Le processus conduisant à la prolifération et différenciation des cellules souches hématopoïétiques (CSH) en cellules de toutes les lignées sanguines s’appelle l’hématopoïèse. Bien que l'engagement des CSH soit régi par les cytokines, les facteurs de transcription, les modificateurs épigénétiques et la niche des CSH, notre groupe a constaté que leur engagement vers la lignée érythroïde dépendait aussi du métabolisme de la glutamine. La glutaminolyse contribue à la biosynthèse des nucléotides de novo ainsi qu’à la production de l'alpha-kétoglutarate (αKG), intermédiaire métabolique du cycle TCA (Oburoglu et al. 2014). Il est cependant important de noter que la différenciation érythroïde est un processus unique, où chaque cellule fille est structurellement et fonctionnellement différente de sa cellule mère. Chaque division définit un stade de différenciation précis avec un dernier cycle de division produisant un réticulocyte énucléé. Ainsi, nous avons émis l'hypothèse que les réseaux métaboliques mobilisés dans les progéniteurs érythroïdes changent en fonction du stade de différenciation et que ces réseaux régulent la transition des progéniteurs d'un stade à l'autre.Au cours de ma thèse, j’ai caractérisé les états métaboliques associés aux différents stades de différenciation des progéniteurs érythroïdes. Nous avons ainsi montré qu'aux stades précoces de différenciation érythroïde, avant la différenciation terminale, les progéniteurs hématopoïétiques présentent une activité métabolique accrue avec un niveau de phosphorylation oxydative (OXPHOS) plus élevé. Ces données sont en corrélation avec l'augmentation de la génération de l’αKG à ces stades de différenciation. De plus, nous avons constaté une augmentation de l’OXPHOS de ces progéniteurs en présence d’αKG exogène. Cependant, la différenciation terminale des précurseurs érythroïdes, caractérisée par la perte de la masse mitochondriale et de leur potentiel membranaire, est associée à une diminution du niveau d'OXPHOS. Ainsi, l'administration exogène d’αKG, a fortement atténué la différenciation érythroïde terminale et l'énucléation, sans affecter la différenciation des pro-érythroblastes. Inversement, un antagoniste de l’αKG (diméthyloxalylglycine, DMOG) n'a pas altéré la différenciation terminale ou l'énucléation, malgré l'abrogation de l'OXPHOS dans les érythroblastes.Ces données suggèrent que la production d’αKG et sa contribution à l’OXPHOS perturbent l'énucléation des globules rouges. C'est pourquoi, dans le but de réduire les niveaux intracellulaires d’αKG, nous avons inhibé l’expression de l'isocitrate déshydrogénase I (IDH1), enzyme cytosolique catalysant la conversion de l'isocitrate en αKG. Cependant, comme IDH1 peut catalyser les réactions dans les deux sens, la diminution de son expression pourrait également augmenter les niveaux d’αKG. En effet, nous avons constaté que le knockdown d'IDH1 entraînait une forte atténuation de la différenciation terminale et de l'énucléation des précurseurs érythroïdes. Cet effet est probablement dû à un déséquilibre de la disponibilité des substrats ; ainsi l’administration ectopique de l’αKG ainsi que du citrate renforce l’altération de la différenciation terminale des précurseurs érythroïdes IDH1-/- ainsi que leur énucléation. Cette étude identifie donc un rôle crucial pour le métabolite αKG dans la régulation de la fonction mitochondriale et de l’OXPHOS, processus qui sont une condition sine qua non pour la différenciation des précurseurs érythroïdes au stade proérythroblaste. Nous montrons en outre que la suppression d’OXPHOS et la catalyse d’intermédiaires du TCA, substrats d’IDH1, sont requis pour les phases terminales de la différenciation érythroïde et l'énucléation.En conclusion, les résultats obtenus au cours de ma thèse mettent en évidence la nature dynamique des réseaux métaboliques qui régulent la progression des précurseurs érythroïdes tout au long des différents stades de la différenciation érythroïde. / Hematopoiesis is the process whereby hematopoietic stem cells (HSCs) proliferate and differentiate to all blood cell lineages. While HSC commitment is known to be regulated by cytokines, transcription factors, epigenetic modifiers and the HSC niche, our group found that specification of HSCs to the red cell lineage is dependent on glutamine metabolism. Glutaminolysis contributes to de novo nucleotide biosynthesis and to the generation of the alpha-ketoglutarate (αKG) TCA cycle metabolite (Oburoglu et al. 2014). Importantly though, erythroid differentiation is a unique process as each daughter cell is structurally and functionally different from its parent cell. Each division defines a stage of differentiation with the final division cycle resulting in the production of an enucleated reticulocyte which further matures to a biconcave erythrocyte. Thus, we hypothesized that progenitor metabolic networks change as a function of the erythroid differentiation stage and moreover, that they regulate the transition of progenitors from one stage of differentiation to the next.During my PhD, I assessed the metabolic alterations that occur as a function of the erythroid differentiation stage. We showed that at early stages of human red cell development, prior to terminal differentiation, hematopoietic progenitors exhibited an increased metabolic activity with a significantly higher level of oxidative phosphorylation (OXPHOS). This correlated with the increased generation of αKG and indeed, we found that ectopic αKG directly augmented OXPHOS in these progenitors. However, the terminal differentiation of erythroid precursors, characterized by the loss of mitochondrial mass and membrane potential, was associated with a decreased level of OXPHOS. Notably, ectopic αKG, which did not alter pro-erythroblast erythroid differentiation, severely attenuated terminal differentiation and enucleation. Conversely, an αKG antagonist (dimethyloxalyl glycine, DMOG) did not negatively impact on terminal differentiation or enucleation despite abrogating OXPHOS in erythroblasts.These data suggested that the production of αKG and its subsequent contribution to oxidative phosphorylation perturb red cell enucleation. We therefore downregulated isocitrate dehydrogenase I (IDH1), the cytosolic enzyme that catalyzes the conversion of isocitrate to αKG, by an shRNA approach in an attempt to decrease αKG levels. However, because IDH1 can catalyze both the forward and reverse reactions, its downregulation could also increase αKG levels. Indeed, we found that IDH1 knockdown resulted in a severe attenuation of terminal erythroid differentiation and enucleation. This effect was likely due to an imbalance in substrate availability––both ectopic αKG as well as citrate further decreased polychromatic to orthochromatic erythroblast differentiation and the subsequent enucleation of IDH1-knockdown erythroid precursors. Thus, the present study identifies a crucial role for the αKG metabolite in regulating mitochondrial function and oxidative phosphorylation, processes that are a sine qua non for erythroid precursors at the pro-erythroblast stage. We further show that terminal erythroid differentiation and enucleation requires OXPHOS suppression and the IDH1-mediated enzymatic catalysis of its TCA substrates.To conclude, the results generated during my PhD highlight the dynamic nature of the metabolic networks that regulate the progression of erythroid precursors through the distinct stages of erythroid differentiation.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Recent data from the literature have suggested that some Krebs cycle intermediates could act as potent antioxidant agents, both in vitro and in vivo, against a variety of pro-oxidant agents. However, the mechanism(s) involved in the antioxidant effect of Krebs cycle intermediates are not fully understood. Additionally, there are scarce data in the literature taking into account the in vitro effect of Krebs cycle intermediates during oxidative stress conditions. Thus, the aim of this study was to determine the effect of some Krebs cycle intermediates on lipid peroxidation induced in vitro by different pro-oxidant agents, and the mechanism(s) by which they act. Firstly, we investigated the effect and the mechanism(s) by which malonate and quinolinic acid modulate the thiobarbituric acid- reactive species (TBARS) production in vitro, using rat brain S1 preparations (Article 1). The present results showed that the malonate-induced TBARS production was not changed by potassium cyanide or MK-801. However, the pro-oxidant effect of quinolinic acid was significantly prevented by MK-801. In addition we found that malonate was able to form complexes with iron ions (Fe2+), but these complexes were not able to interfere with in vitro deoxyribose degradation assays. Based on the results presented, we conclude that malonate pro-oxidant activity in vitro seems to be independent of the NMDA receptors activity. Additionally, we suggest that the malonate effect, in these conditions, is due to its ability to form complexes with iron ions, thus modulating an adequate ratio Fe2+/Fe3+ that could cause an increase in free radicals generation. In contrast, the quinolinic acid effect seems to be dependent of the NMDA receptors activation. However, we can not rule out the involvement of iron ions in quinolinic acid toxicity under our assay conditions. An other objective of this study was to investigate the effect of some Krebs cycle intermediates on quinolinic acid- or iron (Fe2+)-induced TBARS production in the rat brain S1 preparations, and the mechanism(s) by which they act (Article 2). The results showed that oxaloacetate, citrate, succinate, and malate were able to significantly prevent both basal and quinolinic acid- or iron-induced TBARS production. However, α-ketoglutarate induced per se a significant increase in basal TBARS production. The addition of potassium cyanide or the heat-treatment of S1 at 100ºC during 10 min completely abolished the antioxidant succinate activity, without change the effect of other Krebs cycle intermediates studied. Except for succinate, all intermediates used in this study were able to form complexes with iron (Fe2+) ions, however only oxaloacetate and α-ketoglutarate significantly prevented deoxyribose degradation induced by hydrogen peroxide. Based on the results presented, we concluded that oxaloacetate, malate, succinate, and citrate could act as antioxidants under basal, and under quinolinic acid- or iron- induced TBARS production, whereas α-ketoglutarate act as a pro-oxidant agent per se. The mechanism(s) by which citrate, malate, and oxaloacetate acts seems to be related to their ability to form complexes with iron (Fe2+) ions, thus modulating the iron redox cycle. In contrast, the succinate antioxidant effect seems to be dependent of the succinate dehydrogenase (SDH) activity. / Dados recentes na literatura têm relatado que alguns intermediários do ciclo de Krebs podem agir como potentes antioxidantes, tanto in vitro, quanto in vivo, em diversos sistemas pró-oxidantes. Porém, o(s) mecanismo(s) através dos qual(is) os intermediários do ciclo de Krebs exercem suas atividades antioxidantes não são completamente entendidas. Considerando a escassez de dados in vitro na literatura a respeito do efeito desses intermediários durante situações de estresse oxidativo, o presente trabalho tem como objetivo determinar o efeito de intermediários do ciclo de Krebs sob a peroxidação lipídica induzida por diferentes agentes pró-oxidantes in vitro, bem como investigar o(s) mecanismo(s) de ação dos mesmos. Primeiramente investigamos o efeito e o(s) mecanismo(s) pelo(s) qual(is) o malonato e o ácido quinolínico modulam a produção de espécies reativas ao ácido tiobarbitúrico (TBARS) em S1 de cérebro de ratos, in vitro (artigo 1). Os resultados obtidos mostraram um aumento na produção de TBARS induzido pelo malonato, o qual não foi modificado pela adição de cianeto de potássio, nem pelo MK-801. Por outro lado, o efeito pró-oxidante do ácido quinolínico foi significativamente prevenido pelo MK-801. Observamos ainda que o malonato foi capaz de formar complexos com íons ferrosos e que esses complexos não foram capazes de interferir nos ensaios da degradação da desoxirribose in vitro. Portanto, com base nos resultados encontrados, concluímos que o efeito pró-oxidante do malonato in vitro parece ser independente da atividade dos receptores NMDA. Os resultados sugerem que o efeito do malonato nessas condições deve-se principalmente a sua capacidade de interagir com íons ferro, modulando uma razão Fe2+/Fe3+ que favorece a geração de radicais livres. Por outro lado, o efeito do ácido quinolínico parece ser devido à ativação dos receptores NMDA. Porém, não podemos excluir a participação dos íons ferro para a toxicidade do mesmo nessas condições. Outro foco deste estudo foi investigar o efeito de alguns intermediários do ciclo de Krebs na produção de TBARS induzida por ácido quinolínico ou ferro em S1 de cérebro de ratos in vitro, bem como investigar o(s) mecanismo(s) de ação dos mesmos (artigo 2). Os resultados mostraram que o oxaloacetato, o citrato, o sucinato e o malato foram capazes de reduzir significativamente a produção de TBARS basal, bem como a induzida por ácido quinolínico ou ferro. Por outro lado, o α-cetoglutarato foi capaz de induzir per se um significativo aumento na produção de TBARS. A adição de cianeto de potássio, bem como o pré-tratamento do S1 por 10 min a 100ºC aboliram completamente o efeito antioxidante do sucinato, sem interferir significativamente no efeito dos demais intermediários estudados. Todos os intermediários estudados, exceto o sucinato, foram capazes de quelar íons ferro, porém somente o oxaloacetato e o α-cetoglutarato foram capazes de prevenir a degradação da desoxirribose induzida por peróxido de hidrogênio. Com base nos resultados obtidos, podemos concluir que o oxaloacetato, o malato o sucinato e o citrato agem como antioxidantes sob condições basais ou em presença do ácido quinolínico ou ferro, enquanto que o α-cetoglutarato age como um agente pró-oxidante per se. O mecanismo pelo qual o citrato, o malato e o oxaloacetato exercem seus efeitos antioxidantes parece ser devido à capacidade desses em interagir com íons ferro modulando o ciclo redox desse. Por outro lado, o efeito do sucinato parece ser devido à atividade da enzima succinato desidrogenase (SDH).

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Previous data from the literature have shown that some Krebs cycle intermediates could act as antioxidant in several models, both in vitro and in vivo. However, the mechanism(s) involved in the antioxidant effect of Krebs cycle intermediates are not fully understood. Additionally, there are scarce data in the literature taking into account the in vitro effect of Krebs cycle intermediates during oxidative stress conditions. Thus, the aim of this study was to determine the effect of some Krebs cycle intermediates on lipid peroxidation induced in vitro by different pro-oxidant agents, and the mechanism(s) by which they act. Furthermore, it was necessary elucidate the mechanisms by which the different pro-oxidants acts under in vitro conditions. The present results showed that the malonate-induced TBARS production was not changed by potassium cyanide or MK-801. However, the pro-oxidant effect of quinolinic acid was significantly prevented by MK-801. In addition we found that both malonate and oxalate were able to form complexes with iron ions (Fe2+). Based on the presented results, we conclude that malonate pro-oxidant activity in vitro seems to be independent of the secondary excitotoxicity via indirect NMDA receptors activation. Additionally, we suggest that both the malonate and oxalate effect, in these experimental conditions, is due to its ability to form complexes with iron ions, thus modulating an adequate ratio Fe2+/Fe3+ that could cause an increase in free radicals generation. In contrast, the quinolinic acid effect seems to be dependent of the NMDA receptors activation. However, we can not rule out the involvement of iron ions in quinolinic acid toxicity under our assay conditions. Another objective of this study was to investigate the effect of some Krebs cycle intermediates against either basal or induced TBARS production, using rat brain S1 preparations and the mechanism(s) by which they act. The results showed that oxaloacetate, citrate, succinate, and malate were able to significantly prevent both basal and quinolinic acid-, iron- or malonate-induced TBARS production. On the other hand, fumarate prevented only malonate-induced TBARS production, without effect under basal conditions. However, α-ketoglutarate induced per se a significant increase in basal TBARS production. The antioxidant activity of fumarate and succinate were completely abolished when S1 was submitted to heat-treatment at 100ºC during 10 min. Likewise, potassium cyanide completely abolished the antioxidant effect of succinate. The effect of other Krebs cycle intermediates studied was unchanged with respect to heat-treatment, or cyanide. Except for succinate and fumarate, all intermediates used in this study were able to form complexes with iron (Fe2+) ions, however only oxaloacetate and α-ketoglutarate significantly prevented deoxyribose degradation induced by hydrogen peroxide. Based on the results presented, we concluded that oxaloacetate, malate, succinate, fumarate and citrate could act as antioxidants under such conditions, whereas α-ketoglutarate acts as a pro-oxidant agent per se. The mechanism(s) by which citrate, malate, and oxaloacetate acts seems to be related to their ability to form complexes with iron (Fe2+) ions, thus modulating the iron redox cycle. In contrast, the succinate and fumarate antioxidant effect seems to be dependent of the some enzymatic system. / Dados prévios da literatura têm mostrado que alguns intermediários do ciclo de Krebs podem agir como antioxidantes em diversos modelos, tanto in vitro, quanto in vivo. Porém, o(s) mecanismo(s) através dos qual(is) esses intermediários exercem suas atividades antioxidantes não são completamente entendidas. Considerando a escassez de dados na literatura a respeito do efeito dos intermediários do ciclo de Krebs durante situações de estresse oxidativo, o presente trabalho teve por objetivo determinar o efeito desses sob a peroxidação lipídica induzida por diferentes agentes pró-oxidantes in vitro, bem como investigar o(s) mecanismo(s) de ação dos mesmos. Além disso, faz-se necessário caracterizar o(s) mecanismos(s) pelo(s) qual(is) os diferentes pró-oxidantes agem nos sistemas in vitro. Os resultados dessa tese mostraram que a atividade pró-oxidante in vitro do malonato não foi modificada pela adição de cianeto de potássio, nem pelo MK-801. Por outro lado, o efeito pró-oxidante do ácido quinolínico foi significativamente prevenido pelo MK-801. Observamos ainda que o malonato, e também o oxalato foram capazes de formar complexos com íons ferrosos. Portanto, com base nos resultados encontrados, concluímos que o efeito pró-oxidante do malonato in vitro parece ser independente da excitotoxicidade secundária, conseqüência da ativação indireta dos receptores NMDA. Os resultados sugerem que o efeito do malonato e do oxalato nessas condições experimentais deve-se principalmente a sua capacidade de interagir com íons ferro, modulando uma razão Fe2+/Fe3+ que favorece a geração de radicais livres. Por outro lado, o efeito do ácido quinolínico parece ser devido à ativação dos receptores NMDA. Porém, não podemos excluir a participação dos íons ferro para a toxicidade do mesmo nessas condições. Outro foco deste estudo foi investigar o efeito de alguns intermediários do ciclo de Krebs na produção de TBARS basal ou induzida por diferentes pró-oxidantes em S1 de cérebro de ratos in vitro, bem como investigar o(s) mecanismo(s) de ação dos mesmos. Os resultados mostraram que o oxaloacetato, o citrato, o sucinato e o malato foram capazes de reduzir significativamente a produção de TBARS basal, bem como a induzida por ácido quinolínico, ferro ou malonato. O fumarato, por sua vez, teve efeito antioxidante somente sobre a produção de TBARS induzida. Por outro lado, o α-cetoglutarato foi capaz de induzir per se um significativo aumento na produção de TBARS. O efeito antioxidante do fumarato e do sucinato foi completamente abolido quando o S1 foi submetido a um prétratamento por 10 min a 100ºC, enquanto que o efeito dos demais intermediários permaneceu inalterado. Da mesma forma, a adição de cianeto de potássio aboliu completamente o efeito antioxidante do sucinato sem interferir significativamente no efeito dos demais intermediários estudados. Todos os intermediários estudados, exceto o sucinato e o fumarato, foram capazes de quelar íons ferro, porém somente o oxaloacetato e o α- cetoglutarato foram capazes de prevenir a degradação da desoxirribose induzida por peróxido de hidrogênio. Com base nos resultados obtidos, podemos concluir que o oxaloacetato, o malato, o sucinato, o fumarato e o citrato agem como antioxidantes sob determinadas condições, enquanto que o α-cetoglutarato age como um agente pró-oxidante per se. O mecanismo pelo qual o citrato, o malato e o oxaloacetato exercem seus efeitos antioxidantes parece ser devido à capacidade desses em interagir com íons ferro modulando o ciclo redox desse. Por outro lado, o efeito do sucinato e do fumarato parece ser devido a alguma atividade enzimática.