Huntington's disease (HD) is a mid-life-onset neurodegenerative disorder characterized by involuntary movements, personality changes and dementia. It progresses to death within 10-20 years after onset. There is currently no cure to treat this fatal disease. In HD patients, the protein huntingtin contains an abnormal expansion of a polyglutamine tract, which leads to the selective death of striatal neurons. The functions of huntingtin, as well as the dysfunctions induced by the mutation are still poorly understood. The first chapter of this thesis describes the state of the art in the study of Huntington's disease: huntingtin (htt), the protein which induces the disease; the hypothesis of toxic pathways induced by the mutant htt and the proposed therapeutic strategies to interfere with them; the available cellular and animal models and finally; the promise of gene therapy for neurodegenerative diseases such as Huntington's disease. The second chapter presents a new cellular model of the developed by infecting primary cultures with lentiviral vectors expressing a mutant htt fragment. This model is characterized by generalized neuronal transgene chronic pathology, expression, neuronal dysfunction and finally cell death. The slow progression in this model allows the study of the cascade of events leading to cell death. These unique characteristics allow the investigation of the pathological events induced by htt expression with analytical techniques on the entire neuronal population. The chapter ends with the description of an experiment showing the neuroprotective effects of ciliary neurotrophic factor (CNTF) and brain-derived neurotrophic factor (BDNF) on thiese neurons. The third chapter describes the delivery of CNTF in transgenic HD mice which express the human full-length htt (YAC72). CNTF was delivered directly to the affected neurons by striatal lentiviral injection. The sustained one-year delivery was not associated with side effects; it was correlated to reduced hyperactivity and a reduction in degenerating neurons. However, a neuroprotective effect was difficult to establish, due to the very mild and subtle pathology occurring in these animals. The fourth chapter considers delivery of BDNF in two genetic rodent HD models: in the mild phenotypic HD mice (YAC128) and a more severe rat model based on the lentiviral delivery of mutant htt in the striatum. No neuroprotective effect of BDNF could be detected in either animal model. The conclusions of this thesis discuss the potential of lentiviral vectors in modeling neurodegenerative diseases in vitro and in vivo, as well as their potential role in the treatment of such diseases. The perspective gained by the developed cellular model toward the understanding of the molecular events induced by mutant huntingtin is discussed, as well as the difficulties in testing neuroprotective approaches on animal models of HD.