Differential effects of glial cell line-derived neurotrophic factor - Nature

Gene Therapy (1999) 6, 1936–1951  1999 Stockton Press All rights reserved 0969-7128/99 $15.00 http://www.stockton-press.co.uk/gt

Differential effects of glial cell line-derived neurotrophic factor (GDNF) in the striatum and substantia nigra of the aged Parkinsonian rat B Connor1, DA Kozlowski1, T Schallert2, JL Tillerson2, BL Davidson3 and MC Bohn1 1 Children’s Memorial Institute for Education and Research, Department of Pediatrics, Northwestern University Medical School, Chicago, IL; Department of Psychology, University of Texas, Austin, TX; and 3Department of Internal Medicine, University of Iowa College of Medicine, Iowa City, IA, USA

Injection of an adenoviral (Ad) vector encoding human glial cell line-derived neurotrophic factor (GDNF) protects dopaminergic (DA) neurons in the substantia nigra (SN) of young rats. As Parkinson’s disease occurs primarily in aged populations, we examined whether chronic biosynthesis of GDNF, achieved by adenovirus-mediated delivery of a GDNF gene (AdGDNF), can protect DA neurons and improve DA-dependent behavioral function in aged (20 months) rats with progressive 6-OHDA lesions of the nigrostriatal projection. Furthermore, the differential effects of injecting AdGDNF either near DA cell bodies in the SN or at DA terminals in the striatum were compared. AdGDNF or control vector was injected unilaterally into either the striatum or SN. One week later, rats received a unilateral

intrastriatal injection of 6-OHDA on the same side as the vector injection. AdGDNF injection into either the striatum or SN significantly reduced the loss of FG labelled DA neurons 5 weeks after lesion (P ⭐ 0.05). However, only striatal injections of AdGDNF protected against the development of behavioral deficits characteristic of unilateral DA depletion. Striatal AdGDNF injections also reduced tyrosine hydroxylase fiber loss and increased amphetamineinduced striatal Fos expression. These results demonstrate that increased levels of striatal, but not nigral, GDNF biosynthesis prevents DA neuronal loss and protects DA terminals from 6-OHDA-induced damage, thereby maintaining DA function in the aged rat.

Keywords: aging; Parkinson’s disease; gene therapy; adenoviral vector; neurodegeneration

Introduction Parkinson’s disease is a neurodegenerative disorder characterized by progressive degeneration of dopaminergic (DA) neurons in the substantia nigra pars compacta (SN) that innervate the striatum. This degeneration results in the loss of DA terminals in the caudate and putamen leading to the clinical symptoms of bradykinesia, rigidity and resting tremor. Current therapies exist which ameliorate these symptoms, however they do not prevent the continuing degeneration of DA neurons. Growth factors hold potential as neuroprotective therapeutic agents to treat various neurodegenerative disorders.1–4 Members of several growth factor families are trophic for DA neurons and signal through different intracellular mechanisms. These include the TGF-␤ superfamily, the neurotrophins, cytokines and mitogenic growth factors.5–7 The most potent growth factor for DA neurons is GDNF.8 GDNF protects and restores DA neurons in several rodent and primate models of Parkinson’s disease when administered to the adult nigrostriatal system.9–15 In addition, GDNF stimulates regenerative growth or axonal sprouting after partial lesions of the DA

Correspondence: MC Bohn, Department of Pediatrics, CMIER, 2300 Children’s Plaza #209, Chicago, IL 60614, USA Received 28 April 1999; accepted 15 July 1999

system and induces stimulatory effects on metabolism and function of DA neurons.16,17 Systemic administration is an inefficient way of delivering neurotrophic factors into the CNS. Although single or repeated injections of GDNF protein directly into the CNS have been reported to be effective in animal studies,9,13,18 continuous targeted delivery of neurotrophic factors to specific neurons in humans remains a challenge. This challenge might be met by gene therapy where increased levels of neurotrophic factor biosynthesis might prevent neuronal cell death and enhance neuronal function in neurodegenerative disorders such as Parkinson’s disease. Adenoviral (Ad) vectors have been shown to deliver transgenes efficiently to the CNS.19–26 Previous research in our laboratory has revealed that AdGDNF injected in the SN or striatum protects DA neurons from 6-OHDA-induced degeneration in the young Fischer-344 rat brain.24,27 Similar neuroprotective effects have been reported in young rats by other groups using adenoviral and adeno-associated viral vectors expressing GDNF.28,29 Parkinson’s disease is a neurodegenerative disorder that primarily affects the aging population. Several studies have reported that injection of recombinant GDNF protein augments locomotor activity and dopaminergic function in the normal aged rat brain.30–33 Intrastriatal GDNF protein administration has also been observed to induce significant recovery of DA nigrostriatal fibers after

The protective action of AdGDNF in aged Parkinsonian rat B Connor et al

MPTP depletion in aged (12 months) mice.34 However, the effectiveness of chronically delivering GDNF via gene transfer to DA neurons in an aged Parkinsonian rat model has not previously been examined. Zigmond and colleagues35 proposed that compensatory events occur within the young rat brain following partial degeneration of the DA component of the nigrostriatal pathway allowing the SN to continue exerting DA control over striatal cell function. Compensatory events may be observed in the aged rat brain due to age-dependent changes that occur in the nigrostriatal and mesolimbic DA pathways.36 Alternatively, the aged brain may be unable to produce compensatory changes in response to either age-dependent or lesion-induced degeneration.37,38 This lack of compensatory response may worsen lesioninduced degenerative alterations in dopaminergic neuronal morphology and function in the aged brain. Therefore, while AdGDNF has been observed to exhibit neuroprotective effects within the young rat brain,24,27,28 it was important to examine whether chronic biosynthesis of GDNF, achieved via adenovirus-mediated delivery of GDNF was able to restore compensatory events, protect DA neurons and maintain DA function in the aged rat brain after partial lesion-induced degeneration of the nigrostriatal pathway. It has also become apparent that the site of GDNF administration is important in determining the degree of functional recovery observed in the damaged or degenerating nigrostriatal DA system.39 Therefore, the present study was undertaken to compare the behavioral and cellular effects of GDNF gene delivery into striatal and mesencephalic sites in aged (20 months) rats with progressive lesions of the nigrostriatal pathway.

Results Statistical analysis Control rats consisted of rats that received either a striatal or SN injection of AdLacZ or AdmuGDNF (a non-biologically active mutant form of human GDNF24) and rats that received no vector injection in either site (untreated). Control rats were separated into one of two control groups – rats that received control striatal vector injections and rats that received control SN vector injections. Untreated rats were placed in both control groups. Repeated-factor, two-way analysis of variance (ANOVA) for group by day of behavior observed (amphetamineinduced rotational behavior and spontaneous exploratory forelimb use) and a one-way ANOVA for group (cell counts and optical density measurements) were performed to determine whether there was any statistical difference between untreated rats and rats that received AdLacZ or AdmuGDNF in each group (striatal vector injection group and SN vector injection group). Statistical analysis revealed that there was no significant difference (P ⬎ 0.05) between untreated rats and rats that received AdLacZ or AdmuGDNF in either control group. Therefore, control rats were grouped into two separate groups for statistical analysis – striatal vector control group (n = 9; striatal injections of AdLacZ, AdmuGDNF and untreated rats) and SN vector control group (n = 10; SN injections of AdLacZ, AdmuGDNF and untreated rats). Behavioral testing Amphetamine-induced rotational asymmetry: In animals with a unilateral lesion of the nigrostriatal DA system,

the injection of drugs that act to release dopamine, such as amphetamine, will induce rotational behavior towards the denervated striatum. The traditional ‘rotational behavior model’40 states that lesioned animals will turn away from the hemisphere where there is greater amphetamine-stimulated dopamine release and greater dopamine receptor stimulation. We examined the effects of 6-OHDA and Ad vector injection on amphetamineinduced rotational asymmetry at different time-points after 6-OHDA lesioning. Amphetamine-induced rotational activity was recorded for 60 min following an i.p. injection of 5 mg DL-amphetamine per kg body weight. The number of clockwise and counter-clockwise turns were counted and expressed as the number of net rotations per min to the lesioned (ipsilateral) hemisphere (net rotations per min to ipsilateral hemisphere = ipsilateral rotations − contralateral rotations/time). Repeatedfactor, two-way ANOVA revealed a significant effect of group in amphetamine-induced ipsilateral rotational behavior (F3,29 = 5.04, P ⬍ 0.006). Tukey–Kramer post hoc comparison tests indicated that ipsilateral rotational behavior was significantly reduced (P ⭐ 0.05) 35 days after lesioning in rats injected with AdGDNF in the striatum compared with control groups and rats injected with AdGDNF in the SN (Figure 1). In addition, at 35 days after lesion, we observed that rats injected with AdGDNF in the SN exhibited a significant increase in amphetamine-induced ipsilateral rotational behavior compared with control rats (P ⭐ 0.05).

Spontaneous exploratory forelimb use: The spontaneous exploratory forelimb use test is a non-drug-induced test of forelimb locomotor function which has been shown to

Figure 1 AdGDNF injected into the striatum significantly protects against the development of amphetamine-induced ipsilateral rotational behavior observed in control rats after 6-OHDA lesioning. Rats were injected with DL-amphetamine (5 mg/kg i.p.) and their behavior recorded for 60 min. Baseline amphetamine rotation tests were performed 7 days before lesioning (−7) and the results used to assign the side of 6-OHDA lesioning (see Materials and methods). 35 days after lesioning, rats injected with AdGDNF into the striatum exhibited a significant reduction in ipsilateral rotational behavior compared with control groups and rats injected with AdGDNF into the SN (*P ⭐ 0.05). In contrast, rats injected with AdGDNF into the SN exhibited a significant increase in ipsilateral rotational behavior 35 days after lesion compared with control groups (**P ⭐ 0.05).

1937


1938

correlate with DA depletion in the lesioned hemisphere.87 Following a unilateral 6-OHDA lesion of the nigrostriatal system, rats preferentially use the forelimb ipsilateral to the side of the lesion to initiate and terminate weightshifting movements during rearing and exploration along vertical surfaces. To examine this, rats were videotaped for 5 min in a Plexiglas chamber before and 7, 14, 21 and 28 days after 6-OHDA lesioning. Spontaneous exploratory forelimb use was scored using forelimb asymmetry analysis. The number of ipsilateral, contralateral and both paw placements performed against the chamber wall during vertical/lateral explorations and the paw used to land following exploration were quantified during slowmotion playback of the videotaped session. The results are presented as percent of total ipsilateral forelimb use. Rats injected with AdmuGDNF in either the striatum or the SN were removed from forelimb use analysis due to an overall lack of exploratory movement at all timepoints examined and weight loss. In addition, at each time-point examined, rats which exhibited less than five forelimb placements were removed from analysis. Both striatal and SN vector control rats exhibited a consistent use of the ipsilateral forelimb (⬎20% total ipsilateral forelimb use) from 7 to 28 days after lesion compared with baseline measurements. Rats injected with AdGDNF in the SN also displayed an increased preference for ipsilateral forelimb use. While this deficit gradually increased and became more apparent at both 21 and 28 days after lesion (Figure 2), it was not significantly different from the deficit seen in control groups (P ⬎ 0.05). In contrast, one-way ANOVA revealed that rats injected with AdGDNF in the striatum exhibited a significant reduction in the preferential use of the ipsilateral forelimb 21 days (F3,14 = 3.5, P ⬍ 0.04) and 28 days (F3,15 = 3.6, P ⬍ 0.03) after 6-OHDA lesioning compared with control groups and rats injected with AdGDNF in the SN (Figure 2).

Neuronal protection The protective effects of AdGDNF on DA neuronal cell somata in the SN were examined 35 days after 6-OHDA lesioning. The numbers of TH-IR and FG-labeled DA neurons were determined in the SN using the NeuroLucida System optical fractionator method (Table 2). The protective effect of AdGDNF in rats receiving SN vector injections was compared with rats receiving striatal vector injections. The optical fractionator method was also Table 1 Schedule of surgeries and behavioral tests Day pre (−) or post 6-OHDA lesion Amphetamine test Forelimb use FG and Ad injection 6-OHDA lesion Forelimb use Amphetamine rotation Forelimb use Forelimb use Forelimb use Amphetamine rotation Forelimb use Death

−7 −7 −7 0 7 14 14 21 28 35 35 35

Figure 2 AdGDNF injected in the striatum prevents the preferential use of the ipsilateral forelimb after 6-OHDA lesioning. Baseline spontaneous exploratory forelimb tests were performed 7 days before lesioning (−7). (a) Rats injected with AdGDNF in the SN displayed an increased preference for ipsilateral forelimb use over time. This deficit was not significantly different from that seen in control groups (P ⬎ 0.05). (b) In contrast, rats injected with AdGDNF in the striatum exhibit a significant reduction in the preferential use of the ipsilateral forelimb 21 and 28 days after 6OHDA lesioning compared with control groups (*P ⭐ 0.05).

used to determine the number of TH-IR neurons coexpressing FG. Bilateral striatal injections of the retrograde tracer, FG before 6-OHDA lesioning were used to label the subpopulation of DA neurons in the SN that project to the lesion site. Because the same coordinates were used for FG and 6-OHDA injections, FG labeled DA neurons were most susceptible to the 6-OHDA lesion. FG labeling is a more accurate measure of neuronal protection than TH-IR as it does not rely on the expression of TH in injured neurons. The disappearance of FG-labeled nigral neurons indicates the true loss of structural integrity and cell death.41 In order to confirm that FG protected neurons maintained DA phenotype (ie TH-IR), SN


1939 Table 2 The mean number ± s.e.m. of FG-positive and TH-IR neurons detected throughout the entire rostrocaudal extent of the lesioned and unlesioned SN 35 days after 6-OHDA lesioning Condition

Unlesioned side FG-positive

Untreated AdLacZ in striatum AdLacZ in SN AdmuGDNF in striatum AdmuGDNF in SN AdGDNF in striatum AdGDNF in SN

2389 ± 588 2000 ± 618 1037 ± 249 1741 ± 658 1926 ± 469 3071 ± 267 2619 ± 276

TH-IR 3500 ± 567 3222 ± 609 2296 ± 249 2259 ± 280 3203 ± 66 4185 ± 131 2889 ± 450

Lesioned side FG-positive 667 ± 139 759 ± 383 370 ± 129 518 ± 218 778 ± 64.1 2023 ± 322 1365 ± 247

TH-IR 2333 ± 481 1907 ± 382 1111 ± 96 1129 ± 129 2000 ± 96 3129 ± 359 2073 ± 351

Values for each animal represent the mean ± s.e.m. cell count obtained using non-biased optical fractionator stereological methods in every 6th section throughout the midbrain. Values for the overall coefficient of error ranged from 0.05–0.1.

sections were double-labeled for FG and TH using immunocytochemical techniques. We observed that approximately 31.78 ± 6.9% of TH-IR DA neurons in the unlesioned SN were labeled with FG (Figure 3). The percentage of FG-positive neurons in the lesioned SN compared with the unlesioned SN was examined 35 days after 6-OHDA lesioning. When comparing the number of FG-positive neurons in the lesioned hemisphere as a percentage of the number of FG-positive neurons in the unlesioned hemisphere, only 34 ± 4.1% and 32.5 ± 4.8% FG-positive neurons remained in the lesioned hemisphere of SN vector control and striatal vector control rats, respectively, 35 days after 6-OHDA lesioning (Figures 4 and 5). In contrast, AdGDNF injection resulted in a significant increase in the percentage of FG-positive neurons remaining in the 6-OHDA lesioned hemisphere (F3,29 = 5.9, P ⬍ 0.002). Rats injected with AdGDNF in the SN had 52.6 ± 9.2% FG-positive neurons remaining in the lesioned hemisphere while rats injected with AdGDNF in the striatum had 66.1 ± 8.8% FG neurons remaining (Figures 4 and 5). Tukey–Kramer post hoc comparison tests between treatment groups revealed a significant increase in the percentage of FG neurons observed in rats injected with AdGDNF in either the SN or the striatum compared with control groups (P ⭐ 0.01). However, there was no significant difference in the percentage of FG neurons in the lesioned hemisphere between rats injected

with AdGDNF in the striatum to rats injected with AdGDNF in the SN (P ⬎ 0.05). Within the lesioned hemisphere of control rats, small FG-positive cells were observed within and around the SN (Figure 4). These were not apparent to the same degree in the lesioned hemisphere of AdGDNF injected rats and were not included in stereological cell counts. Furthermore, these small FG-positive cells did not co-express TH-IR and therefore were likely to be microglia or dying neurons. The percentage of TH-IR neurons in the lesioned SN compared with the unlesioned SN was also examined 35 days after 6-OHDA lesioning. In agreement with FGpositive cell counts, AdGDNF injection resulted in a significant increase in the percentage of TH-IR neurons remaining in the lesioned hemisphere (F3,29 = 4.7, P ⬍ 0.008; Figure 6). Tukey–Kramer post hoc comparison tests between treatment groups revealed a significant increase in the percentage of TH-IR neurons observed in rats injected with AdGDNF in the striatum compared with control groups (P ⭐ 0.05). However, in contrast to FG-positive cell counts, AdGDNF injected in the SN did not significantly protect TH-IR neurons in the lesioned SN (P ⬎ 0.05). As discussed above, this may reflect the lack of sensitivity of TH-IR as an indicator of neuronal viability compared with FG-labeling in the lesioned SN. To determine whether there was any difference in the extent of neuroprotection by AdGDNF along the rostro-

Figure 3 Striatal injections of the retrograde tracer, FG labels a subpopulation of TH-IR DA neurons in the SN. In the unlesioned SN, approximately 31.78 ± 6.9% of TH-IR DA neurons (A) are labeled with FG (B). The arrows represent TH-IR neurons which are also labeled with the FG tracer. The section was taken from the unlesioned SN (region III). Both (A) and (B) are taken from the same section; (A) TH-IR taken using brightfield and (B) FG taken using UV fluorescence. Scale bar, 100 ␮m.


1940

Figure 4 AdGDNF injection in either the striatum or the SN prevents the degeneration of FG-positive DA neurons in the SN. Sections are from the SN (region III) in either the unlesioned (A and B) or the lesioned hemisphere (C and D) of control rats or from the lesioned hemisphere of rats injected with AdGDNF in either the striatum (E and F) or the SN (G and H). Many large, FG-positive cells (ie DA neurons) remain in the lesioned hemisphere in rats injected with AdGDNF in either the striatum (F, arrow) or the SN (H, arrow). In contrast, fewer large FG-positive cells survive in the control rats. Within and around the lesioned SN of control rats (C and D), small FG-positive cells were observed (D, *) which may be dying neurons or microglia which secondarily take up FG after degeneration of DA neurons. Scale bars: A, C, E and G = 100 ␮m; B, D, F and H = 50 ␮m.

caudal axis of the lesioned SN, we divided the SN into three regions: anterior/MT (sections anterior to the medial terminal nucleus of the accessory optic tract, MT, through to the section in which the MT is visible), III (sections in which fibers of the oculomotor nerve pass through the SN/VTA) and posterior (posterior to III). For each treatment group (striatal vector control, SN vector control, AdGDNF in the striatum and AdGDNF in the SN), the number of FG-positive neurons in the lesioned hemisphere was determined. AdGDNF significantly protected FG-labeled DA neurons in all three of the regions examined (Ant/MT region; F3,33 = 5.45, P ⬍ 0.003: region III; F3,92 = 10.1, P ⬍ 0.0001: posterior region; F3,126 = 12.2, P ⬍ 0.0001). However, a difference in the extent of neuroprotection along the rostrocaudal axis of the lesioned SN was observed between rats injected with AdGDNF either in the SN or the striatum (Figure 7). Tukey–Kramer post hoc comparison tests between each treatment group indicated that rats which received AdGDNF in the SN displayed significant protection in the Ant/MT region of the

SN compared with control groups (P ⭐ 0.05). In contrast, injection of AdGDNF to the striatum significantly protected FG-positive neurons in region III and the posterior region of the SN (P ⭐ 0.05) but not the Ant/MT region of the SN compared with control groups. In addition, rats injected with AdGDNF in the striatum exhibited a greater degree of protection in region III of the lesioned SN than rats injected with AdGDNF in the SN (P ⭐ 0.05).

Effect of GDNF on DA fibers The reduction in behavioral impairment observed in rats with striatal AdGDNF injections may be correlated with an increase of DA nerve terminals in the lesioned striatum. In order to determine whether AdGDNF protected DA nerve terminals, the extent of TH-IR fiber innervation (lesion size) was examined throughout the lesioned striatum at 240 ␮m intervals (all regions determined anterior or posterior from lesion site). TH-IR fiber innervation was greatly reduced in the lesioned striatum around the site of 6-OHDA injection. However, the area of reduced TH-


1941

Figure 5 AdGDNF significantly prevents the degeneration of FG-positive DA neurons in the lesioned SN 35 days after 6-OHDA lesioning. A significant increase in the percentage of FG-positive neurons was observed in the lesioned SN in rats injected with AdGDNF in either the striatum or the SN compared with control groups (*P ⭐ 0.01). There was no significant difference in the percentage of FG-positive neurons between rats injected with AdGDNF in the striatum versus the SN (P ⬎ 0.05).

Figure 6 AdGDNF significantly prevents the degeneration of TH-IR neurons in the lesioned SN 35 days after 6-OHDA lesioning. A significant increase in the percentage of TH-IR neurons was observed in the lesioned SN in rats injected with AdGDNF in the striatum compared with rats injected with AdGDNF in the SN and control groups (*P ⭐ 0.05).

IR fiber innervation (lesion size) was significantly decreased in rats injected with AdGDNF in the striatum compared with control groups (480 ␮m anterior: ANOVA, F3,24 = 5.8, P ⬍ 0.004; 240 ␮m anterior: ANOVA, F3,24 = 4.89, P ⬍ 0.008; lesion site: ANOVA, F3, 24 = 5.4, P ⬍ 0.005; 240 ␮m posterior: ANOVA, F3,24 = 9.73, P ⬍ 0.0002; 480 ␮m posterior: ANOVA, F3,24 = 4.4, P ⬍ 0.01) suggesting that AdGDNF injection in the stria-

Figure 7 AdGDNF significantly protects FG-positive DA neurons in all three regions examined (Ant/MT, sections anterior to the medial terminal nucleus of the accessory optic tract; MT, through to section in which the MT is visible; III, sections in which fibers of the oculomotor nerve pass through the SN/VTA; post, posterior to III). Rats injected with AdGDNF into the SN display a significant protection in the Ant/MT region of the SN compared with control groups (*P ⭐ 0.05). In contrast, rats injected with AdGDNF in the striatum display a significant protection in region III and the posterior region of the SN compared with control groups (**P ⭐ 0.05). In addition, rats injected with AdGDNF in the striatum exhibit a significantly greater degree of protection in region III of the lesioned SN than rats injected with AdGDNF in the SN (**P ⭐ 0.05).

tum protects TH-IR fibers or stimulates neuronal sprouting of these in the denervated striatum (Figure 8). Activation of the transcription factor, c-fos is associated with increased D1 receptor stimulation. In the striatum, DA receptor-activated expression of Fos protein has been used as a marker of DA-mediated postsynaptic responses [42–45]. Normal rats express Fos after injection of an indirect DA agonist, such as amphetamine.46,47 The ability of amphetamine to induce Fos in the striatum is thought to be primarily caused by its ability to release endogenous DA which, via activation of D1 receptors, triggers a transduction cascade culuminating in the expression of the c-fos gene in striato-nigral cells.48–51 In order to determine whether AdGDNF injection maintained or increased Fos expression in the lesioned striatum, rats were injected with amphetamine (5 mg/kg i.p.) and then killed 2 h after drug injection. The number of Fos-IR neurons in the lesioned hemisphere was compared as a percentage of the number of Fos-IR neurons in the unlesioned hemisphere at the lesion site and approximately 700 ␮m posterior to the lesion (400 ␮m anterior to Ad vector striatal injection). Amphetamine induced high levels of Fos expression in the striatum of normal rats. In rats with 6-OHDA lesions, amphetamine-induced Fos expression was observed in both the lesioned and unlesioned striatum. However, the level of Fos-IR was significantly reduced in the lesioned compared with the unlesioned hemisphere. Amphetamine requires intact DA terminals to increase c-fos expression via activation of D1 receptors by endogenous DA.51 A reduction in the percentage of Fos-IR neurons, reflecting a decrease in functional DA teminals was seen in both control groups and rats injected with AdGDNF in the SN at the lesion site (striatal vector control = 68.5 ± 5.9%; SN vector control = 70.5 ± 5.7%; AdGDNF in SN = 79.7 ± 16.7%) and 700 ␮m posterior to the lesion site (striatal vector control = 86 ± 11.9%; SN vector control = 95.3 ± 10.2%; AdGDNF


1942

in SN = 57.6 ± 16.2%; Figure 9). In contrast, rats injected with AdGDNF in the striatum exhibited a significant increase in the percentage of Fos-IR neurons at both the lesion site (112.5 ± 16.7%; F3,25 = 3.13, P ⬍ 0.04) and 700 ␮m posterior to the lesion site (128 ± 20.8%; F3,26 = 3.4, P ⬍ 0.03) compared with control groups and rats injected with AdGDNF in the SN (Figure 9).

Figure 8 (a) An increase in TH-IR fibers are seen in the lesioned striatum in rats injected with AdGDNF in the striatum (C) compared with rats injected with AdGDNF in the SN (B). (A) represents the extent of TH-IR fiber innervation in the unlesioned striatum. Each area has been captured 400 ␮m from the visible needle tract in a section taken from the site of 6OHDA injection in the anterior striatum. The sections have been stained for TH-IR using DAB (see Materials and methods). Scale bar, 50 ␮m. (b) Lesion size is significantly reduced in rats injected with AdGDNF in the striatum compared with control groups (*P ⭐ 0.05). Lesion size represents the area of reduced TH-IR density in the lesioned striatum.

Transgene expression GDNF transgene expression was examined 6 weeks after vector injection using an antibody specifically directed against human GDNF. In order to visualize the injection site, AdGDNF and AdmuGDNF vectors were ‘spiked’ with 25% AdLacZ which was visualized using X-Gal histochemistry. In rats injected with AdGDNF or AdmuGDNF, in either the striatum or the SN, blue nuclei were observed surrounding the injection site, but were not quantified. In addition, rats injected with AdGDNF in either the striatum or SN exhibited a GDNF-IR penumbra extending from the site of vector injection, indicating the persistence of human GDNF transgene protein until the end of the experiment (Figure 10). Within this penumbra, GDNF-IR neurons were observed, indicating the presence of AdGDNF transgene expressing cells (Figure 10). As GDNF is a secreted protein, this penumbra likely represents the area over which GDNF is secreted at high levels from transgene expressing cells. Six weeks after vector injection, six out of seven rats injected with AdGDNF in the striatum exhibited AdGDNF transgene expression. In these six rats, the GDNF-IR penumbra had a mean total area of 12.3 ± 2.84 mm2 (mean ± s.e.m.) surrounding the vector injection site. Within the GDNF penumbra, a mean total of 15 397.3 ± 5583.4 GDNF-IR cells was detected. GDNF-IR cells were also detected in the corpus callosum and globus pallidus of some rats. Transgene expression was also observed in four of seven rats injected with AdGDNF in the SN 6 weeks after vector injection. GDNF-IR neurons were observed throughout much of the rostrocaudal extent of the SN. In the four rats expressing GDNF-IR, the GDNF-IR penumbra had a mean total area of 3.58 ± 1.1 mm2 and within the GDNF penumbra, a mean total of 12 568.4 ± 9962.2 GDNF-IR cells was detected. Rats that did not show GDNF-IR were not removed from the study since a vector injection site was identified in each rat and no difference in host immune response was observed compared with rats exhibiting GDNF-IR. We believe that the inability to visualize GDNF-IR in these rats is due to the low sensitivity of GDNF immunocytochemistry. Supporting this, we have observed that GDNF Elisa is more sensitive than GDNF-IR in studies of monkey caudate injected with this same vector (unpublished observations, Kozlowski, DA et al). In addition, we have previously reported that GDNF expression is down-regulated with time after injection of AdGDNF into the rat brain.24,27 Endogenous GDNF is retrogradely transported within specific regions of the CNS.12,52 In addition, retrograde transport of Ad vectors from the striatum to the SN has previously been reported.20,27,53 Therefore, in rats injected with AdGDNF in the striatum, the retrograde transport of AdGDNF from the striatum to the SN was examined. Using immunocytochemistry, we were unable to detect retrogradely transported GDNF in cells within the SN in rats that received a striatal injection of AdGDNF. However, GDNF protein has previously been detected by


1943

Figure 9 (a) The level of amphetamine-induced Fos-IR in the lesioned striatum. Amphetamine-induced Fos-IR is observed in the unlesioned striatum (A). A reduction in the level of Fos-IR is observed in the lesioned striatum of both control rats (B) and rats injected with AdGDNF in the SN (D). In contrast, the level of Fos-IR in the lesioned striatum of rats injected with AdGDNF in the striatum (C) is greatly increased compared with the unlesioned striatum. These sections have been taken from the site of 6-OHDA injection in the anterior striatum and have been stained for Fos-IR using DAB (see Materials and methods). Scale bar, 100 ␮m. (b) Rats injected with AdGDNF in the striatum exhibit a significant increase in the percentage of Fos-IR cells at both the lesion site and 700 ␮m posterior to the lesion site compared with rats injected with AdGDNF in the SN and control groups (*P ⭐ 0.05). In contrast, a reduction in Fos-IR was observed both in rats injected with AdGDNF in the SN and in control rats.

Elisa in the SN at 1 week after the injection of AdGDNF into the striatum.27 GDNF may have been present in the SN in our study at levels too low to detect by immunocytochemisty. In agreement with AdGDNF transgene expression, 5bromo-4-chloro-3-indolyl-␤-galactopyranoside (X-Gal) staining of rats injected with AdLacZ revealed blue nuclei 6 weeks after vector injection. In rats injected with AdLacZ in the striatum (n = 3), the majority of blue nuclei was detected within an area 1.0 to 1.8 mm2 of the injection site. As reported with AdGDNF transgene expression, blue nuclei were also observed within the

corpus callosum and globus pallidus of some rats. Six weeks after striatal AdLacZ vector injection, 15 900 ± 2206 blue nuclei were detected throughout the striatum. Blue nuclei were observed in three of four rats injected with AdLacZ in the SN 6 weeks after vector injection. In these rats, the majority of blue nuclei was detected within an area 0.4 to 0.6 mm2 of the injection site. Blue cells were observed throughout much of the rostrocaudal extent of the SN. In contrast to striatal AdLacZ transgene expression, 2580 ± 287.5 blue nuclei were detected in rats injected with AdLacZ in the SN 6 weeks after vector injection.


1944

Figure 10 GDNF transgene protein was detected by immunocytochemistry 6 weeks after vector injection both in rats injected with AdGDNF in the striatum (A) and in rats injected with AdGDNF in the SN (B). Six weeks after vector injection, a GDNF-IR penumbra is observed extending from the injection site in rats injected with AdGDNF in either the striatum (A) or the SN (B). Within this penumbra, numerous GDNF-IR cells are observed (C, arrow). Scale bar, 250 ␮m (A, B); 50 ␮m (C).

Host immune response The degree of infiltrating cells into the vector injection site was assessed 6 weeks after FG and Ad injection, using H&E staining (Table 3). While cellular infiltrates were observed in the unlesioned striatum injected only with FG, the lesioned striatum that was injected with FG, 6-OHDA and Ad vector showed increased levels of infiltration. In rats that received vector injection in the striatum, the typical score for each section was between 1 and 4. However, there were no significant differences in the degree of cellular infiltration between rats striatally injected with either AdLacZ or AdmuGDNF and rats injected with AdGDNF (Mann–Whitney, P ⬎ 0.5). Rats that were injected in the SN had levels ranging between 1 and 3. As observed with striatal vector injections, rats Table 3 Host immune response to adenoviral vector injection Group AdLacZ in striatum AdLacZ in SN AdmuGDNF in striatum AdmuGDNF in SN AdGDNF in striatum AdGDNF in SN

Inflammatory rating 3 ± 0.2 2 ± 0.2 2 ± 0.3 2 ± 0.33 4 ± 0.4 2 ± 0.3

Rating scale: 1 = no/few inflammatory cells; 2 = mild inflammatory response; 3 = moderate inflammatory response; 4 = moderate/severe inflammatory response; 5 = severe inflammatory response.

that were injected in the SN had no significant differences in the degree of cellular infiltration between either AdLacZ or AdmuGDNF injected rats and rats injected with AdGDNF (Mann–Whitney, P ⬎ 0.5). The Fischer-344 rat is not the ideal host to characterize host immune response due to the low level of immune response displayed by this species. A more detailed characterization of host immune responses to the same vectors used in this study has recently been described in the non-human primate striatum.54,55

Effect of AdGDNF on body weight Previous studies have indicated that the injection of recombinant GDNF protein can cause a reduction in body weight in rats15,32,33,56,57 In order to determine whether the injection of AdGDNF in either the striatum or the SN affected body weight in aged (20 months) rats, weights were recorded every 7th day throughout the experiment. As the aged rats in this study displayed poor recovery from both set of surgeries and exhibited symptoms of dehydration, all rats were injected with 5 cc Ringers lactate solution (subcutaneous) per day for 3 weeks after 6-OHDA lesioning to aid recovery and prevent body fluid loss. Following the first set of surgeries, in which rats received bilateral injections of FG, as well as Ad vectors or no additional injection, rats lost an average of 38.2 g (8.3% of body weight; baseline = 459 ± 76g, 1 week after vector/FG injection = 421 ± 5.31 g). Following the second surgery, in which the rats were unilaterally injected with 6-OHDA, the rats lost an average of 17.5 g (4.2% of body weight; 1 week after 6-OHDA lesion = 402 ± 5.2 g). Following this the average weight in all treat-


ment groups except rats injected with AdmuGDNF in the SN increased or were maintained at a constant level until death. In contrast to the other treatment groups, rats injected with AdmuGDNF in the SN continued to lose weight until 3 weeks after lesion (average weight loss 22.5 g, 6% of body weight; baseline = 467 ± 24 g, 1 week after vector/FG injection = 402 ± 18 g, 1 week after 6OHDA lesion = 395 ± 30 g, 2 weeks after 6-OHDA lesion = 380 ± 55 g, 3 weeks after 6-OHDA lesion = 352 ± 101 g) after which average body weight increased until death. No significant differences in weight were observed among any of the other treatment groups, including rats injected with AdGDNF (F6,19 = 1.22, P = 0.34).

Discussion Age-related reductions in neurotransmitter synthesis and alterations in receptor levels may contribute to ongoing degenerative and behavioral deficits in Parkinson’s disease. Multiple functional changes of the nigrostriatal dopaminergic system, similar to those observed in the Parkinsonian brain, have been described during normal aging in both animals and humans.58–61 Compensatory events may be observed in the aged brain due to agedependent degenerative changes occurring in the nigrostriatal and mesolimbic pathways. Alternatively, the aged brain may be unable to produce compensatory changes in response to either age-dependent or lesion-induced neuronal degeneration, worsening lesion-induced degenerative alterations in dopaminergic neuronal morphology and function in the aged brain.35–38 As Parkinson’s disease occurs primarily in aged populations, we examined whether increasing neurotrophic factor levels would protect the nigrostriatal system in the aged rat when challenged with the neurotoxin, 6-OHDA. Furthermore, possible differential effects of injecting AdGDNF in the striatum or in the SN were examined. We observed that increased levels of striatal GDNF biosynthesis prevents DA neuronal loss and protects DA terminals from 6-OHDA-induced damage, thereby maintaining DA function in the aged rat. Injections of AdGDNF into either the SN or striatum were similarly effective in protecting DA neurons in the SN of the aged rat brain. In contrast, only the striatal injections of AdGDNF were effective in preventing the development of amphetamine-induced rotational behavior and reducing the preference for ipsilateral forelimb use. Furthermore, rats injected with AdGDNF in the striatum, but not in the SN, exhibited a reduction in TH fiber loss and an up-regulation of amphetamine-induced Fos expression in the lesioned striatum. This suggests that increased levels of striatal GDNF biosynthesis not only protected DA cell bodies against the 6-OHDA challenge, but also increased nigrostriatal input to the lesioned striatum. The delivery of 6-OHDA intrastriatally results in the rapid destruction of DA terminals in the striatum, followed by progressive degeneration of DA soma in the SN over a period of several weeks.41 This degeneration is preceded and accompanied by cellular atrophy and a partial loss of TH expression, yielding an animal model which closely resembles the early stages of Parkinson’s disease in humans in which a portion of the nigrostriatal projection remains intact. These remaining neurons may serve as a substrate for regeneration and functional recovery.39 In a partial 6-OHDA lesion model, GDNF

could exert its neurotrophic effects both at the level of DA cell bodies in the SN or at the level of DA axon terminals in the striatum. Injection of AdGDNF into the striatum will produce GDNF expression at both the terminals and the somata of the nigrostriatal pathway via retrograde transport.27 In contrast, AdGDNF injected into the SN will produce GDNF mainly near DA somata. Striatal injections of AdGDNF before lesioning may have prevented the initial loss of DA terminals in the striatum or, though less likely, may have interfered with the up-take of 6-OHDA so that cells were not subsequently susceptible to degeneration. Injections of AdGDNF either in the SN or striatum of young Fischer-344 rats have previously been shown to protect DA neurons from 6-OHDA-induced degeneration.24,27,28 Extending these observations, AdGDNF significantly protected FG-labeled DA neurons throughout the entire rostrocaudal extent of the lesioned SN in the aged rat brain. While the level of neuroprotection observed in aged rats receiving striatal AdGDNF injections was similar to that previously reported in young rats using the same lesion paradigm,27 aged rats injected with AdGDNF in the SN exhibited a lower magnitude of neuroprotection in the lesioned SN compared with young rats.24 This either reflects a reduction in compensatory responses35 or an alteration in the distribution and/or expression of GDNF receptors in the aged rat brain. The GDNF receptor complex consists of two components, GFR␣-1 and the Ret receptor, both of which are expressed by nigral DA neurons.62–64 While the expression of GFR␣1 and Ret have not been examined in the aged rat brain, age-related alterations in the distribution or expression of the GDNF receptor complex may reduce the protective effects of GDNF in the lesioned SN. A difference in the degree of neuroprotection was also observed along the rostrocaudal axis of the lesioned SN between rats injected with AdGDNF either in the SN or striatum, possibly reflecting differences in the site of vector injection. AdGDNF SN injections were made within the MT region of the SN (5.3 mm posterior and 1.8 mm lateral to bregma), suggesting the increased level of protection seen in the anterior/MT region of the SN in rats with AdGDNF SN injections was due to local protective effects. In contrast, the pattern of neuronal protection observed throughout the SN in rats with striatal AdGDNF injections possibly reflects retrograde transport of AdGDNF from the striatum to the SN. The site of striatal vector injection may be important for maximum protection of both DA cell bodies in the SN and DA terminals in the striatum. Kirik and colleagues65 reported an increased denervation effect and a greater loss of TH-IR cells in the lesioned SN when 6-OHDA injections were placed close to the CPu-globus pallidus junction (‘preterminal’ site) compared with injection sites inside the CPu (‘terminal’ site). In our study, striatal vector injections were made 0.1 mm posterior and 3.2 mm lateral from bregma, corresponding to the ‘terminal’ injection site. Therefore, placement of striatal AdGDNF injections close to the CPu-globus pallidus junction in future studies may extend the degree of neuroprotection. In addition to the neuroprotective effects of AdGDNF within the lesioned SN, striatal injections of AdGDNF reduced TH fiber loss and upregulated amphetamineinduced Fos expression in the lesioned striatum. This suggests that increased levels of striatal GDNF protected

1945


1946

DA terminals in the aged rat brain. Previous studies have reported that GDNF protein16,66 or AdGDNF delivered into the striatum28 restores DA innervation of the striatum compared with control treatments. However, previous research in our laboratory reported that striatal injections of AdGDNF did not affect mean TH-IR fiber density measurements in the lesioned striatum when examined 42 days after lesion.27 In contrast, this study and a recent study by Bilang-Bleuel and colleagues28 observed a significant reduction in TH-IR fiber loss in rats injected with AdGDNF in the striatum when striatal THIR levels were examined at 35 and 21 days after lesion, respectively. Therefore, spontaneous recovery at a longer survival time may account for the lack of protection or induction of DA terminal sprouting in the striatum reported by Choi-Lundberg and colleagues.27 The protection of DA terminals by increased levels of striatal GDNF is further supported by the observation that striatal injections of AdGDNF significantly increased the expression of the transcription factor, Fos, in the lesioned striatum following amphetamine injection. This suggests that denervation-induced changes in striatal Fos activation, induced either by direct or indirect D1 receptor activation, can be upregulated by striatal injections of AdGDNF. The mechanism by which AdGDNF up-regulated D1 receptor activation may be through GDNF protecting DA terminals from degeneration or by stimulating regeneration or sprouting. Striatal injections of AdGDNF also prevented the development of amphetamine-induced rotational behavior. Single intranigral injections of GDNF (10 ␮g) have previously been shown to enhance basal and stimulus-evoked DA overflow and whole tissue neurochemical levels 3 weeks after administration in aged Fischer-344 rats.31 Therefore, the reduction in ipsilateral rotational behavior combined with the increased expression of Fos protein observed in rats injected with AdGDNF in the striatum may reflect an enhanced amphetamine-induced release of DA in the denervated striatum. Alternatively, AdGDNF may have direct effects on striatal neurons thus altering post-synaptic responses to DA. In contrast to the reduction in ipsilateral rotational behavior observed in rats injected with AdGDNF in the striatum, we observed that rats injected with AdGDNF in the SN exhibited a significant increase in amphetamine-induced rotational behavior 35 days after lesion compared with control groups. The mechanisms underlying the increase in rotational behavior is unclear, although our data would suggest that it results from an increase in DA in the SN in the absence of an increase in DA in the striatum. The rate of amphetamine-induced rotations observed in aged rats was comparable to that seen in our previous study using young Fischer-344 rats.27 However, previous studies using different strains of young rats have reported higher levels of ipsilateral rotations per minute in response to amphetamine administration than observed in this study.67 This may reflect a strain and sex difference in amphetamine-induced rotational behavior68 or more likely, a difference in the extent of rotational asymmetry between partial and complete 6-OHDA lesions. Furthermore, aged rats displayed a reduction in movement during both rotational behavior and spontaneous exploratory forelimb use compared with young rats. Aged rats also exhibited an altered time course of forelimb asymmetry compared with our previous

study.27 Although different time-points and a revised method of scoring were used in the previous study, in otherwise untreated adult-aged rats with a wide range of partial dopamine depletion there is no increase in the degree of asymmetry over time.69 In conclusion, GDNF gene therapy has potential use in the clinical treatment of Parkinson’s disease. As Parkinson’s disease is a progressive disorder, we feel that both protective and rescue paradigms are important as approaches to this disease. The protective GDNF paradigm aims at preventing further loss of cells and function while the rescue paradigm aims at reversing damage to DA neurons. Our results indicate that, with regard to preventing DA neuronal loss and protecting DA terminals from degeneration, the striatum is a more desirable site for GDNF therapy than the SN. This finding is relevant for the application of gene therapy to patients with Parkinson’s disease. However further technological advances are required to realize the potential of gene therapy for Parkinson’s disease. Recent developments in lentiviral and adeno-associated viral vectors have resulted in transgene expression within the CNS being detected for greater than 6 months after injection without cytotoxic effects.70–74 Furthermore, research is being undertaken to improve the duration of adenoviral transgene expression and reduce host immune response. Currently, new generation Ad vectors are being developed with deletions in either the E2a or E2b region,75–79 or E4 region80–82 or total gutting of viral sequences except for the inverted terminal repeats and packaging signal.83–85 While these new generation Ad vectors have not been tested yet in the CNS, they would hopefully result in increased transgene expression and reduced cytotoxic effects.86 Furthermore, the development of vectors harboring genes driven by specific cellular promoters would provide the potential of limiting expression of a transgene to one phenotypically defined cell population. Although, it is not known how closely neurotoxininduced lesions mimic the state of diseased neurons in humans with Parkinson’s disease, the findings in this study suggest that GDNF gene delivery near the terminals of DA neurons has the potential of ameliorating cellular and functional consequences that occur when the nigrostriatal pathway is compromised in aged humans.

Materials and methods Vectors Adenoviral vectors with early gene E1 and E3 deletions were used. Vectors encoding either human GDNF preproprotein (AdGDNF), a non-biologically active mutant form of human GDNF with a 12 amino acid deletion in the mature protein (AdmuGDNF) or the reporter enzyme ␤-galactosidase with a nuclear localization signal (AdLacZ) were prepared and titered as previously described.24 Both AdLacZ and AdmuGDNF were used to control for possible effects of the adenoviral vector. In addition, AdmuGDNF was used as a deletion mutation to control for AdGDNF. Viral stocks had infectious particles (plaque forming units, p.f.u.), total particle concentrations and particle ratios (total particles per infectious particles) as shown in Table 4.


Table 4 Titers and particle ratios of adenoviral vectors p.f.u./ml AdGDNF 5 × 1010 AdmuGDNF 2.2 × 1010 AdLacZ 1.0 × 1010

Total particles/ml

Particle ratio

0.47 × 1012 0.55 × 1012 0.92 × 1012

9.4 25.2 91.8

Animals Twenty-month-old Fischer 344 male rats (400–500 g; NIA/Harlan Sprague–Dawley, Indianapolis, IN, USA) were housed at the Children’s Memorial Institute for Education and Research, Northwestern University vivarium and were treated in accordance with institutional and NIH guidelines. The rats were kept in a 12h dark/light environment and food and water were provided ad libitum. Rats were randomly assigned to receive either AdGDNF (striatal injection n = 7, SN injection n = 7), AdLacZ (striatal injection n = 3, SN injection n = 4), AdmuGDNF (striatal injection n = 3, SN injection n = 3) or no treatment (n = 3), 7 days before 6-OHDA injection. Surgical procedures All surgeries were performed using an i.p. injection of 149 mg chloral hydrate and 31 mg sodium pentobarbital per kg body weight. As previously described,27 rats received bilateral striatal injections of Fluorogold (FG; Fluorochrome, Englewood, CO, USA) to label a subpopulation of DA neurons retrogradely within the SN that projected to the site of the lesion. Injection of adenoviral vector at the same time as injection of FG has previously been shown not to effect the retrograde transport and distribution of the FG tracer in the rat brain.24 Following FG injection, one group of rats was injected unilaterally in the striatum, while a second group of rats was injected unilaterally into the SN with either AdLacZ, AdmuGDNF or AdGDNF. An additional group of rats underwent surgery, but received no vector injection into either the striatum or the SN. AdGDNF, AdmuGDNF and AdLacZ were diluted in sterile 20% sucrose in PBS. In order to visualize the injection site, the AdmuGDNF and AdGDNF vectors were ‘spiked’ with 25% of AdLacZ vector which can be visualized in the tissue using X-Gal histochemistry. 2 ␮l of AdLacZ vector or 2.8 ␮l of AdGDNF or AdmuGDNF vector spiked with AdLacZ were unilaterally injected into either the striatum or the SN at 0.5 ␮l/min using a 10 ␮l Hamilton syringe with a 30-gauge needle at the following coordinates: striatum: 0.1 mm posterior and 3.2 mm lateral to bregma and 5.0 mm ventral to the surface of the dura; SN: 5.3 mm posterior and 1.8 mm lateral to bregma and 7.4 mm ventral to the surface of the dura. After each injection, the needle was left in place for 5 min and then withdrawn at 1 mm/min. Vector titers were matched for total particles at approximately 1–2 × 109. Seven days later, rats were again anesthetized and placed in stereotaxis. A unilateral progressive 6-OHDA lesion of nigrostriatal system was generated using the rat model of Parkinson’s disease described by Sauer and Oertel.41 As previously described,27 6-OHDA (diluted to 16 ␮g in 2.8 ␮l of 0.2 mg/ml ascorbic acid in 0.9% sterile saline) was injected unilaterally into the striatum on the

same side as the vector injection and at the same coordinates as the FG injection. The rats were killed 35 days after 6-OHDA lesioning (42 days after vector injection). Before death, the rats were injected with DL-amphetamine (5 mg/kg i.p. injection) in order to examine amphetamine-induced Fos expression. Two hours after amphetamine injection, the rats were deeply anesthetized with 1 cc of sodium pentobarbital and transcardially perfused with 200 ml 0.9% saline followed by 500 ml 4% paraformaldehyde in 0.1 m phosphate buffer (PB), pH = 7.4. The brains were removed and post-fixed in 4% paraformaldehyde in PB overnight. The brains were then cryoprotected in 30% sucrose in PB. Two rats, both injected with AdGDNF in the SN, died 5 days before the planned sacrifice date. They were included in the study and their brains were removed and post-fixed in 4% paraformaldehyde in 0.1 m phosphate buffer (PB), pH = 7.4 for 1 week before being placed in 30% sucrose in PB. Frozen coronal sections (40 ␮m) were processed with a sliding microtome in six sets (equaling a distance of 240 ␮m between each section in one set) and stored in cryoprotective solution at −20°C.

Behavioral testing Amphetamine-induced rotation test: Rats were injected with DL-amphetamine sulfate (5 mg/kg i.p. injection) and placed in a plastic bowl (depth 18 cm; diameter 38 cm) for 60 min. Rotational behavior was recorded on video tape and rated. Baseline amphetamine rotation tests were performed 7 days before 6-OHDA lesions and the results used to assign the side of the subsequent 6OHDA lesion. Rats exhibiting net clockwise turns were lesioned in the left striatum, while rats exhibiting net counter-clockwise turns were lesioned in the right striatum. Following the 6-OHDA lesion, amphetamineinduced rotational behavior was tested 14 and 35 days after lesion (Table 1). The number of clockwise and counter-clockwise turns was counted and expressed as the number of net rotations per minute to the lesioned (ipsilateral) hemisphere. Spontaneous exploratory forelimb use: Spontaneous exploratory forelimb use was scored using forelimb asymmetry analysis modified from previous work.17,87–91 Rats were videotaped for 5 min in a 1 cubic foot Plexiglas chamber before and after 6-OHDA lesioning (Table 1). Movements were analyzed by an experimenter blind to the treatment groups. The videotape was played in slow motion to quantify weight-bearing movements performed by the forelimbs during exploration; specifically, the number of ipsilateral (to lesion), contralateral and both paw placements performed against the chamber walls during vertical/lateral explorations and during landing following exploration of the walls were quantified. The percentage of total ipsilateral forelimb use was determined by first calculating the percentage of ipsilateral wall (ipsilateral wall placement/ipsilateral wall + contralateral wall + both wall placement × 100) and ipsilateral land (ipsilateral land placement/ipsilateral land + contralateral land + both land placement × 100) movements. The percentage of ipsilateral wall and ipsilateral land placements were then added together and divided by two. This calculation assured that the wall and land movements were weighted evenly and resulted in a measure

1947


1948

of ipsilateral forelimb use relative to total forelimb movement. At each time-point examined, rats were removed from analysis if they exhibited less than five forelimb placements. The quantification of total ipsilateral forelimb use in this study was modified from the method of quantification used in our previous study27 as the current method is considered more reliable and sensitive to slight variations in DA levels in partially depleted rats.69

Histology X-gal histochemistry: AdLacZ transgene expression was visualized by histochemistry with X-Gal (5-bromo-4chloro-3-indoyl-b-D-galactopyranside). Sections (40 ␮m thick, every 240 ␮m) were first incubated for 10 min at 4°C in phosphate buffered saline (PBS) with 2 mm MgCl2, followed by 10 min incubation at 4°C in PBS with 2 mm MgCl2, 0.01% sodium desoxycholate and 0.02% NP40. The sections were then incubated for 4–6 h at 37°C in 25 mm K3Fe(CN)6, 25 mm K4Fe(CN)6, 2 mm MgCl2 and 1 mg/ml X-gal (Gibco BRL, Grand Island, NY, USA) in PBS. The number of blue nuclei was counted every 240 ␮m, summed and multiplied by six to obtain an estimate of the total number of transgene expressing cells. GDNF- and Fos-immunocytochemistry: Paraformaldehyde-fixed substantia nigra and striatal sections (40 ␮m) were incubated in 0.3% hydrogen peroxide for 15 min. Sections were blocked for 60 min at room temperature in 3% normal goat serum (NGS) and 2% bovine serum albumin (BSA) in PBS containing 0.05% TritonX100 (TX, Sigma, St Louis, MO, USA). The sections were then incubated overnight at room temperature in either a biotinylated polyclonal rabbit anti-GDNF antibody (R&D Systems, Minneapolis, MN, USA) diluted 1:250 in 1% NGS, 1% BSA and 0.05% TX in PBS or a polyclonal antiFos antibody (Santa Cruz, Santa Cruz, CA, USA; cat No. sc-52-G) diluted 1:1000 in 1% NGS, 1% BSA and 0.05% TX in PBS. The sections were then incubated for 2.5 h at room temperature in a biotinylated goat anti-rabbit secondary antibody (Vector Laboratories, Burlingame, CA, USA) diluted 1:500 in 1% NGS and 1% BSA in PBS. For immunocytochemical visualization, the sections were incubated for 2 h in avidin-peroxidase conjugate (Vectastain ABC Elite, Vector Laboratories, Burlingame, CA, USA) in PBS, followed by 2–3 min in 0.4 mg/ml diaminobenzidine (DAB), 0.8 g/ml nickel sulfate, 0.005% H2O2 in 50 mm sodium acetate, 10 mm imidazol, pH = 7.0. Between each step, sections were washed several times in either PBS with 0.05% TX or PBS alone. Tyrosine hydroxylase-immunocytochemistry: Paraformaldehyde-fixed substantia nigra and striatal sections (40 ␮m) were visualized for tyrosine hydroxylase using a polyclonal rabbit anti-tyrosine hydroxylase (TH) antibody (Chemicon, Temecula, CA, USA) diluted 1:2000 in 1% NGS and 0.3% TX in TBS. The sections were incubated for 24 h at room temperature in the TH primary antibody. This was followed by incubations in a biotinylated goat anti-rabbit secondary antibody (Vector Laboratories) diluted 1:500 in 1% NGS and 0.1% TX in TBS and then avidin-biotin-peroxidase complex (Vectastain ABC Elite, Vector Laboratories). The reaction was visualized using DAB as a chromagen with nickel sulfate enhancement. Between each step, sections were

washed several times in either TBS with 0.1% TX or TBS alone.

Antibody control studies: A preabsorption control study was performed to verify the specificity of the anti-GDNF primary antibody used. Anti-GDNF primary antibody (R&D Systems) was incubated with recombinant human GDNF protein (R&D Systems) at a ratio of 10:1 (protein:antibody) by weight for 3 h at room temperature. In order to control for non-specific binding, a series of sections was incubated with the antibody/protein solution as described above. Preabsorption of the antiGDNF primary antibody with GDNF protein resulted in a lack of immunostaining, confirming the specificity of the primary antibody. In addition, the specificity of individual primary antibodies were confirmed for each immunocytochemical stain by omitting incubation of sections with the primary antibody. Host immune responses The degree of host immune response to Ad vector injection was determined using H&E staining. The intensity of inflammation throughout the SN and the striatum was scored by an observer blind to the treatment groups and injection site. A rating scale of 1–5 based on intensity of the accumulation and the extent of spread of inflammatory infiltrates was used. A score of 1 was the least severe, with sparse inflammatory cells concentrated only along the injection needle tract, and a score of 5 the most severe in which the needle tract was heavily populated by inflammatory cells that spread into surrounding neural tissue. An average score was obtained for each rat. The degree of host immune response was compared in rats receiving SN vector injections with rats receiving striatal Ad vector injections. Morphometry Substantia nigra cell counts: Non-biased stereological cell counts of FG-positive cells, TH-IR cells and cells that coexpressed both FG and TH were performed by an observer blind to the treatment groups. Cell counts were performed in every 6th section throughout the rostrocaudal extent of the midbrain DA cell groups (SN (A9), VTA (A10) and A8) and within the dorsal to ventral aspect of the SN 5 weeks after 6-OHDA lesioning using Optical Fractionator stereological methods (StereoInvestigator Software, MicroBrightField, Colchester, VT, USA). A section every 240 ␮m was examined. Cell counts throughout the entire structures were conducted using a Nikon Inverted microscope (Eclipse TE300) equipped with a motorized stage driven in the X, Y and Z planes, a cooled CCD camera and a pentium computer equipped with the NeuroLucida Morphometry and Stereology System. Cell counts were taken at 20× magnification and for each section seven counting frames were examined (frame size: 75/75 ␮m; spacing between frames: 250/250 ␮m). Unbiased estimates of cell numbers for FG-positive cells, TH-IR cells and cells which co-expressed both FG and TH were obtained for the entire extent of the SN using the optical fractionator method.92 GDNF transgene expression: This stain resulted in the visualization of GDNF-immunoreactive (IR) cell bodies as well as an area of immunoreactivity surrounding these


cells similar to a ‘penumbra’. Using the NeuroLucida System (MicroBrightField, Colchester, VT, USA), the area and volume of the GDNF-IR penumbra was analyzed by drawing a contour around the stained area of each section and reconstructing the stained penumbra area using the Area and Volume Estimator function of the NeuroLucida Morphometry System. Total cell counts of GDNF-IR neurons within the GDNF penumbra were also made. The number of GDNF-IR cells were counted in every 6th section, summed and multiplied by 6 to obtain an estimate for the total number of transgene expressing cells. The total number of cells within the GDNF penumbra were corrected for section thickness using the Abercrombie method.93

Area of striatal denervation: The area of reduced TH-IR (lesion size) in the lesioned striatum was evaluated using NIH Image (1.60) software on a MacIntosh 8600/300 computer. Sections were directly scanned using a Path Scan Enabler to hold the glass slide for scanning with the SprintScan 35 slide scanner (Polaroid, Cambridge, MA, USA). Adobe Photoshop was used to acquire the images in grey-scale which were then imported into NIH Image. Pseudocolor was used to enhance image resolution and define areas of reduced TH-IR in the lesioned striatum. A level of color intensity was set to represent normal levels of TH-IR. From this level, the area of reduced TH-IR was determined visually, outlined and quantified using NIH Image. For each rat, the area of reduced TH-IR in the lesioned side was quantified every 240 ␮m rostro-caudally throughout the striatum. The data are expressed as the mean area of reduced TH-IR in the lesioned striatum. Fos-immunoreactivity: Quantitative analysis of striatal Fos expression was performed using the NeuroLucida System. For each rat, the number of Fos-IR nuclei was determined within randomly selected areas (100 ␮m2), selected every 250 ␮m in a total sample area 4 mm2 surrounding the injection site. The number of Fos-IR nuclei was determined at the lesion site and approximately 700 ␮m posterior to the lesion site (400 ␮m anterior to Ad vector striatal injection). Striatal Fos expression was analyzed in both the lesioned and unlesioned hemisphere. The number of Fos-positive nuclei in the lesioned hemisphere was expressed as a percentage of the number of Fos-positive nuclei in the unlesioned hemisphere. Statisical analysis Cell counts and optical density measures were analyzed by applying a one-way ANOVA for group (striatal vector control group, SN vector control group, striatal AdGDNF group and SN AdGDNF group) using SuperAnova Software. Behavior, rotation and forelimb use were analyzed using a repeated-factor two-way ANOVA for group by day of behavior observed (repeated measure). Comparison of the efficacy of AdGDNF injection site was evaluated using a one-way ANOVA for group. When ANOVA showed significant differences, pair-wise comparisons between means were tested by applying the Tukey–Kramer post hoc comparison test. For analysis of the degree of host immune response the Mann–Whitney non-parametric test was used to compared means between treatment groups.

Acknowledgements This work was supported by NIH grants NS31957 and HD33531, The Medical Research Institute Council of Children’s Memorial Hospital, the Carver Foundation and The National Parkinson Foundation Center of Excellence in Parkinson’s Disease at the University of Pittsburgh Seed Money Program partially funded by the Tuchman Foundation, Princeton, New Jersey. AdGDNF and AdmuGDNF were the generous gifts of Genetic Therapy Inc/Novartis. The authors thank Dr Cristina Backman for assistance with surgeries, Dr Pauline Chou for assistance with analysis of pathology and Mr Richard Anderson (University of Iowa, Gene Transfer Vector Core) for preparation of vector stocks. We also thank Drs James Surmeier and Derek Choi-Lundberg for critical reading of the manuscript. B Connor is the recipient of a New Zealand Neurological Foundation Post-Doctoral Fellowship.

References 1 Bohn MC, Choi-Lundberg DL. Neurotrophic factor gene therapy for neurodegenerative disease. Gene Ther Molec Biol 1998; 1: 265–277. 2 Choi-Lundberg DL, Bohn MC. Applications of gene therapy to neurological diseases and injuries. In: Quesenberry PJ, Stein GS, Forget B, Weissman S (eds). Stem Cell Biology and Gene Therapy. Wiley-Liss: New York, 1998, pp 503–553. 3 Connor B, Dragunow M. The role of neuronal growth factors in neurodegenerative disorders of the human brain. Brain Res 1998; 27: 1–39. 4 Lindsay RM. Neurotrophic growth factors and neurodegenerative diseases: therapeutic potential of the neurotrophins and ciliary neurotrophic factor. Neurobiol Aging 1994; 15: 249–251. 5 Bohn MC, Choi-Lundberg DL. Gene therapies for Parkinson’s disease. In: Chiocca EA, Breakfield XO (eds). Gene Therapy for Neurological Disorders and Brain Tumors. Humana Press: Totowa, NJ, 1997, pp 377–395. 6 Unsicker K. GDNF: a cytokine at the interface of TGF-betas and neurotrophins. Cell Tiss Res 1996; 286: 175–178. 7 Krieglstein K, Suter-Crazzolara C, Unsicker K. Development of mesencephalic dopaminergic neurons and the transforming growth factor-beta superfamily. J Neural Transmis Suppl 1995; 46: 209–216. 8 Lin L-FH et al. A glial cell-line derived neurotrophic factor for midbrain dopaminergic neurons. Science 1993; 260: 1130–1132. 9 Beck KD et al. Mesencephalic dopaminergic neurons protected by GDNF from axotomy-induced degeneration in the adult brain. Nature 1995; 373: 339–341. 10 Hoffer BJ et al. Glial cell line-derived neurotrophic factor reverses toxin-induced injury to midbrain dopaminergic neurons in vivo. Neurosci Lett 1994; 182: 107–111. 11 Kearns C, Gash D. GDNF protects nigral dopaminergic neurons against 6-hydroxydopamine in vivo. Brain Res 1995; 672: 104–111. 12 Tomac A et al. Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo. Nature 1995; 373: 335–339. 13 Sauer H, Rosenblad C, Bjorklund A. Glial cell line-derived neurotrophic factor but not transforming growth factor beta-3 prevents delayed degeneration of nigral dopaminergic neurons following striatal 6-hydroxydopamine lesion. Proc Nat Acad Sci USA 1995; 92: 8935–8939. 14 Bowenkamp KE et al. Glial cell line-derived neurotrophic factor supports survival of injured midbrain dopaminergic neurons. J Comp Neurol 1995; 355: 479–489. 15 Lapchak PA, Miller PJ, Collins F, Jiao S. Glial cell line-derived neurotrophic factor attenuates behavioral deficits and regulates nigrostriatal dopaminergic and peptidergic markers in 6-hydroxydopamine-lesioned adult rats: comparison of intraventricular and intranigral delivery. Neuroscience 1997; 78: 61–72.

1949


1950

16 Rosenblad C, Martinez-Serrano A, Bjorklund A. Intrastriatal glial cell line-derived neurotrophic factor promotes sprouting of spared nigrostriatal dopaminergic afferents and induces recovery of function in a rat model of Parkinson’s disease. Neuroscience 1998; 82: 129–137. 17 Lindner MD et al. Implantation of encapsulated catecholamine and GDNF-producing cells in rats with unilateral dopamine deletions and Parkinsonian symptoms. Exp Neurol 1995; 132: 62–76. 18 Gash DM et al. Functional recovery in Parkinsonian monkeys treated with GDNF. Nature 1996; 380: 252–255. 19 Ghodsi A et al. Extensive beta-glucuronidase activity in murine central nervous system after adenovirus-mediated gene transfer to brain. Hum Gene Ther 1998; 9: 2331–2340. 20 Akli S et al. Transfer of a foreign gene into the brain using adenovirus vectors. Nat Genet 1993; 3: 224–228. 21 Bajocchi G, Feldman SH, Crystal RG, Mastrangeli A. Direct in vivo gene transfer to ependymal cells in the central nervous system using recombinant adenovirus vectors. Nat Genet 1993; 3: 229–234. 22 Davidson BL et al. A model system for in vivo gene transfer into the central nervous system using an adenovirus vector. Nat Genet 1993; 3: 219–223. 23 Le Gal La Salle G et al. An adenovirus vector for gene transfer into neurons and glia in the brain. Science 1993; 259: 988–990. 24 Choi-Lundberg DL et al. Dopaminergic neurons protected from degeneration by GDNF gene therapy. Science 1997; 275: 338–341. 25 Davidson BL, Bohn MC. Recombinant adenovirus: a gene transfer vector for study and treatment of CNS disease. Exp Neurol 1997; 144: 125–130. 26 Horellou P, Sabate O, Buc-Caron M-H, Mallet J. Adenovirusmediated gene transfer to the central nervous system for Parkinson’s disease. Exp Neurol 1997; 144: 131–138. 27 Choi-Lundberg DL et al. Behavioral and cellular protection of rat dopaminergic neurons by an adenoviral vector encoding glial cell line-derived neurotrophic factor. Exp Neurol 1998; 154: 261–275. 28 Bilang-Bleuel A et al. Intrastriatal injection of an adenoviral vector expressing glial cell-line derived neurotrophic factor prevents dopaminergic neuron degeneration and behavioral impairment in a rat model of Parkinson’s disease. Proc Natl Acad Sci USA 1997; 94: 8818–8823. 29 Mandel R, Spratt S, Snyder R, Leff S. Midbrain injection of recombinant adeno-associated virus encoding rat glial cell linederived neurotrophic factor protects nigral neurons in a progressive 6-hydroxydopamine-induced degeneration model of Parkinson’s disease in rats. Proc Natl Acad Sci 1997; 94: 14083– 14088. 30 Bowenkamp K, Lapchak P, Hoffer B, Bickford P. Glial cell linederived neurotrophic factor reverses motor impairment in 16– 17 month old rats. Neurosci Lett 1996; 211: 81–84. 31 Hebert MA, Gerhardt GA. Behavioural and neurochemical effects of intranigral administration of glial cell line-derived neurotrophic factor on aged Fischer 344 rats. J Pharmacol Exp Ther 1997; 282: 760–768. 32 Lapchak P, Miller P, Jiao S. Glial cell line-derived neurotrophic factor induces the dopaminergic and cholinergic phenotype and increases locomotor activity in aged Fischer-344 rats. Neuroscience 1997; 77: 745–752. 33 Emerich DF et al. Alleviation of behavioral deficits in aged rodents following implantation of encapsulated GDNF-producing fibroblasts. Brain Res 1996; 736: 99–110. 34 Date I et al. GDNF administration induces recovery of the nigrostriatal dopaminergic system both in young and aged parkinsonian mice. NeuroReport 1998; 9: 2365–2369. 35 Zigmond M et al. Compensatory responses to partial loss of dopaminergic neurons: studies with 6-hydroxydopamine. In: Schneider J, Gupta M (eds). Current Concepts in Parkinson’s Disease Research. Hogrefe and Huber: Go¨ttingen, 1993, pp 99– 140. 36 Demarest KT, Riegle GD, Moore KE. Characteristics of dopami-

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54 55

56

nergic neurons in the aged male rat. Neuroendocrinology 1980; 31: 222–227. Unnerstall JR, Long MM. Differential effects of the intraventricular administration of 6-hydroxydopamine on the induction of type II beta-tubulin and tyrosine hydroxylase mRNA in the locus coeruleus of the aging Fischer 344 rat. J Comp Neurol 1996; 364: 363–381. Schallert T. Aging-dependent emergence of sensorimotor dysfunction in rats recovered from dopamine sustained early in life. In: Joseph JA (ed). Central Determinants of Age-Related Decline in Motor Function. Annals of New York Academy of Science: New York, 1988, pp 108–119. Bjorklund A, Rosenblad C, Winkler C, Kirik D. Studies on neuroprotective and regenerative effects of GDNF in a partial lesion model of Parkinson’s disease. Neurobiol Dis 1997; 4: 186–200. Ungerstedt U. Striatal dopamine release after amphetamine or nerve degeneration revealed by rotational behavior. Acta Physiol Scand 1971; 82: 49–68. Sauer H, Oertel WH. Progressive degeneration of nigrostriatal dopamine neurons following intrastriatal terminal lesions with 6-hydroxydopamine a combined retrograde tracing and immunocytochemical study in the rat. Neuroscience 1994; 59; 401–415. Liu FC, Graybiel AM, Dunnett SB, Braughman RW. Intrastriatal grafts derived from fetal striatal primordia. II. Reconstitution of cholinergic and dopaminergic systems. J Comp Neurol 1990; 295: 1–14. Morgan JI, Curran T. Stimulus-transcription coupling in the nervous system: involvement of the inducible proto-oncogenes fos and jun. Ann Rev Neurosci 1991; 14: 421–451. Cenci M et al. Dopaminergic transplants normalize amphetamine and apomorphine-induced Fos expression in the 6-hydroxydopamine lesioned striatum. Neuroscience 1992; 46: 943–957. Labandeira-Garcia J et al. Time course of striatal changes induced by 6-hydroxydopamine lesion of the nigrostriatal pathway, as studied by combined evaluation of rotational behavior and striatal Fos expression. Exp Brain Res 1996; 108: 69–84. Robertson HA, Peterson MR, Murphy K, Robertson GS. D1dopamine receptor agonists selectively activate striatal c-fos independent of rotational behavior. Brain Res 1989; 503: 346–349. Graybiel AM, Moratalla R, Robertson HA. Amphetamine and cocaine induces drug-specific activation of the c-fos gene in striosome-matrix compartments and limbic subdivisions of the striatum. Proc Natl Acad Sci USA 1990; 87: 6912–6916. Robertson GS, Vincenty SR, Fibiger HC. D1 and D2 dopamine receptors differentially regulate c-fos exprssion in striatonigral and striatopallidal neurons. Neuroscience 1992; 49: 285–296. Paul ML, Graybiel AM, David JC, Robertson HA. D1-like and D2-like dopamine receptors synergistically activate rotation and c-fos expression in the dopamine-depleted striatum in a rat model of Parkinson’s disease. J Neurosci 1992; 12: 3729–3742. Konradi C, Cole RL, Heckers S, Hyman SE. Amphetamine regulates gene expression in rat striatum via transcription factor CREB. J Neurosci 1994; 14: 5623–5634. Berretta S, Robertson HA, Graybiel AM. Dopamine and glutamine agonists stimulate neuron-specific expression of Fos-like protein in the striatum. J Neurophysiol 1992; 68: 767–777. Matheson CR et al. Glial cell line-derived neurotrophic factor (GDNF) is a neurotrophic factor for sensory neurons: comparison with the effects of the neurotrophins. J Neurobiol 1997; 32: 22–32. Byrnes AP, Rusby JE, Wood MJA, Charlton HM. Adenovirus gene transfer causes inflammation in the brain. Neuroscience 1995; 66: 1015–1024. Bohn MC et al. Adenoviral-mediated transgene expression in non-human primate brain. Hum Gene Ther 1999; 10: 1175–1184. Lawrence MS et al. Inflammatory responses and their impact on beta-galactosidase transgene expression following adenovirus vector delivery to the primate caudate nucleus. Gene Therapy 1999; 6: 1368–1379. Martin D et al. Intranigral or intrastriatal injections of GDNF – effects on monoamine levels and behavior in rats. Euro J Pharmacol 1996; 317: 247–256.


57 Lapchak P et al. Pharmacological activities of glial cell linederived neurotrophic factor (GDNF): preclinical development and application to the treatment of Parkinson’s disease. Exp Neurol 1997; 145: 309–321. 58 Morgan D, Finch C. Dopaminergic changes in the basal ganglia: a generalized phenomenon of aging in mammals. Ann NY Acad Sci 1998; 515: 145–160. 59 Felten D et al. Age-related decline in the dopaminergic nigrostriatal system: the oxidative hypothesis and protective strategies. Ann Neurol 1992; 32: S133–S136. 60 Roth GS, Joseph JA. Cellular and molecular mechanisms of impaired dopaminergic function during aging. Ann NY Acad Sci 1994; 719: 129–135. 61 Hubble JP. Aging and the basal ganglia. Neurol Clin 1998; 16: 649–657. 62 Yu T et al. Expression of GDNF family receptor components during development: implications in the mechanisms of interaction. J Neurosci 1998; 18: 4684–4696. 63 Trupp M, Belluardo N, Funakoshi H, Ibanez CF. Complementary and overlapping expression of glial cell line-derived neurotrophic factor (GDNF), c-ret proto-oncogene, and GDNF receptor-alpha indicates multiple mechanisms of trophic actions in the adult rat CNS. J Neurosci 1997; 17: 3554–3567. 64 Glazner GW, Mu X, Springer JE. Localization of glial cell linederived neurotrophic factor receptor alpha and c-ret mRNA in rat central nervous system. J Comp Neurol 1998; 391: 42–49. 65 Kirik D, Rosenblad C, Bjorklund A. Characterization of behavioral and neurodegeneration changes following partial lesions of the nigrostriatal dopamine system induced by intrastriatal 6-hydroxydopamine in the rat. Exp Neurol 1998; 153: 259–277. 66 Shults CW, Kimber T, Martin D. Intrastriatal injection of GDNF attenuates the effects of 6-hydroxydopamine. NeuroReport 1996; 7: 627–631. 67 Schwarting R, Huston J. The unilateral 6-hydroxydopamine lesion model in behavioral brain research analysis of functional deficits, recovery and treatments. Prog Neurobiol 1996; 50: 275– 331. 68 Brass CA, Glick SD. Sex difference in drug-induced rotation in two strains of rats. Brain Res 1981; 223: 229–234. 69 Tillerson JL, Castro SL, Zigmond MJ, Schallert T. Motor rehabilitation of forelimb use in unilateral 6-hydroxydopamine (6OHDA) rat model of Parkinson’s disease. Soc Neurosci Abstr 1998; 24: 1720. 70 Naldini L et al. Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proc Natl Acad Sci USA 1996; 93: 11382–11388. 71 Kafri T et al. Sustained expression of genes delivered directly into liver and muscle by lentiviral vectors. Nat Genet 1997; 17: 314–317. 72 Kaplitt MG et al. Long-term expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994; 8: 148–154. 73 McCown TJ et al. Differential and persistent expression pattern of CNS gene transfer by an adeno-associated virus (AAV) vector. Brain Res 1996; 713: 99–107. 74 Lo WD et al. Adeno-associated virus-mediated gene transfer to the brain: duration and modulation of expression. Hum Gene Ther 1999; 10: 201–213.

75 Yang Y et al. Inactivation of E2a in recombinant adenoviruses improves the prospect for gene therapy in cystic fibrosis. Nat Genet 1994; 7: 362–369. 76 Fang B et al. Lack of persistence of E1-recombinant adenoviral vectors containing a temperature-sensitive E2a mutation in immunocompetent mice and hemophilia B dogs. Gene Therapy 1996; 3: 217–222. 77 Amalfitano A, Chamberlain JS. Isolation and characterization of packaging cell lines that coexpress the adenovirus E1, DNA polymerase, and preterminal proteins: implications for gene therapy. Gene Therapy 1997; 4: 258–263. 78 Amalfitano A, Begy CR, Chamberlain JS. Improved adenovirus packaging cell lines to support the growth of replication-defective gene-delivery vectors. Proc Natl Acad Sci USA 1996; 93: 3352–3356. 79 Amalfitano A et al. Production and characterization of improved adenovirus vectors with E1, E2b and E3 genes deleted. J Virol 1998; 72: 926–933. 80 Crouzet J et al. Recombinational construction in Escherichia coli of infectious adenoviral genomes. Proc Natl Acad Sci USA 1997; 94: 1414–1419. 81 Armentano D et al. Characterization of an adenovirus gene transfer vector containing an E4 deletion. Hum Gene Ther 1995; 6: 1343–1353. 82 Armentano D et al. Effect of the E4 region on the persistence of transgene expression from adenovirus vectors. J Virol 1997; 71: 2408–2416. 83 Parks RJ et al. A helper-dependent adenovirus vector system: removal of helper virus by cre-mediated excision of the viral packaging signal. Proc Natl Acad Sci USA 1996; 93: 13565–13570. 84 Kochanek S et al. A new adenoviral vector: replacement of all viral coding sequences with a 26 kb of DNA independently expressing both full-length dystrophin and beta-galactosidase. Proc Natl Acad Sci USA 1996; 93: 5731–5736. 85 Fisher KJ et al. Recombinant adenovirus deleted of all viral genes for gene therapy therapy of cystic fibrosis. Virology 1996; 217: 11–22. 86 Morsy MA et al. An adenoviral vector deleted for all viral coding sequences results in enhanced safety and extended expression of a leptin transgene. Proc Natl Acad Sci USA 1998; 95: 7866–7871. 87 Schallert T, Jones TA, Norton DA. A clinically relevant unilateral rat model of Parkinsonian akinesia. J Neural Transplant Plast 1992; 3: 332–333. 88 Schallert T, Lindner MD. Rescuing neurons from trans-synaptic degeneration after brain damage: helpful, harmful, or neutral in recovery of function? Can J Psychol 1990; 44: 276–292. 89 Jones TA, Schallert T. Overgrowth and pruning of dendrites in adult rats recovering from neocortical damage. Brain Res 1992; 581: 156–160. 90 Jones TA, Schallert T. Use-dependent growth of pyramidal neurons after neocortical damage. J Neurosci 1994; 14: 2140–2152. 91 Kozlowski DA, James DC, Schallert T. Use-dependent exaggeration of neuronal injury after unilateral sensorimotor cortex lesions. J Neurosci 1996; 16: 4776–4786. 92 Coggeshall RE, Lekan HA. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J Comp Neurol 1996; 364: 6–15. 93 Abercrombie M. Estimation of nuclear population from microtome sections. Anat Rec 1946; 94: 239–247.

1951