Introduction

Parkinson’s disease (PD) is a chronic, progressive neurodegenerative disorder characterized by the massive degeneration of dopaminergic neurons in the substantia nigra (SN) pars compacta. This nigral neuronal loss leads to a striatal dopamine (DA) deficiency, which is considered to underlie the most overt symptoms of the disease.

Several phenotypical features of human PD can be mimicked in rodents through the intracerebral injection of neurotoxins to study DA-related functions and subsequent assessment of the efficacy of various symptomatic/neuroprotective modulatory approaches [1]. A widely used experimental acute model of PD relies on the use of 6-hydroxydopamine (6-OHDA), which selectively destroys catecholaminergic neurons through mitochondrial damage [2]. The neurochemical and histopathological alterations obtained following unilateral injection of 6-OHDA into the rat SN pars compacta have been extensively characterized using invasive histological and electrophysiological analysis (for review see [3]). On the behavioral–motor level, characteristic gait disturbances are easily assessable by utilizing tests that examine side bias, such as amphetamine-induced rotation tests and spontaneous motor tests [4].

The recent development of positron emission tomography (PET) scanners designed for laboratory animals has enabled further detailed study of rodent models for human neurodegenerative diseases by providing in vivo insights on biochemical/molecular processes and allowing non-invasive intrasubject follow-up of neuroprotective/neuromodulatory approaches. In vivo studies of the nigrostriatal dopaminergic projection have been performed in the unilateral 6-OHDA rat model using radiotracers with a high affinity for the plasma membrane dopamine transporter (DAT) [5]. The reduction in striatal DAT binding as a measure of presynaptic dopaminergic denervation in PD patients is established with PET and single photon emission computed tomography (SPECT) [6], although there are some restrictions on the use of DAT as a biomarker in pathophysiological studies [7].

[18F]-FDG (2-[18F]-fluoro-2-deoxy-D-glucose) is considered to be a marker of cerebral glucose consumption based on neuronal entrapment and accumulation of [18F]-FDG-6-PO4, indicating neuronal viability [8]. In PD patients, specific cortico-subcortical metabolic alterations have been described based on direct regional analysis [9] or network analysis approaches [10]. PD patients were found to have metabolic covariance patterns characterized by lentiform and thalamic hypermetabolism associated with regional metabolic decrements in the lateral premotor cortex, the supplementary motor area, the dorsolateral prefrontal cortex and the parieto-occipital association regions [11]. [18F]-FDG imaging also contributes clinically to the differential diagnosis of Parkinsonian disorders [12]. Although a number of in vivo imaging studies on dopaminergic neurotransmission in animal models of PD have been published [13–15], no information is as yet available on the in vivo glucose metabolic changes in the unilateral 6-OHDA lesion rat model, which would allow further biochemical translational validation of cortico-subcortical alterations observed in humans.

The aim of this study was, therefore, to characterize the unilateral 6-OHDA lesion rat model in vivo, through multitracer glucose metabolism and dopamine transporter imaging, in relation to in vitro tyrosine hydroxylase (TH) histology and behavioral measurements (limb-use asymmetry test) in the same animals.

Materials and methods

Animals

Experiments were conducted on 18 female Wistar rats (body weight range at the start of the experiment 194–304 g, age range 12–17 weeks). All animals were housed three to a cage, at an average room temperature of 22°C and a 12/12-h light/dark cycle. Food and water were given ad libitum. The research protocol was approved by the local Animal Ethics Committee and was according to European Ethics Committee guidelines (decree 86/609/EEC).

Nigrostriatal 6-OHDA lesion

All surgical procedures were performed under ketamine [60 mg/kg interperitoneal (i.p.)] and medetomidine (0.4 mg/kg) anesthesia using aseptic procedures. The rats assigned to the 6-OHDA-lesioned group (n = 8) were placed in a stereotactic head frame (Stoelting, Wood Dale, IL), and a small hole was drilled in the skull in the appropriate location using Bregma as reference. The neurotoxic lesions of the right substantia nigra were performed by injecting 24 μg 6-OHDA (Sigma, St. Quentin-Fallavier, France) at the following coordinates: anterio-posterior (AP), −4.8 mm; lateral (L), 2.1 mm; dorsoventral (DV), 7.2 mm. 6-OHDA was dissolved in 3 μl of 0.9% sterile NaCl containing 0.1% ascorbic acid. After the injection, the needle was left in place for an additional 10 min before being slowly withdrawn from the brain. No surgery was performed on the control animals (n = 10).

Image acquisition

DAT images were obtained in 6-OHDA and control animals (n = 7 and n = 9 respectively; two animals died) using the radioligand [18F]-FECT {2′-[18F]-fluoroethyl-(1R-2-exo-3-exe)-8-methyl-3-(4-chlorophenyl)-8-azabicyclo(3.2.1)-octane-2-carboxylate} [16]. [18F]-FECT binds selectively to the DAT with high affinity KI = 9 nM with respect to PE2I (D. Guilloteau, Tours, France, personal communication). The synthesis of the [18F]-FECT PET radiotracer was performed according to the procedure described by Wilson et al. but using 2-[18F]-fluoroethyltrifluoromethanesulfonate instead of 2-[18F]-fluoroethylbromide [16]. Metabolic images were obtained using [18F]-FDG in eight 6-OHDA rats and ten controls. FDG was prepared by a standard synthesis module (IBA, Louvain La Neuve, Belgium).

Small animal PET imaging was performed using an LSO detector-based FOCUS 220 tomograph (Siemens/Concorde Microsystems, Knoxville, TN), which has a transaxial resolution of 1.35 mm full-width at half-maximum. Data were acquired in list mode in a 128 × 128 × 95-matrix with a pixel width of 0.475 mm and a slice thickness of 0.796 mm. The coincidence window width was set at 6 nsec. Prior to small animal PET imaging, the rats were anesthetized with an i.p. injection of 50 mg/kg sodium pentobarbital (Nembutal; Ceva Sante Animale, Brussels, Belgium). Tail veins were catheterized to enable the infusion of the labeled ligands ([18F]-FECT 27 ± 2 MBq; [18F]-FDG 21 ± 3 MBq; specific activity range 53–760 GBq/μmol; injection volume 500 μl). All rats were breathing spontaneously throughout the entire experiment. Imaging studies of the 6-OHDA-lesioned rats were performed 6–11 weeks post-lesioning. [18F]-FECT measurements were obtained during a 40-min interval, starting 3 h post-injection. For [18F]-FDG, acquisitions were performed for 40 min, starting 30 min post-injection after overnight fasting. The acquisition timing rationale and kinetics of both [18F]-FECT and [18F]-FDG in rats have been described previously [17].

For quantification purposes, sinograms were reconstructed using filtered backprojection (FBP). No corrections were made for attenuation or scatter, since the rat brain size is relatively small compared to that of humans (approximately brain length of humans vs. rats 15 vs. 2.5 cm).

Behavioral testing

Following small animal PET imaging, the limb-use asymmetry test, a modified version of the cylinder test for rats [18] was performed to assess gait disturbances in 6-OHDA-lesioned rats (n = 7, 24 weeks post-lesioning; range 19–33 weeks). For comparison, a subgroup of five controls was sufficient to detect significant limb-use asymmetry in the 6-OHDA model, since intersubject variability in the controls was expected to be small [19]. Forelimb use during explorative activity was analyzed by videotaping rats in a transparent glass cylinder (diameter 20 cm, height 38 cm). A mirror was placed behind the cylinder at an angle to enable the rater to record forelimb movements when the animal was turned away from the camera. The cylindrical shape encourages vertical exploration of the walls. No habituation to the cylinder prior to videotaping was allowed. The test was performed between 0900 and 1800 hours after an overnight fasting to increase behavioral activity. To stimulate rats that showed little or no tendency to explore, the following methods were used in the order presented: (1) turning the lights in the room down for a few seconds; (2) taking the rat out of the cylinder and cleaning the cylindrical environment before putting the animal back [20]. All scoring was done by an experimenter blinded to the condition of the animal using computerized digital video recordings with slow motion. The number of wall contacts performed independently with the left and right forepaw were counted up to a total number of 20 wall contacts per rat and session. Only supporting contacts were counted, i.e. full appositions of the paw with open digits to the cylinder wall. Data were expressed as the percentage use of the impaired forelimb relative to the total number of wall contacts (%I).

Immunohistochemistry

At the completion of the in vivo experiments, the remaining dopaminergic neurons in the SN pars compacta of six 6-OHDA rats (one animal died) were visualized by in vitro tyrosine hydroxylase staining. Rats were sacrificed by sodium pentobarbital overdose and perfused transcardially with 300 ml phosphate buffered saline (PBS), followed by 300 ml 4% paraformaldehyde at a flow rate of 30 ml/min before decapitation. Their brains were carefully removed from the skull and post-fixed in 4% paraformaldehyde solution overnight. Coronal brain sections measuring 50 μm in thickness were cut with a vibratome (Leica; Microsystems, Wetzelar, Germany), collected free floating and stored at 4°C in PBS buffer containing 0.1% sodium azide. Every fifth section of the SN was then treated with 3% hydrogen peroxide (H2O2) to block endogenous peroxidase activity and incubated overnight with primary rabbit anti-TH antibody (Chemicon, Temecula, CA; dilution of 1:5000) in 10% normal swine serum. After three rinses with PBS-0.1% Triton X-100, the sections were incubated with biotinylated swine anti-rabbit secondary antibody for 30 min at room temperature, followed by incubation with Strept-ABC-HRP complex (DAKO, Glostrup, Denmark). Before visualization of the TH-immunoreactive neurons with 3,3-diaminobenzidine (DAB), H2O2 was added to run the reaction to completion. Sections were mounted on gelatinized slides, dehydrated in increasing ethanol concentrations and coverslipped using mounting fluid.

The number of TH-immunopositive neurons in the SN pars compacta was counted using an unbiased stereological method: the optical fractionator (StereoInvestigator; MicroBrightField, Colchester, VT). Quantification was performed on 6-OHDA-lesioned rats on average 29 weeks post-lesioning (range 24–37 weeks), with a total of seven to nine sections for each animal. The remaining dopaminergic cells on the lesioned side are expressed as the percentage of the cell numbers on the contralateral (unlesioned) side (%TH).

Small animal PET data processing and statistical parametric mapping

Parametric DAT binding index (BI) images were constructed by reference to the cerebellum: (tissue/average activity concentration cerebellum) −1. [18F]-FDG data were count normalized to the whole-brain uptake. No difference in whole-brain uptake was observed between 6-OHDA-lesioned rats and controls (data not shown). To obtain maximal use of image information without a priori knowledge, we analyzed the images on a voxel-by-voxel basis using SPM2 (Statistical Parametric Mapping; Wellcome Department of Cognitive Neurology, London, UK; http://www.fil.ion.bpmf.ac.uk/spm/). The procedures for spatial normalization and its validation have been described previously [17]. This methodology allows results to be reported in coordinates directly corresponding to the Paxinos coordinate system using Bregma as reference [21]. Realistic voxel dimensions are 0.2 × 0.2 × 0.2 mm.

SPM analysis was carried out using a categorical subject design (conditions: disease vs. controls) on parametric DAT and [18F]-FDG images of 6-OHDA-lesioned rats in comparison to the respective control data. Spatially normalized images were masked to remove extracerebral signal that would disrupt the global normalization for [18F]-FDG. All images were smoothed with an isotropic Gaussian kernel of 1.6 realistic mm. SPM analysis of parametric DAT data was performed without global normalization and with an absolute analysis threshold of –Infinity. For [18F]-FDG data, proportional scaling was applied (scaled to 50 ml/100 g/min), with an analysis threshold of 0.8 of the mean image intensity. For statistical analysis, T-maps data were interrogated at a peak level of p = 0.001 (uncorrected) and extent threshold kext > 200 voxels (1.6 mm3), unless indicated otherwise. Only clusters with p < 0.05 corrected were withheld as significant.

In addition, a voxel-based correlation analysis between metabolism and Parkinsonian markers was performed within the 6-OHDA-lesioned group with the following covariates: DAT impairment, %I and %TH. DAT values were determined by a predefined volume-of-interest (VOI) analysis based on striatal volume definition on the Paxinos atlas. The VOI characteristics have been presented elsewhere [17]. For DAT impairment, the affected-to-non-affected side BI ratio was obtained and expressed as the percentage of normal values.

As a supplement to the SPM analysis, spatially normalized [18F]-FDG and parametric DAT images of 6-OHDA-lesioned rats and controls were also evaluated using a predefined VOI approach, especially to determine asymmetry indices and changes herein.

General statistics

Conventional statistics were carried out using Statistica v 6.0 (Statsoft, Tulsa OK). VOI data as well as behavioral and histological outcomes were analyzed using unpaired/paired t tests. Significance was accepted at the 95% probability level.

Results

Behavioral and histological outcome

Individual behavioral and histological data are detailed in Table 1. Quantitative analysis of TH-positive neurons in the pars compacta of the substantia nigra showed an average unilateral destruction of 98.1% (range 96%–99%) in the 6-OHDA-lesioned rats. Figure 1 illustrates the histology of the SN pars compacta in an animal with 98.9% loss of TH-positive neurons.

Table 1 Percentage of intact dopaminergic neurons after unilateral infusion of 6-OHDA, as measured by caudate-putamen [18F]-FECT binding in vivo using small animal PET (%DAT), and in vitro with TH-immunostaining (%TH), versus the percentage use of the impaired forelimb during the limb-use asymmetry test (%I)
Full size table
Fig. 1
figure 1

Photomicrograph of a section through the substantia nigra, processed for tyrosine hydroxylase (TH)-immunohistochemistry, from an animal subjected to unilateral 6-hydroxydopamine (6-OHDA) injection into the right substantia nigra. The arrow denotes the lesioned right side (magnification 5×)

Full size image

Dopamine-depleted rats showed significant severe limb-use asymmetry during explorative behavior compared to the control animals (t test, 2 vs. 44%, p = 3 × 10−8). The difference between the limb-use asymmetry outcome of control animals in comparison to a perfect symmetric result (50%) was not significant (single t test, p = 0.11).

Categorical [18F]-FECT analysis

Representative transverse slices of the mean [18F]-FECT binding index of 6-OHDA-lesioned rats and controls are presented in Fig. 2. DAT binding was strongly decreased in the affected caudate-putamen of 6-OHDA-lesioned rats, as shown in Fig. 3. The VOI range of the right caudate-putamen binding index, expressed as a percentage of normal values, was between 0.9 and 17.5%. Regional VOI-based [18F]-FECT binding indices of 6-OHDA-lesioned rats and controls are displayed in Table 2.

Fig. 2
figure 2

Series of axial sections (ventral to dorsal, slice interval 1.0 mm) showing mean [18F]-FECT {2′-[18F]-fluoroethyl-(1R-2-exo-3-exe)-8-methyl-3-(4-chlorophenyl)-8-azabicyclo(3.2.1)-octane-2-carboxylate} binding index (a) and mean glucose metabolism (b) of 6-OHDA-lesioned rats (top row) in comparison to control data (bottom row) as well as the co-registered rat brain magnetic resonance atlas (c). Color bars indicate binding indices for the dopamine transporter (DAT) and relative intensities for fluoro-2-deoxy-D-glucose (FDG). Single arrows denote glucose hypometabolism, double-headed arrow indicates increased relative brain metabolism

Full size image
Fig. 3
figure 3

Coronal brain sections showing significant decreased DAT availability in the caudate-putamen of 6-OHDA-lesioned rats (n = 7; figure given at p < 8 × 10−5 corrected for multiple comparisons) compared to the control group (n = 9). Significance at the voxel level is shown with a t statistic color scale. No contralateral changes in DAT availability were observed at p height < 0.001 uncorrected. The distance between the sections is 1.00 mm with the position relative to Bregma (positive values for sections anterior to Bregma) given above the sections in the left column

Full size image
Table 2 Regional [18F]FECT binding indices and [18F]-FDG uptake values of dopamine-depleted rats and controls obtained using a predefined volume-of-interest approach
Full size table

The mean intensity of the deficit using the voxel-based approach was 96.2% (4.2 SD) at the Paxinos coordinate peak maximum (x,y,z) = (−2.8, 0, −4.8) (p < 1 × 10−10 corrected for multiple comparisons). The average right caudate-putamen binding indices at this Paxinos coordinate peak maximum were 0.4 (0.4 SD) and 10.0 (0.3 SD) for the 6-OHDA-lesioned rats and controls, respectively, while the left caudate-putamen values were 9.4 (1.5 SD) and 10.6 (0.7 SD) at (x,y,z) = (2.8, 0.6, −4.8), respectively. No contralateral changes in DAT availability were observed at p height < 0.001 (uncorrected), which was confirmed using the VOI-based analysis approach (Table 2).

The reduction in DAT availability included the ipsilateral nucleus accumbens (−80.1%, p < 1 × 10−4 corrected,) and the ipsilateral substantia nigra (−74.7%, p < 3 × 10−4 corrected). Cluster peak locations and p values of SPM analysis are shown in Table 3.

Table 3 Peak locations for the clusters in the group comparison of [18F]-FECT and [18F]-FDG data
Full size table

Categorical [18F]-FDG analysis

Representative transverse slices of the mean glucose metabolism of 6-OHDA-lesioned rats and controls are depicted in Fig. 2.

Voxel-based analysis revealed a significant glucose hypometabolism in the ipsilateral cortex of 6-OHDA-lesioned rats in comparison to controls, as shown in Fig. 4. The cluster in the ipsilateral cortex at p height < 0.001 (cluster p < 0.05 corrected) encompasses the insular cortex (agranular, dysgranular and granular part), the association cortex (temporal and pariental part), the ectorhinal cortex, the ventral area of the secondary auditory cortex, the primary (forelimb and hindlimb region and barrel field) and secondary somatosensory cortex, the primary and secondary visual cortex and the superior anterior part of the premotor cortex. Glucose hypometabolism was impaired the most intensely in the primary visual cortex (−6.3%, p = 4 × 10−6).

Fig. 4
figure 4

Coronal brain sections showing overlays on the regions with statistically significant decreased relative brain metabolism in the 6-OHDA group (n = 8; values given at p height < 0.001) compared to the control group (n = 10). The cluster shown encloses most regions of the frontal and sensory-motor cortex

Full size image

Relative metabolism was increased in one cluster covering the contralateral ectorhinal cortex and SN, albeit at a lower significance level. Only after a small volume correction in the SN [sphere of 1 mm at (x,y,z) = (4.8,−8.2,−6.4)] did this cluster reach significance (+3.4%, p = 0.001). Cluster peak locations and p values of SPM analysis are shown in Table 3.

In the VOI analysis, relative [18F]-FDG uptake values were significantly reduced in the ipsilateral caudate-putamen, lateral globus pallidus, nucleus accumbens, pons, hippocampus and cerebral cortex of dopamine-depleted rats as compared to the contralateral side. However, significance was only found in the ipsilateral sensory-motor cortex and the contralateral SN of dopamine-depleted rats in comparison to control data. VOI-based relative [18F]-FDG uptake values of 6-OHDA-lesioned rats and controls are given in Table 2.

Correlation analysis between metabolism and dopaminergic impairment

Within the 6-OHDA group, correlation analysis revealed a positive relationship between the degree of DAT impairment and the change in metabolic activity in the ipsilateral hippocampus (p = 3 × 10−5; Fig. 5). Cerebellar metabolism was inversely correlated to the affected to non-affected caudate-putamen DAT binding ratio (p < 3 × 10−4; Fig. 5). Cluster peak locations and p values are shown in Table 4. No correlations were found with %TH and %I as covariate.

Fig. 5
figure 5

Coronal brain sections showing overlays on the regions where a statistically significant positive (red, figure given at p < 0.001) and negative (blue, figure given at p height < 0.005) correlation between relative metabolic alterations and DAT impairment were observed in the 6-OHDA lesioned group (n = 7)

Full size image
Table 4 Peak locations for the correlation clusters (p height < 0.001, kext > 200)
Full size table

Discussion

The acute 6-OHDA animal model is a widely accepted model for exploring some of the pathological signaling events that are specific to PD. The neurochemical and morphological changes induced by 6-OHDA in the rat nigrostriatal system have been extensively characterized. In addition to a strong decline in DAT levels, 6-OHDA reduces the level of the rate-limiting enzyme of dopamine synthesis, tyrosine hydroxylase, as well as striatal dopamine and its metabolite levels [22]. However, although 6-OHDA neurotoxicity provokes dopaminergic molecular alterations comparable to those seen in PD, the pathological hallmark of human sporadic PD, Lewy body formation, has not been convincingly demonstrated in this rat model, a lack that is regarded as a major shortcoming of the 6-OHDA model [23]. Furthermore, the acute nature and large intensity of the 6-OHDA lesion is at odds relative to the human condition in terms of time progression and severity. Since multiple metabolic defects have been observed in human PD, it is of importance to understand to which level the acute dopaminergic lesion in this 6-OHDA model may modulate subcortico-cortical metabolic networks and to what extent this is similar to the human situation.

In this study, we demonstrated that the unilateral intranigral lesion produced with 6-OHDA caused a severe metabolic impairment in the ipsilateral sensory-motor cortex of 6-OHDA-lesioned rats as compared to both the contralateral hemisphere and the control rats, while metabolism was relatively increased in the contralateral midbrain comprising the substantia nigra. Moreover, regional glucose metabolism was significantly lower in the ipsilateral caudate-putamen, nucleus accumbens, lateral globus pallidus, pons and hippocampus of 6-OHDA-lesioned rats as compared to the contralateral hemisphere. This glucose metabolism pattern shows several similarities compared to [14C]-2-deoxyglucose autoradiography data in this model. In the 6-OHDA lesion rat model, reduced rCMR glucose levels have been described in the primary motor, sensory and auditory cortex ipsilateral to the lesion as compared to the contralateral hemisphere [24]. This reduced glucose metabolism in the auditory and somatosensory cortex of 6-OHDA-lesioned rats may be in line with a lateralized sensory neglect that has been behaviorally observed by several authors in this model, contralateral to the lesion side after auditory and tactile stimuli (for overview see [4]). However, several authors additionally found an increase in glucose metabolism in the ipsilateral globus pallidus and ipsilateral habenula [25–27]. These latter regions are relatively small compared to the microPET resolution and influenced by partial volume effects (e.g. inscatter from the large caudate-putamen), which may thus have damped possible changes in this study.

Normal caudate-putamen glucose metabolism ipsilateral to the 6-OHDA lesion as compared to controls, which was found even under these circumstances of severely impaired dopaminergic neurotransmission, is consistent to ex vivo literature data [24, 27]. Although one would hypothesize that glucose metabolism would be decreased after nigrostriatal deafferentiation due to toxin injection, as this is seen in patients [25], other factors, such as neuroinflammation, may compensate rCMR glucose decrease. Neuroinflammatory activity has been demonstrated in vivo in this model using [11C]-PK91115 PET [28] and is likely to counteract to a certain extent cellular metabolic decreases by FDG uptake in reactive activated microglia. To clarify this, further direct correlational studies between microglial activation and glucose metabolism in this rodent model are needed.

The hemispheric differences in cerebral metabolism, observed in most regions of 6-OHDA-lesioned rats, not only point to a globally reduced ipsilateral metabolism, but also to opposite metabolic effects in both hemispheres. However, we found significant differences only in the contralateral SN of dopamine-depleted rats in comparison to control data. In this context, the increased neuronal activity found in the contralateral SN might reflect a plasticity phenomenon with recruitment of parallel motor circuits in order to compensate for the functional deficit of the ipsilateral striatocortical motor loop. This adaptation was not accompanied by measurable changes in contralateral DAT density.

The metabolic cortical alterations, found in the 6-OHDA lesion rat model, are in good agreement to human PD. With respect to early human PD, a pattern of reduced premotor metabolism was observed, the degree of which relates to the severity of motor impairment [29]. Whereas in early disease increased metabolism in the striatum is observed, in advanced PD both decreased striatal and cortical metabolic activity are present, including motor, premotor and sensory cortical areas [30]. Decreased cortical glucose metabolism correlates with psychometric/cognitive performance [31].

In the 6-OHDA model, intranigral toxin administration is known also to deplete dopaminergic neurons of the ventral tegmental area (VTA) [3], projecting to the hippocampus [32]. This is in accordance with our correlation analysis, where the positive correlation with DAT impairment indicates the loss of ipsilateral hippocampal function. However, in humans, an increase in hippocampal activity is seen in advanced PD [33].

An inverse relationship between the degree of dopaminergic impairment and cerebellar metabolism is in line with human findings of cerebellar hyperactivity in PD patients, as shown by several [18F]-FDG PET studies [12]. This cerebellar hypermetabolism is a common physiopathological feature of PD and has been interpreted as compensatory on the dysfunctional basal ganglia loop system in PD. This association is confirmed by the normalization of cerebellar hypermetabolism after deep brain stimulation of the subthalamic nucleus (DBS-STN) [34].

In this study, we also found that the striatal uptake of [18F]-FECT was sensitive in detecting nigrostriatal hypofunction provoked by toxin administration. These in vivo small animal PET data are in good agreement with autoradiographic DAT binding values of iodinated and fluorine-labeled markers of the intact and 6-OHDA-lesioned rat caudate-putamen [35, 36] as well as with the loss of TH-immunoreactive neurons in the SN pars compacta, as all 6-OHDA-treated animals revealed more than 95% loss. However, the animal with a residual DAT activity in the right caudate-putamen of 17.5% compared to controls died before TH staining.

In addition, 6-OHDA-induced dopaminergic lesioning results in behavioral motor impairments mimicking human PD [4]. Tests assessing the motor behavior have been widely applied to estimate the extent of neuronal loss following stereotactical 6-OHDA lesioning. In our study, dopamine-depleted rats showed significant severe limb-use asymmetry during explorative behavior compared to the subset of control animals. However, no within-group correlations were found between the limb-use asymmetry test outcome and the metabolic alterations. This in contrast to DAT impairment as a covariate, which was, as mentioned earlier, correlated to ipsilateral hippocampal and cerebellar glucose metabolism. This distinction may be explained by the fact that DAT imaging enables the measurement of continuous mild to severe dopaminergic lesions and has a more robust quantitative outcome, whereas the cylinder test is unable to consistently detect motor abnormalities (particularly in low dose toxin). Furthermore, the striatal DAT binding index is likely to show a higher predictability of nigral TH positive cell loss compared to behavioral outcomes. Given the limited data, the absence of correlations with %TH as a covariate must be interpreted carefully.

The limb-use asymmetry test and TH-immunostaining were performed 19–37 weeks following the administration of 6-OHDA intranigrally, which is a time interval of 10–26 weeks from the [18F]-FDG/[18F]-FECT small animal PET imaging. Putative alterations of the dopaminergic lesions occurring within that time interval, which could affect the correlation analysis, are unlikely as it has been shown that the lesions induced by 6-OHDA result in irreversible damage to the dopaminergic system [37], leading to stable motor deficits over time.

In comparison to human studies, the animals in this study received a relatively high dose of radiotracer. Because the average weight of humans versus rats has a ratio of approximately 300, care needs to be taken not to induce pharmacologic effects. Andringa and coworkers recently reported that 80 MBq of [123I]-FPCIT corresponds to a quantity of 4 nmol/kg in mice, which is 500–1000 less than that required to occupy 50% of dopamine transporters in vivo [38]. Considering approximately similar pharmacological characteristics for [18F]-FECT and the lower activity used in this study, occupancies that are well below the pharmacological threshold were obtained (1–10% occupancy).

We used female Wistar rats, and it is known that the menstrual cycle may influence receptor expression. However, to the best of our knowledge, DAT and metabolic activity as used in this study are not significantly influenced by the menstrual cycle. In humans, second-order effects of gender have been reported in some studies for indirect markers such as [18F]-FDG [39]. Moreover, as reported previously, a regional intersubject variability below 7% for [18F]-FDG and [18F]-FECT uptake was found in similar female Wistar rats, indicating no confounding effects of the menstrual cycle [17].

The use of anaesthetics may significantly change tracer kinetics and uptake. It has been reported that chloral hydrate and barbiturates reduce cerebral glucose metabolism in rats [40] and that isoflurane alters the amount of DAT expressed in humans [41]. We have used barbiturate anaesthesia to preclude any effect on DAT. As only relative [18F]-FDG measurements were performed and the control group underwent exactly the same procedure, we have assumed that possible differential effects of anaesthesia procedure could be neglected.

In addition, functional differences between 6-OHDA-lesioned rats and the controls are unlikely to be related to experimental artifacts, such as the absence of intranigral insertion of the needle in controls, as it has previously been demonstrated that [18F]-FDG uptake in the injected region is similar between controls and sham operated animals [42].

Conclusion

In vivo cerebral mapping of 6-OHDA lesioned rats using dual tracer 18F-FDG and 18F-FECT small-animal PET points to a hypofunctional network incorporating the sensory-motor cortex and hippocampus, with a compensatory role for the cerebellum. Metabolically, this model shows strong functional correspondence to cortico-subcortical impairment in Parkinson’s disease patients, and thus adds further proof to the validity of the 6-OHDA lesion model to mimic multiple aspects of human disease.