Tesi sul tema "Mitofusins"

Le projet de doctorat porte sur l'étude structurale des mitofusines (Mfn1/2 chez l'homme et Fzo1 chez la levure) en utilisant principalement des méthodes basées sur la modélisation telles que la dynamique moléculaire ou les méthodes de prédiction de structure basées sur l'intelligence artificielle (principalement AlphaFold). Les mitochondries forment un réseau complexe à l'intérieur des cellules, subissant des événements continus de fusion et de fission. Ces processus façonnent la dynamique mitochondriale et sont essentiels pour l'entretien, la fonction, la distribution et l'héritage des mitochondries. La morphologie de ces dernières répond donc aux changements physiologiques constants de la cellule. Les larges GTPase impliquées dans l'ancrage et la fusion des membranes externes de mitochondrie sont des protéines transmembranaires appelées mitofusines. Les mitofusines Mfn1 et Mfn2 se trouvent chez les mammifères. Fzo1 (Fuzzy Onion 1) est l'homologue unique de Mfn1/2 chez la levure Saccharomyces cerevisiae. La fusion de la membrane interne mitochondriale et l'organisation des crêtes sont médiées par l'OPA1 humaine (Atrophie Optique 1) et la Mgm1 de la levure (Maintenance du Génome Mitochondrial 1). La dysfonction de la fusion mitochondriale est liée à plusieurs troubles neurodégénératifs, tels que Parkinson, Alzheimer et la maladie de Huntington. En effet, il a été montré que les mutations dans Mfn2 induisent le développement et la progression de dystrophies musculaires, telles que la maladie de Charcot-Marie-Tooth de type 2A, la forme la plus courante de la maladie CMT axonale. Le mécanisme exact par lequel les mitofusines contribuent à la dysfonction mitochondriale, ainsi que le mécanisme moléculaire exact de la fusion, ne sont pas encore entièrement compris. Dans l'ensemble, la fusion mitochondriale joue un rôle important dans la CMT2A, il est donc d'une importance capitale de comprendre pleinement le processus au niveau moléculaire. Les structures de Mfn1 et Mfn2 ont étés partiellement résolue, le domaine transmembranaire étant exclu, mais aucune structure résolue n'est disponible pour Fzo1. Fzo1 est intégré à membrane externe de mitochondrie avec ses deux domaines transmembranaires, exposant les parties N- et C-terminales vers le cytosol et une boucle vers l'espace intermembranaire. Du côté N-terminal, on trouve deux domaines de répétitions en heptad (HRs), HRN (présent uniquement chez la levure) et HR1, flanquant un domaine GTPase. Un troisième domaine HR, HR2, se trouve dans la partie C-terminale. Certains modèles de Fzo1 ont été construits avec comme template la protéine bactérienne de type dynamin-like (BDLP). BDLP est impliquée dans le remodelage des membranes et existe sous deux états conformationnels, une version compacte fermée qui passe à une structure étendue ouverte lors de la liaison au GTP, sur laquelle les modèles construits étaient basés. L'objectif du doctorat est de mettre à jour le modèle de Fzo1 construit en 2017, en travaillant dans un premier temps le domaine transmembranaire à l'aide de dynamiques moléculaires à plusieurs échelles. Un autre projet a consisté à étudier l'hélice amphipathique du domaine HR1 de Mfn1 (MfnA-AH), à tester ses capacités de liaison à la membrane. Initialement, nous avons utilisé des simulations gros grains, établissant ainsi une base solide pour évaluer la capacité prédictive de la famille de champs de force MARTINI. En utilisant d'autres simulations réalisées avec la pénétratine, nous avons pu fournir une analyse comparative des interactions AH-membranes dans les champs de force MARTINI. Mfn1-AH a ensuite été caractérisé plus en détail à l'aide de simulations tout-atomiques
The Phd project is the structural study of mitofusins (Mfn1/2 in humans and Fzo1 in yeasts) using mainly modeling-based methods such as molecular dynamics or structure prediction methods based on artificial intelligence (mainly AlphaFold). This project is a part of an ANR (MITOFUSION) shared between different partners (Laboratoire de Biochimie Théorique: Antoine Taly, Marc Baaden, Laboratoire des Biomolécules: Patrick Fuchs, Laboratoire de Biologie Moléculaire et Cellulaire des Eucaryotes: Mickaël Cohen, Institut de Psychiatrie et Neurosciences de Paris: David Tareste) whose goal is to understand the structure-function relationships of the mitofusin. Mitochondria form a complex network inside the cells, undergoing continuous fusion and fission events. These processes shape mitochondrial dynamics and are essential for the maintenance, function, distribution and inheritance of mitochondria. The morphology of the latter therefore respond to the ever-changing physiological changes of the cell. The large GTPase involved in the tethering and fusion of the mitochondrial outer membranes (OM) are transmembrane proteins called mitofusins. The mitofusins Mfn1 and Mfn2 can be found in mammals. Fzo1 (Fuzzy Onion 1) is the unique mitofusin homologue in Saccharomyces cerevisiae. The mitochondrial inner membrane fusion and cristea organisation is mediated by human OPA1 (Optic Atrophy 1) and yeast Mgm1 (Mitochondrial Genome Maintenance 1). Mitochondrial fusion dysfonction is related to several neurodegenerative disorders, such as Parkinson, Alzheimer and Huntingtion diseases. As a matter of fact, research has shown that mutations in Mfn2 induce the development and progression of muscular dystrophies, such as Charcot-Marie-Tooth Type 2A, the most common form of axonal CMT disease. The exact mechanism by which the mitofusins contributes to mitochondria dysfunction as well as the exact molecular fusion mechanism is not fully understood yet. Overall, mitochondrial fusion plays an important role in CMT2A, it is thus of paramount importance to get a full understanding of the process at the molecular level. The structure of both Mfn1 and Mfn2 was partially solved, the transmembrane domain being excluded, and no solved structure are available for Fzo1. With our ANR partners, we decided to work on the yeast version of Mitofusin (named Fzo1) as it is a good model (of homology with human Mfn1 and Mfn2) as yeast are convenient hosts for testing how other protein partners are involved in the process (e.g. Ugo1). Fzo1 is embedded in the mitochondrial OM as it possesses two transmembrane domains, exposing N- and C- terminal portions towards the cytosol and a loop towards the intermembrane space. On the N-terminal side can be found two coiled-coil heptad repeats (HRs) domains, HRN (in yeast only) and HR1, flanking a GTPase domain. A third coiled-coil heptad repeats domain HR2 is on the C-terminal portion. Some models of Fzo1 were built based on the mitofusin related bacterial dynamin-like protein (BDLP). BDLP is involved in membrane remodelling and exists in two conformational states, a closed compact version which changes to an opened extended structure, upon GTP-binding, on which the built models were based. The goal of the PhD is to update the model of Fzo1 built in 2017, by working on the transmembrane domains using multiscale molecular dynamics, and then update the overall structure using artificial intelligence methods. An other project consisted in studying the amphipathic helix of HR1 domain of Mfn1 (MfnA-AH), test its membrane binding capabilities. Initially, we employed coarse-grained simulations, establishing a robust foundation for evaluating the predictive capacity of the MARTINI family of force fields. Using other simulations ran with the penetratin, we were able to provide a comparative analysis for the AH-membranes interactions in the MARTINI force-fields. The Mfn1-AH was then further characterized using all-atom simulations

Les mitochondries sont des organites dynamiques qui fusionnent et se divisent continuellement. Cette dynamique est requise pour la biogenèse mitochondriale, la fonction et la dégradation. Les relations entre les OXPHOS, la dynamique et les mécanismes assurant la modulation de la dynamique restent largement inconnus. Nous avons étudié grâce à un essai de fusion in vivo, les relations entre la fusion et les OxPhos dans des cellules de levure portant des mutations ponctuelles dans le gène mitochondrial ATP6 qui sont associés à des maladies chez l’homme. Nous montrons que les défauts des OxPhos provoquent des défauts de fusion de la membrane interne mitochondriale mais pas de la membrane externe. L'inhibition sélective de la fusion de la membrane interne peut être mimée par les ionophores qui dissipent le potentiel de membrane interne, mais pas par des inhibiteurs des phosphorylations oxydatives. Nous montrons une inhibition dominante de fusion qui pourrait être un mécanisme d'exclusion des mitochondries dysfonctionnelles du réseau mitochondrial pour les adresser à la mitophagie. Ces résultats indiquent que les défauts de fusion pourraient contribuer à la pathologie des maladies provoquées par des mutations de l'ADNmt. De plus ces résultats impliquent que dans des cellules, l'inhibition de la fusion dominante pourrait permettre l'exclusion des mitochondries dysfonctionnelles du réseau mitochondrial. La fusion mitochondriale implique de nombreuses protéines de la superfamille des dynamines. Si ces protéines ont été identifiées, les mécanismes moléculaires permettant la fusion restent indéterminés. Dans le but de comprendre ces mécanismes, nous avons choisi de caractériser les protéines Mitofusine 1 et 2, essentielles à la fusion des membranes externes mitochondriales. Ces protéines sont composées de deux domaines coiled-coil et un domaine N-terminal GTPase et des domaines hydrophobes prédits pour être des segments transmembranaires. Après la détermination des activités GTPase des mitofusines, nous avons reconstitué les mitofusines ou des fragments des mitofusines dans des liposomes afin d'étudier leur capacité à fusionner ces liposomes. Les mitofusines, permettent la fusion des liposomes contenant des cardiolipides. Étonnamment, ces événements sont indépendants de la présence du GTP mais nécessitent Mg2+ dans la solution. En utilisant la microscopie électronique, nous montrons que les mitofusines 1 et 2 induisent une déformation des liposomes. Cette capacité permettant de créer localement des régions très courbes (et donc fusogènes) ouvre un nouvel angle pour comprendre les mécanismes moléculaires de la fusion mitochondriale
Mitochondria are dynamic organelles that continuously fuse and divide. This dynamic is required for mitochondrial biogenesis, function and degradation. The cross-talk between OXPHOS and dynamics and the mechanisms ensuring modulation of dynamics remain largely unraveled. We have investigated the relationship between fusion and OXPHOS in yeast cells carrying point mutations in the mitochondrial ATP6 gene that are associated to human diseases. We show that OXPHOS defects provoke severe defects of inner membrane, but not outer membrane fusion. Selective inhibition of inner membrane fusion can be recapitulated by ionophores that dissipate the inner membrane potential, but not by inhibitors of OXPHOS. We show a dominant inhibition of fusion that further provides a mechanism for the exclusion of defective mitochondria from the functional mitochondrial network, a pre-requisite for their selective targeting to mitophagy. These results suggest that defects of fusion could contribute to the pathology of diseases caused by mtDNA mutations. Moreover, these results imply that in cells, inhibition of dominant fusion could allow the exclusion of dysfunctional mitochondria mitochondrial network. Mitochondrial fusion involves many proteins of the superfamily of dynamin. If these proteins have been identified, the molecular mechanisms of fusion remain undetermined. In order to understand these mechanisms, we choose to characterize Mitofusin 1 and 2 proteins, essential for outer mitochondrial membrane fusion. These transmembrane proteins are consisting of two coiled-coil domains and one N-terminal GTPase domain. We have characterized GTPase activity of Mitofusin and reconstituted Mitofusins or fragments of Mitofusins into liposomes to study their capacity to fuse these liposomes. Full-length mitofusins can fuse liposomes containing cardiolipins. Surprisingly, these fusion events are independent of GTP but require Mg2+ in the buffer. Using electron microscopy, we show that mitofusin 1 and 2 induce local deformation of liposomes. This capacity of mitofusins to locally create highly curved (and thus fusogenic) membrane regions opens a new angle to understand the molecular mechanisms of mitochondrial fusion

Les mitochondries sont des organelles très dynamiques qui subissent des phénomènes de fission et de fusion constants de leurs membranes extérieures et intérieures. Ces processus sont essentiels pour le maintien des fonctions mitochondriales essentielles telles que la phosphorylation oxydative ou la signalisation du calcium. D’un point de vue moléculaire, la fusion et la fission mitochondriale dépendent tous les deux des grandes GTPases de la famille des protéines de type dynamine. Les dynamines qui favorisent l’attachement et la fusion des membranes mitochondriales extérieures sont appelés les mitofusines.La mitofusine de la levure Fzo1 est une GTPase transmembranaire située dans la membrane externe de la mitochondrie. Son oligomérisation favorise l’attachement suivi de la fusion de la membrane externe mitochondriale. Fzo1 a été proposé récemment comme une protéine d’attachement potentielle entre les peroxysomes et les mitochondries lorsqu’elle est surexprimée. Cependant, on ignore si Fzo1 est présent sur les membranes peroxysomales dans les cellules sauvages ou si cette localisation extra-mitochondriale est une conséquence de la surexpression. De plus, nous ne savons toujours pas comment le Fzo1 peroxysomal et le Fzo1 mitochondrial interagissent dans ces contacts et quel est leur rôle dans la cellule. Durant ma thèse, j’ai pu prouver que Fzo1 se trouve réellement aux peroxysomes dans des conditions physiologiques et oligomérise avec le Fzo1 mitochondrial créant ainsi des contacts Fzo1-Fzo1 entre les peroxysomes et les mitochondries que nous appellerons maintenant des contacts « Permit Fzo1-dépendants ». On a découvert que ces contacts sont modulés par les niveaux de Fzo1 qui sont étroitement régulés par la ligase ubiquitine appelée Mdm30 mais aussi en fonction des niveaux de désaturation des acides gras dans la cellule. D’un point de vue fonctionnel et après avoir écarté plusieurs possibilités, nous avons trouvé que le rôle des contacts Permit Fzo1-dépendants est de réguler la fusion mitochondriale à travers le cycle glyoxylate, un processus qui permet aux cellules de convertir des composés unitaires de C2 en précurseurs de C4 pour la biosynthèse des acides aminés et des glucides. Nous avons découvert que les contacts Permit Fzo1-dépendants permettent le transfert mitochondrial des produits intermédiaires du cycle de glyoxylate pour stimuler la fusion mitochondriale. Ces résultats révèlent ainsi une réponse des organelles aux changements de désaturation des acides gras et aux besoins métaboliques de la cellule pour réguler la fusion mitochondriale.Enfin, les résultats obtenus au cours de ma thèse ont enrichi nos connaissances sur les contacts entre organelles et nous ont permis de prouver que Fzo1 est localisé sur les membranes mitochondriales et peroxysomales dans les cellules de type sauvage de levure. Nos études montrent également que les contacts Permit Fzo1-dépendants sont modulés en fonction des besoins de la cellule car ils jouent un rôle crucial dans l’entretien de la fusion mitochondriale en créant un raccourci possible pour les produits intermédiaires du cycle du glyoxylate pour atteindre les mitochondries lorsque cela est nécessaire
Mitochondria are highly dynamic organelles that undergo constant fission and fusion of their outer and inner membranes. These processes are critical to maintain essential mitochondrial functions such as oxidative phosphorylation or calcium signaling. On a molecular basis, mitochondrial fusion and fission both depend on large GTPases of the Dynamin-Related Protein (DRP) family. The DRPs that mediate attachment and fusion of mitochondrial outer membranes are called the Mitofusins. The yeast mitofusin Fzo1 is located in the mitochondrial outer membrane. Its oligomerization promotes mitochondrial tethering followed by mitochondrial outer membrane fusion. Fzo1 has recently been proposed as a potential tether between peroxisomes and mitochondria when overexpressed. However, whether Fzo1 is present on peroxisomal membranes in WT cells or whether this extra-mitochondrial localization is a consequence of overexpression is unknown. In addition, we still don’t know how peroxisomal and mitochondrial Fzo1 mediate these contacts and their purpose in the cell. In my thesis, we were able to prove that Fzo1 naturally localizes to peroxisomes and oligomerizes with the mitochondrial Fzo1 thus creating Fzo1-Fzo1 contacts between peroxisomes and mitochondria which we will now call “Fzo1-mediated permit” contacts. We found that these contacts are modulated by Fzo1 levels which are tightly regulated by an SCF ubiquitin ligase called Mdm30 but also depending on fatty acid desaturation levels in the cell. From a functional standpoint, we found that the role of Fzo1-mediated permit contacts is to regulate mitochondrial fusion through the glyoxylate cycle, a process which allows cells to convert C2 unit compounds to C4 precursors for amino acid and carbohydrate biosynthesis. We discovered that Fzo1-mediated permit contacts allow the mitochondrial transfer of early byproducts of the glyoxylate cycle to stimulate mitochondrial fusion. In fine, the results obtained during my thesis enriched our knowledge on organelle contacts and allowed us to prove that Fzo1 is localized on both mitochondrial and peroxisomal membranes in wild type cells. Our studies also show that Fzo1-mediated permit contacts are modulated according to the cell’s needs as they play a crucial role in upkeeping mitochondrial fusion by providing a possible shortcut for byproducts of the glyoxylate cycle to reach mitochondria when direly needed

Mitofusin 1 and 2 are outer-mitochondrial membrane proteins that have been shown to be involved in fusion. Mitofusin 2 has also been associated with apoptosis and development. When Mfn1 and Mfn2 were each conditionally knocked out from the cerebellum, Purkinje cells in Mfn2 deficient cerebellum during development had undergone neurodegeneration. Mutations in Mfn2 have also been associated with the Charcot Marie Tooth Type 2A (CMT2A). We want to asses the effect Mfn2 and Mfn1 might have on the development of other regions of the brain such as the telencephalon. We generated Mfn1 and Mfn2 conditional knockouts in the telencephalon by crossing them with Foxg1 Cre - a cre expressed in the telencephalon. We found that Mfn1 deficient mice have lost their corpus callosum at the midline, but survive over 6 months with a decrease in progenitor cells postnatally. Mfn2 deficient mice die between P9 and P12 with a decrease in progenitor cells postnatally and a decrease in number of neurons in the cortex. Therefore, our results suggest that Mfn1 and Mfn2 play a significant role in the development of the telencephalon.

Les mécanismes moléculaires impliqués dans la fusion membranaire ont été amplement étudiés au cours des trente dernières années. Notre compréhension actuelle de ce phénomène est principalement basée sur des résultats obtenus par (1) le développement de modèles physiques décrivant la fusion des membranes biologiques, (2) l’étude mécanistique et structurale des protéines de fusion membranaire des virus à enveloppe et (3) l’étude des évènements de fusion intracellulaire médiés par les protéines SNARES dans les cellules eucaryotes. La découverte du complexe SNARE fut l’aboutissement de travaux interdisciplinaires qui ont exigés un large éventail de techniques tel que la génétique de la levure, l’électrophysiologie, la biologie moléculaire, la biochimie cellulaire, la biophysique expérimentale et l’imagerie. Tirant parti des paradigmes et techniques biophysiques qui ont émergés de ces études, nous avons examiné les fonctions et mécanismes de régulation des domaines « coiled-coil » dans les processus de fusion intracellulaire impliquant des protéines de la famille des Longin-SNAREs ou des Mitofusines, deux machineries protéiques de fusion dont le mode d’action exact reste encore peu clair. La conception exacte des mécanismes moléculaires de la fusion membranaire requiert la reconstitution in vitro des protéines de fusion dans un large spectre d’environnement membranaire avec des propriétés biophysiques définies et facilement modulables. Idéalement, ces systèmes membranaires devraient permettre à l’expérimentateur de contrôler la composition lipidique et protéique, ainsi que la topologie membranaire, afin de rendre compte de l’importante variabilité observée entre les différents compartiments de fusion cellulaire. La reconstitution dans des liposomes offre une incroyable flexibilité avec la possibilité de faire varier la plupart des paramètres clefs et de créer un environnement minimal dans lequel les facteurs solubles et/ou membranaires peuvent être ajoutés, seuls ou en combinaison, pour dévoiler leur rôle avec clarté. Nous avons mis au point des systèmes in vitro de reconstitution de protéines dans des plateformes membranaires artificielles pour nos deux systèmes d’études (les deux protéines Longin-SNAREs TI-VAMP et Sec 22b, ainsi que les domaines « coiled-coil » des Mitofusines) et nous avons réalisé des expériences biochimiques pour caractériser le mode d’action de ces protéines. L’objectif à long-terme de ce projet est de comparer les mécanismes moléculaires des machineries de fusion associés aux protéines SNAREs et Mitofusines, et ainsi de dévoiler des similitudes structurelles et fonctionnelles entre (1) leur protéines de fusion principales et (2) leur facteurs régulateurs
The molecular mechanisms involved in membrane fusion have been extensively studied for the past thirty years. Our current understanding of this phenomenon is mainly based on results obtained by (i) the development of physical models describing the fusion of membranes, (ii) structural and mechanistic investigations on fusion proteins of enveloped viruses and (iii) studies of SNARE protein-mediated intracellular fusion events of eukaryotic cells. Discovery of the SNARE complex was the outcome of interdisciplinary works which involved a wide range of techniques including yeast genetics, electrophysiology, molecular biology, cell-free biochemistry, adhesion/fusion biophysics and imaging. Taking advantage of the paradigms and biophysical techniques that emerged from these studies, we investigated the function and regulation of coiled-coil domains in intracellular fusion processes involving Longin-SNAREs or Mitofusins, two fusion protein machineries whose exact mode of action still remains unclear. A comprehensive understanding of the molecular mechanisms of membrane fusion requires the in vitro reconstitution of fusion proteins into a wide variety of membrane environments with defined and tunable biophysical properties. Ideally, these membrane systems should allow the experimentalists to control the lipid and protein composition as well as the membrane topology, to account for the variability observed across cellular fusing compartments. Reconstitution into liposomes offers amazing flexibility with the capacity to vary most of these relevant parameters, and to create a minimal environment in which membrane and/or soluble factors can be added, one at a time or in combination, to reveal their role with clarity. We have set up the in vitro reconstitution of proteins into various artificial membrane platforms for both systems (the Longin-SNAREs TI-VAMP and Sec22b and the coiled-coil domains of Mitofusins) and performed biochemical assays to gain insight into how these proteins execute their functions. The long-term goal of this project is to compare the molecular mechanisms of SNARE and Mitofusin fusion machineries and thus reveal structural and functional similitudes between (i) their core fusion proteins, and (ii) their regulatory factors

A restrição calórica (RC) estende a expectativa de vida de muitos organismos por mecanismos ainda em estudo. Entre os vários efeitos fisiológicos da RC encontra-se o aumento na biogênese mitocondrial, dependente de óxido nítrico (NO•), sintetizado pela enzima óxido nítrico sintase endotelial (eNOS). Um dos indutores fisiológicos mais potentes da eNOS é a insulina, cujos níveis plasmáticos são consideravelmente reduzidos nos organismos em RC. O objetivo deste trabalho foi investigar os mecanismos associados ao aumento da sinalização por NO• durante a RC in vivo e in vitro, e as conseqüências celulares do aumento de massa mitocondrial no que diz respeito à longevidade e capacidade respiratória celulares. Submetemos camundongos Swiss fêmeas à RC de 40% e observamos um considerável aumento tecido-específico na fosforilação basal de Akt e eNOS em músculo esquelético, tecido adiposo visceral e cérebro, os quais também apresentaram maior massa mitocondrial. A associação entre a sinalização por insulina, NO• e biogênese mitocondrial foi adicionalmente confirmada em um grupo de camundongos tratados com o desacoplador mitocondrial dinitrofenol (DNP), que também reduz a insulinemia e aumenta a longevidade em camundongos. Para o estudo mecanístico deste fenômeno, usamos soros de ratos Sprague-Dawley submetidos à RC de 40% ou alimentados ad libitum (AL) em cultura celular de células vasculares da musculatura lisa (VSMC), reproduzindo um protocolo descrito para RC in vitro. O uso do soro RC aumentou a fosforilação do receptor de insulina e Akt, a expressão de eNOS e nNOS (forma neural da NOS) e a fosforilação de eNOS, o que se refletiu em maior liberação de nitrito (NO2) no meio de cultura. Inibindo-se a Akt, todos os efeitos promovidos pela RC na sinalização por NO• foram revertidos. Ao se imunoprecipitar do soro a adiponectina, citocina conhecida por aumentar a sensibilidade à insulina, aumentada durante a RC, os efeitos do soro RC na via de sinalização de insulina foram abolidos e, conseqüentemente, os efeitos na sinalização por •NO foram prevenidos. Neurônios de células granulosas de cerebelo, que não expressam eNOS, apenas nNOS, foram cultivados com os soros AL ou RC, e também apresentaram considerável aumento na sinalização por •NO. Estas alterações induziram a biogênese mitocondrial e capacidade respiratória, e foram associadas à maior longevidade celular. Os mesmos efeitos mitocondriais foram observados em células secretoras de insulina, INS1, entretanto a secreção de insulina em resposta à glicose tornou-se inibida, por um mecanismo desconhecido, porém associado a reduzidos níveis intracelulares de espécies oxidantes, moléculas-chave para a secreção de insulina; e à alteração da morfologia mitocondrial, provavelmente devido à maior expressão de mitofusina-2 (Mfn-2). Ao se nocautear a Mfn-2, houve um aumento na geração de EROs e as células em RC passaram a secretar insulina a níveis comparáveis aos das células controle. Concluímos que durante a RC a maior sensibilidade à insulina aumenta a atividade de eNOS, via Akt, associada à maior biogênese mitocondrial. A adiponectina é uma molécula-central nestes eventos. A expressão de nNOS também é afetada, por mecanismos desconhecidos. O aumento de biogênese mitocondrial eleva a capacidade respiratória celular e impacta positivamente a longevidade in vitro. A alteração da morfologia mitocondrial associa-se a alterações na produção de oxidantes intracelulares e mudanças na secreção de insulina.
Calorie restriction (RC) is known to extend the lifespan in many organisms, and its mechanisms of action are still under investigation. Enhanced mitochondrial biogenesis driven by nitric oxide (•NO), synthesized by the endothelial nitric oxide synthase (eNOS), is proposed to be a CR central effect. Insulin is one of the most potent physiological activators of eNOS. However, plasmatic insulin levels are dramatically reduced in organisms under CR. The goal of this work was uncover the mechanisms associated with enhanced •NO signaling during CR, in vivo and in vitro, as well as the cellular consequences of increased mitochondrial mass, regarding lifespan and reserve respiratory capability. Female Swiss mice were submitted to 40% of CR. A tissue-specific (skeletal muscle, abdominal adipose tissue and brain) increment in basal Akt and eNOS phosphorylation, which was related to enhanced mitochondrial biogenesis, was observed. Indeed, this association was also verified in tissues from mice treated with low doses of a mitochondrial uncoupler, dinitrophenol (DNP). To unveil the mechanism behind the insulin signaling effects on •NO levels, serum from Sprague-Dawley rats submmited to 40% of CR was used to culture in VSMC cells, an in vitro CR protocol. CR sera enhanced insulin receptor (IR) and Akt phosphorylation, as well as nitrite (NO2-) accumulation in the culture media, the expression of eNOS and nNOS (neural NOS isoform) and eNOS phosphorylation. The effects of CR sera were reversed by Akt inhibition. The immunoprecipitation of serum adiponectin, a cytokine known to improve peripheral insulin sensitivity, also reversed the CR serum effects on insulin and •NO signaling. Cerebellar neurons, which do not express eNOS, just nNOS, were also cultured with CR or AL serum and also presented striking increments in •NO signaling, associated with mitochondrial biogenesis, increased reserve respiratory capability and lifespan extension. The mitochondrial effects promoted by CR were also observed in insulin secreting cells (INS1). However, under the CR condition, insulin secretion stimulated by glucose was impaired. The likely explanations are reduced mitochondrial reactive oxygen species (ROS) generation, or the alteration in mitochondrial morphology, associated, in our model, with enhanced mitofusin-2 expression (Mfn-2). In cells which the Mfn-2 was knocked down, insulin secretion in CR and AL groups was responsive to glucose at the same level, and the intracellular oxidants levels were much higher. Overall, CR improves •NO signaling due to enhanced insulin sensitivity, through Akt, and results in mitochondrial biogenesis. Adiponectin is a key molecule in this phenomenon. Increments in mitochondrial mass enhance the cellular reserve respiratory capability and lifespan. Mitochondrial morphology alterations are associated with possible decreases in ROS generation and impaired insulin release, maintained the low levels of plasmatic insulin.

Les mitochondries sont des organites intracellulaires délimités par deux membranes. Remarquablement dynamiques, elles fusionnent et fissionnent en permanence. Au cours de ma thèse je me suis intéressée aux mécanismes de cette dynamique et à sa pertinence physiologique. De nouveaux tests de fusion nous ont permis de montrer que la fusion des membranes mitochondriales interne et externe est effectuée par deux machineries aux besoins énergétiques différents. Nous avons également montré que la protéolyse d’OPA1, facteur de la fusion de la membrane interne, est régulée par le potentiel de membrane et catalysée par une métalloprotéase intramitochondriale. Nous avons mis en évidence une interaction physique entre OPA1 et les Mitofusines, facteurs de la fusion de la membrane externe, dont nous avons cherché d’autres partenaires par crible double hybride. Par ailleurs, nous avons montré que les défauts d’oxydation phosphorylante n’influent que faiblement sur la morphologie mitochondriale. Enfin, nous avons observé le maintien de mitochondries filamenteuses contenant plusieurs nucléoïdes à tous les stades de la mitose.

ABSTRACT Mitochondria are the energy producing organelles in eukaryotic cells providing ATP through oxidative phosphorylation (OXPHOS). Mitochondria are highly dynamic and undergo fission, fusion and move into the cell along the microtubules to generate the mitochondrial network. Mitochondrial dynamics play a critical role in the control of organelle shape, size, number, function and quality control of mitochondria. It is regulated by several GTPases that play an important role in fusion and fission processes. In mammals, mitochondrial fusion is controlled by Mitofusin 1 (Mfn1), Mitofusin 2 (Mfn2) and Optic atrophy protein 1 (Opa1), while mitochondrial fission is regulated by Dynamin related protein 1 (Drp1). The aim of this study is to understand how mitochondrial distribution in neuronal cells is affected by mitochondria function and/or morphology. We use Drosophila melanogaster , whose genome contains homologs for all mitochondrial fusion and fission proteins, as a modelorganism to study how loss of fusion and fission protein modify the axonal distribution and motility of mitochondria. We demonstrate that loss of Marf (Mitochondrial associated regulatory factor, homologous to human mitofusins) or Opa1 causes an accumulation of mitochondria in the soma, a defect in the axonal distribution of mitochondria, a severe depletion of mitochondria in neuromuscular junctions (NMJs) and reduced mitochondrial motility. Simultaneous loss of Drp1 rescues the Opa1 phenotype very robustly while loss of Marf essentially does not. Viability data however show the opposite trend. The expression of Marf RNAi or Opa1 RNAi cause lethality, and so does the double down regulation of Opa1 and Drp1. Conversely individuals expressing Marf RNAi and Drp1 RNAi simultaneously survive and are comparable to the controls. We then examined possible alterations of mitochondrial function by analyzing the mitochondrial respiratory capacity, the activity of the respiratory chain complexes and ATP production capacity. The data show that individuals where Marf, Opa1 or simultaneously Opa1 and Drp1 are down-regulated display severe alterations in mitochondrial function, while there are no obvious energy defects in individuals in which the expression of Marf and Drp1 is simultaneously reduced. Collectively our results obtained suggest that mitochondrial morphology is important for a homogeneous distribution of mitochondria along the axon and their transport to synapses and that these mechanisms are independent of mitochondria function.
RIASSUNTO I mitocondri sono organelli essenziali per la cellula e la loro funzione primaria è di produrre energia sottoforma di ATP. I mitocondri sono organelli altamente dinamici:processi di fusione e fissione delle membrane mitocondriali ne controllano la forma, la lunghezza e il numero e un equilibrio tra i due meccanismi è fondamentale per una corretta morfologia mitocondriale. Numerose proteine sono coinvolte nei processi di fusione e fissione mitocondriale: Mitofusina 1 e Mitofusina 2 (Mfn1 e Mfn2) e Optic atrophy 1 (Opa1) regolano i processi di fusione mitocondriale, mentre Dynamin-related protein 1 (Drp1)mediala fissione. Drosophila possiede il gene mitochondrial assembly regulatory factor (MARF), espresso in modo ubiquitario ed omologo al gene MFN2. Nel tessuto muscolare la riduzione di espressione di Marf induce frammentazione e alterazione della morfologia del mitocondrio. Inoltre, mutanti di Marf mostrano una severa deplezione dei mitocondri nelle giunzioni neuromuscolari (NMJs) ed un’alterazione della morfologia della giunzione caratterizzata dall’aumento nel numero e da una riduzione nella dimensione dei bottoni sinaptici. Un altro aspetto della dinamica mitocondriale, oltre ai processi di fusione e fissione, è la motilità dei mitocondri, che deve essere altamente regolata soprattutto in cellule come i neuroni. Il trasporto mitocondriale e la continua ridistribuzione dei mitocondri lungo l’assone è essenziale per il mantenimento dell’integrità assonale e delle normali funzioni della cellula. Studi hanno messo in evidenza come la mancanza di mitocondri a livello delle giunzioni neuromuscolari in Drosophila comprometta la trasmissione sinaptica e come difetti nel trasporto mitocondriale assonale siano implicati nello sviluppo di disordini neurologici e malattie neurodegenerative (Chan, 2006). Lo scopo di questo lavoro è quello di capire il ruolo della morfologia e della funzione mitocondriale nella distribuzione intracellulare dei mitocondri nei neuroni. Per fare questo abbiamo utilizzato Drosophila melanogaster, organismo modello efficace per l’analisi della funzione genica, inclusa quella di geni responsabili di patologie umane. L’analisi della morfologia mitocondriale è stata effettuata utilizzando linee di Drosophilache esprimono in vivo un transgene per RNA interference e che permette di ridurre l’espressione di geni endogeni coinvolti nei processi di fusione e fissione mitocondriale, quali Marf, Opa1 e Drp1. Abbiamo inoltre creato linee che esprimono contemporaneamente i trangeni per RNAi di Marf e Drp1 o Opa1 e Drp1, con lo scopo di bilanciare i meccanismi di fusione e/o fissione. Ci siamo soffermati in particolare sullo studio di due aspetti principali, la morfologia e la funzionalità mitocondriale, per capire se difetti nella morfologia e nella funzionalità mitocondriale siano collegate e concorrano insieme allo sviluppo di patologie.Numerose patologie neurodegenerative sono infatti caratterizzate da alterazioni del trasporto mitocondriale e spesso questo è associato a difetti nella morfologia e nella funzionalità mitocondriale. Per studiare la morfologia mitocondriale, le linee UAS-RNAi sono state incrociate con una linea che contiene il promotore ELAV per l’espressione tessuto-specifica nei neuroni ed esprime una GFP mitocondriale. Abbiamo analizzato la morfologia dei mitocondri, sia nel corpo cellulare sia negli assoni e la distribuzione mitocondriale in assoni lunghi come i motoneuroni e assoni corti come quelli del nervo ottico e la distribuzione mitocondriale nella giunzione neuromuscolare.I risultati ottenuti mostrano che frammentazione dei mitocondri e alterazione della distribuzione mitocondriale assonale in individui in cui sia ridotta l’espressione di proteine di fusione. Inoltre si osserva una diminuzione della percentuale dei mitocondri mobili e del numero assoluto dei mitocondri anterogradi e retrogradi. Questi dati dimostrano che vi è una stretta correlazione tra morfologia mitocondriale e distribuzione dei mitocondri, in particolare in assoni lunghi. Inoltre analizzando le linee Marf RNAi Drp1 RNAi e Opa1 RNAi Drp1 RNAi, nelle quali gli eventi di fusione e fissione ridotti ma sono in equilibrio tra loro, si osserva un miglioramento la morfologia, la distribuzione e il trasporto mitocondriale assonale in modo particolare nel caso di Opa1 e non nel caso di Marf. Abbiamo cercato di capire quindi se in questi individui vi fossero alterazioni delle funzionalità mitocondriali attraverso l’analisi della capacità respiratoria mitocondriale, dell’attività dei complessi della catena respiratoria e della capacità di produzione di ATP. I risultati ottenuti dimostrano che morfologia e funzionalità mitocondriale non sempre sono collegate tra loro hanno effetti diversi nella modulazione della distribuzione mitocondriale assonale. In conclusione possiamo affermare che solamente la morfologia e la dimensione del mitocondrio sembrano essere essenziali per la corretta distribuzione mitocondriale assonale.

Mitochondria in all cell types undergo frequent fission and fusion events, and these dynamics determine the overall morphology of the organelle in cells. Two important GTPases have been recently identified that regulate mitochondrial membrane activity, a dynamin related protein (DRP1) required for fission, and the novel fusion GTPase, Mitofusin2. Mitofusin2 is an outer mitochondrial membrane protein and, like other GTPases involved in membrane fusion events, the N-terminal GTPase domain is exposed to the cytosol, such that it could interact with and recruit potential cytosolic proteins. The work documented in this thesis aims towards understanding the specific role of this unique GTPase in regulating mitochondrial fusion events, and to identify potential interacting proteins that work together with Mfn2 to carry out this complex biochemical event. Two independent approaches were taken to identify interacting proteins, both a yeast two-hybrid screen and affinity chromatography using recombinant bacterial expressed Mfn2 protein as bait. (Abstract shortened by UMI.)

Mitofusina 2 (Mfn2) es una proteína de fusión mitocondrial. Las evidencias experimentales demuestran una menor expresión génica de Mfn2 en músculo esquelético de sujetos obesos y pacientes diabéticos de tipo 2. Al inhibir la expresión de Mfn2 en células en cultivo se reduce el consumo de oxígeno, el potencial de membrana mitocondrial y la oxidación de glucosa, indicando un papel relevante de Mfn2 en la biología mitocondrial así como en la fisiopatología de la obesidad y/o la diabetes de tipo 2. Además, se han relacionado mutaciones en el gen de Mfn2 con la neuropatía de Charcot-Marie-Tooth de tipo 2 y se le ha descrito un papel antiproliferativo en desordenes vasculares. Consecuentemente, el estudio de la regulación de la expresión génica de Mfn2 es de notable interés.
En esta tesis, se ha clonado el promotor humano de Mfn2 y se ha determinado que carece de caja TATA y se localiza en una isla CpG, lo que lo hace susceptible de ser regulado por metilación. En músculo esquelético posee al menos 6 inicios de transcripción en una ventana de 171 pares de bases. Estudios con genes reporteros han determinado que Sp1 podría tener un papel relevante dirigiendo a la maquinaria basal de transcripción para formar el complejo de preiniciación. La expresión de Mfn2 es mayor en tejidos con requerimientos energéticos elevados. Ensayos con genes reporteros han permitido identificar factores de transcripción que podrían ser responsables de la expresión dependiente de tejido.
Se ha estudiado con detalle la inducción de Mfn2 en músculo esquelético y tejido adiposo marrón con la exposición al frío y tratamiento con agonistas de los receptores beta-3-adrenérgicos, condiciones ambas caracterizadas por un elevado gasto energético. El aumento de expresión de Mfn2 en estas condiciones es mediado por PGC-1-alfael cual coactiva a ERR-alfa que se une al promotor de Mfn2 en forma de monómero entre las bases -413 y -408. El incremento del potencial de membrana mitocondrial asociado a la expresión de PGC-1-alfa es inhibido por la represión de la expresión de Mfn2. Mfn2 podría ser un efector crucial en la activación mitocondrial inducida por PGC-1-alfa.
Se han aportado nuevas evidencias de la relevancia de Mfn2 en el control de la oxidación mitocondrial de substratos al mostrar una mayor expresión de Mfn2 en fibras musculares oxidativas que en fibras glucolíticas. La transcripción de Mfn2 aumenta en respuesta a incrementos en la concentración de calcio intracelular. Nuestros resultados sugieren que MEF2, el cual se une y activa al promotor de Mfn2, podría ser un efector de la cascada de señalización iniciada por el calcio citosólico.
Los datos obtenidos indican que la expresión de Mfn2 en músculo esquelético se incrementan en condiciones de elevado gasto energético lo cual estimula la oxidación de substratos y provoca incrementos en el potencial de membrana mitocondrial. Dependiendo de la demanda celular de ATP, el potencial de membrana mitocondrial se disipa para producir ATP o generar calor.

ENGLISH
Mitofusin 2 (Mfn2) is a mitochondrial fusion protein. Mfn2 have been related in several diseases such as neuropathy of Charcot-Marie-Tooth type 2A, vascular proliferative disorders and type 2 diabetes and obesity. Consequently, the mechanisms that regulate Mfn2 gene expression are of relevance.
The human Mfn2 promoter was cloned, showing it is located in a CpG island and lacks TATA box. That makes Mfn2 promoter susceptible to be regulated by methylation. In skeletal muscle transcription is initiated in at least 6 points in a 171 bp window. Sp1 may direct the basal machinery to form a preinitiation complex in human Mfn2 promoter.
Mfn2 is highly expressed in tissues with elevated energy demand. Gene reporter assays allowed to identify some transcription factors that may mediate tissue specific expression.
Mfn2 was induced with cold and -3-adrenergic receptor agonist exposure, conditions associated with enhanced energy expenditure. The cold induced coactivator PGC-1regulated Mfn2 expression requiring the integrity of an estrogen-related receptor alpha (ERR-alpha-binding element located at -413/-408. ERR-alpha also activated the transcriptional activity of the Mfn2 promoter and the effects were synergic with those of PGC-1-alphaRepression of Mfn2 reduces oxygen consumption, glucose oxidation and mitochondrial membrane potential. Using knock down strategy was showed that Mfn2 cold play a crucial role in mitochondrial energization by PGC-1-alpha
Mfn2 showed higher expression in slow twitch oxidative muscle fibres than in fast twitch glucolitic fibres. This is a new evidence of Mfn2 as regulator of substrate oxidation. We proved that Mfn2 gene expression was regulated by calcium. High intracellular calcium levels in oxidative muscle fibres could be the initiator of the cascade that results in higher Mfn2 expression. One of the downstream effectors of calcium is MEF2 that binds and activates Mfn2 promoter.
Data provided indicated that Mfn2 expression was increased in skeletal muscle when energy demand is high. The increases in Mfn2 expression supposed a higher mitochondrial membrane potential, this mitochondrial membrane potential could be dissipated coupled to ATP synthesis (oxidative muscle) or uncoupled to produce heat (cold exposure) depending the energetic requirements of the cell in each case.

Els mitocondris són orgànuls citoplasmàtics que tenen un paper fonamental en múltiples processos biològics com l’oxidació de substrats i la producció d’ATP, la senyalització cel•lular, l’apoptosi, el control del cicle cel•lular i l’homeòstasi del calci. Els mitocondris són orgànuls dinàmics, que pateixen canvis de morfologia regulats per processos de fusió i de fissió. Existeix un equilibri entre ambdós processos que és indispensable per a la correcta funció mitocondrial. Les proteïnes que participen directament en la fusió mitocondrial en mamífers són les mitofusines (Mfn1 i Mfn2), localitzades a la membrana mitocondrial externa i OPA1, situada a la membrana mitocondrial interna. Diferents estudis han demostrat que la proteïna Mfn2, a més de promoure la fusió dels mitocondris, també està implicada en la interacció entre els mitocondris i el reticle endoplasmàtic i que participa en la regulació del cicle cel•lular i del metabolisme mitocondrial. Per altra banda, l’expressió de Mfn2 es troba disminuïda en múscul esquelètic en situacions de resistència a la insulina, com l’obesitat o la diabetis de tipus 2, que a la vegada es caracteritzen per una alterada activitat mitocondrial. En base a aquestes observacions, l’objectiu principal de la present tesi doctoral ha estat estudiar els efectes de la modulació de l’expressió de Mfn2 sobre el metabolisme i la bioenergètica mitocondrial en múscul esquelètic. Amb aquest propòsit hem expressat una forma truncada de Mfn2 (hMfn2Δ614-757) o bé hem reprimit l’expressió de Mfn2 endògena en el model cel•lular C2C12 i en múscul esquelètic de ratolí. Per dur a terme aquest objectiu hem generat 3 models de ratolí diferents: el model d’expressió transitòria de la forma hMfn2Δ614-757; el model de repressió transitòria de Mfn2 i el ratolí knockdown de Mfn2. Els dos primers models han estat generats mitjançant la tècnica de l’electrotransferència d’ADN en múscul esquelètic. La sobreexpressió de la forma hMfn2Δ614-757 en cèl•lules C2C12 diferenciades incrementa el consum d’oxigen mitocondrial en situació basal i també en desacoblar la cadena de transport d’electrons de la síntesi d’ATP, suggerint una major capacitat respiratòria dels miotubs que expressen la hMfn2Δ614-757. En múscul esquelètic de ratolí, l’expressió d’aquesta forma de Mfn2 causa una estimulació de la taxa d’oxidació de glucosa així com un increment de la Respiratory Control Ratio (RCR). La inducció del metabolisme mitocondrial observada en sobreexpressar la forma hMfn2Δ614-757 no és deguda a un augment de la massa mitocondrial, sinó a un increment en l’expressió i l’activitat d’alguns dels complexes de la cadena respiratòria mitocondrial. La repressió de Mfn2 en miotubs C2C12 produeix un increment en la respiració no associada a la producció d’ATP o proton leak i una disminució en el potencial de membrana mitocondrial. Aquests resultats indiquen que la repressió de Mfn2 provoca el desacoblament de la cadena de transport d’electrons i la síntesi d’ATP, suggerint una disminució de l’eficiència de la fosforilació oxidativa. Els músculs dels ratolins knockdown de Mfn2 presenten una reducció de la taxa d’oxidació de glucosa i de la Respiratory Control Ratio. A més, la repressió de Mfn2 disminueix l’activitat del complex IV de la cadena respiratòria. En conjunt aquests resultats suggereixen que la disminució de l’expressió de Mfn2 origina una disfunció del sistema de transport electrònic mitocondrial. També cal remarcar que els ratolins knockdown de Mfn2 presenten una major susceptibilitat a desenvolupar resistència a la insulina en resposta a l’envelliment o a una dieta rica en greixos. La disfunció mitocondrial i l’augment en la producció d’espècies reactives d’oxigen (ROS) observats en el múscul esquelètic d’aquests ratolins podrien explicar aquesta major susceptibilitat.
Mitochondria are cellular organelles that play a fundamental role in many cellular functions, such as substrates oxidation, ATP production, apoptosis and calcium economy. Mitochondria are dynamic organelles that can fuse and divide; the balance between both processes is crucial for a correct mitochondrial function. The most relevant proteins described to date involved in the regulation of mitochondrial fusion are mitofusins 1 and 2 (Mfn1 and Mfn2, respectively) and OPA1. Substantial data indicates that Mfn2 is also a key regulator of cell cycle and mitochondrial metabolism. On the other hand, Mfn2 expression is reduced in skeletal muscle of obese subjects and type 2 diabetic patients, situations characterized by altered mitochondrial activity. Based on these observations, the main objective of this thesis was the study of the metabolic role of Mfn2 in skeletal muscle. We have studied the metabolic effects caused by the manipulation of Mfn2 expression in mice skeletal muscle in vivo. By means of DNA electrotransfer technologies, we have expressed a truncated Mfn2 mutant in skeletal muscle and we have also repressed endogenous Mfn2 expression with microRNAs. We have also generated a skeletal muscle Mfn2 knockout mouse model (Mfn2 KO). The expression of truncated Mfn2 mutant in tibialis stimulated glucose oxidation and increased the Respiratory Control Ratio (RCR). It also increased the expression of subunits Cox4 of OXPHOS complex IV and Atp5a1 of complex V. We observed these metabolic effects in absence of changes in mitochondrial content. The repression of Mfn2 in mice skeletal muscle caused a marked reduction in the expression of subunit Cox4 of OXPHOS complex IV, accompanied with a 20% of decrease in COX activity. In this case we neither observed differences in mitochondrial content. Skeletal muscle from Mfn2 KO mice showed a decrease in glucose oxidation and in the RCR. In addition, Mfn2 KO mice showed a higher susceptibility to develop insulin resistance in response to aging or a high fat diet. Mitochondrial dysfunction and the increased ROS production observed in skeletal muscle of these mice could explain this higher susceptibility.

Thesis (Ph.D.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you.
Mitofusin-1 (Mfn-1) and mitofusin-2 (Mfn-2) are membrane-embedded mechanoenzymes involved in the remodelling and merging of the mitochondrial biomembrane. In differentiated cardiac myocytes, mitochondria occupy a third of the cell's volume and express both Mfn-1 and Mfn-2. The present thesis was aimed at exploring the roles of Mfn-1 and Mfn-2 specifically in cardiac myocytes using loss-offunction approaches in mice. We individually ablated either Mfn-1 or Mfn-2 specifically in cardiac myocytes. Ultramicroscopic analysis conducted in hearts of Mfn-1 KO or Mfn-2 KO mice revealed significant alterations in mitochondrial structure. Nevertheless, these knockout mice had normal heart function and a normal lifespan. Furthermore, Mfn-1 and Mfn-2 deficient mitochondria exhibited normal respiratory function in vitro. We also tested the susceptibility of Mfu-1 and Mfu-2 mitochondria against stress and unexpectedly found that the absence of these proteins conferred resistance to mitochondrial permeability transition (MPT). MPT reflects the loss of membrane integrity in mitochondria and is strongly associated with cell death. Using isolated adult cardiac myocytes we were able to demonstrate that the cell death in either Mfu-1 KO or Mfn-2 KO cells was delayed, consistent with the idea that MPT is attenuated in the absence of these proteins. We also utilized Mfn-2 KO mice to demonstrate that loss of Mfn-2 was associated with protection against cardiac ischemia/reperfusion injury, a stress model strongly linked to MPT. This work suggested for the first time that both Mfu-1 and Mfu-2 have important roles in the process of MPT. To incorporate these novel findings in context with the well-known role of mitofusins in membrane merging, I propose a working model where mitochondrial membrane fusion proceeds through formation of transient lipidic pores that compromise mitochondrial membrane integrity and serve as hotspots for MPT in conditions of stress. Lastly, we generated and characterized mice double-knockout (DKO) for Mfu-1 and Mfu-2. These mice are born in the expected ratios but undergo aberrant cardiac remodelling during the first week of their life and eventually succumb. The DKO mitochondria present multiple morphological and molecular abnormalities. This latter work illustrates that Mfn-1 and Mfn-2 operate interchangeably to regulate the early postnatal development of cardiac myocytes.

Durante el ictus, los niveles elevados de glutamato extracelular, el principal neurotransmisor excitador en el sistema nervioso central, aumentan y se unen al receptor de NMDA promoviendo la excitotoxicidad. Esta situación genera que un flujo excesivo de Ca2+ pase a través del receptor de NMDA aumentando sus concentraciones intracelulares activando toda una serie de cascadas de señalización que pueden actuar directamente sobre las mitocondrias, promoviendo varios programas de muerte celular. Las mitocondrias son orgánulos muy dinámicos que constantemente se fusionan y dividen, cambiando de forma y localización. El equilibrio entre la fisión y la fusión es importante para la función mitocondrial y es un regulador clave de la progresión de la muerte celular. Las proteínas que conforman el proceso de fusión mitocondrial y de fisión están formadas por un grupo de GTPasas. La fusión de la membrana mitocondrial interna está mediada por OPA1. Dos mitofusinas (Mfn1 y 2) median la fusión de la membrana externa mitocondrial. Por otro lado, la fisión mitocondrial está mediada por Drp1, una proteína citoplasmática que es reclutada a la superficie mitocondrial después de ser modificada a nivel post-traduccional. Los problemas relacionados con la dinámica mitocondrial se han descrito en múltiples enfermedades neurodegenrativas así como también en cáncer y sida. En esta Tesis hemos demostrado que Mfn2 juega un papel muy importante en el mantenimiento de la funcionalidad mitocondrial y la supervivencia neuronal. En excitotoxicidad es la única proteína de la maquinaria de fusión/fisión que disminuye su expresión en dos modelos in vitro distintos y también en un modelo in vivo. Hemos identificado dos fases de fragmentación mitocondrial en excitotoxicidad. La primera y que cursa en poco tiempo depende de la activación y reclutamiento de Drp1 mientras que la fase tardana que se inicia 4 horas después del insulto, parece depender de la reducción de los niveles proteicos de Mfn2. Esta fase tardía es irreversible y parece condenar a la neurona. La disminución de Mfn2 genera alteraciones en la homeóstasis del Ca2+, disminuye el potencial de membrana mitocondrial, empeora la comunicación entre las mitocondrias y el retículo endoplasmático, aumenta la traslocación de Bax a las mitocondrias y favorece la liberación del citocromo c. Hemos encontrado que la reducción en la expresión de Mfn2 durante la excitotoxicidad se da a nivel transcripcional y depende del factor de transcripción MEF2 que regula a Mfn2 a nivel basal en neuronas. En excitotoxicidad, MEF2 es degradado y causa la disminución de Mfn2. Este contexto hace pensar que Mfn2 podría ser una diana a tener en cuenta para futuros desarrollos de drogas y poder, de este modo, incrementar la pequeña ventana terapéutica que existe actualmente para el ictus.
Mitochondrial fusion and fission is a dynamic process critical for the maintenance of mitochondrial function and cell viability. During excitotoxicity neuronal mitochondria are fragmented, but the mechanism underlying this process is poorly understood. Here, we show that Mfn2 is the only member of the mitochondrial fusion/fission machinery whose expression is reduced in in vitro and in vivo models of excitotoxicity. Whereas in cortical primary cultures, Drp1 recruitment to mitochondria plays a primordial role in mitochondrial fragmentation in an early phase that can be reversed once the insult has ceased, Mfn2 downregulation intervenes in a delayed mitochondrial fragmentation phase that progresses even when the insult has ceased. Downregulation of Mfn2 causes mitochondrial dysfunction, altered calcium homeo- stasis, and enhanced Bax translocation to mitochondria, resulting in delayed neuronal death. We found that transcription factor MEF2 regulates basal Mfn2 expression in neurons and that excitotoxicity- dependent degradation of MEF2 causes Mfn2 downregulation. Thus, Mfn2 reduction is a late event in excitotoxicity and its targeting may help to reduce excitotoxic damage and increase the currently short therapeutic window in stroke.

Mitochondria are essential organelles for cellular homeostasis. Their main function is to produce energy: mitochondrial respiration provides most of the ATP required for endoergonic reactions. Furthermore, they regulate levels and transients of cytosolic Ca2+ and are crucially involved in apoptosis, aging and oxidative stress (Dimmer and Scorrano, 2006). As much as 20% of the mitochondrial surface is in close contact with the endoplasmic reticulum (ER). This organization is important for the generation of high Ca2+ microdomains required to activate mitochondrial Ca2+ uptake under certain conditions (Rizzuto et al., 1998a). The sites where ER and mitochondria are juxtaposed form the so called mitochondria-associated membranes (MAMs), crucial for lipid and Ca2+ traffic between the two organelles and are also involved in cell death (de Brito and Scorrano, 2010). Molecular mechanisms responsible for this ER-mitochondria juxtaposition are largely unknown, but it is thought that structural changes of either organelle could regulate the interaction (Pitts et al., 1999; Simmen et al., 2005). The shape of mitochondria is determined by the equilibrium between fusion and fission processes, controlled by a family of “mitochondria-shaping proteins”. In mammalian cells, mitochondrial fusion depends on mitofusin 1 and 2 (Mfn1 and 2) in the outer mitochondrial membrane and on the inner membrane protein optic atrophy 1 (Opa1) (Liesa et al., 2009). Fission requires the additional step of translocation of dynamin-related protein 1 (Drp1) from the cytosol to mitochondria where it presumably docks on the protein fission-1 (hFis1), its adaptor in the outer membrane. Oligomerization of Drp1 is believed to provide the mechanical force to constrict mitochondrial membranes and to fragment the organelle (Hinshaw, 1999b; Smirnova et al., 2001). Translocation of Drp1 to mitochondria depends on its dephosphorylation by calcineurin and other phosphatases (Cereghetti et al., 2008). Mitofusins are dynamin-related GTPases which control mitochondrial fusion and morphology. In the past, the presence of two mammalian mitofusins raised the question of their functional divergence. Nowadays many findings support the idea that Mfn1 and 2 do not display redundant functions. Mfn2 has roles which are not shared by Mfn1, such as the control of mitochondrial oxidation (Bach et al., 2005b) and anti-proliferative function (Chen et al., 2004a). Moreover, while mitochondrial fusion mediated by the inner mitochondrial membrane GTPase OPA1 requires Mfn1 (Cipolat et al., 2004b), Mfn2 appears to be involved in the regulation of endoplasmic reticulum (ER)-mitochondria tethering (de Brito and Scorrano, 2008a). MFN2 has also been involved in several diseases including the peripheral Charcot-Marie-Tooth type IIa (CMTIIa) neuropathy (Zuchner et al., 2004a). Both Mfn1 and Mfn2 are essential for embryonic development and mice deficient in either gene die in mid-gestation, but Mfn2-/- embryos display also deficient placentation (Chen et al., 2003h).. However, post-placentation ablation of either Mfn in the mouse leads to two completely different phenotypes: Mfn1-/- mice are viable, while Mfn2-/- ones day between P1 and P17 as a consequence of massive cerebellar degeneration (Chen et al., 2007b). It is unclear whether the different phenotypes observed in vivo and in vitro can be ascribed to functional differences or simply to different expression patterns of the two Mfns . In addition, it should be noted that all vertebrates possess two Mfn, whereas only one Mfn is retrieved in almost all invertebrates reigns, with the exception of D.melanogaster. The fruitfly possesses 2 mitofusins, Fuzzy onion (Fzo) and Mitochondrial Assembly Regulatory Factor (Marf). Fzo was the first identified mediator of mitochondrial fusion (Hales and Fuller, 1997a). In spermatocytes Fzo mediates the fusion of mitochondria into two giant organelles that form a structure called Nebenkern, which is required to give energy to the flagellum of the spermatid. In fzo mutant flies, mitochondria fail to fuse and wrap each other as many fragmented organelles, giving the impression of ‘fuzzy onions’ when viewed at the electron microscopy. Altough Fzo was able to promote mitochondrial elongation, it was expressed restrictely to spermatids. This raised the question of how mitochondrial fusion was regulated in other tissues of the fruitfly. In 2002 the second mitofusin homologue Marf was identified in D.melanogaster (Hwa et al., 2002). Unlike Fzo, DMarf is expressed ubiquitously and shares 64% of homology with both mammalian mitofusins. Thus, Marf can be likely considered the single “functional” invertebrate Mfn and Drosophila melanogaster is therefore a useful model organism to investigate phylogenesis of higher vertebrate Mfns. In this thesis we set to answer to the following questions: does this model organism represent the evolutionary turning point from one to two mitofusins? Which Mfn is functionally closer to DMarf? To answer to these questions, we first tried to understand in which extension DMarf could complement mammalian mitofusins. To obtain a wider phylogenetic view, we extended our analysis on mitofusins 1 and 2 from the vertebrate Xenopus laevis (XMfn1 and XMfn2) and on mitofusin homologue from the yeast S.cerevisae (Fzo1p). The expression of Fzo1p, DMarf, XMfn1 and XMfn2, induced mitochondrial elongation in Mfn1-/- or Mfn2-/- MEFs, rescuing the fragmented phenotype caused by the absence of the Mfns and indicating that they can substitute for both Mfns. Unlike Mfn1-/- MEFs, Mfn2-/- MEFs have altered ER morphology (de Brito and Scorrano, 2008b). This defect is recovered only after expression of hMfn2 or by an ER-targeted variant of hMfn2 (hMfn2YFFT), independently from mitochondrial shape. In fact Mfn2 is partially localized on ER and is enriched in MAMs. To investigate whether DMarf could complement Mfn2 also in ER shape regulation, we cotransfected Mfn2-/- MEFs with DMarf-V5 and ER-targeted yellow fluorescent protein (ERYFP). DMarf was able to rescue ER shape in Mfn2-/-, thus displaying a functional overlap with Mfn2. We then turned our attention to an in vivo analysis of DMarf function in Drosophila. DMarf was essential for viability as shown by ubiquitous, neuronal and muscle specific downregulation of the protein, which were all lethal. In larvae depleted of DMarf mitochondria clumped in the perinuclear regions of neuronal cell bodies and muscle tissues. In addition, neuromuscular junctions (NMJs) of DMarf-RNAi individuals were severely depleted of mitochondria compared to control larvae. Given that Mitofusin-2 regulates ER morphology and tethering to mitochondria, we also investigated the effect of DMarf knockdown on ER architecture which was seriously compromised upon DMarf knockdown in muscle tissues, with loss of the sarcomeric organization of the ER. To investigate the effect of overexpression of hMfns in Drosophila, we characterized transgenic lines expressing hMfn1, hMfn2 and hMfn2R94Q, one of the most frequent mutations associated with CMT2A (Zuchner et al., 2004b). Aggregation of mitochondria in neuronal cell bodies and muscle and elongated or clumped organelles inside axons were observed upon overexpression of all hMfns. However, NMJs analysis revealed a major difference between hMfn1, hMfn2 and hMfn2R94Q expressions: while NMJs of hMfn2 and hMfn2R94Q overexpressing individuals were depleted of mitochondria, hMfn1 expression did not alter organelles distribution in the nervous system and mitochondria were retained inside junctions. Muscle ER organization was not affected in wild-type hMfns expressing larvae; interestingly, ER morphology was found altered in muscle tissues only of hMfn2R94Q individuals, showing an aspect never investigated in the study of Charcot-Marie-Tooth pathogenesis and suggesting ER morphology alteration as one of the putative factors contributing to the pathobiology of the disease. Given that DMarf was able to complement both mfns in mitochondrial morphology and Mfn2 in ER shape regulation, we tried to understand to which extent hMfn1 and -2 could surrogate Marf in Drosophila. Simultaneous expression of Mfn2 and Marf-RNAi transgenes in the nervous system or in muscle resulted in partial survival to adulthood,. while Mfn1 or Mfn2R94Q were unable to recover the DMarf-RNAi and the flies died as pupae,. However, mitochondrial clusters were still present in neuronal cell bodies and muscle and mitochondria were lacking in NMJs of individuals simultaneously expressing DMarf-RNAi and hMfn2. Thus, although Mfn2 rescued DMarf depletion lethality, this could not be ascribed to a recovery of mitochondrial morphology and distribution. We therefore assessed recovery of ER organization in individuals simultaneously expressing DMarf-RNAi and hMfns. HMfn1 had no effect on ER alteration caused by DMarf knockdown. On the contrary, hMfn2 expression recovered muscle ER architecture both when expressed ubiquitously or only in the muscle . In conclusions, in this Thesis we demonstrate that Fzo1p, DMarf, XMfn1 and XMfn2 expression could rescue mitochondrial morphology in Mfn1 -/- and Mfn2-/- MEFs. Moreover DMarf specifically complements ER shape in Mfn2-/- MEFs. Thus, mitofusins role in regulation of mitochondrial dynamics is conserved between vertebrates and invertebrates. On the other hand, in vivo experiments show that Mfn2, but not Mfn1 rescues the lethal phenotype of DMarf knock-down in Drosophila melanogaster. the rescue of DMarf-RNAi by hMfn2 seemed correlate with the correction of ER organization, whereas mitochondrial morphology and distribution were not restored. We can therefore speculate that Marf is functionally closer to Mfn2 than Mfn1, which may have diverged later during mammalian evolution. Finally, factors other than the disruption of mitochondrial distribution and morphology in the nervous system could explain the lethal phenotype of DMarf knockdown and in perspective be involved in the pathogenesis of CMT2a
I mitocondri sono organelli essenziali per l’omeostasi cellulare. La loro funzione primaria è di produrre energia : la respirazione mitocondriale fornisce la maggior parte di ATP necessaria per le reazioni endoergoniche. Inoltre, essi regolano i livelli e i transienti di calcio citosolico e hanno un ruolo cruciale nei processi di apoptosi, invecchiamento e stress ossidativo (Jouaville et al., 1995; Wang, 2001). Il 20% della superficie mitocondriale è in stretto contatto con il reticolo endoplasmico (RE). Questa disposizione è importante per la generazione di microdomini ad alta concentrazione di calcio necessari in certe condizioni per l’attivazione dell’uniporto mitocondriale del Ca2+. I siti di stretto contatto tra il RE e i mitocondri formano le cosiddette “membrane associate ai mitocondri” (MAMs), che sono cruciali per il trasporto di lipidi e Ca2+ tra i due organelli e hanno un ruolo anche nel processo di morte cellulare (Rizzuto et al., 1998). Sebbene i meccanismi molecolari alla base di questa stretta vicinanza tra il RE e i mitocondri siano in larga parte ignoti, si ritiene che tale interazione possa essere regolata da cambiamenti morfologici dei due organelli (Pitt set al., 1999; Simmen et al., 2005). La forma del reticolo mitocondriale è determinata dall’equilibrio tra eventi di fusione e fissione, controllati da una famiglia di “proteine di morfologia mitocondriale”. In cellule di mammifero, la fusione mitocondriale è controllata dalle proteine mitofusina-1 (MFN1) e mitofusina-2 (MFN2) nella membrana esterna e da OPA1 nella membrana mitocondriale interna (Olichon et al., 2002). Nel nostro laboratorio è stato dimostrato che OPA1 promuove la fusione dei mitocondri solo in presenza di mitofusina 1 (Cipolat et al., 2004). La fissione richiede il passaggio aggiuntivo della traslocazione della proteina Drp1 dal citosol ai mitocondri, dove si ancora a hFis1, il suo adattatore molecolare nella OMM. L’oligomerizzazione di Drp1 fornisce la forza meccanica per costringere le membrane mitocondriali fino alla frammentazione dell’organello (Hinshaw et al., 1999; Smirnova et al., 2001). La traslocazione di Drp1 ai mitocondri dipende dalla sua defosforilazione ad opera della fosfatasi calcineurina e di altre fosfatasi (Cereghetti et al., 2008). Le mitofusine sono GTPasi appartenenti alla famiglia delle dinamine che controllano la fusione e la morfologia dei mitocondri. In passato, la presenza di due mitofusine nei mammmiferi ha sollevato la questione sulla loro divergenza funzionale. Oggi molti risultati supportano l'idea che Mfn1 e 2 non abbiano funzioni ridondanti. Mfn2 ha ruoli che Mfn1 non esercita, come il controllo dei processi di ossidazione mediati dai mitocondri (Bach et al., 2005) e la funzione anti-proliferativa (Chen et al., 2004). Inoltre, mentre la fusione mitocondriale indotta dalla proteina della membrana interna OPA1 richiede Mfn1, (Cipolat et al., 2004), Mfn2 sembra essere coinvolta nella regolazione della giustapposizione reticolo endoplasmatico (RE)-mitocondri (de Brito e Scorrano, 2008a). MFN2 è stata associata inoltre alla neuropatia di Charcot-Marie-Tooth di tipo IIa (CMTIIa) (Züchner et al., 2004b). Sia Mfn1 e Mfn2 sono essenziali per lo sviluppo embrionale e topi deficienti in entrambi muoiono prematuramente, ma (Chen et al, 2003) embrioni Mfn2-/- mostrano anche difetti di placentazione. Tuttavia, l'ablazione post-placentazione di una singola Mfn nel topo porta a due fenotipi completamente diversi: i topi Mfn1-/ - sono vitali, mentre topi Mfn2-/ - muoiono tra P1 e P17 come conseguenza della massiccia degenerazione cerebellare (Chen et al. , 2007). Non è chiaro se i diversi fenotipi osservati in vivo e in vitro possano essere attribuiti a differenze funzionali o semplicemente ai diversi schemi di espressione delle due Mfn. Inoltre, tutti i vertebrati possiedono due Mfn, mentre solo una Mfn viene riscontrata nella maggior parte degli invertebrati, con l'eccezione di D.melanogaster. Il moscerino della frutta possiede due mitofusine, ‘Fuzzy onion’ (Fzo) e Mitochondrial Assembly Regulatory Factor (Marf). Fzo è stato il primo mediatore della fusione mitocondriale individuato (Hales e Fuller, 1997). In spermatociti Fzo media la fusione dei mitocondri in due giganti organelli che formano una struttura chiamata Nebenkern, la quale serve a dare energia al flagello degli spermatidi. Nel mutante di fzo i mitocondri non riescono a fondersi e si avvolgono l'un l'altro come molti organelli frammentati, dando l'impressione di 'cipolle increspate' se osservate al microscopio elettronico. Anche se Fzo è in grado di promuovere l'allungamento mitocondriale, è espresso solo negli spermatidi. Ciò ha sollevato la questione di come la fusione dei mitocondri potesse venire regolata in altri tessuti del moscerino della frutta. Nel 2002 il secondo omologo delle mitofusine Marf è stato identificato in D.melanogaster (Hwa et al., 2002). A differenza di Fzo, DMarf è espressa ubiquitariamente e condivide il 64% di omologia con entrambe le mitofusine dei mammiferi. Quindi, Marf può essere considerato la singola mitofusina "funzionale" e Drosophila melanogaster è dunque un organismo modello utile per studiare la filogenesi di MFN nei vertebrati superiori. Le domande a cui abbiamo cercato di dare risposta in questa tesi sono le seguenti: questo organismo modello rappresenta il punto di svolta evolutiva da uno a due mitofusine? Quale Mfn è funzionalmente più vicina a DMarf? Al fine di rispondere a queste domande, abbiamo innanzitutto cercato di capire in quale estensione DMarf potesse complementare le mitofusine dei mammiferi. Per ottenere una visione più ampia dal punto di vista filogenetico, abbiamo esteso la nostra analisi alle mitofusine 1 e 2 del vertebrato Xenopus laevis (XMfn1 e XMfn2) e all’omologo di mitofusina di lievito S.cerevisae (Fzo1p). L'espressione di Fzo1p, DMarf, XMfn1 e XMfn2, induce allungamento dei mitocondri in Mfn1-/ - o Mfn2-/ - MEF, recuperando il fenotipo di frammentazione causato dalla mancanza delle singole Mfn e indicando che esse possono sostituire entrambe le Mfn. A differenza di Mfn1-/ - MEF, Mfn2-/ - MEF mostrano una morfologia del reticolo endoplasmico alterata (de Brito e Scorrano, 2008b). Questo difetto è recuperato solo dopo l'espressione di hMfn2 o di una variante hMfn2 specificamente localizzata al RE (hMfn2YFFT), indipendentemente dalla forma mitocondriale. Mfn2 infatti è parzialmente localizzata nel RE e si trova arricchita nelle MAMs. Per capire se DMarf potesse complementare Mfn2 anche nella regolazione della morfologia del RE, abbiamo co-trasfettato Mfn2-/ - MEFs con Dmarf con il tag V5 e una proteina fluorescente gialla localizzata nel RE (ERYFP). DMarf riesce a recuperare la morfologia del RE in Mfn2-/ -, mostrando quindi una più forte conservazione funzionale con Mfn2. Abbiamo poi rivolto la nostra attenzione ad un analisi in vivo della funzione di DMarf in Drosophila. DMarf è essenziale per la vitalità, poichè il silenziamento ubiquitario o specificatamente a carico del sistema nervoso o del tessuto muscolare della proteina è letale. In larve DMarf-RNAi i mitocondri si aggregano nelle regioni perinucleari dei corpi cellulari neuronali e dei tessuti muscolari. Inoltre, giunzioni neuromuscolari degli individui DMarf-RNAi risultano gravemente depauperati dei mitocondri rispetto alle larve di controllo. Dato che mitofusina-2 regola la morfologia del reticolo endoplasmico e il suo avvicinamento ai mitocondri, abbiamo studiato l’effetto del silenziamento di Dmarf sull’architettura del RE, la quale è risultata essere seriamente compromessa nei tessuti muscolari mostrando perdita dell’organizzazione del RE lungo i sarcomeri in seguito al silenziamento della proteina. Per studiare l'effetto della sovraespressione di hMfn in Drosophila, abbiamo caratterizzato le linee transgeniche esprimenti hMfn1, hMfn2 e hMfn2R94Q, una delle mutazioni più frequenti associate al CMT2A (Züchner et al., 2004a). Aggregazione dei mitocondri in corpi cellulari neuronali e nel muscolo e organelli di forma allungata o aggregati all'interno di assoni sono stati osservati in seguito a sovraespressione di tutte le hMfns. Tuttavia, l'analisi delle giunzioni neuromuscolari ha rivelato una grande differenza tra le espressioni hMfn1, hMfn2 e hMfn2R94Q: mentre le giunzioni di individui sovraesprimenti hMfn2 e hMfn2R94Q erano depauperate dei mitocondri, l’espressione di hMfn1 non ha modificato la distribuzione degli organelli nel sistema nervoso e mitocondri erano mantenuti all'interno delle giunzioni. L’organizzazione del RE nel muscolo non è stata influenzata dall’espressione di hMfns wild-type; la morfologia del RE è invece alterata in tessuti muscolari di individui esprimenti hMfn2R94Q, mettendo in luce un aspetto non ancora indagato nello studio della patogenesi di Charcot-Marie-Tooth e suggerendo l'alterazione della morfologia del RE come uno dei plausibili fattori che possono contribuire alla patogenesi della malattia. Dato che DMarf è stato in grado di complementare sia entrambe le Mfn nella morfologia mitocondriale e sia Mfn2 nella regolazione della morfologia del RE, abbiamo cercato di capire in che misura hMfn1 e -2 potessero complementare DMarf in Drosophila. L'espressione simultanea dei transgeni MFN2 e DMarf-RNAi nel sistema nervoso o nel tessuto muscolare provocano parziale sopravvivenza fino all'età adulta, mentre Mfn1 o Mfn2R94Q non sono riusciti a recuperare il fenotipo letale del DMarf-RNAi. Tuttavia, gli aggregati mitocondriali erano ancora presenti nei corpi delle cellule neuronali e nei tessuti muscolari e le giunzioni neuromuscolari di individui esprimenti contemporaneamente DMarf-RNAi e hMfn2 prive di mitocondri. Quindi, anche se MFN2 recupera la letalità dovuta al silenziamento di DMarf, questo non può essere attribuito a un recupero della morfologia e della distribuzione mitocondriale. Abbiamo quindi valutato il recupero dell’organizzazione del RE in individui esprimenti simultaneamente Dmarf-RNAi e hMfn. HMfn1 non ha avuto alcun effetto sull’ alterazione del RE causata da Dmarf-RNAi. Al contrario, hMfn2 espressione ha recuperato l’architettura del RE nel muscolo sia quando espressa ubiquitariamente o solo nel tessuto muscolare. In conclusione, in questa tesi abbiamo dimostrato che l’ espressione di Fzo1p, DMarf, XMfn1 e XMfn2 è in grado di recuperare la morfologia mitocondriale in Mfn1 - / - e Mfn2-/ - MEF. Inoltre DMarf complementa specificamente la morfologia del RE in Mfn2-/ - MEF. Quindi, il ruolo delle mitofusine nella regolazione della dinamica mitocondriale è conservato tra vertebrati e invertebrati. Tuttavia, esperimenti in vivo dimostrano che Mfn2, ma non Mfn1 recupera il fenotipo letale dovuto al silenziamento di DMarf in Drosophila melanogaster. Il recupero di DMarf-RNAi da parte di hMfn2 sembra correlare con la correzione dell’ organizzazione del RE, mentre la morfologia mitocondriale e la distribuzione non sono vengono ripristinati. Possiamo quindi ipotizzare che Marf sia funzionalmente più vicina ai Mfn2 rispetto a Mfn1, la quale potrebbe essere comparsa più tardi nel corso dell'evoluzione dei mammiferi. Infine, altri fattori,oltre l’alterazione della morfologia e distribuzione mitocondriale nel sistema nervoso potrebbero spiegare la letalità causata dal silenziamento di DMarf e essere quindi coinvolti nella patogenesi della CMT2A.

Blood vessels distribute nutrients and oxygen to every single cell in the body. Endothelial cells define the vessel wall, and thus they are ideally located to crucially modulate nutrient availability and act as metabolic gatekeepers of the organism. In recent years, mitochondrial dynamics has emerged as a bioenergetic adaptation process to cellular metabolic demands. Mitofusins are GTPase-like proteins implicated in external mitochondrial membrane fusion. Our hypothesis is that mitochondrial fusion in endothelial cells is implicated in energy balance and metabolic control. In order to address this hypothesis, we generated mice lacking either Mitofusin 1 (Mfn1) or Mitofusin 2 (Mfn2) into adulthood by breeding a tamoxifen-inducible endothelial Cre line (PdgfbiCreERT2) with Mfn1 or Mfn2 floxed animals (hereafter called Mfn1ΔEC and Mfn2ΔEC respectively). Mfn2iΔEC mice showed a progressive reduction (25%) in body weight when compared to control counterparts. Intestinal nutrient absorption, food intake and locomotor activity were unaltered in knockout mice. However, enhanced energy expenditure and a shift towards lipid oxidation was observed, while the thermogenesis capacity was not different between groups. Consistent with this phenotype, Mfn2iΔEC mice exhibited lower fat mass and improved glucose tolerance and insulin sensitivity in the face of unaltered insulin release. Collectively, these results indicate that loss of Mfn2 in endothelial cells causes a lean phenotype as the consequence of enhanced lipid metabolism. However, endothelial Mfn1 deletion did not alter systemic metabolism. Upon high-fat diet administration, Mfn2iΔEC mice showed complete resistance to its obesogenic effects. In concordance with lower body weight due to reduced adiposity, mutant mice exhibited improved glucose homeostasis. Moreover, induction of endothelial Mfn2 ablation in established obesity reduced body weight to standard diet control levels and improved metabolic alterations. Interestingly, Mfn1iΔEC mice do not show any metabolic alteration when fed high-fat diet. Aged Mfn2iΔEC mice preserved young-like health-span parameters. Indeed, mutant mice exhibited improved age-associated physiological parameters such as kidney function or anaemia. Diverse motor and cognitive parameters were also preserved in old Mfn2i∆EC mice. Collectively, our results indicate that Mfn2 in endothelial cells is implicated in systemic energy homeostasis control as well as in ageing progression in mice.

Thesis (Ph.D.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you.
Mitochondrial injury and apoptosis promote organ failure after ischemic acute kidney injury (AKI), a common cause of morbidity and mortality. In these studies, we propose that hexokinase (HK), mitofusin 2 (MFN2) and Bax, key mitochondrial associated proteins, modulate apoptotic cell death and organ function after ischemia. In the kidney, HKI and HKII isoforms both possess mitochondrial localization sequences. In vivo ischemia reduced murine proximal tubule HKII content and caused mitochondrial HKII dissociation. In cultured renal epithelial cells, expression of HKI or II significantly improved survival after ATP depletion, an in vitro model of ischemia, without preventing Bax activation or reducing mitochondrial fragmentation, a determinant of organelle sensitivity to injury. HKII over-expression increased mitochondrial associated HKII during stress and decreased mitochondrial Bax accumulation, a major cause of outer membrane permeabilization and apoptosis, suggesting that HK improves renal cell survival by antagonizing Sax-mediated injury. Deficiency of MFN2, a pro-fusion protein, caused mitochondrial fragmentation in primary proximal tubule cells without altering baseline or maximal oxygen consumption rate, or cell apoptosis. However, MFN2 deficiency significantly increased mitochondrial Bax accumulation and exacerbated mitochondrial outer membrane injury after stress. In the mouse, whole kidney MFN2 knockout caused severe mitochondrial fragmentation in renal epithelial cells. However, despite a small (20%) decrease in nephron number compared to littermate controls, newborn knockouts exhibited normal tubular and organ function. Surprisingly, proximal tubule specific MFN2 knockouts were also protected from renal ischemia. Although histologic injury scores as well as levels of apoptosis and necrosis, were similar, renal function and animal survival were significantly higher in proximal tubule specific MFN2 knockout mice at 24 and 48 hours post-ischemia. Interestingly, cortical oxidant stress was halved while cortical proliferation was nearly 4 times higher in proximal tubule knockouts compared to control, suggesting that MFN2 deficiency promotes organ recovery and survival after ischemia by enhancing proximal epithelial cell growth. While HK and MFN2 modulate Bax-mediated mitochondrial injury and apoptosis, "off-target" effects of MFN2 on renal cell proliferation ameliorate ischemia-reperfusion injury. These studies highlight the role of Bax-mediated mitochondrial injury in ischemic organ failure and suggest new targets for both attenuating injury and promoting organ recovery.

Les mitochondries sont des organites dynamiques dont la morphologie dépend de l’équilibre fusion/fission de leurs membranes. Ce processus essentiel à la survie cellulaire est nommé dynamique mitochondriale et sa dérégulation est associée à des troubles neurologiques. Cependant les mécanismes précis régissant la dynamique mitochondriale ne sont pas élucidés. Cette thèse porte sur la protéine Fzo1p, une grande GTPase de la superfamille des Dynamin-related-Protein. C’est un élément clé impliqué dans la fusion mitochondriale de la membrane externe de la levure. Sa structure et sa dynamique ont été étudiées par modélisation et simulations de dynamiques moléculaires tout-atome dans une bicouche lipidique solvatée. Le modèle structural obtenu tient compte de données expérimentales, de template structuraux, et de modèles ab initio du domaine transmembranaire de Fzo1p. Ce modèle a été validé expérimentalement par mutagenèse dirigée. Des permutations de charges ont confirmé des ponts salins à longue distance prédits dans le modèle. En outre, des mutations ont montré que les domaines coiled-coil de Fzo1p, contrairement à sa partie N-terminale, sont indispensables à sa fonction. L’ensemble des résultats expérimentaux et in silico met en évidence l’implication des domaines charnières dans le changement conformationnel de Fzo1p, ainsi que des résidus critiques affectant sa stabilité. Les précisions atomiques obtenues sur l’interaction de Fzo1p avec le GDP permet de formuler des hypothèses sur le mécanisme moléculaire de la catalyse du GTP pour la fusion membranaire; voire à la compréhension de la dynamique mitochondriale
Mitochondria are dynamic organelles whose morphology is determined by fusion and fission of their membranes. This essential process is known as mitochondrial dynamics. Defects in mitochondrial dynamics are associated with neurological disorders making the investigation of physiological relevance. However, the precise sequence of events that lead mitochondrial dynamics are still not well characterised. Fzo1p, a large GTPase of the Dynamin-Related Proteins superfamily, is a key component in mitochondrial outer membrane fusion in yeast. During this PhD project I built a model of the protein Fzo1p. The structure and dynamics of the model was investigated through molecular modelling and all-atom molecular dynamics simulation in a fully hydrated lipid bilayer environment. The Fzo1p structural model integrates information from several template structures, experimental knowledge, as well as ab initio models of the transmembrane segments. The model is validated experimentally through directed mutagenesis, for instance charge-swap mutations confirm predicted long-distance salt bridges. A series of mutants indicate that coiled-coil domains are required for protein function at variance with its N-terminal region. Overall, the experimental and in silico approaches pinpoint the hinge domains involved in the putative conformational change and identifies critical residues affecting protein stability. Finally, key Fzo1p-GDP interactions provide insights about the molecular mechanism of membrane fusion catalysis. The model provides insight on atomic level and proposes a structure that will be instructional to understanding mitochondrial membrane fusion

The mitofusin 1 and 2 (MFN and MFN2) proteins reside in the outer mitochondrial membrane and have been shown to regulate mitochondrial network architecture by mediating tethering and fusion of mitochondria. Mitochondria normally form a tubular and branched reticular network dynamically regulated by a balance of fusion and fission events. Absence of either Mfn1 or Mfn2 results in a fragmented mitochondrial network. Züchner et al. previously described mutations in the gene mitofusin 2 (MFN2) as the cause of the major autosomal-dominant, axonal form of Charcot-Marie-Tooth neuropathy (CMT2A). CMT type 2 (CMT2) is characterized by chronic axonal degeneration of peripheral nerves leading to the loss of functional nerve fibers. Mutations in MFN2 are the most common cause of CMT2, and in Chapter 2 we report the results from a genetic screen of MFN2 in a CMT2 patient cohort. The original finding that mutations in MFN2 cause CMT2A led to investigations focused on deficiencies of mitochondrial fusion and transport, specifically in the context of long axonal processes affected in CMT. While some experimental work supports disrupted mitochondrial transport in the etiology of CMT2A, other studies on CMT2A patient fibroblasts and cell models suggest abnormal mitochondrial fusion and dynamics do not underlie the etiology of this. In the first half of Chapter 3, we present some of our initial investigations prior to de Brito and Scorrano’s report published in 2008 regarding a novel role for Mfn2 in tethering the endoplasmic reticulum (ER) to mitochondria. In Mfn2 null mouse embryonic fibroblasts (MEFs) regions of contact between mitochondria and the endoplasmic reticulum (ER) are significantly reduced. These regions of contact are thought to form specialized subdomains of the ER, called mitochondrial associated membranes (MAM). Besides observing a fragmented ER network in Mfn2 knockout (KO) mouse embryonic (MEF) cells, de Brito and Scorrano presented several lines of evidence which suggest that the underlying pathogenic mechanism in CMT2A stems from disrupted ER-mitochondria. As this observation had not been replicated in the literature, we describe our attempts to replicate these finding in the last half of Chapter 3. The MAM represents a sub-domain of the ER in close association with the mitochondrial outer membrane. The movement of phosphatidylserine (PS) from the MAM domains of the ER to mitochondria and its subsequent decarboxylation to phosphatidylethanolamine (PE) by the enzyme PS decarboxylase (Pisd) has been well characterized and is known to depend on the existence of an outer mitochondrial membrane protein. As PE has curvature inducing and fusogenic biophysical characteristics, a deficiency in PE would be an attractive mechanism contributing to the morphological and fusion defects observed in Mfn2 null cell models. We hypothesized that loss of Mfn2 would lead to specific decreases in mitochondrial and cellular levels of PE. Chapter 4 describes experiments designed to test this hypothesis. We observed significantly lower levels of PE in Mfn2 null cells, yet observe similar changes in Mfn1 null cells. Likewise, other lipid species such as ether linked PE (ePE) are decreased. To investigate how CMT2A mutations in MFN2 influence cellular phospholipid profiles, we then profiled cellular phospholipids of CMT2A patients and control lymphoblasts. We hypothesized that mutations in MFN2 would result in decreased levels of PE. In Chapter 5, we report the results of a phospholipid screen which reveal changes in ePE in CMT2A patient lymphoblasts, without the drastic decreases in PE previously observed in Mfn2 null lines. In conclusion, our data indicates an important role for both mitofusins in the mitochondrial synthesis of PE. In the context of CMT2A mutations, ePE levels are specifically reduced. Future studies may reveal how deficiencies in ePE might have important functional consequences in the pathogenesis of CMT2A.

Les mitochondries sont la principale source d’énergie dans les neurones. Les défauts mitochondriaux participent à l’apparition de maladies neurodégénératives, cependant ils peuvent être contrés par un système de contrôle qualité. Le but de ma thèse a été de déterminer si ce système est dérégulé dans la maladie de Huntington (MH) et la sclérose latérale amyotrophique (SLA) et si sa restauration est neuroprotectrice, en utilisant principalement des modèles drosophile. La MH, caractérisée par une atteinte des neurones du striatum, est due à la protéine Huntingtin mutée (mHtt). Nous avons montré que la mHtt induit une accumulation des mitochondries dans la rétine. Ceci pourrait être dû à un défaut de la mitophagie, un mécanisme qui permet l’élimination des mitochondries défectueuses et qui est orchestré par la protéine PINK1. De manière intéressante, la surexpression de PINK1 corrige le phénotype pathologique des drosophiles exprimant la mHtt. Je me suis aussi intéressé à la SLA, chez laquelle les motoneurones dégénèrent, plus exactement au gène TDP-43 qui est un contributeur majeur à la maladie. Nous avons montré que la surexpression de TDP-43 dans les neurones de drosophiles entraîne une fragmentation des mitochondries liée à une sous-expression du gène mitofusin. Ce dernier contrôle le processus de fusion entre les mitochondries saines et endommagées et donc l’intégrité de cet organite. La surexpression de Mitofusin améliore les défauts locomoteurs et l’activité neuronale altérée chez les drosophiles exprimant TDP-43. Nos résultats montrent l’importance du contrôle qualité mitochondrial dans la pathogenèse de ces maladies, et que de le renforcer pourrait être bénéfique
Mitochondria are the main energy source in neurons. Mitochondrial defects contribute to the development of neurodegenerative diseases, however they can be countered by a quality control system. The purpose of my thesis has been to determine if this system is dysregulated in Huntington’s disease (HD) and in amyotrophic lateral sclerosis (ALS) and if restoring it can be neuroprotective, by mainly using Drosophila models. HD, which is characterized by loss of striatal neurons, is caused by the mutant Huntingtin protein (mHtt). We showed that mHtt induces the accumulation of mitochondria in the retina. This could be due to a defect in mitophagy, a mechanism which allows the elimination of defective mitochondria and which is orchestrated by the protein PINK1. Interestingly, PINK1 overexpression ameliorates the abnormal phenotype of flies expressing mHtt. I also got interested in ALS, in which motor neurons degenerate, and mainly in the TDP-43 gene which is a major contributor to the disease. We showed that TDP-43 overexpression in Drosophila neurons leads to fragmentation of mitochondria due to decreased expression levels of the mitofusin gene. The latter controls the fusion process between healthy and damaged mitochondria and therefore the organelle integrity. We show that Mitofusin overexpression ameliorates locomotor defects and abnormal neuronal activity in flies expressing TDP-43. Our results show the importance of mitochondrial quality control in the pathogenesis of these diseases, and that reinforcing it can be beneficial

Ce travail initié en lien avec le centre de référence des Pathologies Rares de la Résistance à l’Insuline et de l’Insulino-Sensibilité, s’intéresse à l’étude de la physiopathologie de formes rares de syndromes lipodystrophiques. Parmi ceux-ci, la lipomatose de Launois-Bensaude (LLB) est caractérisée par la présence de masses lipomateuses associées à des troubles métaboliques. Nous avons investigué la plus grande cohorte de patients atteints de LLB due au variant p.Arg707Trp du gène MFN2, codant la mitofusine 2, une protéine de la fusion mitochondriale. Nous avons également étudié une patiente présentant des symptômes compatibles avec une LLB porteuse d’un nouveau variant p.Glu943Glyfs*22 de LIPE, codant la lipase hormono-sensible, enzyme-clé de la lipolyse adipocytaire. Les caractéristiques de ces patients porteurs de variants de MFN2 et LIPE sur le plan clinico-biologique et tissulaire, permettent de mieux définir les particularités de la LLB au sein des syndromes lipodystrophiques. Nous avons isolé les cellules souches adipocytaires (ASC) des lipomes et utilisé ce modèle pour évaluer le retentissement des variants sur la différenciation et les fonctions adipocytaires. Les études morphologiques (microscopie optique et électronique) et fonctionnelles (immuno-histochimie, expression génique et protéique, lipolyse et respiration mitochondriale) des lipomes et/ou des ASC, montrent des altérations adipocytaires multiples avec un phénotype thermogénique des adipocytes LLB-MFN2. La lipodystrophie associée à MFN2 pourrait provenir d’une altération de la balance de différenciation adipocytaire blanche/beige
This work, initiated in cooperation with the rare diseases reference center ‘Pathologies de la Résistance à l’Insuline et de l’Insulino-Sensibilité’, focuses on the pathophysiology of rare lipodystrophic syndromes. Among them, Launois-Bensaude lipomatosis, also called multiple symmetric lipomatosis (MSL), is characterized by upper-body lipomatous masses and frequent metabolic alterations. We have investigated the largest reported series of patients with MSL due to the MFN2 p.Arg707Trp variant. MFN2 encodes mitofusin 2, a protein involved in mitochondrial fusion. Additionally, a patient with clinical symptoms consistent with MSL, harboring a new p. Glu943Glyfs*22 variant of LIPE, encoding hormone-sensitive lipase, a key enzyme in the lipolysis pathway, has also been studied. The clinical, biological and adipose tissue characteristics of patients carrying MFN2 and LIPE variants, allow for a better definition of MSL within the lipodystrophic syndromes. We have isolated adipose-derived stem cells (ASC) from lipomas and used this cellular model to assess the impact of variants on adipocyte differentiation and functions. Morphological (optic and electronic microscopy) and functional studies (immunohistochemistry, gene and protein expression, lipolysis, and mitochondrial respiration) on pseudo-lipomas and/or on ASC show numerous adipose dysfunctions and highlight the thermogenic phenotype of adipocytes from MFN2-MSL patients. This MFN2-related lipodystrophy could result from a misbalance of white and beige adipocyte differentiation

Charcot-Marie-Tooth diseases (CMTs) are the most common hereditary pathologies of the peripheral nervous system. The dominant subtype CMT2A, one of the most frequent, is caused by mutations in MFN2 gene, coding for mitofusin 2, a dynamin-like GTPase located in the outer membrane of mitochondria. MFN2 is involved in several cellular processes, as the fusion of mitochondrial membranes, the transport of mitochondria along axons and the tethering of mitochondria with endoplasmic reticulum. To date more than 80 mutations associated with CMT2A and complicated variants of CMT2 have been described. They are mainly point mutations with dominant transmission, but recently some semi-dominant and recessive mutations have been reported. With the aim to identify new causing mutations and possibly identify a genotype-phenotype correlation, in the first part of this project it was performed a genetic analysis of MFN2 in 25 patients with diagnosis of axonal CMT. This analysis was conducted primarily by direct sequencing of all exons of MFN2 gene. Subsequently, in subjects where no mutations were detected, it was carried out a multiplex ligation-dependent probe amplification (MLPA), a relatively new PCR-based technique that enables the identification of complex gene rearrangements. The genetic analysis of patients with CMT2A allowed the identification of the disease causing mutation in five probands. In four patients the mutation was found in heterozygosis, while one patient affected by an uncommonly severe axonal CMT resulted a compound heterozygote for two MFN2 mutations (p.[K38del]+[T362M]). In all, four mutations (E329del, A738V, R94P, K38del) never reported in literature were found. The analysis of MFN2 by MLPA did not revealed large gene rearrangements in patients without point mutations in MFN2, thus meaning that probably other genes are responsible for the disease in such patients and confirming the genetic heterogeneity of CMT neuropathies. The understanding of the mechanisms by which the mutated forms of MFN2 lead to neurodegeneration is limited by the fact that few mouse models for CMT2A were successfully developed. Since zebrafish turned useful to study many neurological and neuromuscular disorders, we decided to use it to investigate the function of MFN2 and its role in CMT2A disease. Using the morpholino technique, mitofusin 2 was knocked-down in developing embryos, which resulted severely motor-impaired in accordance with the loss of limbs motility observed in CMT2A patients. Investigations performed on the neuromuscular system of morphants, demonstrated that larvae exhibit misshapened motorneurons and a reduction of neuromuscular junctions. As in human pathology, possibly because their incorrect innervation, muscular fibres appear hypotrophic and are reduced in diameter. These results, validated by the rescue of morphants with human MFN2 mRNA, confirm the essential role of mitofusin 2 in motorneurons development and suggest that the zebrafish could be a very useful tool to study in living embryos the effects of mutations identified in CMT2A patients. Therefore, with the aim to study the mutations found in our patients, we developed a method to evaluate the effect of human mutations in MFN2 using zebrafish embryos as a model. For this reason we first concentrated on the analysis of R94Q, which is a well studied mutation found frequently in CMT2A patients with severe phenotype. By injecting the mutated human MFN2 mRNA alone and together with the morpholino against zebrafish mfn2, we demonstrated that, in the first stages of zebrafish development, R94Q is, at least in part, loss of function. This agrees with some studies performed in cells that demonstrated that this mutant allele of MFN2 cannot restore the shape of mitochondria in knock-out cells. Moreover, a previously reported R94Q knock-in mouse displayed no phenotype in heterozygous state, while in homozygosity is lethal, but survives longer than the complete knock-out, thus confirming the hypothesis that R94Q mutated mitofusin 2 has a reduced activity. However a pure loss of function mechanism was recently excluded in mouse by the finding that a transgenic model over-expressing this mutation develops movement defects and alterations in mitochondria distributions resembling CMT2A disease by the age of 5 months. Our data, together with that found in literature, suggest that R94Q, even if it has a reduced functionality, may have also a gain of function activity that is disclosed phenotypically later during development. Since our method allows the analysis of MFN2 mutations effect only during the first stages of development, we will need a stable transgenic zebrafish model to confirm the mechanism of action of R94Q and analyse the effect of MFN2 mutations found in the patients we investigated. Altogether the results obtained with this work contributed to clarify the possible genotype-phenotype relations involved in CMT2A disease. Moreover it was proposed zebrafish as a new tool to dissect the molecular mechanism by which mitofusin 2 mutations lead to CMT2. Being easily amenable to drug screening, the zebrafish model could be not only a useful complement to the studies performed in mouse, but it may also contribute to identification of effective pharmacological compounds for the treatment of Charcot-Marie-Tooth disease.
Le Charcot-Marie-Tooth (CMT) sono le più comuni patologie ereditarie del sistema nervoso periferico. Il sottotipo CMT2A, uno dei più frequenti, è causato da mutazioni nel gene MFN2, codificante mitofusina 2, una proteina GTPasica, per struttura simile alla dinamina, localizzata a livello della membrana esterna dei mitocondri. MFN2 risulta implicata in diversi processi cellulari, quali la fusione della membrana mitocondriale esterna dei mitocondri, il trasporto di questi lungo gli assoni e il tethering tra mitocondri e reticolo endoplasmatico. Ad oggi sono state identificate più di 80 mutazioni associate alla CMT2A e a varianti complicate di CMT2. Tali mutazioni sono nella maggior parte dei casi mutazioni puntiformi a trasmissione autosomica dominante, ma recentemente sono state identificate anche alcune mutazioni a trasmissione semi-dominante o recessiva. Con lo scopo di identificare nuove mutazioni in mitofusina 2 e individuare una possibile correlazione genotipo-fenotipo, nella prima parte di questo progetto si è proceduto a condurre un’analisi genetica del gene in 25 pazienti con diagnosi di CMT assonale. Tale analisi è stata condotta primariamente mediante sequenziamento diretto degli esoni di cui il gene MFN2 è composto. Successivamente, nei soggetti risultati negativi all’analisi per sequenziamento, si è proceduto ad eseguire un’analisi mediante la tecnica della multiplex ligation-dependent probe amplification (MLPA), la quale consente di identificare eventuali riarrangiamenti genici, non identificabili mediante sequenziamento. L’analisi genetica dei pazienti ha permesso di identificare la mutazione causativa in cinque probandi. In quattro di questi sono state trovate mutazioni in eterozigosi, mentre un paziente affetto da una grave forma di CMT assonale è risultato essere un eterozigote composto per due mutazioni in MFN2 (p.[K38del]+[T362M]). Delle mutazioni così rilevate, quattro (E329del, A738V, R94P, K38del) non risultavano prima descritte. La frequenza delle mutazioni di MFN2 nel campione indagato (20%) è in accordo con i dati di letteratura. L’analisi di MFN2 mediante MLPA non ha invece permesso di rilevare alcun riarrangiamento genico. Attualmente lo studio dei meccanismi tramite cui mutazioni in MFN2 inducono neurodegenerazione è limitato dal fatto che solo pochi modelli di CMT2A sono stati sviluppati con successo in topo. Considerando che lo zebrafish si è recentemente distinto come un buon modello per lo studio di molte malattie neurodegenerative, si è deciso di investigare la funzione di mitofusina 2 e il suo ruolo nella patogenesi della CMT2A in tale organismo. Mitofusina 2 è stata quindi silenziata durante lo sviluppo di zebrafish usando un oligo-morfolino antisenso, il quale viene direttamente iniettato in uova fecondate. L’analisi del fenotipo dei morfanti ha permesso di evidenziare come, in assenza di mfn2, gli embrioni presentino evidenti alterazioni fenotipiche ed in particolare risultino avere gravi problemi di movimento, in accordo col fenotipo osservato nei pazienti. Ulteriori analisi hanno evidenziato che il sistema neuromuscolare delle larve è gravemente compromesso, con motoneuroni morfologicamente alterati e una riduzione del numero di giunzioni neuromuscolari correttamente formate. Inoltre, come nei pazienti, le fibre muscolari risultano ipotrofiche e con diametro ridotto, probabilmente a seguito della denervazione dei muscoli dei somiti. Tali risultati, validati da esperimenti di rescue con il trascritto umano di MFN2, confermano il ruolo fondamentale di mitofusina 2 nello sviluppo dei motoneuroni e suggeriscono che lo zebrafish può essere uno strumento utile per lo studio in vivo degli effetti delle mutazioni trovate nei pazienti con CMT2. Per tale motivo si è proceduto a mettere a punto un metodo che consentisse di valutare l’effetto delle mutazioni in MFN2 utilizzando lo zebrafish come modello. Ci si è concentrati in primo luogo sull’analisi della mutazione R94Q, la più frequentemente identificata nei pazienti e per la quale sono disponibili svariate informazioni riguardanti il suo meccanismo molecolare, ottenute sia da indagini in vitro che in vivo. Mediante esperimenti di iniezione dell’mRNA umano mutato, sia in presenza che in assenza del morfolino contro mfn2, si è potuto dimostrare che nei primi stadi dello sviluppo di zebrafish, R94Q è, almeno in parte, loss of function. Tale risultato è in accordo con alcuni dati ottenuti da studi in vitro che dimostrano che tale allele mutato di MFN2 non è in grado di ripristinare la forma del network mitocondriale in cellule derivate da topi knock-out per Mfn2. Inoltre una linea di topi knock-in per R94Q non mostra alcun segno patologico quando la mutazione è in eterozigosi, mentre è letale, ma tardivamente rispetto al knock-out, in omozigosi. Tuttavia un solo meccanismo di loss of function non è in grado di spiegare il fatto che una linea transgenica di topi in cui R94Q è sovraespressa nel sistema nervoso mostri un fenotipo compatibile con la CMT2A a partire dai 5 mesi di vita. I dati così ottenuti, confrontati con quelli di letteratura, suggeriscono che R94Q, pur avendo funzionalità ridotta, potrebbe avere anche un effetto di gain of function, il quale però è in grado di manifestarsi solo dopo lo sviluppo embrionale. Va segnalato che il metodo messo a punto, permette di evidenziare alterazioni solo nei primi stadi dello sviluppo e per tale motivo sarà necessario sviluppare delle linee transgeniche stabili di zebrafish per confermare il meccanismo d’azione di R94Q. Questo potrà permettere di ampliare l’indagine anche valutando gli effetti delle mutazioni in MFN2 rilevate con la presente indagine. Nel complesso i dati ottenuti con questo lavoro hanno contribuito a chiarire la possibile correlazione genotipo-fenotipo in pazienti affetti da CMT2A. Inoltre è stato possibile identificare zebrafish come un nuovo strumento per l’analisi del meccanismo d’azione di mutazioni in MFN2 associate a CMT2. Dato che zebrafish è un organismo modello che si presta meglio di altri a esperimenti di drug screening, la validazione di un sistema modello per la CMT2A in zebrafish potrebbe in futuro contribuire all’identificazione di nuovi farmaci efficaci per il trattamento di tale patologia.

Magíster en Bioquímica, área de especialización Bioquímica de Proteínas Recombinantes
Memoria de Título de Bioquímico
La hormona GLP-1 es una reguladora importante de la homeostasis de la glucosa, favoreciendo su metabolismo en varios tejidos a través de la activación de la vía del AMP cíclico-proteína kinasa A. En las células de músculo liso vascular (VSMC), el control del metabolismo de la glucosa es esencial para la mantención de la función y el fenotipo celular, ya que al favorecer su oxidación en las mitocondrias, se previene la aparición de un fenotipo desdiferenciado, característico de patologías cardiovasculares. Un mecanismo que promueve la actividad mitocondrial es su acoplamiento con el retículo endoplásmico (RE), pues se favorece el traspaso de metabolitos y Ca2+ desde el RE a la mitocondria. Estructuralmente, el acoplamiento entre ambos organelos depende de Mitofusina-2 (Mfn-2). Existe evidencia de que una disminución en el acoplamiento entre RE y mitocondrias en VSMC conduce a la aparición de un fenotipo proliferativo, mientras que el uso de GLP-1 previene el desarrollo de este fenotipo. Además, datos de nuestro grupo de trabajo sugieren que esta hormona potencia la actividad mitocondrial. Basados en estos antecedentes, quisimos estudiar si GLP-1 era capaz de promover el acoplamiento entre RE y mitocondrias en la línea celular vascular A7r5, y evaluar el posible rol de PKA y de Mfn-2 en el fenómeno. A través de inmunofluorescencia indirecta con marcación de RE y mitocondrias, se observó un aumento en la colocalización entre ambos organelos en células estimuladas con GLP-1 (100 nM, 3 h). Además, mediante el uso de la sonda fluorescente Rhod-FF, sensible a Ca2+ mitocondrial, se determinó que la pre-incubación de estas células con GLP-1 favorecía la entrada de Ca2+ reticular a la mitocondria. En forma paralela, se determinó que el tratamiento por 3 h con GLP-1 aumentó los niveles de Mfn-2, efecto que se perdió cuando las células se pre-incubaron con el inhibidor de PKA, H-89. Finalmente, se determinó que el incremento en la colocalización entre RE y mitocondrias y el aumento en la entrada de Ca2+ reticular a las mitocondrias en respuesta al tratamiento con GLP-1, se inhibían en células pre-tratadas con H-89. Así, concluimos que la incretina GLP-1 promueve el acoplamiento funcional entre RE y mitocondrias en células de la musculatura lisa vascular, a través de un mecanismo que requiere de la activación de PKA, y presumiblemente, del aumento de la proteína Mfn-2.
The hormone GLP-1 is an important regulator of glucose homeostasis that promotes its metabolism on several tissues through a cyclic AMP-protein kinase A pathway. In vascular smooth muscle cells (VSMC) glucose metabolism is involved in the control of both cellular function and phenotype, given that promoting its oxidation in the mitochondria prevents the appearance of VSMC undifferentiated phenotype, a hallmark of cardiovascular pathologies. Mitochondrial activity is activated by their coupling with the endoplasmic reticulum (ER), which enhances metabolite and Ca2+ transfer from the ER to the mitochondria. Structurally, the coupling between these organelles depends on Mitofusin-2 (Mfn-2). A decrease on ER-mitochondria coupling in VSMC leads to a proliferative phenotype, while GLP-1 treatment prevents its appearance. Besides, data from our work group suggests that this hormone enhances mitochondrial activity. Based on this background, we evaluated whether GLP-1 modulated functional coupling between ER and mitochondria in the vascular cell line A7r5, and the involvement of PKA and Mfn-2 on this phenomenon. Inmunofluorescence analysis of ER and mitochondria stained cells revealed that GLP-1 (3 h, 100 nM) treatment increased colocalization of these two organelles. Furthermore, by using the mitochondrial Ca2+ sensitive fluorescence probe, Rhod-FF, we determined that pre-incubation of these cells with GLP-1 enhanced reticular Ca2+ entry into the mitochondria. Moreover, the treatment with GLP-1 by 3 h increased Mfn-2 levels, an effect that was prevented by the pre-incubation with the PKA inhibitor, H-89. Finally, the increment of ER and mitochondria colocalization and the increase on reticular Ca2+ entry into the mitochondria in response to GLP-1 stimulation were both abolished in H-89 pre-treated cells. Therefore, we concluded that the hormone GLP-1 promotes ER and mitochondria coupling in VSMC, through a mechanism that requires PKA activation, and presumably, an increment of Mfn-2 levels.
Fondecyt

Les mitochondries forment un réseau très dynamique remodelé par deux processus antagonistes appelés : fusion et fission mitochondriales. Chez l’homme, une altération de ces processus, sont à l’origine de nombreuses maladies qui affectent essentiellement le système nerveux. L'objectif principal des travaux de ma thèse était de caractériser l'impact d'un déséquilibre entre la fusion et la fission mitochondriale dans le contexte d'une neuropathie héréditaire : la maladie de Charcot-Marie-Tooth de type 2A (CMT2A), qui est causée par des mutations dominantes dans la mitofusine MFN2. Dans le but d’étudier les mécanismes à l’origine de cette maladie, j’ai développé le premier modèle drosophile de CMT2A en exprimant dans les neurones de mouches quatre allèles de mitofusine retrouvés fréquemment chez les patients. De manière surprenante, les différents allèles altèrent très différemment la morphologie mitochondriale. En effet, alors que les mutations associées au domaine GTPase inhibent la fusion et agrègent les mitochondries, les mutations du domaine dit HB1 induisent au contraire un excès de fusion. J’ai pu ensuite déterminer que l’agrégation des mitochondries et l’excès de fusion, conduisent de manière commune à un défaut de transport des mitochondries au niveau des synapses et à une altération du métabolisme oxydatif associée à une accumulation de mutation dans l’ADN mitochondrial. Chez les drosophiles exprimant des allèles dominants actifs de mitofusine, tous ces dysfonctionnements disparaissent lorsqu’on augmente la fission suggérant que la pathogénicité des allèles du domaine HB1 résulte d’un déséquilibre de la balance entre fusion et fission en faveur de la fusion
Mitochondria form a dynamic network remodeled by two antagonistic processes called mitochondrial fusion and fission. While mitochondrial fusion creates interconnections between mitochondria, mitochondrial fission result in fragmentation. These processes are mediated by Dynamin-related GTPases, the outer-membrane fusion protein mitofusin, and the fission factor DPR1.The main aim of my resaearch was to characterize the impact of an imbalance between mitochondrial fusion and fission in neurons in the context of a severe hereditary neuropathy called Charcot-Marie-Tooth type 2A (CMT2A). Indeed, this disease is caused by dominant mutations in the mitofusinMFN2.In order to dissect the mechanisms by which these mutations alter mitofusin properties and neuronal function, we developed four drosophila models of CMT2A expressing the two most frequent substitutions (R94Q, R364W) and two others localizing to similar domains (T105M, L76P). The four alleles resulted in mitochondrial depletion at neuromuscular junctions, decreased oxidative metabolism, increased mtDNA mutations, and impaired locomotion that were associated with aberrant mitochondrial morphology. Interestingly, while GTPase domain-associated mutations (R94Q, T105M) aggregate unfused mitochondria, mutations within helix bundle 1 (R364W, L76P) unexpectedly enhance mitochondrial fusion, as demonstrated by rescue of mitochondrial morphology and locomotion provided by the DRP1 fission factor. In conclusion, we show that both dominant negative and dominant active forms of mitofusin can cause CMT2A, and propose for the first time that excessive mitochondrial fusion drives CMT2A pathogenesis in a large number of patients

Il est bien accepté dans des modèles en culture que les dynamiques mitochondriales et le remodelage des crêtes régulent le fonctionnement mitochondrial sous diverses conditions de stress, particulièrement l’apoptose et la famine. Malgré la quantité impressionnante de recherche effectuée dans ce domaine, on en connait encore très peu au sujet de l’importance des dynamiques mitochondriales et du remodelage de la structure mitochondriale sous des conditions physiologiques. Dans les années 1960, Hackenbrock a démontré que des mitochondries isolées adoptent des conformations internes distinctes selon l’état métabolique. D’après ses observations, il a prédit que les changements ultrastructurels de la mitochondrie régulent la production fonctionnelle de l’organite. Cependant, il n’est pas évident que ces changements ultrastructuraux suivent bien les changements métaboliques in vivo dans des conditions physiologiques. De plus, le métabolisme hépatique nécessite une adaptation constante de la production bioénergétique et biosynthétique de la mitochondrie suite aux changements de l’état anabolique/catabolique de la cellule hépatique. Toutefois, le fonctionnement de ce processus est encore largement inconnu. Dans cette étude, nous apportons les premières descriptions quantitatives in vivo de la réponse adaptative du réticulum mitochondrial aux transitions métaboliques du foie. Grâce à un modèle hépatique de souris postprandiale et une analyse cryo- microscopie électronique (cryo-EM) quantitative, nous montrons que, 5 heures après un repas, la voie mTORC1 est bloquée, le réseau mitochondrial se fragmente, la densité des crêtes diminue et la capacité respiratoire des mitochondries chute. Ces changements sont accompagnés d’une augmentation parallèle de la longueur des contacts mitochondrie-réticulum endoplasmique (MERCs), qui contrôle les échanges de calcium et de phospholipides entre les deux organites. De plus, ces évènements sont associés à l’expression transitoire de deux fragments C-terminaux (CTFs) inconnus jusqu’à présent provenant de la protéine Optic atrophy-1 (OPA1), une GTPase qui régule les dynamiques des crêtes mitochondriales et des mitochondries. Grâce à un protocole in vitro, nous montrons que ces CTFs proviennent d’un nouveau clivage d’OPA1, appellé clivage-C, qui élimine l’activité d’OPA1 en la coupant. Plus important encore, nous montrons que le clivage-C nécessite la présence de Mitofusin-2 (MFN2), une protéine clé dans la régulation de la fusion mitochondriale et dans la génèse des MERCs, mais pas la présence de l’homologue Mitofusin-1 (MFN1), ce qui confirme le lien entre le remodelage des crêtes et l’assemblage des MERCs.
It is well established in cultured models that mitochondrial dynamics and cristae remodeling regulate mitochondrial function under different stress conditions, such as starvation and apoptosis. Despite the tremendous amount of research in this field, relatively little is known about the significance of mitochondrial dynamics and ultrastructure remodeling under normal physiological conditions in vivo. In the 1960’s, Hackenbrock demonstrated that isolated mitochondria adopt distinct internal conformations under different metabolic states. Based on these observations, he predicted that mitochondrial ultrastructural changes regulate the organelles functional output. However, whether these ultrastructural changes also accompany metabolic transitions in vivo, under physiological conditions, is not known. Further, hepatic metabolism requires mitochondria to adapt their bioenergetic and biosynthetic output to the ever-changing anabolic/catabolic state of the liver cell, but the wiring of this process is still largely elusive. In this study, we provide the first in vivo quantitative description of the adaptive response of the mitochondrial reticulum to hepatic metabolic transitions. Using a postprandial mouse liver model and quantitative cryo-EM analysis we show that at 5 hours after feeding the mTORC1 signaling is blocked, the mitochondria network fragments, the cristae density decreases and the mitochondrial respiratory capacity drops. These changes are accompanied with a parallel increase in the mitochondria-ER contact (MERCs) lengths, which control calcium and phospholipids fluxes between the two organelles. Further, these events are associated with the transient expression of two previously unidentified C-terminal fragments (CTFs) of Optic atrophy-1 (OPA1), a mitochondrial GTPase that regulates cristae and mitochondrial dynamics. Using an in vitro assay, we show that these CTFs originate from a novel OPA1 processing, termed C-cleavage that eliminates OPA1 activity by breaking off the GTPase. Importantly, we show that C-cleavage requires the presence of Mitofusin-2 (MFN2), a key regulator of mitochondria fusion and MERCs biogenesis, but not that of its homolog Mitofusin-1 (MFN1), thereby linking cristae remodeling to MERCs assembly.

Mitochondria form a tubular, reticulated network which shape is controlled by opposing fusion and fission events (Bereiter-Hahn and Voth, 1994). The mitofusins 1 and 2 (Mfn1 and Mfn2) are conserved, dynamin-like GTPases embedded in the outer mitochondrial membrane (OMM) that mediate mitochondrial fusion in coordination with OPA1 (Rojo et al., 2002; Santel and Fuller, 2001; Wong et al., 2000). Mitochondrial shaping proteins have pleiotropic functions. In particular, while Mfn1 seems primarily involved in organellar docking and fusion, Mfn2 is enriched at contact sites between ER and mitochondria where it is implicated in the formation of molecular linkers that are capable of organelles tethering (Chen et al., 2012; de Brito and Scorrano, 2008). Recent works attributed to these points of close contact between the OMM and the nearby ER, called MAMs (mitochondria-associated ER-membranes) or MERCs (mitochondria-ER contacts), an important role in the propagation of cellular signals, including those that control lipid metabolism, calcium (Ca2+) homeostasis and cell death (Rowland et al., 2012; Rizzuto et al., 1998; Vance, 1990). Indeed, aberrations in ER-mitochondria juxtaposition have been described in cellular models of different neurodegenerative diseases, including Alzheimer's, Huntington's and Parkinson's disease (Krols et al., 2016; Calì et al., 2013; Ottolini et al., 2013; Area-Gomez et al., 2012; Calì et al., 2012; Panov et al., 2002). Although the exact cause for neuronal loss is not clear Parkin, an E3-ubiquitin ligase mutated in familiar Parkinson's Disease (PD) is selectively recruited to dysfunctional mitochondria and promotes their elimination via autophagy, a process known as mitophagy (Narendra et al., 2008). PINK1, a protein kinase, also a PD related gene, is required for Parkin recruitment and stress induced mitophagy (Ziviani et al., 2010). In several model systems, Parkin selectively ubiquitinates the mitochondrial outer membrane profusion protein Mfn1 and Mfn2 and fly homologue Marf (Tanaka et al., 2010; Ziviani et al., 2010; Gegg et al., 2010). Accordingly, lack of Parkin or PINK1, which operates upstream Parkin in the same pathway, results in impaired ubiquitination of Mfn and increased levels of Mfn (Ziviani et al., 2010). Given that Parkin affects Mfn steady state and ubiquitination levels, we propose to (i) address the ubiquitination levels of Mfn2 and whether Parkin downregulation affects it; (ii) investigate whether Parkin regulates ER-mitochondria tethering by impinging on Mfn2 steady state and ubiquitination levels; (iii) evaluate the physiological significance of ER-mitochondria interaction in an in vivo animal model of Parkinson's disease. Our hypothesis is that Parkin dependent Mfn ubiquitinatination controls ER-mitochondria tethering, thus impinging on Ca2+ transfer and Ca2+ homeostasis, which dysregulation has been described in a number of molecular pathways leading to PINK1 and Parkin loss of function dependent neurodegeneration (Calì et al., 2013; Ottolini et al., 2013; Calì et al., 2012). In order to address the previously listed hypothesis we analysed the pattern of ubiquitination of Mfn2 in mouse embryonic fibroblasts (MEFs) upon downregulation of Parkin. To this aim, we (i) immunoprecipitated Mfn2 with specific anti Mfn2 antibody and performed western blotting analysis with specific anti HA antibody in cells overexpressing HA tagged Ubiquitin; (ii) measured the degree of tethering between ER and mitochondria in control and Parkin downregulating cells. We used two independent approaches to measure ER-mitochondria tethering: we first measured the percentage of ER co-localizing with mitochondria by using Mander's coefficient of co-localization upon volume-rendered 3D reconstruction of z-axis stacks of confocal images of cells expressing organelles targeted fluorescence probes (mito-RFP and ER-YFP, respectively) (Rizzuto et al., 1998). Secondly, we took advantage of a FRET based probe (Naon et al., 2016) to measure ER-mitochondria proximity. In this sensor, called FEMP, FRET intensity is inversely proportional to the distance between the two fluorophores (mito-YFP and ER-CFP) that are appropriately targeted to the two compartments. (iii) We investigated the physiological significance of ER-mitochondria tether in an in vivo animal model of PD that lacks PINK1 expression. To this aim we used the fruitfly Drosophila melanogaster, which has many advantages. First, fly mutants deriving from loss of function mutations of PINK1 have been extensively characterized and cause a robust phenotype represented by age-related degeneration of DA neuron loss and locomotor deficits (Poole et al., 2008; Clark et al., 2006; Park et al., 2006; Yang et al., 2006; Wang et al., 2006). Secondly, a variety of genetic modifications and epistasis experiments can be easily performed in vivo to dissect molecular pathways. Our results showed that Parkin downregulation reduced Mfn ubiquitination and ER-mitochondria tethering in MEFs. Interestingly, we found that the pattern of Mfn2 ubiquitination and ER-mitochondria tethering is also impaired in CMT type 2A disease-associated Mfn2 mutations (Mfn2R94Q, Mfn2P251A and Mfn2R280H respectively). Although indirectly, these findings strongly suggested that ubiquitination of Mfn2, rather than its steady state levels, is important in the regulation of ER-mitochondria tethering. To identify the precise site of Mfn2 ubiquitination and directly link lack of ubiquitination with impaired ER-mitochondria tether, we took advantage of a bioinformatics approach to identify among species-highly conserved lysine (K) residues. We identified twenty Lysine residues that were conserved between human, mouse and fly. We compared these residues with those identified by a mass spectrometry-based study published in 2014 (Bingol et al., 2014) that identified Parkin-dependent ubiquitination sites. We identified six Lysine residues that were likely to represent good candidates for Parkin-dependent ubiquitination of Mfn2. We generated non-ubiquitinatable mutants for those sites by substituting Lysine (K) with Arginine (R), a common procedure to impair ubiquitination and investigated the pattern of ubiquitination of the non-ubiquitinatable Mfn2 mutants by western blotting. Expression of non-ubiquitinable mutant K416R resulted in impaired Mfn2 ubiquitination. Of note this mutant was also unable to correct ER-mitochondrial contacts when expressed in Mfn2 KO MEFs and only partial restored ER-mitochondrial Ca2+ transfer. In summary, our results provided strong evidences that Mfn2 ubiquitination is a prerequisite for ER-mitochondria physical and functional interaction and that K416 in the HR1 domain of Mfn2 is a genuine site for Parkin dependent ubiquitination. A number of studies have shown impaired Ca2+ homeostasis in cellular models lacking PINK1 or Parkin (Heeman et al., 2011; Sandebring et al., 2009). Although it is not clear why dopaminergic neurons specifically degenerate in PD, it is tempting to hypothesis that impaired Ca2+ homeostasis resulting from impaired Ca2+ cross talk at ER-mitochondria interface could lead or contribute to degeneration. Elegant studies have shown that artificial tether between ER and mitochondria can be used to modulate Ca2+ transfer (Csordas et al., 2010; Csordas et al., 2006). With that in mind, we addressed whether expressing an ER-mitochondria synthetic linker in a well-established in vivo Drosophila model of PINK1 loss of function could ameliorate PINK1 KO phenotypes by impinging on ER-mitochondria cross talk. We therefore generated a number of fly lines expressing the synthetic linker driven by a neuron-specific driver in the fly wing neurons. This linker was generated by Csordas et al. (Csordas et al., 2006) and consists of a monomeric fluorescent protein (RFP) fused to the outer mitochondrial membrane targeting sequence at the N terminus and fused to the ER targeting sequence at the C terminus. We could observe a well-defined and easily quantifiable RFP-fluorescence spots throughout the L1 vein of the fly wing that perfectly matched the morphology seen when expressing mito-GFP or ER-GFP alone in the wing neurons (Vagnoni and Bullock, 2016), which indicated that the synthetic linker was appropriately expressed. Interestingly, we found an amelioration of PINK1 KO climbing ability upon expression of the artificial synthetic linker. This result strong indicates that restoration of proper ER-mitochondrial communication in PINK1 KO background can be beneficial in ameliorating the phenotype associated to an in vivo animal model of PD, paving the way for novel approaches for medical intervention.
I mitocondri formano un network reticolare e tubulare la cui forma è controllata da eventi opposti di fusione e fissione (Bereiter-Hahn and Voth, 1994). Le mitofusine 1 e 2 (Mfn1 e Mfn2), sono delle GTPasi dynamin-like incorporate nella membrana mitocondriale esterna (OMM, outer mitochondrial membrane) e mediano la fusione mitocondriale in cooperazione con OPA1 (Rojo et al., 2002; Santel and Fuller, 2001; Wong et al., 2000). Le shaping protein mitocondriali hanno una funzione pleiotropica. In particolare, mentre la Mfn1 sembra principalmente coinvolta nel docking e nella fusione di organelli, la Mfn2 è arricchita nei punti di contatto tra ER e mitocondri dove è implicata nella formazione di collegamenti molecolari che sono capaci di produrre un'interazione tra gli organelli (Chen et al., 2012; de Brito and Scorrano, 2008). Lavori recenti attribuiscono a questi punti di stretto contatto tra l'OMM e il vicino ER, chiamati MAMs (mitochondria-associated ER-membranes) o MERCS (mitochondria-ER contacts), un importante ruolo nella propagazione del segnale cellulare, incluso quello che controlla il metabolismo lipidico, l'omeostasi del calcio (Ca2+) e la morte cellulare (Rowland et al., 2012; Rizzuto et al., 1998; Vance, 1990). Un'anomalia nella comunicazione tra ER e mitocondri è stata descritta in vari modelli cellulari di differenti malattie neurodegenerative, che includono la malattia di Alzheimer, Huntington e Parkinson (Krols et al., 2016; Calì et al., 2013; Ottolini et al., 2013; Area-Gomez et al., 2012; Calì et al., 2012; Panov et al., 2002). Tuttavia la causa esatta che induce la perdita neuronale non è ancora conosciuta. Parkin, una E3-ubiquitina ligasi mutata nelle forme familiari di malattia di Parkinson (PD, Parkinson's disease) è selettivamente reclutata sui mitocondri disfunzionali e promuove la loro eliminazione tramite autofagia, un processo conosciuto come mitofagia (Narendra et al., 2008). PINK1, una proteina chinasica e gene associata alla PD, è richiesto per il reclutamento di Parkin e per la mitofagia indotta da stress (Ziviani et al., 2010). Nei diversi sistemi modello, Parkin ubiquitina selettivamente le proteine ancorate sulla membrana mitocondriale esterna che promuovono la fusione (Mfn1 e Mfn2) e il loro omologo in Drosophila (Marf) (Tanaka et al., 2010; Ziviani et al., 2010; Gegg et al., 2010). Di conseguenza, l'assenza di Parkin o PINK1, il quale opera a monte di Parkin nella stessa pathway, causa un'alterazione nell'ubiquitinazione della Mfn e un aumento dei livelli di Mfn (Ziviani et al., 2010). Dato che Parkin altera i livelli basali e di ubiquitinazione della Mfn, noi abbiamo proposto di (i) valutare i livelli di ubiquitinazione della mitofusina ed analizzare se la downregolazione di Parkin ha effetti su questi livelli; (ii) investigare se Parkin regola il legame tra ER e mitocondri andando ad agire sui livelli stazionari o di ubiquitinazione della Mfn2; (iii) valutare il significato fisiologico dell'interazione ER-mitocondri in un modello animale in vivo di malattia di Parkinson. La nostra ipotesi è che l'ubiquitinazione della Mfn Parkin-dipendente controlli il legame ER-mitocondri, interferendo così con il trasferimento e l'omeostasi di Ca2+, la cui alterazione è stata descritta in numerose pathway molecolari che causano neurodegenerazione associata alla perdita di funzionalità  di PINK1 e Parkin (Calì et al., 2013; Ottolini et al., 2013; Calì et al., 2012). Al fine di affrontare la precedente lista di ipotesi abbiamo analizzato i livelli di ubiquitinazione della Mfn2 in fibroblasti embrionali di topo (MEFs: mouse embryonic fibroblasts) in seguito alla downregolazione di Parkin. A questo scopo, abbiamo (i) immunoprecipitato la Mfn2 con lo specifico anticorpo anti Mfn2 ed eseguito il western blotting con lo specifico anticorpo anti HA in cellule overesprimenti l'ubiquitina taggata HA; (ii) misurato i livelli di connessione tra ER e mitocondri nelle cellule di controllo e in cellule con Parkin downregolato. Abbiamo utilizzato due approcci indipendenti per misurare questa connessione: per prima cosa abbiamo misurato la percentuale di co-localizzazione dell'ER con i mitocondri usando il coefficiente di co-localizzazione di Mander in seguito alla ricostruzione volumetrica 3D delle immagini confocali lungo l'asse z di cellule esprimenti le sonde fluorescenti bersaglio degli organelli (rispettivamente, mito-RFP and ER-YFP) (Rizzuto et al., 1998). In secondo luogo, abbiamo sfruttato una sonda basata su FRET (Naon et al., 2016) per misurare la vicinanza dell'ER con i mitocondri. In questo sensore, chiamato FEMP, l'intensità  di FRET è inversamente proporzionale alla distanza tra i due fluorofori (mito-YFP e ER-CFP) che sono opportunamente indirizzati ai due compartimenti. (iii) Abbiamo studiato il significato fisiologico dell'interazione ER-mitocondrio in un modello animale in vivo di PD privo dell'espressione di PINK1. A questo scopo abbiamo usato la Drosophila melanogaster, che ha molti vantaggi. Innanzitutto i mutanti di Drosophila, derivati da mutazioni che causano la perdita di funzionalità  di PINK1, sono stati ampiamente caratterizzati ed inducono un fenotipo robusto rappresentato dalla perdita dei neuroni DA e deficit locomotori correlati all'età  (Poole et al., 2008; Clark et al., 2006; Park et al., 2006; Yang et al., 2006; Wang et al., 2006). In secondo luogo, in vivo possono essere facilmente eseguiti un'ampia varietà  di modifiche genetiche ed esperimenti di epistasi per comprendere più approfonditamente le pathway molecolari. I nostri risultati mostrano che in MEFs la downregolarione di Parkin riduce l'ubiquitinazione della Mfn e l'interazione tra ER e mitocondri. Abbiamo osservato inoltre che le mutazioni della Mfn2 associate a CMT di tipo 2A (rispettivamente Mfn2R94Q, Mfn2P251A e Mfn2R280H) causano l'alterazione dei livelli di ubiquitinazione della Mfn2 ed una diminuzione nell'interazione tra ER e mitocondri. Sebbene indirettamente, questi risultati suggeriscono fortemente che l'ubiquitinazione della Mfn2, piuttosto che i livelli stazionari, è importante nella regolazione dell'interazione tra ER e mitocondri. Per identificare il sito preciso di ubiquitinazione della Mfn2 e correlare direttamente la mancanza dell'ubiquitinazione con la riduzione dell'interazione ER-mitocondri, abbiamo sfruttato un approccio bioinformatico che ci ha permesso di individuare i residui di lisina (K) altamente conservate nelle varie specie. Abbiamo identificato venti residui di lisina che sono conservati tra uomo, topo e Drosophila. Abbiamo confrontato questi residui con quelli descritti in uno studio basato sull'utilizzo della spettrometria di massa per identificare i siti di ubiquitinazione dipendenti da Parkin pubblicato nel 2014 (Bingol et al., 2014). Abbiamo identificato sei residui di lisina che potrebbero rappresentare dei buoni candidati per l'ubiquitinazione Parkin-dipendente della Mfn2. Abbiamo generato i mutanti non ubiquitabili per questi siti sostituendo la lisina (K) con l'arginina (R), una procedura comune per bloccare l'ubiquitinazione e studiato il pattern di ubiquitinazione dei mutanti Mfn2 non ubiquitinabili mediante western blotting. L'espressione del mutante non ubiquitinabile K416R ha provocato un'alterazione dell'ubiquitinazione della Mfn2. Da notare che questo mutante non è stato in grado di ripristinare i contatti ER-mitocondri quando reintrodotto in MEF Mfn2 KO ed ha restaurato solo parzialmente il trasferimento di Ca2+ ER-mitocondriale. In breve, i nostri risultati hanno fornito prove evidenti che l'ubiquitinazione di Mfn2 è un prerequisito per l'interazione fisica e funzionale dei mitocondri con l'ER e che la K416 nel dominio HR1 della Mfn2 è un vero e proprio sito per l'ubiquitinazione Parkin-dipendente. Vari studi hanno osservato in diversi modelli cellulari privi di PINK1 o Parkin un'alterata omeostasi del Ca2+ (Heeman et al., 2011; Sandebring et al., 2009). Sebbene non sia chiaro il motivo per il quale nel PD degenerino specificatamente i neuroni dopaminergici è allettante ipotizzare che un deficit nell'omeostasi del Ca2+, risultante da un alterato scambio di Ca2+ nell'interfaccia ER-mitocondrio, possa portare o contribuire alla degenerazione. Eleganti studi hanno dimostrato che un legame artificiale tra ER e mitocondri può essere usato per modulare il trasferimento di Ca2+ (Csordas et al., 2010; Csordas et al., 2006). Tenendo questo a mente, ci siamo occupati del fatto che l'espressione di un legante sintetico tra ER e mitocondri in un modello in vivo di Drosophila PINK1 loss of function potesse migliorare il fenotipo dei PINK1 KO incidendo sulla comunicazione ER-mitocondriale. Per questo motivo abbiamo generato un certo numero di linee di Drosophila esprimenti il linker sintetico guidato da uno specifico driver neuronale nei neuroni delle ali. Questo linker è stato generato da Csordas et al. (Csordas et al., 2006) e consiste in una proteina monomerica fluorescente (RFP) fusa all'N terminale con la sequenza bersaglio della membrana mitocondriale esterna e fusa al C-terminale con la sequenza bersaglio dell'ER. Abbiamo potuto visualizzare e quantificare i punti di fluorescenza RFP lungo la vena L1 nell'ala della Drosophila dimostrando una corrispondenza con la morfologia osservata a seguito dell'espressione di mito-GFP o ER-GFP sui neuroni dell'ala (Vagnoni e Bullock, 2016), indice del fatto che l'espressione del linker sintetico era appropriata. E' interessante notare che abbiamo osservato un miglioramento dell'abilità  di arrampicata della Drosophila PINK1 KO in seguito all'espressione del linker sintetico artificiale. Questo risultato indica che il ripristino della corretta comunicazione tra ER e mitocondri nel background PINK1 KO può essere utile per migliorare il fenotipo associato ad un modello animale in vivo di PD, aprendo la strada a nuovi approcci per l'intervento medico.

碩士
國立中興大學
生物醫學研究所
93
The accumulated evidences show that the dynamic behaviors (fission and fusion) of mitochondria are crucial for many cellular functions. Mitochondrial fusion serves to mix and to unify the mitochondrial compartment, which are important in cellular aging, development, energy dissipation and mitochondrial DNA inheritance. Mitofusins, proteins that are expressed on mitochondrial outer membrane, have been identified as key components of mitochondrial fusion machinery in mammalian cells. Moreover, these proteins were shown to mediate cell growth and development. However, the mechanism how mitofusins affect cancer cell growth is still unclear. In this study, we investigated the expression of mitofusins in non-small cell lung cancer. Our results showed that mitofusins were overexpressed in 60% of tumor specimens by RT-PCR, in situ hybridization, Western blot and immunohistochemistry. Besides the predicted hMfn1 (84 kDa), we also detected an extra protein with molecular weight about 100 kDa by Western blot. Although we highly expect that it may be due to a post-translational modification. However, it awaits further determination.

We have investigated the role of mitofusin proteins in mitochondrial fusion and Charcot-Marie-Tooth disease Type 2A (CMT2A). Mitofusins (Mfn1 and Mfn2) are required for mammalian mitochondrial fusion. In structure-function analysis, we have identified loss-of-function mutations in mitofusin GTPase and heptad-repeat domains that disrupt homotypic and heterotypic domain interactions. Mutations in Mfn2 cause CMT2A, a progressive peripheral neuropathy. We have functionally characterized Mfn2 disease mutations and find that wild-type Mfn1, but not Mfn2, can efficiently complement nonfunctional CMT2A alleles to restore mitochondrial fusion. This finding demonstrates the importance of Mfn1-Mfn2 heterooligomers and suggests that Mfn1 expression is important in determining the cell-type specificity of CMT2A. To study the consequences of an Mfn2 CMT2A allele in vivo, we generated transgenic mice that express Mfn2 T105M in motor neurons. These animals demonstrate gait impairments due to distal muscle loss, axonopathy and altered mitochondrial morphology and distribution in motor neurons. In a second approach, we have generated CMT2A knock-in mice by replacing the endogenous genomic Mfn2 with Mfn2 alleles L76P or R94Q. Preliminary characterizations suggest that heterozygous animals have no disease symptoms, but homozygous Mfn2 R94Q animals are severely affected. Together, these mouse models provide means to assess the pathology of Mfn2 CMT2A alleles and the role of mitochondrial dynamics in vivo.

Mitofusin 2 (MFN2), an outer mitochondrial membrane protein expressed in virtually all human tissues, is a multi-faceted protein known to affect mitochondrial morphology, metabolism, tethering, and movement as well as overall cell cycle progression. Most intriguing among its characteristics is its ability to bind to Ras and Raf, upstream effectors in the MAPK/ERK pathway. Conditional knockout (cKO) of renal proximal tubule MFN2 in vivo showed a post-ischemic protective effect. While the two day survival of control mice was only 28%, an unexpected 86% of the MFN2 cKO mice were alive at two days post-ischemia. This is likely explained by MFN2's ability to bind and sequester Ras at baseline. Because the MFN2 deficient mice did not sequester as much Ras, renal proximal tubule cells were able to proliferate at a greater rate and restore organ function more quickly. Immunoprecipitation studies confirm a strong interaction between Ras and MFN2 in resting cells but a weaker one immediately following ischemic insult, even in cells replete with MFN2. These results suggest that blocking the MFN2-Ras interaction may be a novel method to treat acute kidney injury. A small peptide mimicking Ras to block MFN2 could be feasible. This should grant ischemic tissue an increased propensity to regenerate healthy cells while leaving non-ischemic tissue completely unaffected. Such a therapeutic agent would be novel in the treatment of acute kidney injury and may have uses in other tissues as well due to MFN2's widespread expression profile.

Das Herz ist physiologisch auf einen fein regulierten und ausgeglichenen bioenergetischen Energiehaushalt angewiesen, um auf akute Belastungssituationen adäquat reagieren zu können und oxidativen Stress zu vermeiden. Ca2+ reguliert zentral sowohl die zyklischen Kontraktions-/Relaxationsprozesse (ECC) als auch unmittelbar den mitochondrialen Metabolismus. Der ECC liegt in den Kardiomyozyten die Ca2+- Freisetzung durch die RyR2 zu Grunde; die IP3 Rezeptoren des sarkoplasmatischen Retikulums (SR) führen davon unabhängig zu einer Ca2+ Freisetzung aus dem SR. Diese IP3R vermittelten Signale werden in den räumlich nahe gelegenen Mitochondrien zum Teil über den mRyR1 in die mitochondriale Matrix aufgenommen und stimulieren dort langfristig die oxidative Phosphorylierung und den Erhalt der antioxidativen Kapazität. Die enge räumliche Nähe zwischen SR und Mitochondrien wird durch Strukturproteine wie Mitofusin 2 (Mfn2) ergänzt, die das SR mit der äußeren Mitochondrienmembran koppeln und so die Ca2+-Interaktion beeinflussen. Ziel der Arbeit war, den Effekt von Mfn2 Defizienz auf die IP3 induzierte mitochondriale Ca2+-Regulation in Kardiomyozyten zu evaluieren. Dazu erfolgten Fluoreszenzfärbungen an adulten isolierten Ventrikelkardiomyozyten kardiospezifischer Mfn2 Knock-Out (KO) Mäusen bzw. deren wildtypischen Geschwistertieren (WT). Erhobene Parameter umfassten das mitochondriale Ca2+, das mitochondriale Membranpotenzial, die mitochondriale Superoxidbildung und mitochondriale ATP-Gehalt. Die Ergebnisse bestätigten eine Signalachse, bei der die Stimulation von isolierten murinen Kardiomyozyten mit dem IP3 Agonisten ET-1 zu einer mitochondrialen Ca2+ Aufnahme führte, dem Erhalt des mitochondrialen Membranpotenzials diente und der ATP Gehalt stiegt. Bei induzierter kardiospezifischer Ablation von Mfn2 geht diese SR-mitochondriale Interaktion verloren, und es entstand ein energetisches Defizit sowie eine verminderte Superoxidbildung. Bei beta-adrenerger Stimulation mit Isoproterenol (ISO) resultierte in WT zwar eine mitochondriale Ca2+-Aufnahme, allerdings ein Abfall des ATP-Gehaltes. In den Mfn2 defizienten Kardiomyozyten zeigte sich eine Steigerung des ATP-Gehaltes auch auf beta-adrenerge Stimulation, die einen energetischen Kompensationsmechanismus in den Mfn2 KO Tieren vermuten lässt. Dies identifiziert Mfn2 als kritische Strukturkomponente für die basale bioenergetische Adaptation der durch IP3R-mRyR1 vermittelten Signalachse unter physiologischen Bedingungen
Under physiological conditions the heart needs a finely tuned bioenergetic adaptation system to adequately match sudden changes in the workload and to avoid oxidative stress. Ca2+ regulates the excitation-contraction-coupling (ECC) as well as the mitochondrial metabolism. The ECC is based on the release of Ca2+ via the RyR2 while the IP3 receptor (IP3R) releases Ca2+ independently from the sarcoplasmatic reticulum (SR). The signals from the latter are taken up by the surrounding mitochondria via the mRyR1 channel to stimulate both the basal oxidative phosphorylation and the antioxidative capacity. The close functional relationship between mitochondria and SR is affected by membrane-coupling proteins like mitofusin 2 (Mfn2) that may influence the Ca2+ transmission. This work aimed at evaluating the effect of Mfn2 deficiency on the IP3-induced mitochondrial calcium regulation in cardiomyocytes. Mitochondrial Ca2+ uptake, membrane potential, redox state and ATP generation were monitored in isolated ventricular cardiomyocytes of cardio-specific mitofusin 2 Knock-out (KO) mice and their wildtype littermates (WT) via fluorescent staining using laser scanning confocal microscopy. The results show that stimulation with the IP3 agonist ET-1 led to mitochondrial calcium uptake, ATP generation and maintained mitochondrial membrane potential. The cardio-specific loss of the tethering protein Mfn2 resulted in an energetic deficit and decreased levels of superoxide. Beta adrenergic receptor activation with isoproterenol (ISO) in WT resulted in a mitochondrial calcium uptake but decreased ATP content, while leading in Mfn2 KO cardiomyocyte to increased levels of ATP, pointing probably towards an energetic compensatory mechanism. Taken together these results propose Mfn2 as a critical structural component that affects under physiological conditions the privileged SR-mitochondrial metabolic feedback mechanism via IP3R and mRYR1 to maintain normal cardiac function and bioenergetics

碩士
國立中興大學
生命科學院碩士在職專班
98
Mitochondria are key players in several cellular functions including growth, division,energy metabolism, and apoptosis. Mitochondrial dysfunction contributes to a number of human disorders and may aid cancer progression. Although normal mitochondrial morphology is important for the function and physiology of cells, defects in mitochondrial dynamics may cause severe diseases. Fusion and fission of mitochondria regulate their morphology and distribution. Mitofusin-2（Mfn-2）is an outer mitochondrial membrane protein. Recent observations indicate that Mfn2 is a multifunctional membrane protein that participates in cell proliferation and metabolism and that it is required for normal endoplasmic reticulum morphology. The roles of Mfn-2 in cancers are largely unknown. In this study, which aims at defining the role of Mfn-2 in lung cancer, we analyzed the expression of Mfn-2 in non-small cell lung cancer（NSCLS） using immunohistochemistry . Mfn-2 expression that was detected in the cytoplasm in average immunostaining of Mfn2 was 80％ for tumor and 20％ for non-tumor（Pearson Chi-Square；p＜0.05）.We analyzed the expression of Mfn-2 in tumor and non-tumor pairs by using western blot. Increased expression of Mfn-2 in malignant cells compared to the non-tumor cells suggests that Mfn-2 expression may play a role in NSCLS aggravation. It may be a novel marker for diagnosis and treatment of NSCLS.

The recent surge in Type 2 Diabetes (T2D) has renewed interest in the study of cellular metabolism – which mitochondria tightly control. Previous work has shown mitochondrial dysfunction plays a critical role in the development of metabolic diseases, such as T2D. The pancreatic β-cell synthesizes and secretes insulin in vivo in response to diverse fuel signals such as glucose, fatty acids, and amino acids; failure or loss of β-cell mass is a hallmark of T2D. Pancreatic β-cell mitochondria are dynamic organelles living a life of fusion, fission, and movement collectively called mitochondrial dynamics. Mitochondrial fusion is impaired in obesity and models of obesity, while basal secretion of insulin is elevated. Previous studies demonstrate that hyperinsulinemia alone is sufficient to induce insulin resistance, yet the relationship between mitochondrial morphology and basal insulin secretion has not yet been studied. Here, we investigated the link between loss of mitochondrial fusion and insulin secretion at basal glucose concentrations by reducing the expression of mitofusin 2 (Mfn2), which controls mitochondrial morphology and metabolism. We found that forced mitochondrial fragmentation caused increased insulin secretion at basal glucose concentrations. In addition, fragmentation of mitochondria enhanced the secretory response of islets to palmitate at nonstimulatory glucose concentrations and increased fatty acid uptake and oxidation in a cell model of pancreatic β-cells. We developed unique solutions to challenges posed by the measurement of mitochondrial dynamics via confocal microscopy by using novel image analysis techniques, including a novel method of mitochondrial segmentation. This technique also revealed novel biology of brown adipose tissue mitochondria dependent on their localization within the cell. Our findings demonstrate that changes to mitochondrial dynamics in the β-cell can lead to increased insulin secretion at basal glucose concentrations. These data support the possibility that hyperinsulinemia and the downstream outcome of insulin resistance can be initiated by altered mitochondrial function in the β-cell independently of other tissues. By uncovering a new process that governs basal insulin secretion, we provide novel targets for regulation, such as mitochondrial morphology or fatty acid induced insulin secretion that may present new approaches to treatment of diabetes.

碩士
國立臺灣大學
分子醫學研究所
105
Tissue-specific stress responses are protective mechanisms against proteotoxic stress and could be regulated in a non-autonomous fashion. Inhibition of mitochondrial respiration or proteostasis triggers systemic mitochondrial unfolded protein response (UPRmt), and recently serotonin and the FLP-2 neuropeptide had been shown to be important for this regulation. Here we report that disrupting mitochondrial dynamics in the neurons, by silencing the mitochondrial fusion gene fzo-1, induced UPRmt and mitochondrial fragmentation in the intestine. Acetylcholine, tyramine, glutamate and neuropeptides were required to mediate non-autonomous UPRmt. Our data suggest that tyramine signals derived from the RIM and RIC neurons target neurons that express the TYRA-3 tyramine receptor. Strikingly, neuropeptides, but not neurotransmitters, are important for non-autonomous regulation of mitochondrial dynamics in non-neural tissues. Consistent with previous studies linking UPRmt and bacterial defense, fzo-1 mutants showed avoidance to bacterial food. We are now exploring the neural mechanisms that link mitochondrial dynamics to non-autonomous UPRmt regulation and pathogen avoidance.

Neutrophils, contributing to approximately 40%-70% of white blood cells in mammals, are the most abundant type of leukocytes in human circulation. As critical effector cells in innate immunity, neutrophils form the first line of host defense against microbes and are the first immune cells recruited to an inflamed tissue. The pathogen phagocytosis, release of reactive chemicals and proteases, and formation of extracellular traps are the key weapons of neutrophils in host defense. However, neutrophils also contribute to collateral tissue damage when performing their antimicrobial functions. The destructive potential of neutrophils requires the tight regulation of their activation and recruitment. In this study, we found that miR-223 in epithelial cells controls neutrophil response to inflammation through regulating the activation of NF-kB. As fast moving cells, neutrophils rely on glycolysis for energy production. The function of mitochondria in neutrophil motility is unknown. We demonstrated that mitochondria play an indispensable role in neutrophil migration: the biogenesis of mitochondria, mitochondrial ROS and the interaction between mitochondria and ER are all involved in maintaining the movement of neutrophils.

Non-alcoholic fatty liver disease (NAFLD) comprises a range of liver lesions from simple steatosis to non-alcoholic steatohepatitis (NASH) and remains a major cause of mortality when progressing to cirrhosis and hepatocellular carcinoma (HCC). Although major risk factors relate with the metabolic syndrome, including cross-talk between the liver and the skeletal muscle, the biological mechanisms of disease are not entirely known. Therefore, a better understanding of NAFLD pathogenesis may help in finding novel targeted therapies for patients with liver damage. In that regard, we have shown that morbid obese patients with NAFLD exhibit increased microRNA-34a (miRNA/miR-34a) expression, p53 acetylation and apoptosis, as well as decreased Sirtuin-1 expression, in more severe disease stages, suggesting that miRNAs actively contribute to disease progression. In turn, further evidence also supports a role for mitochondrial dysfunction in NAFLD pathogenesis. In particular, mitochondrial fusion protein mitofusin 2 (Mfn2) is decreased in human NASH, with its liver-specific ablation in mice leading to a NASH-like phenotype. The main goal of the work presented in this thesis was to dissect the role of mitochondria-targeting miRNAs in human and experimental models of NAFLD, elucidating their functional role in disease progression and potential as therapeutic targets. First, we explored the role of miR-34a in skeletal muscle dysfunction associated to NASH development. Our results showed that miR-34a is activated in skeletal muscle of both human and experimental models of NASH, leading to inhibition of Sirtuin-1 and AMP-activated protein kinase (AMPK). This resulted in impairment of insulin signalling and dysfunctional mitochondrial dynamics, including abnormally downregulated levels of Mfn2. Further, functional studies established a direct association between miR-34a- and palmitic acid (PA)-induced muscle cell deregulation, with downstream AMPK activation being able to restore muscle homeostasis. Finally, muscle miR-34a expression and Mfn2 protein levels correlated with hallmarks of human NAFLD and its progression. We next sought to identify miRNAs directly targeting liver Mfn2, which we showed to be consistently downregulated in multiple, complementary diet-induced NAFLD mice models. miR-222-3p was identified and validated as a direct Mfn2-binding miRNA, in vitro. Strikingly, miR-222-3p inhibition in two complementary diet-induced NASH mice models led to significantly increased Mfn2 expression levels, paralleling decreased hepatic steatosis, inflammation, liver injury and oxidative stress. Finally, we evaluated expression levels of both miR-34a and miR-222-3p in the liver of a large cohort of NAFLD patients, which progressively increased with disease severity and, further, correlated with the grades of steatosis, lobular inflammation and fibrosis. Overall, our results highlight the key function of mitochondria-targeting miRNAs in metabolic syndrome-associated NAFLD. In particular, muscle miR-34a- and liver miR-222-3p-induced mitochondrial dysfunction, involving Mfn2 downregulation, appear amenable to effective therapeutic targeting. This is particularly relevant considering that both miR-34a and miR-222-3p associate to a worse disease prognosis in NAFLD patients