The unfolded protein response (UPR) is activated by protein misfolding stress in the endoplasmic reticulum (ER). The UPR is a transcriptional program that aims to maintain ER folding capacity, where imbalances between protein load and processing ability is termed ER stress. Signal transduction of the UPR begins with 3 ER-resident transmembrane sensors: PERK, IRE1 and ATF6. All sensors initiate downstream signaling cascades which culminate in improved protein folding, transcriptional upregulation of genes encoding ER chaperones, and mechanisms to reduce translational and transcriptional ER load, therefore re-establishing ER homeostasis. The signaling cascades of each sensor are distinct but cooperative, and involve a significant amount of crosstalk, feedback and overlap. Indeed, there are many pathological and physiological conditions have an effect on ER protein burden, and therefore on activation of the UPR. Increases in protein load in professional secretory cells, hypoxic conditions in a tumor mass, obesity all induce cause changes in the ER folding environment. Although we understand how the UPR contributes to relieve ER stress under acute conditions (e.g. pharmacological treatment) much less is understood about the contributions to physiological processes and chronic stress conditions. Our overall goal was to understand how the UPR is activated during physiological settings, the mechanisms it uses to maintain folding capacity under these setting and the specific components responsible for adapting the response to various stresses. We first decided to understand a chronic stress from a transgenic approach. By creating a knockout mouse, the genetic deletion functions as a stress and we can understand its physiological role. By compounding two genetic deletions in UPR components (ATF6α and p58IPK) we provide evidence for the developmental role these components play. Homozygous deletion ATF6α bears no gross histological phenotype yet causes synthetic lethality when combined with p58IPK deletion. This also reveals that the UPR is able to adapt to genetic impairment of protein folding in vivo. Next, to better understand these chronic states, we established an experimentally tractable chronic stress treatment in vivo. Our treatment suppressed ATF6α dependent chaperone expression through an mRNA degradative mechanism, which led to long term changes in UPR expression. We determined that chronic conditions can change the sensitivity of the UPR to ER stress, potentially as an adaptive consequence. We also showed that sensitivity to ER stress can be changed during chronic stress. Finally we simulated the UPR in a computational ordinary differential equation (ODE) model in order to determine how various stresses and component interactions determine the output of the UPR. We built a series of equations to describe the UPR signaling network, entrained it on experimental data and refined it through the use of transgenic knockout cells. Our model was robust enough to recreate experimental measurements of UPR components when tested in parallel with knockout cells. We found that stress sensitivity is dependent on the crosstalk and negative feedback connections of the UPR. This study has enhanced our understanding of activation of the UPR under non-acute settings. It demonstrates that the UPR is a signaling hub with a broad output range that is capable of handling a variable degree of insults because of the intrinsic properties of the signaling network. This provides a better understanding for the contributions of the UPR to physiological stresses and certain chronic diseases.

La Sclérose Latérale Amyotrophique (SLA) est une maladie neurodégénérative incurable caractérisée par la perte progressive des motoneurones. Récemment, des mutations dans le gène UBQLN2, codant pour l'ubiquiline 2, ont été identifiées dans des formes héréditaires de SLA et de SLA avec Démence Fronto-Temporale (SLA-DFT). Ces mutations induisent la formation d'agrégats positifs à l'ubiquiline 2 dans la moelle épinière de patients atteints de SLA et dans le cerveau de patients atteints de SLA-DFT, ainsi que dans d'autres formes familiales et dans des formes sporadiques de la maladie. En raison de l'implication de l'ubiquiline 2 dans la dégradation protéique intracellulaire, son rôle semble être déterminant dans la pathogénèse des différents types de SLA. Afin d'étudier les mécanismes physiopathologiques de la maladie et d'envisager des solutions thérapeutiques, nous avons créé un modèle murin de SLA et de SLA-DFT liées à l'UBQLN2 qui récapitule l'ensemble des caractéristiques associées à la maladie, suite à l'injection intracérébroventriculaire de vecteurs viraux associés à l'adénovirus (AAV) de sérotype 10, renfermant l'ADNc muté de l'ubiquiline 2. De plus, nos analyses de spectrométrie de masse ont révélé une perte d'interaction de l'ubiquiline 2 mutée avec une protéine indispensable pour le maintien du cytosquelette des neurones, l'?-spectrine. Enfin, nous avons observé que le Bleu de Méthylène augmente la quantité d'inclusions engendrées par la surexpression de l'ubiquiline 2 mutée. Cette étude a donc permis d'obtenir un paradigme solide pour l'étude de l'implication de l'ubiquiline 2 dans la SLA et pourra servir à l'élaboration de stratégies thérapeutiques efficaces. / Amyotrophic Lateral Sclerosis (ALS) is an incurable neurodegenerative disease characterized by progressive loss of motor neurons (MNs). Recently, mutations in the UBQLN2 gene, encoding ubiquilin 2, have been identified in hereditary forms of ALS and ALS with FrontoTemporal Dementia (ALS-FTD). These mutations induce the formation of ubiquilin-2-positive aggregates in the spinal cord of patients with ALS and in the brain of patients with ALS-FTD, as well as in other familial forms and in sporadic forms of ALS and ALS-FTD. The role of ubiquilin 2 seems to be crucial in the pathogenesis of the different types of ALS, due to its involvement in proteostasis. To understand the physiopathology of the disease and to envision therapeutic strategies, we have generated an ubiquilin 2-linked ALS and ALS-FTD mouse model which recapitulated disease associated phenotypes, after intracerebroventricular (ICV) injection of Adeno Associated Virus serotype rh10 (AAV10) vectors, containing the cDNA encoding mutated ubiquilin 2. Furthermore, mass spectrometry analysis revealed a loss of interaction between the mutant protein and one protein necessary for cytoskeleton maintenance of neurons, the α-spectrin. Finally, we have observed that Methylene Blue increased the number of inclusions in mice expressing the mutant ubiquilin 2. In conclusion, this study presents a solid paradigm to study the involvement of ubiquilin 2 in ALS and can be used for the development of effective therapeutic strategies.

The accumulation of amyloid-ß; protein (Aß) in Alzheimer's disease (AD) is a well known pathological event. Decreasing the production or increasing the degradation of Aß; is therefore thought to serve as a potential therapeutic intervention in AD. Recent in vitro and in vivo studies have suggested that certain proteases may be involved in the catabolism of Aß; and defects in the degradation of Aß; could contribute to AD disease progression. Studies implicating the homing of monocytes to regions of CNS damage have led to the idea that it may be possible to use genetically modified monocytes to carry exogenous genes of interest into the brain or other organs for the purposes of gene therapy. To determine the time course of monocyte recruitment into the brain during the neurodegenerative damage characteristic of Alzheimer's disease, we used transplanted GFP labeled bone marrow monocytes to characterize the kinetics that peripheral monocytes display once injected into the circulation. We determined the half life of bone marrow derived monocytes after one injection into the peripheral circulation, and found this time to be 1.5 hours post injection. We also examined the effects of the APP+PS1 transgene on the recruitment of peripheral monocytes and showed that these cells are actively recruited to the brains in AD transgenic mouse models compared to non transgenic mice. As an approach to increase expression of NEP in a transgenic mouse model of AD, we developed an ex vivo gene therapy method utilizing bone marrow monocytes from GFP mice. These monocytes were transfected with a NEP construct designed to express either a secreted form of NEP or a form which lacks any enzyme activity. Monocytes were administered through a microvascular port twice a week for two months and we observed recruitment of bone marrow-derived monocytes into the CNS. In addition, we found significant reductions in both Aß and Congo red staining in the NEP-S injected mice only. These studies show that putting monocytes together with an amyloid degrading enzyme such as neprilysin offers a powerful novel therapeutic tool for the treatment of AD.

L’immunoglobuline A (IgA) est l’immunoglobuline la plus abondamment synthétisée chez les mammifères. Ses propriétés ambivalentes l’impliquent non seulement dans des fonctions de protection contre les agents pathogènes mais aussi dans des phénomènes de tolérance immunitaire vis-à-vis des germes commensaux du microbiote. Toutefois, les IgA peuvent développer des propriétés pathogènes. Dans la première partie de mon travail de thèse, nous avons étudié les effets pathogènes de l’IgA. Les dépôts d’IgA sur le mésangium sont la caractéristique de l’IgAN. La physiopathologie de cette maladie est mal connue. L’hypothèse d’un défaut de glycosylation de l’IgA est souvent retenue ; ce défaut peut être la cause de sa polymérisation et de son antigénicité, il peut aussi favoriser le clivage du récepteur CD89. Nous avons analysé l’effet du défaut d’affinité de la région variable des IgA, de la substitution de la chaîne légère ainsi que de l’association des IgA à leur récepteur, le CD89 sur l’induction des lésions et le dysfonctionnement rénal chez quatre modèles murins différents générés au laboratoire et suivis pendant 12 mois. Nous avons également étudié les propriétés physico-chimiques des IgA de 28 patients ayant une dysglobulinémie et de 28 IgA produites par des hybridomes ; la relation entre ces propriétés et la capacité des IgA à se déposer a été observée. Dans une seconde partie, nous avons étudié l’aspect immunomodulateur et les propriétés antiinflammatoires conférées par l’IgA humaine surexprimée chez un modèle murin de lupus systémique (souris MRL/lpr). Dans la dernière partie du travail, nous avons contribué à la caractérisation d’un modèle de souris transgénique exprimant l’IgA de classe 2 et à l’étude de l’effet de signalisation médiée par cette IgA2 sur le développement des populations lymphocytaires. L’ensemble de ces travaux a montré l’effet pathogène des IgA naturelles ayant une faible affinité sur le développement de la néphropathie à IgA ; ainsi les analyses des IgA des patients et des hybridomes montrent que c’est la stabilité moléculaire de préférence au profil de glycosylation qui joue un rôle crucial dans leur capacité de dépôt. L’expression des IgA humaines chez les souris lupiques a considérablement prolongé leur durée de vie et a ralenti la survenue de l’auto-immunité et de l’atteinte rénale ce qui témoigne du rôle anti-inflammatoire des IgA. L’étude du modèle murin exprimant l’IgA2 humaine a montré que la signalisation via l’IgA2 joue un rôle inhibiteur sur le développement précoce de certaines sous-populations de cellules B. L’ensemble de ces résultats montrent la multitude d’effets de l’IgA lui permettant d’intervenir d’une part dans la pathogenèse d’une maladie complexe (l’IgAN) et d’autre part dans la protection de l’auto-immunité, témoignant de la complexité des interactions mises en jeu et du caractère régulateur de cette immunoglobuline. / Immunoglobulin A (IgA) is the most synthetized immunoglobulin in mammals. IgA has ambivalent properties: it is implicated in the mechanisms of defense against pathogens but also in the immune tolerance of commensal microbiota. However, IgA can develop pathogenic properties. In the first part of my thesis, we studied the pathogenic effects of IgA. IgA deposits are the main characteristic of IgA nephropathy (IgAN). IgAN physiopathology is not yet clearly understood. The hypothesis of a glycosylation defect is strongly adapted. This defect can be due to IgA polymerization or antigenicity. It can also induce shedding of CD89 (IgA Fc receptor) or other factors. We studied the effect of variable region altered affinity, the light chain substitution and the association of IgA with CD89 on the development of kidney lesions and impairment of kidney function in four mouse models followed up during 12 months. In addition, we studied the physico-chemical properties of 28 IgA purified from patients with dysglobulinaemia and 28 chimeric IgA produced by hybridomas. The effect of these properties on the propensity of IgA for mesangial deposition was explored. In the second part, we studied the immunomodulatory and anti-inflammatory properties conferred by the overexpression of human IgA in a mouse model with systemic lupus (MRL/lpr model). In the last part, we contributed to the characterization of a transgenic mouse model producing IgA class 2 and to the study of the effect of IgA2-mediated signaling on B lymphocyte development. Altogether, obtained results show the pathogenic effect of low affinity-IgA on the development of IgA nephropathy. In addition, different analyses showed that molecular stability but not glycosylation profile is the determining factor for IgA deposition. On the other hand, IgA expression in lupus-prone mice extended their survival, delayed the onset of auto-immunity and ameliorated kidney functions in these animals which supports IgA anti-inflammatory properties. The study of IgA2-mediated signaling in the transgenic model showed the inhibitory effect of IgA2 on the early development of several B cell sub-populations. All of these results show the multiple effects of IgA which contribute on one hand to the pathogenesis of a complex disease (IgAN) and on the other hand to protection from autoimmunity, demonstrating the complexity of interactions and the regulatory character of this immunoglobulin.

Hearing impairment is the most common sensory deficit worldwide, affecting at least one child in every one thousand born. Gene therapy targeting the inner ear offers promise for treatment of genetic forms of hearing loss. Many genetic forms of deafness are congenital and gene therapies in these cases would require treatment prior to inner ear maturation. Included in this category is the dominant-negative R75W mutation in GJB2 which encodes connexin 26, a gap junction protein expressed in the supporting cells of the organ of Corti. RNA interference (RNAi)-based therapeutics offer promise for treating dominant-negative diseases. Our goal has been the in vivo application of RNAi-therapy to the GJB2-R75W transgenic mouse, a model of severe-to-profound dominant-negative hearing loss. Here we describe our efforts to identify a therapeutic, a suitable delivery route, and an optimal delivery vector. We have designed and optimized siRNA to achieve robust silencing of the mutant transgene in vitro and have prepared artificial miRNA constructs for in vivo application. We have determined to use the embryonic otocyst microinjection technique as the route for therapeutic delivery and have successfully utilized this technique to study the tropism and safety of several viral vector (adeno-associated virus 2/1, early- and late-generation adenoviruses, and bovine adeno-associated virus). For the first time we have characterized viral tropism for cochlear supporting cells following in utero delivery to their progenitor cells in the developing cochlea and identified bovine adeno-associated virus as a safe vector for gene delivery to the supporting cells of the cochlea. We have also described two previously unreported phenotypes in the GJB2-R75W transgenic mouse model: skin disease and cataracts. Both can be caused by dominant connexin mutations in humans. Our work shows that although gene therapy is not simple, powerful tools are in place for treating dominant forms of hereditary hearing loss.

Millions of people suffer from tinnitus, a disorder for which there is currently no effective treatment or cure. My dissertation work provides insight into the neural correlates of this pervasive hearing disorder and examines how a newly emerging therapy, transcranial magnetic stimulation (TMS), affects the central auditory system in the generation of the tinnitus percept. This work has a multifold focus of: i) developing and modeling the function of a miniature magnetic coil that can be used for TMS in rodents, ii) establishing a reliable mouse model of tinnitus that can be used for assessing TMS treatment-induced changes, iii) measuring the behavioral alterations and neural changes induced by TMS throughout the auditory system in mice with tinnitus, and iv) to assay underling molecular changes in the auditory cortex (AC) related to TMS and tinnitus. Chapter 1 gives an overview of the current research on tinnitus and TMS. Chapter 2 establishes a reliable neural and behavioral assay of verifying tinnitus in a mouse model and provides further evidence that the underlying hyperactivity associated with tinnitus is initiated in the brainstem following reduced afferent input. The remainder of the dissertation examines the modulation of tinnitus in the auditory central nervous system using a miniature TMS coil. Chapter 3 of the dissertation details the creation and evaluation of a rodent-sized TMS coil, which could increase the overall effectiveness and applicability for human treatment. TMS is currently an FDA approved treatment of depression and has been shown to decrease tinnitus perception in human clinical trials, albeit with variable results. There have been few published studies of tinnitus modulation by TMS using animal models and therefore little is known about the molecular and neural bases of this potential tinnitus treatment. TMS is thought to be therapeutic because the magnetic flux generated from the electromagnetic coil induces an electric field in the brain, altering ion flow and subsequently neural function, as the excitation and inhibition of cortical networks become synchronized to the magnetic pulse. Chapter 4 demonstrates that TMS with our custom-designed miniature rodent coil can successfully reduce behavioral evidence of tinnitus in a mouse model, mainly through activating inhibitory networks in the AC. It also shows that presynaptic activity is altered in the upper layers of the AC responsible for intralaminar processing and sound perception. Finally, chapter 5 describes an in-depth proteomic analysis of over 3000 proteins from the AC, which shows that TMS and noise-induced tinnitus alter the expression of several key proteins and pathways that play a critical role in cortical excitatory and inhibitory activation. The results of this work are also important because they are the first animal model to demonstrate neural changes during TMS-treated tinnitus, creating a paradigm that can be used for optimizing parameters to improve clinical outcomes in human trials.

The feasibility of the Alb-uPA/SCID/Bg chimeric mouse as a model for Hepatitis C Virus (HCV) infection was assessed experimentally by (1) the infection and treatment with another hepatotropic virus, Hepatitis B Virus (HBV) and (2) the infection of the model with HCV and the subsequent treatment of that infection with a pro-apoptotic factor (BID) targeted to infected hepatocytes. In the former, the infected mouse responded favorably, and in the manner of human patients, to a standard imunoglobulin therapy. In the latter, HCV-infected hepatocytes were successfully targeted for cell death, with repeated doses of Adenovirus-delivered BID being the most effective at inhibiting virus spread. Efficacy and toxic side-effects of BID treatment could be reconciled by modulating the timing between doses, the most effective tested being three doses of BID at 7-day intervals. Analyses of chimeric model production were undertaken to improve the quality of human hepatocyte engraftment (typically only 25-35% of mice receiving grafts are currently used experimentally). Minor variations in success rates were experienced with respect to donor age or health status, or the age of recipient mice within an operational window of 5 to 13 days from birth. The greatest obstacle to useful engraftment (aside from technical challenges) was deemed to be the genetic/cellular integrity of the recipient mouse. This conclusion was based on variable engraftment success with ‘healthy’ donor cell preparations and a consideration of variability in immune deficiency arising in mice within a SCID/Bg mouse colony. / Experimental Surgery

Prader-Willi syndrome (PWS) is a rare, neurodevelopmental disorder with multiple features caused by hypothalamic deficiency, including infantile failure to thrive, hyperphagia leading to obesity, growth hormone deficiency, hypogonadism, and central adrenal insufficiency. Other features of PWS including global developmental delay, hypotonia, pain insensitivity, gastrointestinal dysfunction, and psychiatric disorders are caused by deficits in other regions of the nervous system. PWS is caused by the loss of a subset of paternally-expressed genes on chromosome 15, which includes NDN and MAGEL2. Necdin and Magel2 are both members of the melanoma antigen (MAGE) family of proteins and are expressed throughout development, particularly in the nervous system. This thesis describes experiments that examine the loss of function of necdin and Magel2 in mice and their potential roles in the pathogenesis of PWS. Targeted inactivation of Ndn and Magel2 in mice has aided in determining how loss of function of these proteins affects the development and function of the nervous system. Loss of necdin causes reduced axonal outgrowth and neuronal differentiation in the central and peripheral sensory nervous systems. I examined the autonomic nervous system in Ndn-null embryos and identified a defect in the migration of the most rostral sympathetic chain ganglion and consequently increased neuronal cell death and reduced innervation of target tissues supplied by this ganglion. Reduced axonal outgrowth was observed throughout the sympathetic nervous system in Ndn-null embryos although no gross deficits in the parasympathetic and enteric nervous systems were identified. Loss of Magel2 causes reduced fertility and abnormal circadian rhythm patterns in mice. I further identified an altered response to stress, a delayed response to insulin-induced hypoglycemia, a reduced stimulated growth hormone response, and lower thyroid hormone levels in Magel2-null mice, indicative of deficits in multiple hypothalamic-pituitary axes. The findings presented in this thesis support a role for necdin and Magel2 in the development and function of the nervous system. The data also indicates that these MAGE proteins play a key role in multiple features of PWS, including endocrine deficiencies and autonomic dysfunction

Of the thousands of dividing progenitor cells (PCs) generated daily in the adult brain only a very small proportion survive to become mature neurons through the process of neurogenesis. Identification of the mechanisms that regulate cell death associated with neurogenesis would aid in harnessing the potential therapeutic value of PCs. Apoptosis, or programmed cell death, is suggested to regulate death of PCs in the adult brain as overexpression of B-cell lymphoma 2 (Bcl-2), an anti-apoptotic protein, enhances the survival of new neurons. To directly assess if Bcl-2 is a regulator of apoptosis in PCs, this study examined the outcome of removal of Bcl-2 from the developing PCs in the adult mouse brain. Retroviral mediated gene transfer of Cre into adult floxed Bcl-2 mice eliminated Bcl-2 from developing PCs and resulted in the complete absence of new neurons at 30 days post viral injection. Similarly, Bcl-2 removal through the use of nestin-induced conditional knockout mice resulted in reduced number of mature neurons. The function of Bcl-2 in the PCs was also dependent on Bcl-2-associated X (BAX) protein, as demonstrated by an increase in new neurons formed following viral-mediated removal of Bcl-2 in BAX knockout mice. Together these findings demonstrate that Bcl-2 is an essential regulator of neurogenesis in the adult hippocampus.

Immune thrombocytopenic purpura (ITP) is an autoimmune disease characterized by autoantibody-mediated platelet destruction. To test the efficacy of novel sulfur compounds as alternative treatments for ITP, we used a mouse model of passive immune thrombocytopenia (PIT). Using this model, the platelet nadir could not be maintained, with platelet counts rising after day 4, despite daily anti-platelet antibody administration. We examined reticulated platelet counts by flow cytometry, and found increased thrombopoiesis in the bone marrow to be at least partially responsible for this platelet rebound. Consequentially, two improved mouse models of PIT were developed, where the platelet rebound is circumvented. The first model employs sublethal total body gamma-irradiation in combination with daily antibody administration, while the second model employs gradual escalation of the daily antibody dose. Finally, we show that none of the tested candidate compounds show efficacy in elevating platelet counts in vivo, likely due to their limited solubility.