Retinopathy of prematurity, a disease characterised by aberrant retinal vascular development in premature neonates, is a leading cause of blindness and visual impairment in childhood. This work sought to examine differences in the susceptibility of inbred rat strains to oxygen-induced retinopathy, a model of human retinopathy of prematurity. The overriding aim was to identify genetic factors in rats that might be generalisable to humans. Newborn rats of six different strains were exposed to alternating cycles of hyperoxia and relative hypoxia for fourteen days. Rats were removed to room air and killed for analysis immediately, to assess oxygen-induced retinal vascular attenuation, or four days later to evaluate the extent of hypoxia-induced vasoproliferation. Whole flat-mounted retinae were stained with fluorophore conjugated isolectin GS-IB4, and measurement of vascular area was conducted using fluorescence microscopy and video-image analysis. A hierarchy of susceptibility to the inhibitory effects of cyclic hyperoxia and relative hypoxia on postnatal retinal vascularization was identified for the rat strains studied. Susceptibility to vascular attenuation was predictive of the subsequent risk of vascular morphological abnormalities. Cross-breeding experiments between susceptible and resistant strains demonstrated that the susceptible phenotype was dominantly inherited in an autosomal fashion. These studies confirmed an association between ocular pigmentation and retinopathy risk, however the finding of differential susceptibility amongst albino rat strains implicated factors in addition to those associated with ocular pigmentation. Quantitative real-time reverse transcription-polymerase chain reaction was used to compare the retinal expression of angiogenic factor genes in susceptible and resistant strains with the aim of identifying a genetic basis for the strain difference. Eight angiogenic factor genes were selected for study: vascular endothelial growth factor (VEGF); VEGF receptor 2; angiopoietin 2; Tie2; pigment epithelium-derived factor; erythropoietin; cyclooxygenase-2 and insulin-like growth factor-1. The most notable difference between strains was the expression of vascular endothelial growth factor (VEGF) during the cyclic hyperoxia exposure period - higher VEGF expression was associated with relative resistance to retinopathy. Other differences in retinal angiogenic factor gene expression between strains, such as higher expression of VEGF receptor 2 and angiopoietin 2 in resistant strains, appeared to be secondary to those in VEGF. Following cyclic hyperoxia, the expression pattern of angiogenic factor genes changed - messenger RNA levels of hypoxia-induced genes, including VEGF, VEGF receptor 2, angiopoietin 2 and erythropoietin, were significantly higher in those strains with larger avascular areas, than in those strains that were relatively resistant to retinopathy. These findings provide firm evidence for hereditary risk factors for oxygen-induced retinopathy in the rat. Differences in the regulatory effects of oxygen on VEGF expression appear to be central to the risk of retinopathy. The potential relevance of these hereditary factors is discussed in the context of the human disease.

Lipoprotein lipase (LPL) is the main enzyme hydrolyzing triglyceride-rich lipoproteins in plasma. Proteoglycan-bound LPL on the vascular endothelium represent the functional pool of active enzyme. LPL is regulated in a tissue specific manner according to metabolic demands. Rapid regulation of LPL activity is necessary to provide free fatty acids for storage or energy production. This regulatory mechanism appears to be post-translational and requires synthesis of other protein/proteins. Recently it was demonstrated that angiopoietin-like protein 4 (ANGPTL4) is involved in the metabolism of plasma triglycerides and that it is able to inhibit LPL activity in vitro. These properties were linked to the N-terminal coiled-coil domain of ANGPTL4 (ccd-ANGPTL4), but the mechanism for the inhibition was not known. The aim of this thesis was to investigate the molecular mechanism for inhibition of LPL by ccd-ANGPTL4, to characterize regions in ccd-ANGPTL4 that are important for inactivation of LPL and to study the role of ANGPTL4 for regulation of LPL in vivo. Binding of ccd-ANGPTL4 to LPL was demonstrated by several methods, including surface plasmon resonance. The interaction was transient and resulted in conversion of the enzyme from catalytically active dimers to inactive monomers with decreased affinity for heparin. We have shown that ANGPTL4 mRNA in rat adipose tissue turns over rapidly and that changes in the ANGPTL4 mRNA abundance were inversely correlated to LPL activity, both during the fed to fasted and the fasted to fed transitions. We conclude that ANGPTL4 is a fasting-induced controller of LPL in adipose tissue, acting extracellularly on the native conformation of LPL in an unusual fashion, like an unfolding molecular chaperone. Site directed mutagenesis was used to explore regions in ccd-ANGPTL4 important for inactivation of LPL, and for binding of ANGPTL4 to heparin. Others had shown that ccd-ANGPTL4 forms higher oligomers. Structure prediction analyses demonstrated that the coiled-coil domain of ccd-ANGPTL4 probably forms three consecutive α-helices with strong hydrophobic faces, and that there are clusters of positively charged residues both on the helices and in intervening sequences. We made replacements of hydrophobic residues, positively charged residues, cysteine residues and negatively charged residues in ccd-ANGPTL4. In addition, helix-breaking proline residues were introduced in all three helices. We found that hydrophobic residues are important for oligomer formation. The higher oligomers appeared to be stabilized by disulfide bonds, but cysteines are not crucial for oligomerization. Introduction of Pro-residues in the first and second helix prevented formation of higher oligomers and reduced the ability of ccd-ANGPTL4 to inactivate LPL. We found that negatively charged residues in ccd-ANGPTL4 are important for inactivation of LPL. A heparin binding site was localized in the C-terminal end of ccd-ANGPTL4 (amino acid residues 114-140). To investigate whether LPL is differently processed in different depots of adipose tissue we measured the levels of LPL mRNA, protein and activity in omental and subcutaneous adipose tissue in human subjects undergoing elective surgery. Our results show that, although the expression level of LPL was higher in subcutaneous adipose tissue, the specific LPL activity (ratio of activity over the LPL protein mass) was higher in omental adipose tissue. Interestingly, the levels of ANGPTL4 mRNA were lower in omental compared to subcutaneous adipose tissue in most of the studied subjects. This difference can possibly explain the higher specific activity of LPL in omental adipose tissue and indicated that ANGPTL4 is involved in regulation of LPL activity also in humans. LPL produced by macrophages in the artery wall promotes local accumulation of lipids in these cells, and thereby plays an important role in development of atherosclerosis. The known association between type 2 diabetes and atherosclerosis forwarded us to study production of LPL by THP-1 macrophages under hyperglycemic conditions and under treatment with a peroxisome proliferator-activated receptor delta (PPARδ) agonist (GW501516). We found that LPL activity (but not LPL mass) produced by macrophages was decreased by GW501516. The loss of LPL activity coincided with increased level of ANGPTL4 mRNA, indicating that the agonist regulates LPL activity through expression of ANGPTL4. This effect was even more pronounced in cells grown under hyperglycemic conditions. Our data suggest that a suitable PPARδ agonist, like GW501516, may have protective effects against development of atherosclerosis in subjects with diabetes type 2.

The acute respiratory distress syndrome (ARDS) and its less severe form, acute lung injury (ALI), are among the leading causes of morbidity and mortality in critically ill patients. Commonly induced by conditions associated with severe pulmonary inflammation, ALI results in disruption of the lung alveolar-capillary membrane barrier and resultant pulmonary edema associated with a proteinaceous alveolar exudate. Sepsis is another frequent and often fatal clinical condition for patients in the intensive care unit. It is characterized by a combination of infection and systemic inflammatory response syndrome (SIRS). Current effective treatment strategies for both ALI/ARDS and sepsis are lacking. We first examined the potential therapeutic role of mesenchymal stromal cells (MSCs) alone or together with the vasculoprotective factor, angiopoietin-1 (ANGPT1), for treatment of experimental ALI in mice. MSCs significantly reduced LPS (lipopolysaccharide)-induced pulmonary inflammation, as reflected by cell counts in bronchoalveolar lavage (BAL) fluid and pro-inflammatory cytokine levels in both BAL fluid and lung parenchymal homogenates. More importantly, administration of MSCs transfected with human ANGPT1 plasmid (MSCs-pANGPT1) completely reversed LPS-induced permeability in the lung (i.e., ALI). A follow-up study showed that MSCs remained effective in rescuing mice with LPS-induced ALI; however, the additional benefit from ANGPT1 was no longer observed. To further evaluate MSC-based therapy in a more clinically relevant model of acute injury, the cecal-ligation-and-puncture (CLP) model for sepsis was employed. Our results demonstrated that MSCs can reduce both systemic and pulmonary inflammation, as well as renal and liver dysfunction/injury, as reflected by plasma urea and bilirubin levels, in septic mice. Most notably, MSCs reduced sepsis-associated mortality from 45% to 24%. Our data demonstrate the feasibility and effectiveness of MSC- and MSC-based gene therapy for experimental ALI and sepsis, and provide the basis for the development of an innovative approach for the prevention and treatment of clinical ALI/ARDS and sepsis.

The acute respiratory distress syndrome (ARDS) and its less severe form, acute lung injury (ALI), are among the leading causes of morbidity and mortality in critically ill patients. Commonly induced by conditions associated with severe pulmonary inflammation, ALI results in disruption of the lung alveolar-capillary membrane barrier and resultant pulmonary edema associated with a proteinaceous alveolar exudate. Sepsis is another frequent and often fatal clinical condition for patients in the intensive care unit. It is characterized by a combination of infection and systemic inflammatory response syndrome (SIRS). Current effective treatment strategies for both ALI/ARDS and sepsis are lacking. We first examined the potential therapeutic role of mesenchymal stromal cells (MSCs) alone or together with the vasculoprotective factor, angiopoietin-1 (ANGPT1), for treatment of experimental ALI in mice. MSCs significantly reduced LPS (lipopolysaccharide)-induced pulmonary inflammation, as reflected by cell counts in bronchoalveolar lavage (BAL) fluid and pro-inflammatory cytokine levels in both BAL fluid and lung parenchymal homogenates. More importantly, administration of MSCs transfected with human ANGPT1 plasmid (MSCs-pANGPT1) completely reversed LPS-induced permeability in the lung (i.e., ALI). A follow-up study showed that MSCs remained effective in rescuing mice with LPS-induced ALI; however, the additional benefit from ANGPT1 was no longer observed. To further evaluate MSC-based therapy in a more clinically relevant model of acute injury, the cecal-ligation-and-puncture (CLP) model for sepsis was employed. Our results demonstrated that MSCs can reduce both systemic and pulmonary inflammation, as well as renal and liver dysfunction/injury, as reflected by plasma urea and bilirubin levels, in septic mice. Most notably, MSCs reduced sepsis-associated mortality from 45% to 24%. Our data demonstrate the feasibility and effectiveness of MSC- and MSC-based gene therapy for experimental ALI and sepsis, and provide the basis for the development of an innovative approach for the prevention and treatment of clinical ALI/ARDS and sepsis.

Non transmis / Not transmitted

No description available.

L’angiogenèse est la formation de nouveaux vaisseaux sanguins à partir d’un réseau vasculaire existant. C’est un phénomène essentiel pour des processus physiologiques et pathologiques. L’activation des cellules endothéliales est contrôlée par plusieurs facteurs de croissance. Le VEGF et son récepteur le VEGFR-2 ont été prouvés comme étant spécifiques et critiques pour la formation des vaisseaux sanguins alors que Tie2, le récepteur auquel se lie l’Ang-1, est requis aussi bien dans le développement vasculaire que dans l’angiogenèse tumorale. Il est connu que l’activation de Tie2 est nécessaire à la stabilisation finale de la vascularisation en inhibant la perméabilité vasculaire induite par le VEGFR-2. Nous avons premièrement découvert que le facteur de croissance pro-angiogénique, l’Ang-1 contrecarre les effets de perméabilité cellulaire induits par le VEGF en inhibant la production de NO dans les cellules endothéliales. Cet effet inhibiteur de Tie2 intervient directement au niveau de l’activité de l’enzyme eNOS. Suite à l’activation de Tie2 par l’Ang-1, eNOS devient fortement phosphorylé sur la Thr497 après la phosphorylation et l’activation de la PKCζ. Nos résultats suggèrent que l’inhibition, par Tie2, de la perméabilité vasculaire durant l’angiogenèse serait due, en partie, à l’inhibition de la production de NO. Deuxièmement nous avons pu distinguer entre deux modes de migration cellulaire endothéliale induits par l’Ang-1 et le VEGF. À l’opposé du VEGF qui promeut une migration individuelle aléatoire, l’Ang-1 induit une migration collective directionnelle. Dans cette étude, nous avons identifié la β-caténine comme un nouveau partenaire moléculaire de la PKCζ. Cette association de la PKCζ à la β-caténine amène le complexe de polarité Par6-aPKC et le complexe des jonctions d’adhérences cellulaires à interagir ensemble à deux localisations différentes au niveau de la cellule endothéliale. Au niveau des contacts intercellulaires, le complexe PKCζ/β-caténine maintien la cohésion et l’adhésion cellulaire nécessaire pour le processus migratoire collectif. Ce complexe se retrouve aussi au niveau du front migratoire des cellules endothéliales afin d’assurer la directionalité et la persistance de la migration endothéliale en réponse à l’Ang-1. D’une manière intéressante, lors de l’inhibition de la PKCζ ou de la β-caténine on assiste à un changement du mode de migration en réponse à l’Ang-1 qui passe d’une migration directionnelle collective à une migration individuelle aléatoire. Ce dernier mode de migration est similaire à celui observé chez des cellules endothéliales exposées au VEGF. Ces résultats ont été corroborés in vivo par une polarité et une adhésion défectueuses au cours de la vasculogenèse chez le poisson zèbre déficient en PKCζ. En résumé, Ang-1/Tie2 module la signalisation et les réponses biologiques endothéliales déclenchées par le VEGF/VEGFR-2. L’identification des mécanismes moléculaires en aval de ces deux récepteurs, Tie2 et VEGFR-2, et la compréhension des différentes voies de signalisation activées par ces complexes moléculaires nous permettra de mettre la lumière sur des nouvelles cibles thérapeutiques pour le traitement des maladies angiogéniques. / Angiogenesis is the formation of new blood vessels from a pre-existing vascular network. It is an essential mechanism for many physiological and pathological conditions. Also, the general mechanism in both conditions remains the same. VEGF and its receptor VEGFR-2 have been proven to be specific and critical for blood vessel formation. The Angiopoietin-1 receptor, Tie2, is required for vascular development as well as in tumor angiogenesis. It is known that the activation of Tie2 is required for vascular stabilization by inhibiting vascular permeability induced by VEGFR-2. First, we found that the pro-angiogenic growth factor, Ang-1 counteracts the effects of VEGF-induced permeability by inhibiting NO production by endothelial cells. This inhibitory effect of Tie2 acts directly on eNOS activity. Following activation of Tie2 by Ang-1, eNOS becomes highly phosphorylated on the inhibitory site, the Thr497, following PKCζ phosphorylation and activation. Our results suggest that the inhibition by Tie2 of vascular permeability during angiogenesis is due, in part, to the inhibition of NO production. In our second study we distinguished between two types of endothelial cell migration induced by Ang-1 and VEGF. At the opposite of Ang-1 that induced collective and directional cell migration, VEGF promoted individual and random cell motility. We identified β-catenin as a new molecular partner of PKCζ. This association of PKCζ with β-catenin brings the Par6-aPKC polarity complex and the adherens junctions complex to interact with each other at two different locations in endothelial cells. PKCζ/β-catenin complex is located specifically at cell-cell contacts to maintain cohesion and cell adhesion necessary for the collective migration process. This complex was located also at the leading edge of endothelial cells during migration to ensure the directionality and the persistence of migration in response to Ang-1. In addition, inhibition of PKCζ or β-catenin switched the migration mode, in response to Ang-1, from directional and collective to a more random and individual cell migration which resembles the type of migration of endothelial cells exposed to VEGF. These results were confirmed in vivo by aberrant cell polarity and cell adhesion defects of tip cell during vascular sprouting of intersegmental vessels in PKCζ deficient zebrafish embryos. In summary, Ang-1/Tie2 modulates endothelial cell signaling and biological responses induced by VEGF/VEGFR-2. The identification of molecular mechanisms involved in the action of these two receptors, VEGFR-2 and Tie2, and the understanding of the different signaling pathways activated by these molecular complexes will allow us to identify new therapeutic targets for the treatment of angiogenic diseases treatment.

Le vieillissement vasculaire est caractérisé par une dysfonction de l’endothélium. De nombreux facteurs de risque cardiovasculaire tels que l’obésité et l’hypertension prédisposent l’endothélium à un stress oxydant élevé aboutissant à une dysfonction endothéliale, celle-ci étant communément accompagnée d’une diminution de la biodisponibilité du monoxyde d’azote. Bien que la fonction endothéliale soit un déterminant majeur de la prédiction du risque cardiovasculaire des patients, son évaluation individuelle reste très limitée. En conséquence, il existe un intérêt scientifique grandissant pour la recherche de meilleurs biomarqueurs. L’Angiopoiétine like-2 (angptl2), une protéine identifiée récemment, joue un rôle pro-inflammatoire et pro-oxydant dans plusieurs désordres causés par une inflammation chronique allant de l’obésité à l’athérosclérose. L’inflammation et un stress oxydant accru ont été établis comme des mécanismes sous-jacents à l’apparition d’une dysfonction endothéliale, c’est pourquoi ce travail met l’accent sur le rôle de l’angptl2 dans la dysfonction endothéliale. Plus précisément, ce travail vise à: 1) déterminer les effets aigus de l’angptl2 sur la fonction endothéliale, 2) caractériser la fonction endothéliale et la contribution des différents facteurs relaxants dérivés de l'endothélium (EDRF) dans plusieurs lits vasculaires, et ce, dans un modèle de souris réprimant l’expression de l’angptl2 (knock-down, KD), et 3) examiner si l'absence d'expression angptl2 protège contre la dysfonction endothéliale induite par un régime riche en graisses (HFD) ou par perfusion d'angiotensine II (angII) chez la souris. Dans la première étude, l’incubation aigue avec de l’angptl2 recombinante induit une dysfonction endothéliale dans les artères fémorales isolées de souris de type sauvage (WT), probablement en raison d’une production accrue d'espèces réactives oxygénées. Les artères fémorales de souris angptl2 KD présentent une meilleure fonction endothéliale en comparaison aux souris WT, vraisemblablement par une plus grande contribution de la prostacycline dans la vasodilatation. Après 3 mois d’une diète HFD, les principaux EDRF respectifs des artères fémorales et mésentériques sont conservés uniquement dans les souris angptl2 KD. Cette préservation est associée à un meilleur profil métabolique, une moindre accumulation de triglycérides dans le foie et des adipocytes de plus petite taille. De plus, l’expression de gènes inflammatoires dans ces tissus adipeux n’est augmentée que chez les souris WT. Dans la seconde étude, l’absence d’angptl2 résulte en une production accrue de monoxyde d’azote dans les artères cérébrales isolées par rapport à celles des souris WT. La perfusion chronique d’angII provoque, seulement chez les souris WT, une dysfonction endothéliale cérébrale probablement par le biais d’une augmentation de la production d’espèces réactives oxygénées, probablement dérivé des NADPH oxydase 1 et 2, ainsi que l'augmentation des facteurs constricteurs dérivés de l’endothélium issus de la cyclo-oxygénase. En revanche, l’apocynine réduit la dilatation cérébrale chez les souris KD traitées à l’angII, ce qui suggère le recrutement potentiel d’une voie de signalisation compensatoire impliquant les NADPH oxydases et qui aurait un effet vaso-dilatateur. Ces études suggèrent fortement que l’angptl2 peut avoir un impact direct sur la fonction endothéliale par ses propriétés pro-inflammatoire et pro-oxydante. Dans une optique d’application à la pratique clinique, les niveaux sanguins d’angptl2 pourraient être un bon indicateur de la fonction endothéliale. / Vascular aging is characterized by changes in the endothelium. Common cardiovascular risk factors, including obesity and hypertension, predispose the endothelium to increased oxidative stress, leading to endothelial dysfunction commonly characterized by diminished nitric oxide bioavailability. Although endothelial function can be a major determinant of cardiovascular risk prediction in patients, individual testing is still limited in clinical settings and thus there is increasing scientific interest in finding better biomarkers. Angiopoietin like-2 (angptl2), a recently identified protein, is a pro-inflammatory and pro-oxidative protein involved in chronic inflammatory disorders ranging from obesity to atherosclerosis. As inflammation and increased oxidative stress are established underlying mechanisms by which endothelial dysfunction occurs, this work focuses on the role of angptl2 in endothelial dysfunction, a topic that is largely unexplored. Specifically, this work aims to 1) determine the acute effects of angptl2 on endothelial function, 2) characterize endothelial function and contribution of different endothelium-derived relaxing factors in various vascular beds in a newly generated angptl2 knock-down (KD) mouse model, and 3) examine whether the lack of angptl2 expression protects against endothelial dysfunction induced by either a high-fat diet (HFD) or angiotensin II (angII) infusion in mice. In the first study, we show that a recombinant angptl2 acutely evokes endothelial dysfunction in the femoral artery isolated from wild-type (WT) mice, likely due to increased production of reactive oxygen species. Also in the femoral artery, angptl2 KD mice display better endothelial function compared to WT, which may be a result of greater prostacyclin contribution to vasodilation. After a 3-month HFD, the main respective endothelium-derived relaxing factors in the femoral and mesenteric arteries are preserved in angptl2 KD mice only, which was associated with a better metabolic profile, such as lower total cholesterol-to-high-density lipoprotein and low-density-to-high-density lipoprotein ratios compared to WT mice. After a HFD, KD mice have less triglyceride accumulation in the liver and smaller adipocytes in their mesenteric and epididymal white adipose tissues compared to WT mice, while inflammatory gene expressions in adipose tissues increase in WT mice only. In the second study, we reveal that the lack of angptl2 in KD mice results in greater nitric oxide production compared to WT mice in their isolated cerebral arteries. Chronic infusion of pro-inflammatory and pro-oxidative angII results in cerebral endothelial dysfunction only in WT mice, which is acutely ameliorated with either N-acetylcysteine, apocynin, or indomethacin, suggesting increased reactive oxygen species, likely derived from the NADPH oxidases 1/2, and increased cyclooxygenase-derived endothelium-derived contracting factors. In contrast, apocynin reduces cerebral dilation in angII-treated KD mice, suggesting recruitment of a potential compensatory dilatory NADPH oxidase pathway. These studies are the first to explore angptl2 contribution to endothelial dysfunction in different vascular beds, and strongly suggest that angptl2 can directly impair endothelial function by its pro-inflammatory and pro-oxidative properties. Translating this to the clinical setting, expression levels of angptl2 may be an indicator of endothelial function, and lowering angptl2 levels could become a potential therapeutic approach in the treatment of chronic inflammatory disorders including cardiovascular diseases.

Suite à l’exposition à des facteurs de risque incluant la malnutrition, la dyslipidémie, la sédentarité et les désordres métaboliques, les maladies cardiovasculaires (MCV) sont caractérisées par un état pro-oxydant et pro-inflammatoire, et une dérégulation de l’expression de divers facteurs responsables de l’homéostasie de l’environnement rédox et inflammatoire. L’implication d’enzymes antioxydantes telles que les superoxyde dismutases (SOD) et les glutathion peroxydases (Gpx), ainsi que la contribution de médiateurs pro-inflammatoires tels que l’angiopoietin-like 2 (Angptl2) ont été rapportées dans le cadre des MCV. Toutefois, les mécanismes moléculaires sensibles aux facteurs de risque et menant au développement des MCV sont peu connus. L’épigénétique est un mécanisme de régulation de l’expression génique sensible aux stimuli extracellulaires et pourrait donc contribuer au développement des MCV. La méthylation de l’ADN est un des mécanismes épigénétiques pouvant varier tant de manière gène-spécifique qu’à l’échelle génomique, et la conséquence de tels changements sur l’expression des gènes ciblés dépend du site de méthylation. Puisqu’il a été démontré que des variations au niveau de la méthylation de l’ADN peuvent être associées à divers contextes pathologiques incluant les MCV, le but de nos travaux était d’étudier le lien entre la méthylation de gènes antioxydants et pro-inflammatoires avec leurs répercussions fonctionnelles biologiques en présence de facteurs de risques associés aux MCV, tels que le vieillissement, la dyslipidémie et la sédentarité. Dans la première étude, nous avons observé que dans l’artère fémorale de souris vieillissantes, la méthylation au niveau du promoteur du gène Sod2, codant pour l’enzyme antioxydante superoxyde dismutase de type 2 (SOD2 ou MnSOD), diminue avec l’âge. Ceci serait associé à l’induction de l’expression de MnSOD, renforçant ainsi la défense antioxydante endogène. Le vieillissement étant associé à une accumulation de la production de radicaux libres, nous avons étudié la vasodilatation dépendante de l’endothélium qui est sensible au stress oxydant. Nous avons observé que la capacité vasodilatatrice globale a été maintenue chez les souris âgées, aux dépens d’une diminution des facteurs hyperpolarisants dérivés de l’endothélium (EDHF) et d’une contribution accentuée de la voie du monoxyde d’azote (NO). Nous avons ensuite utilisé deux approches visant à réduire les niveaux de stress oxydant in vivo, soit la supplémentation avec un antioxydant, la catéchine, et l’exposition chronique à de l’exercice physique volontaire. Ces interventions ont permis de prévenir à la fois les changements au niveau de la fonction endothéliale et de l’hypométhylation de Sod2. Cette première étude démontre donc la sensibilité de la méthylation de l’ADN à l’environnement rédox. Dans la deuxième étude, nous avons démontré une régulation de l’expression de l’enzyme antioxydante glutathion peroxydase 1 (Gpx1) en lien avec la méthylation de son gène codant, Gpx1, dans un contexte de dyslipidémie sévère. Nos résultats démontrent que dans le muscle squelettique de souris transgéniques sévèrement dyslipidémiques (LDLr-/-; hApoB+/+), Gpx1 est hyperméthylé, ce qui diminue l’expression de Gpx1 et affaiblit la défense antioxydante endogène. Chez ces souris, l’exercice physique chronique a permis d’augmenter l’expression de Gpx1 en lien avec une hypométhylation transitoire de son gène. Cette étude démontre que le stress oxydant associé à la dyslipidémie sévère altère les mécanismes de défense antioxydante, en partie via un mécanisme épigénétique. De plus, on observe également que l’exercice physique permet de renverser ces effets et peut induire des changements épigénétiques, mais de manière transitoire. La troisième étude avait pour but d’étudier la régulation de l’Angptl2, une protéine circulante pro-inflammatoire, dans le contexte des MCV. Nous avons observé que chez des patients coronariens, la concentration circulante d’Angptl2 est significativement plus élevée que chez des sujets sains et ce, en lien avec une hypométhylation de son gène, ANGPTL2, mesurée dans les leucocytes circulants. Nous sommes les premiers à démontrer qu’en réponse à l’environnement pro-inflammatoire associé à une MCV, l’expression de l’Angptl2 est stimulée par un mécanisme épigénétique. Nos études ont permis d’identifier des nouvelles régions régulatrices différentiellement méthylées situées dans les gènes impliqués dans la défense antioxydante, soit Sod2 en lien avec le vieillissement et Gpx1 en lien avec la dyslipidémie et l’exercice. Nous avons également démontré un mécanisme de régulation de l’Angptl2 dépendant de la méthylation d’ANGPTL2 et ce, pour la première fois dans un contexte de MCV. Ces observations illustrent la nature dynamique de la régulation épigénétique par la méthylation de l’ADN en réponse aux stimuli environnementaux. Nos études contribuent ainsi à la compréhension et l’identification de mécanismes moléculaires impliqués dans le développement du phénotype pathologique suite à l’exposition aux facteurs de risque, ce qui ouvre la voie à de nouvelles approches thérapeutiques. / Following exposure to risk factors including malnutrition, dyslipidemia, physical inactivity and metabolic disorders, cardiovascular diseases (CVD) are characterized by a pro-oxidative and pro-inflammatory state, and a dysregulation in the expression of various factors responsible for the redox and inflammatory environment homeostasis. The implication of antioxidant enzymes, such as superoxide dismutases (SOD) and glutathione peroxidases (Gpx), as well as the contribution of pro-inflammatory mediators such as angiopoietin-like 2 (Angptl2) are well characterized in the context of CVD. However, little is known about the molecular mechanisms sensitive to environmental cues leading to the development of CVD. Epigenetics are mechanisms regulating gene expression that are sensitive to extracellular stimuli and could therefore contribute to the pathogenesis of CVD. DNA methylation is an epigenetic mechanism that can vary both at gene and genomic levels; the consequence of these epigenetic changes on the expression of targeted genes is dependent on the methylation site. Since it has been reported that DNA methylation variations can be associated with diverse pathological conditions including CVD, the goal of our work was to study the link between the methylation of antioxidant and pro-inflammatory genes, and their consequences on biological functions in the context of risk factors associated with CVD, such as aging, dyslipidemia and physical inactivity. In the first study, we observed that in the femoral artery of aging mice, the methylation at the promoter of the Sod2 gene, which codes for the antioxidant enzyme superoxide dismutase, type 2 (SOD2 or MnSOD), decreases with age. This suggests an induction of MnSOD expression and thus a strengthening of the endogenous antioxidant defense. Since aging is associated with an accumulation of free radicals, we studied the endothelium-dependant vasodilation, known to be sensitive to oxidative stress. We observed that, overall, vasodilatory capacity was preserved in aging mice, due to a concomitant decrease in endothelium-derived hyperpolarizing factors (EDHF) and an increased contribution of the nitric oxide (NO) pathway. We then used two in vivo oxidative stress-reducing approaches, namely the supplementation with the antioxidant catechin and chronic exposure to voluntary physical exercise. These interventions prevented the changes in endothelial function and the Sod2 hypomethylation-dependent induction of MnSOD expression. Hence, this first study demonstrates the sensitivity of DNA methylation to the redox environment. In the second study, we demonstrated that the antioxidant enzyme glutathione peroxidase 1 (Gpx1) expression was regulated through the methylation of its coding gene, Gpx1, in the context of severe dyslipidemia. Our results show that in the skeletal muscle of severely dyslipidemic transgenic mice (LDLr-/-; hApoB+/+), Gpx1 is hypermethylated, which in turn decreased Gpx1 expression and weakened the endogenous antioxidant defense. In these mice, chronic physical exercise managed to increase Gpx1 expression, an effect linked with a transient gene hypomethylation. This study demonstrates that oxidative stress associated with severe dyslipidemia alters antioxidant defense mechanisms, partially through an epigenetic mechanism. Moreover, we also observed that physical exercise can revert these changes and can induce epigenetic changes, at least transiently. The goal of the third project was to study Angptl2 regulation, a circulating pro-inflammatory protein, in the context of CVD. We observed that, in coronary patients, circulating Angptl2 concentration is significantly increased in conjunction with hypomethylation of its gene, ANGPTL2, measured in circulating leukocytes. We are the first to show that in response to the pro-inflammatory environment associated with a CVD, Angptl2 expression is stimulated by an epigenetic mechanism. In conclusion, our studies allowed the identification of novel regulatory differentially methylated regions located in genes involved in antioxidant defense, namely Sod2, in the context of aging, and Gpx1 in the context of dyslipidemia and exercise. We also revealed, for the first time, an Angptl2 regulating mechanism dependent on ANGPTL2 methylation in a context of CVD. These observations illustrate the dynamic nature of epigenetic regulation through DNA methylation in response to environmental cues. Our studies therefore contribute to the understanding and identification of molecular mechanisms involved in the development of pathological phenotypes following exposure to risk factors, which opens the way to novel therapeutic strategies.

Malaria remains one of the world's major health problems, especially in developing countries. A better understanding of the pathology and pathophysiology of severe malaria is key to develop new treatments. Different approaches have been used in malaria research including the in vitro co-culture models with endothelial cells and both murine and simian animal models. However these are open to controversy due to disagreement on their representativeness of human disease. Using human post-mortem tissue in malaria research is another important approach but is practically challenging, limiting the availability of post mortem samples from malaria patients. The work in this thesis had two main themes. First I examined the role of the endothelial signalling Angiopoetin-Tie-2 receptor pathway in malaria. Ang-2 has been shown to be a significant biomarker of severe and fatal malaria. I examined the tissue specific expression of proteins from this pathway in post-mortem brain tissues from fatal malaria cases, but found no difference between cerebral malaria and non-cerebral malaria cases. Ang-2 correlated with the severity of malaria in these patients. An attempt to examine the interaction of hypoxia and the Ang-Tie-2 pathway in vitro using a co-culture model of human brain endothelial cells was unsuccessful due to contamination of the cell line. The second part of the thesis aimed to utilise molecular pathology techniques including miRNA and whole-genome microarrays. I have shown for the first time that these can be successfully applied to human post-mortem tissue in malaria. First I used archival tissues to examine the microRNA signature in the kidney of patients with malaria associated renal failure. Second I optimised a protocol to preserve post mortem tissue for molecular pathology, from an autopsy study in Mozambique. Using the subsequent total mRNA transcriptomic data and bioinformatics analysis this work has expanded our knowledge of differential gene expression and the families of genes which are dysregulated in the brain in response to malaria infection.