammatory responses and increased energy substrate uptake and/or utilisation providing better metabolic support for neuronal function following inflammation. It argues for the development of tissue specific small molecule inhibitors of 11β-HSD1 that can cross the blood brain barrier as therapeutic agents against the adverse cognitive effects of systemic inflammation and/or inflammaging.

INTRODUCTION Tuberculosis remains a major public health threat globally and the Human Immunodeficiency Virus (HIV) pandemic afflicting developing and developed countries has resulted in enormous increases in tuberculosis infections worldwide. Researchers have previously documented very low plasma vitamin C levels in patients with active pulmonary tuberculosis. This was attributed to a number of factors including: accelerated turnover of vitamin C, shifts in plasma concentrations, increased collagen formation and tissue repair and decreased vitamin C intake. Vitamin C appears to have a role in steroid-genesis and catecholamine synthesis. Decreased plasma vitamin C levels may therefore impact on the stress hormone response and acute phase reaction of patients with active tuberculosis. AIM The primary aims of the study were to measure plasma vitamin C levels, as well as stress hormone levels and acute phase reaction in patients with newly diagnosed active pulmonary tuberculosis and control patients (without tuberculosis), to determine if there was any relationship between vitamin C levels and the levels of these other variables. METHODS AND MATERIALS This was a prospective study of seventy one (71) consecutive patients admitted to Helen Joseph Hospital (between March and October 2002) with newly diagnosed active pulmonary tuberculosis and eighty nine (89) control patients with medical conditions other than tuberculosis. Demographic, clinical and laboratory data were captured and analyzed using SPSS 7.5 soft-ware. Continuous variables were analyzed using students t-test. Categorical data were analyzed by non parametric analysis and Pearsons linear regression model was used to determine the correlation between vitamin C and the other variables in the two groups. RESULTS There were no differences in race, gender, age, suburb of residence and occupational distributions in the study group with tuberculosis compared to the control group. There were more smokers and consumers of alcohol in the control group (54 and 62 patients respectively) than in the study group (28 and 31 patients respectively). The study patients had lower blood pressure (average 90/40 mmHg versus 100/60 mmHg of controls), higher mean pulse rate (101.87 ± 15.14 beats/minute versus 82.92 ± 8.88 beats/minute, p< 0.01), higher mean temperature (38.66 ± 0.67oC versus 37.14 ± 0.44oC, p< 0.01), and lower body mass index (18.29 ± 3.80 Kg/ M2 versus 23.20 ± 5.35 Kg/ M2, p< 0.01). Laboratory data comparing study group and controls also showed marked differences as follows: White cell count (WCC) 8.68×106 / L ± 5.44 versus 11.00×106 / L ± 4.94, p = 0.01; Haemoglobin 9.56gm / dl ± 1.93 versus 12.92gm / dl ± 2.34, p < 0.01 and platelet count 369.21× 106/L ± 190.71 versus 295.94×106 / L ± 94.64, p = 0.01. White cell vitamin C levels (normal range – 20-40 μg/108 leucocytes) were low in half of the patients in both groups (study patients mean 29.85 ± 28.70μg/108 leucocytes versus controls 31.39 ± 30.24μg/108 leucocytes, p = NS). Plasma vitamin C levels were reduced (normal range 10-20 mg/ml) in both groups but more so in the controls (mean 3.87 ± 2.82 mg/ml versus 4.81 ± 3.21 mg / ml in study patients, p= 0.053). Mean cortisol levels were slightly higher in the study patients (448.11 ± 197.41ηmol/L) than controls (392.70 ± 191.25ηmol/L, p = NS). Norepinephrine levels were slightly higher in the study patients than controls (study patients mean 2531.61 ± 2043.60 ρmol/L versus 2178.98 ± 1719.98 ρmol/L of controls, p = NS). Dopamine levels were higher in the study patients than in the controls (468.42 ± 377.57 ρmol/L in study patients versus 293.37 ± 355.84 ρmol/L in controls, p = 0.01). Epinephrine levels were higher in the controls (control patients mean 680.64 ± 743.78 ρmol/L versus 449.41 ± 380.04 ρmol/L of the study patients, p = 0.03). Ferritin levels were much higher in the study patients compared with controls (study patients mean 3005.87 ± 5023.26 μg/L versus 466.51 ± 1774.76 μg/L of the controls, p<0.01) as were CRP levels (125.91 ± 54.77 mg/L in the study patients versus 77.22 ± 81.17 mg/L in the controls, p=0.01). Mean urine cotinine levels were 16.42 ± 24.26μM/L for controls and 9.28 ± 11.59 μM / L for the study patients (p=0.027). Correlation studies did not show any significant differences between the different variables. There was an inverse correlation between CRP levels and urine cotinine levels in the control group (R squared=0.058 and p= 0.024). DISCUSSION There were no differences in the demographic profile of the two groups. Smoking and alcohol consumption were more common in the control group than in the study patients. Over 90 % of patients in both groups had low plasma vitamin C levels, while half of the patients in each group had low white cell vitamin C levels. The low levels of vitamin C could be due to some of the reasons given above or possibly due to the fact that generally there are low levels in Africans for reasons that are not apparent. The control group had increased mean urine cotinine levels suggesting a possible influence of cigarette smoking on vitamin C homeostasis in these patients. In both groups, the majority of patients had normal cortisol levels as well as normal to high catecholamine levels. Also, Ferritin and CRP levels were much higher in the study group than in the controls. The low levels of vitamin C did not, however, have any relationship with stress hormone levels and acute phase reactants. CONCLUSION This study has reaffirmed low plasma and white cell vitamin C levels in patients with new onset pulmonary tuberculosis but has also found low levels in control patients with diseases other than pulmonary tuberculosis. The study demonstrates adequate stress hormone responses in tuberculosis patients, which was not different from non- tuberculosis patients. Acute phase responses were found to be of higher magnitude in tuberculosis patients than in the controls. There were, however, no correlations between plasma vitamin C and stress hormones or acute phase reactants.

The depression of immune cell function that is typically observed after prolonged exercise is thought to be largely mediated by increased plasma concentrations of stress hormones and cytokines and possibly oxidative stress. The aims of this thesis were to determine the effects of acute and longer term oral antioxidant supplementation on immunoendocrine responses following prolonged exercise. In study 1 (Chapter 3) it was shown that vitamin C ingested acutely before and during prolonged exercise has little or no effect on immunoendocrine responses. Furthermore, the combined ingestion of vitamin C with carbohydrate provides no additional effects compared with carbohydrate alone. However, when vitamin C was supplemented acutely, 2 h prior to, and during prolonged exercise in addition to on the night before (14 h prior) exercise this limited the fall in neutrophil oxidative burst activity (study 2, Chapter 4). This was probably a result of reduced direct oxidative damage to neutrophils with vitamin C supplementation since there were no effects on the cortisol, interleukin-6, leukocytosis or neutrophilia responses. Longer periods of antioxidant supplementation (2 - 4 weeks) may be effective at blunting the cortisol, leukocytosis and neutrophilia responses to prolonged exercise (Chapters 5 and 6) but this had no effect on in vitro measures of neutrophil function. In study 5 (Chapter 7) it was shown that acute pre-exercise dark chocolate (which contains polyphenols) ingestion has some effects on plasma oxidative stress markers and circulating insulin and glucose responses but not the immunoendocrine responses to prolonged exercise.

Biologische Diversität existiert sowohl zwischen mehreren Arten als auch innerhalb einer Art, innerhalb von Populationen und Individuen einer Population. Die intraspezifische Diversität wurde bislang ausgiebig auf der Ebene des Genoms untersucht. Sie ist im Kontext metabolischer Zusammenhänge in Pflanzen bisher kaum untersucht und es existieren nur wenige Veröffentlichungen zu diesem Thema. Uns sind bisher keine Publikationen zu Phytohormonen in Sesam bekannt. Neben dem wissenschaftlichen Interesse an der metabolischen Diversität in Sesam, spielen Stresshormone eine wichtige Rolle in der pflanzlichen Abwehr. Der Phytohormonspiegel im Samen ist unter agronomischen Gesichtspunkten relevant, da es vorkommen kann, dass Sesamsamen spontan auskeimen, während sie sich noch an der grünen Pflanze befinden. Diese Eigenschaft ist unerwünscht, da der wertvolle Samen auf diese Weise verloren geht. Im Rahmen dieser Arbeit wurde die Variation im Phytohormonniveau in 16 Akzessionen mit unterschiedlicher geographischer Herkunft untersucht. In Blättern und Wurzeln konnten ABA, JA, SA und SAG nachgewiesen werden, während GA4 lediglich in Blättern vorkam. Eine der Akzessionen aus Japan („Japan 2“) produzierte JA, SA und SAG in hohem Ausmaß. Hier konnten außerdem hohe Gehalte an Chitinasen festgestellt werden. Chitinasen sind für den Abbau von Chitin, dem Hauptbestandteil der pilzlichen Zellwand, verantwortlich. Eine Charakterisierung der Akzessionen mittels AFLP-Analyse zeigte, dass sich „Japan 2“ genetisch nicht mehr von anderen Akzessionen unterschied, als das Mittel der Unterschiede innerhalb aller gesammelten Proben. Bereits in früheren Untersuchungen unserer Arbeitsgruppe im Rahmen einer ungerichteten Metabolitenanalyse, konnte eine hohe Variabilität bei Sesamakzessionen gezeigt werden (Laurentin et al. 2006). Darüber hinaus, stimmen die Unterschiede im metabolischen Profil der Akzessionen nicht mit dem Grad ihrer genetischen Verwandtschaft überein. Es ist bekannt, dass tageszeitliche Unterschiede viele biologische Prozesse kontrollieren. Wir haben die tageszeitlichen Effekte auf den Phytohormonstatus untersucht und dabei die Unterschiede in Pflanzenorganen berücksichtigt. Tageszeitliche Konzentrationen von ABA, JA, IAA, SA und SAG wurden zu 8 unterschiedlichen Tageszeitpunkten in 3 unabhängigen Replikaten mittels HPLC untersucht. Wir konnten keine statistisch signifikanten Unterschiede erkennen. Die Untersuchungen zeigten jedoch eine Variation in den Phytohormonkonzentrationen in unterschiedlichen pflanzlichen Organen. Sekundäre pflanzliche Metabolite spielen als Resistenzfaktoren gegen Mikroorganismen eine wichtige Rolle. Sesamakzessionen, die diese Substanzen im hohen Ausmaß produzieren, stellen eine wichtige züchterische Ressource da. Um die Variation innerhalb der Akzessionen zu untersuchen, die ein hohes Niveau an sekundären Inhaltsstoffen aufweisen, haben wir die Effekte von 32 Pflanzenextrakten aus Sesamakzessionen gegen phytopathogene Pilze untersucht. Darunter befand sich ein Wurzelpathogen mit Spezialisierung auf Sesam (Macrophomina phaseolina), ein Blattpathogen mit breitem Wirtspflanzenkreis (Alternaria alternata) und Gefäßpathogen (Fusarium oxysporum). Die Diversität der Effekte, die für die unterschiedlichen Akzessionen beobachtet werden konnten, führt zu der Annahme, dass die Resistenzeigenschaften der Pflanzen durch gezielte züchterische Beeinflussung der metabolischen Aktivität verbessert werden können. In weiterführenden Untersuchungen zur Aufreinigung der Substanzen mit inhibitorischer Wirkung wurden Pflanzenextrakte in 80% Ethanol mit verschiedenen organischen Lösungsmitteln fraktioniert. Die meisten inhibitorischen Effekte konnten der Diethylether-Fraktion zugeschrieben werden. Im ersten Schritt wurden 4L des Extraktes hergestellt. Zwei aufgereinigte Lignane aus Sesam wurden gegen M. phaseolina, A. alternata und F. oxysporum getestet. Sesamin zeigte keinen Effekt bis zu einer Konzentration von 5mg/ml, während Sesamol (und 2,4-Dinitrophenol als Kontrolle) einen starken inhibitorischen Effekt aufwies. Für diese Substanzen wurden IC50 Werte ermittelt. Man kann festhalten, dass Sesamol dazu dienen kann, das Wachstum invasiver Pathogene einzuschränken. Durch die Kreuzung von zwei Elternlinien, die in der AFLP-Analyse einen signifikanten Polymorphismus aufwiesen, wurden Inzuchtlinien erzeugt. Die Nachkommen dieser Kreuzung wurden in 5 Generationen selbstbefruchtet. Das so entstandene Set aus RILs wurde mittels AFLP charakterisiert. Alle untersuchten RILs waren Hybride. Dies zeigt, dass während der ersten Kreuzung der Elternlinien keine Selbstung erfolgte. Wie erwartet, spalteten polymorphe AFLP-Marker der Elternlinien in den RILs zufällig auf. Monomorphe Marker fehlten in einigen RILs. Des Weiteren traten neue Marker auf, die zuvor nicht in den Elternlinien festgestellt werden konnten. Das Auftreten neuer Marker kann durch Rekombination zwischen Restriktionsfragmenten erklärt werden, welche die AFLP-Marker begrenzen. Die RILs werden nun von unseren Kooperationspartnern zum Aufbau einer genetischen Karte verwendet (Prof. Sami Doganlar und seine Arbeitsgruppe, Universität Antalya, Türkei).