Delayed infusion of GDNF promotes recovery of motor function in the partial lesion model of Parkinson's disease Deniz Kirik, Biljana Georgievska, Carl Rosenblad and Anders Bjo È rklund Wallenberg Neuroscience Center, Department of Physiological Sciences, Lund University, BMC A11, 221 84 Lund, Sweden Keywords: 6-hydroxydopamine, cell death, glial cell line-derived neurotrophic factor, Parkinson's disease, paw use, sensorimotor behaviour, stepping, stereology, tyrosine hydroxylase Abstract Here we studied the effects of glial cell line-derived neurotrophic factor (GDNF) in a rat model that represents the symptomatic stages of Parkinson's disease. GDNF was infused starting 2 weeks after an intrastriatal 6-hydroxydopamine (6-OHDA) lesion in order to halt the ongoing degeneration of the nigrostriatal dopaminergic neurons. GDNF or vehicle was infused in the striatum or the lateral ventricle via an osmotic minipump over a total 4-week period (2±6 weeks postlesion). Motor function was evaluated by the stepping, paw reaching and drug-induced motor asymmetry tests before the pump infusion was initiated, and was repeated once during (5 weeks postlesion) and twice after the withdrawal of the minipumps (7 and 11 weeks postlesion). We found that within two weeks following the lesion » 40% of the nigral TH-positive neurons were lost. In the vehicle infusion groups there was an additional 20% cell loss between 2 and 12 weeks after the lesion. This latter cell loss occurred mainly in the caudal part of the SN whereas the cell loss in the rostral SN was almost complete within the ®rst two weeks. Ventricular GDNF infusion completely blocked the late degenerating neurons in the caudal SN and had long lasting behavioural effects on the stepping test and amphetamine rotation, extending to 6 weeks after withdrawal of the factor. Striatal infusion affected the motor behaviour transiently during the infusion period but the motor performance of these animals returned to baseline upon cessation of the GDNF delivery, and the delayed nigral cell loss was marginally affected. We conclude that intraventricular GDNF can successfully block the already initiated degenerative process in the substantia nigra, and that the effects achieved via the striatal route, when GDNF is given acutely after the lesion, diminish as the ®bre terminal degeneration proceeds. Introduction Parkinson's disease (PD) affects prominently the dopamine (DA) neurons of the substantia nigra (SN). The disease is a slowly progressive neurodegenerative disorder and data obtained from post mortem analysis of brains from PD patients suggest that a signi®cant portion of the nigral DA neurons still remain alive, albeit in an atrophic and compromised stage (Javoy-Agid et al., 1990; Fearnley & Lees, 1991; Schulzer et al., 1994). The progressive nature of the degenerative process provides a basis for the development of neuroprotective therapies aimed at blocking or slowing down the degenerative process, and stimulating recovery and regeneration of the remaining DA neurons. Glial cell line-derived neurotrophic factor (GDNF) is one of the most potent neurotrophic factors for mesencephalic DA neurons identi®ed so far and it has been shown to protect the DA neurons against toxic or mechanical insults and stimulate axonal sprouting and regrowth from lesioned nigral DA neurons (see Bjo Èrklund et al., 1997; Lapchak et al., 1997a; Gash et al., 1998, for reviews). Although GDNF has been reported to protect the nigral DA neurons almost completely provided that it is given prior to the onset of the cell death process (Winkler et al., 1996; Kearns et al., 1997; Sullivan et al., 1998; Rosenblad et al., 1999; Kirik et al., 2000a; Kirik et al., 2000b; Rosenblad et al., 2000b), the magnitude of morpho- logical and behavioural effects that would be seen if GDNF treatment were initiated in a later phase of the degeneration is not well documented. Intrastriatal injections of 6-OHDA provide a useful model for such an experiment as the cell death in the substantia nigra occurs in a delayed and progressive fashion. Administration of the toxin into the striatum induces an acute terminal axotomy that is followed by a dieback of the axons and a protracted loss of the parent cell bodies within the SN. This cell loss starts » 5±7 days after the lesion and progresses for several weeks after the insult (Sauer & Oertel, 1994; Przedborski et al., 1995; Rosenblad et al., 2000b). The effects obtained by administration of GDNF in vivo depend on the time and site of administration of the trophic factor (Kearns et al., 1997; Kirik et al., 2000a). Rescue of the nigral DA cell bodies can be obtained not only when GDNF is given over the SN (Sauer et al., 1995; Winkler et al., 1996; Kearns et al., 1997; Sullivan et al., 1998; Kirik et al., 2000a; Rosenblad et al., 2000b), but also when it is administered into the striatum or into the cerebrospinal ¯uid (Rosenblad et al., 1999; Kirik et al., 2000a; Kirik et al., 2000b). However, nigral delivery of GDNF does not preserve the DA innervation of the striatum and the functional impairments induced by the intrastriatal 6-hydroxydopamine (6-OHDA) lesion are only marginally affected, showing that preservation of the nigral cell bodies alone, in the absence of a functional striatal innervation, is not suf®cient for behavioural recovery (Winkler et al., 1996; Rosenblad et al., 2000b). In the present study GDNF was delivered in a delayed fashion, starting 2 weeks after the intrastriatal 6-OHDA lesion, to investigate Correspondence: Dr Deniz Kirik, as above. E-mail: Deniz.Kirik@mphy.lu.se Received 3 January 2001, revised 15 February 2001, accepted 19 February 2001 European Journal of Neuroscience, Vol. 13, pp. 1589±1599, 2001 ã Federation of European Neuroscience Societies