Animal models of human pathologies remain invaluable tools for unraveling disease mechanisms and evaluating potential therapeutic strategies. For a number of diseases, the lack of a reliable animal model represents an important limiting step towards the development of efficient treatments. This holds particularly true for Parkinson's disease (PD), a major neurodegenerative disorder for which only symptomatic treatments currently exist. The difficulties encountered by researchers to reproduce PD pathology in animals stem primarily from an incomplete understanding of the disease. Indeed, the cause of the disease remains unknown in 90% of cases, referred to as sporadic or idiopathic. The discovery of familial forms of the disease, however, has led to the development of a large number of transgenic mice models based on genetic modifications that play a direct causative role in a significant proportion of human PD cases. Unfortunately, these transgenic mice fail to recapitulate the robust neurodegeneration of dopaminergic (DAergic) neurons of the substantia nigra pars compacta (SNpc) and concomitant loss of DAergic projections to the striatum, the neuropathological hallmark of the human condition. The lack of nigral pathology severely limits the usefulness of such models for pre-clinical evaluation of potential therapeutics. Viral vector gene delivery tools represent an interesting alternative to classical transgenesis as they allow for targeted and high-level transgene expression in the nigrostriatal system of adult animals. During the course of this thesis we have developed two new viral vector-based rodent models of PD. In our first model, we have used a recombinant adeno-associated virus (rAAV) vector, with a high tropism towards nigral DAergic neurons, to drive overexpression of the parkin-associated endothelin receptor-like receptor (Pael-R) in the SNpc of adult rats. Indeed, accumulation of Pael-R is implicated in the pathogenesis of autosomal-recessive juvenile parkinsonism (AR-JP), a young-onset familial form of PD. We show that insoluble accumulation of Pael-R in rats induces a rapidly progressing degeneration of nigral DAergic neurons and a loss of DAergic fibers and terminals in the striatum. Lesioned animals also displayed spontaneous behavioral abnormalities linked to depletion of striatal dopamine (DA) and persisting up to 6 months post-injection. Chronic accumulation of Pael-R in the nigrostriatal system of adult rats therefore represents a robust and highly reproducible model of PD, recapitulating key pathological and phenotypical features of the human condition. The second model developed was based on nigral delivery of the PD-associated mutant G2019S leucine-rich repeat kinase 2 (LRRK2) protein. Indeed, the G2019S mutation in the LRRK2 gene is the most important genetic determinant of PD, accounting for a significant proportion of both familial and sporadic PD cases. Due to the large size of the LRRK2 coding sequence, an adenoviral system with a high packaging capacity was used to drive expression of the protein. Recombinant adenoviral (rAd) vectors are potentially pro-inflammatory and less efficient tools as rAAV vectors for long-term gene delivery to the SNpc. Nevertheless, through retrograde axonal delivery of rAd-LRRK2 particles, we achieved robust and neuron-specific expression of full-length wild-type or mutant G2019S human LRRK2 in nigral DAergic neurons of adult rats. Expression persisted up to 6 weeks post-injection, with no visible signs of inflammation in the SNpc. We demonstrate that the wild-type form of LRRK2 does not induce neuronal loss when expressed in the SNpc. In contrast, under the same conditions and levels of expression as the wild-type form, the PD-associated G2019S mutation in LRRK2 is sufficient to cause a progressive loss of nigral dopaminergic neurons. This is the first demonstration of frank dopaminergic neuronal degeneration in rodents induced by the expression of G2019S mutant LRRK2. Our data also provide a new rodent model of LRRK2-linked PD which recapitulates one of the cardinal pathological features of the disease. In the absence of a clear understanding of human PD pathogenesis, the development of multiple transgenic models may help to identify common disease mechanisms and drug targets. A potential treatment identified in one model may be effective only in the corresponding subset of PD patients, carrying this particular genetic modification. On the other hand, the pathogenic pathway targeted may also be activated in sporadic PD. Ultimately, whether or not these models stand the test of time will depend on how effective newly-identified drugs will be, not only on AR-JP or LRRK2-linked PD patients, but above all on sporadic PD, which represents the vast majority of PD patients.