The second strategy relies on the neuroprotective effects of the neurotrophin GDNF or its close relative neurturin. These factors have previously been demonstrated to provide substantial preservation of dopaminergic inputs to striatum. Evidence for this effect in animal models and in clinical case reports represent a considerable body of evidence that the strategy is sound. Despite the recent negative trials of GDNF micro-infusion by intraventricular administration (39,95) and by direct administration into the striatum (41), it is possible that local expression of GDNF provided by gene therapy would generate sufficient product to create benefit (96). Local production might also limit the side effects noted with exogenous delivery.
Numerous studies have demonstrated success of viral GDNF therapy in rats. Several viral vectors have been used to deliver GDNF to the striatum and SNc, including adenovirus, AAV, herpes virus, and lentiviruses. Of these, herpes viruses were noted to be problematic, providing only limited benefit and demonstrating significant toxicity related to purification methods (97). The other models have demonstrated the ability to generate stable and sufficient quantities of GDNF, to maintain or restore tyrosine hydroxylase activity, and, in several cases, to improve parkinsonian behavioral correlates (98,99).
Recent studies have adopted more complex models to more nearly approximate PD. In one study, Brizard et al. selected a rat model for progressive degeneration by injecting on SNc with a partial dose of 6-OHDA. This model is thought to more closely approximate the gradual degeneration of PD. They found that, four weeks after lesioning, addition of a lentiviral vector conferring GDNF production (lenti-GDNF) restored dopaminergic innervation of the striatum to near normal levels. This was accompanied by behavioral improvements in a task requiring paw-reaching to obtain food pellets (100). Similar studies have examined performance of complex motor selection tasks in rats pretreated with lenti-GDNF prior to 6-OHDA lesioning (101). These studies suggest treatment success in behavioral models more relevant to human disease.
Several other neurotrophins have been tested for their protective effect on nigrostriatal neurons. A study by Fjord-Larsen et al. demonstrated that in vivo lentiviral delivery of a modified neurturin construct produced neuroprotection of rat nigrostriatal projections. Tyrosine hydroxylase immunoreactive neurons were 91% of the unlesioned side. This was equivalent to the effect of lentiviral GDNF trans-duction (102). A similar study using HSV delivery to compare the effect of GDNF and brain-derived neurotrophic factor (BDNF) in vivo demonstrated that BDNF is less efficient in preserving nigrostriatal neurons or restoring normal behavior than GDNF (103).
GDNF delivery by lentivirus has been demonstrated in primates to produce stable transmission of the GDNF gene in both aged unlesioned monkeys and in young MPTP-lesioned monkeys. PET studies in the aged monkeys, who had received lentiviral GDNF administration to both the striatum and the substantia nigra, demonstrated that putaminal GDNF administration increased fluorodopa uptake by 37% on the treated side. Histological analysis demonstrated stable delivery of GDNF and migration from the injection site into the pallidum. Tyrosine hydroxy-lase immunoreactivity was also increased to 39% and 44% in the caudate and putamen, respectively. The number of tyrosine hydroxylase-reactive cells in the substantia nigra increased 85% on the side of GDNF viral delivery.
Young monkeys were treated with GDNF one week after unilateral administration of MPTP. Treatment with the lenti-GDNF showed significant increases in stri-atal fluorodopa uptake, averaging a three-fold increase, compared to controls. These animals were monitored for several months for performance on a hand-reaching task and for assessment with a modified clinical rating scale. The treated monkeys showed significant improvements both in the clinical assessment and in their speed on the hand-reaching task (104).
The expression of GDNF after viral delivery has been noted specifically to increase the number of tyrosine hydroxylase expressing striatal cells by an average of seven-fold. This suggests that the addition of GDNF may act to confer a dopamin-ergic phenotype to adult striatal cells, in addition to preserving dopaminergic nigral inputs onto the striatum (105). Now established in rodent and primate models, GDNF gene therapy holds considerable promise for the treatment of PD.
Was this article helpful?