While Parkinson's disease has been described nearly 200 years ago, the mechanisms leading to the degeneration of selectively vulnerable populations of neurons, such as dopaminergic neurons in the substantia nigra, remain mostly unknown. Our poor understanding of the disease etiology has dramatically hampered the rational development of therapies interfering with the processes underlying neuronal degeneration. Current therapeutic approaches provide symptomatic relief, but fail to slow down the course of Parkinson's disease. In addition, the diagnosis of Parkinson's disease relies primarily on the clinical assessment of motor symptoms that become detectable only when a large part of the nigral dopaminergic neurons have already degenerated. There is therefore a strong need to better understand the pathological processes underlying early stages of Parkinson's disease to improve the tools for diagnosis as well as to develop effective disease-modifying therapies. Towards these goals, it is important to develop, analyze and validate animal models, which faithfully replicate the human disease. In particular, the induced degeneration of nigral dopaminergic neurons in the adult mammalian brain provides an experimental tool to evaluate in vivo how the disease process leads to motor symptoms. Although neurotoxins such as 6-hydroxydopamine produce a selective degeneration of nigral dopaminergic neurons, their mode of action is clearly different from what takes places in the human pathology. During the last decade, the identification of genes linked to familial inheritance of Parkinson's disease led to the development of novel animal models based on the actual genetic cause of these rare forms of the disease. Alpha-synuclein, an abundant brain protein encoded by a gene involved in autosomal dominant forms of Parkinson's disease, is now considered a primary actor in the pathogenesis of this neurodegenerative disorder. Our laboratory has developed a rat model of Parkinson's disease based on the over-expression of human α-synuclein in the rat substantia nigra using adeno-associated viral vectors. In this model, α-synuclein over-expression progressively leads to a moderate loss of nigral dopaminergic neurons. Remarkably, animals develop motor deficits that are partly due to impaired neurotransmitter release in the remaining dopaminergic axons. This model system, where some nigral dopaminergic neurons survive in presence of overabundant human α-synuclein, can be used to explore early pathologic mechanisms that undermine the nigrostriatal system prior to outright neurodegeneration. The first part of this thesis investigates the detrimental effects of α-synuclein accumulation on the secretory pathway at the level of the endoplasmic reticulum (ER) and Golgi apparatus in the rat substantia nigra. We found that a significant proportion of the nigral neurons expressing α-synuclein display a pathological fragmentation of the Golgi apparatus. This effect has been explored by co-expressing the small GTPase Rab1A, a positive regulator the ER-to-Golgi vesicular trafficking. Interestingly, Rab1A co-expression does not prevent nigrostriatal degeneration but significantly corrects the motor deficits caused by α-synuclein. These results show that enhancing the ER-to-Golgi vesicular trafficking partially protects dopaminergic neurons from the α-synuclein-induced pathology in vivo, by restoring Golgi morphology in the remaining neurons and significantly preserving motor behavior. In order to further investigate the possible implication of the ER in the stress caused by α-synuclein, we evaluated the effects on the pathology induced by A53T α-synuclein of Salubrinal, an inductor of the PERK pathway of the unfolded protein response. Similarly to Rab1A, the Salubrinal treatment robustly reduced Golgi fragmentation and delayed the appearance of motor symptoms, without preventing nigrostriatal degeneration. Taken together, these results indicate that α-synuclein over-expression impairs the ER/Golgi system in rat dopaminergic neurons and thereby leads to neuronal dysfunction in the surviving nigral neurons. The second part of this thesis explores potential biomarkers for early stages of Parkinson's disease using ultra high field 1H magnetic resonance spectroscopy. The technology of magnetic resonance spectroscopy allows for the non-invasive quantification of abundant brain metabolites including metabolic markers and neurotransmitters. We measured metabolite levels in the striatum following nigral injection of an adeno-associated viral vector to express human α-synuclein and induce mild degeneration. The observed changes were compared to similar measurements made following intoxication with 6-hydroxydopamine to induce near complete nigro-striatal denervation. Expectedly, the level of the metabolic marker N-acetyl aspartate considered to reflect neuronal degeneration was significantly decreased only in the 6-hydroxydopamine model and failed to detect the mild neuronal loss induced by α-synuclein. On the other hand, we found an increase in striatal GABA concentration clearly detectable in both models. This result indicates that the measurement of GABA levels by magnetic resonance spectroscopy can be used as a sensitive marker for the striatal effects of α-synuclein accumulation. The increase in GABA levels observed in conditions of mild degeneration suggests that the functional deficits induced by human α-synuclein can have measurable consequences on neurotransmitter levels in the nigrostriatal system before extensive neuronal loss. Therefore, assessments of the neurotransmitter levels by magnetic resonance spectroscopy in the basal ganglia could be a promising approach for the early detection of Parkinson's disease. Overall, this thesis work shows that neurodegeneration is not the only actor in the α-synuclein-induced phenotype. We demonstrate that modifiers of α-synuclein toxicity at the ER/Golgi level mainly improve motor deficits without any effect on α-synuclein-induced degeneration. In addition, the over-expression of α-synuclein has detectable effects on the homeostasis of striatal neurotransmitters. If confirmed in human patients, these results may improve our understanding of the processes leading to the initial motor symptoms and open new avenues for early diagnosis of Parkinson's disease.