OUP user menu

Striatal and cortical pre‐ and postsynaptic dopaminergic dysfunction in sporadic parkin‐linked parkinsonism

Christoph Scherfler, Naheed L. Khan, Nicola Pavese, Louise Eunson, Elizabeth Graham, Andrew J. Lees, Niall P. Quinn, Nicholas W. Wood, David J. Brooks, Paola P. Piccini
DOI: http://dx.doi.org/10.1093/brain/awh150 1332-1342 First published online: 16 April 2004


To investigate striatal and cortical pre‐ and postsynaptic dopaminergic function in parkin‐linked parkinsonism, 13 unrelated patients homozygous or compound heterozygous for parkin mutations were studied with [18F]dopa and [11C]raclopride (RAC) PET. Data were compared with a young‐onset Parkinson’s disease (YOPD) cohort, matched for age, disease severity and duration, but negative for parkin mutations. Significant changes in [18F]dopa uptake and RAC binding potential (BP) were localized in striatum using regions of interest (ROIs) and throughout the entire brain volume with statistical parametric mapping (SPM). As expected, both YOPD and parkin patients showed significant decreases in striatal [18F]dopa uptake; however, in parkin patients, additional reductions in caudate and midbrain were localized with SPM. The RAC‐BP was significantly decreased in striatal, thalamic and cortical areas (temporal, orbito‐frontal and parietal cortex) in parkin compared with YOPD patients. Our [18F]dopa PET findings suggest that, compared with YOPD, parkin disease is associated with more severe and widespread presynaptic dopaminergic deficits. The global decreases in D2 binding found in parkin compared with YOPD patients could be a direct consequence of the parkin genetic defect itself or a greater susceptibility to receptor downregulation following long‐term dopaminergic agent exposure. Cortical reductions in D2 binding may contribute to the behavioural problems reported in parkin patients.

  • parkin gene; PET; [18F]dopa; [11C]raclopride; SPM
  • AADC = aromatic amino acid decarboxylase; AI = asymmetry index; BA = Brodmann area; BP = binding potential; Ki0 = influx rate constant; IPD = idiopathic Parkinson’s disease; PCR = polymerase chain reaction; RAC = [11C]raclopride, ROI = region of interest; SPM = statistical parametric mapping; UPDRS = Unified Parkinson’s Disease Rating Scale; YOPD = young‐onset Parkinson’s disease


Mutations in the gene designated parkin (PARK2, OMIM 602544) recently have been identified in a number of families with autosomal‐recessive juvenile parkinsonism and isolated young‐onset cases (Kitada et al., 1998; Leroy et al., 1998; Abbas et al., 1999). The parkin phenotype, although in many respects similar to that of idiopathic Parkinson’s disease (IPD), is often characterized by early‐onset parkinsonism, foot dystonia, diurnal fluctuations, slow progression, a good response to anticholinergics and exquisite dose sensitivity to l‐dopa, and the frequent occurrence of behavioural and neuropsychiatric complications (Ishikawa and Tsuji, 1996; Hattori et al., 1998; Khan et al., 2002a). Autopsied cases of parkin‐linked parkinsonism have generally shown degeneration of nigral dopaminergic neurons in the absence of Lewy body inclusions (Yamamura et al., 1973; Takahashi et al., 1994; Mori et al., 1998; Maruyama et al., 2000; Van de Warrenburg et al., 2001). The parkin gene on chromosome 6q25–27 is known to code for a ubiquitin ligase involved in protein degradation presumably leading to aberrant accumulation of cell proteins resulting in neuronal dysfunction and death (Matsumine et al., 1997; Shimura et al., 2001).

[18F]Dopa PET allows the study of regional cerebral aromatic amino acid decarboxylase (AADC) activity, a marker of the functional integrity of presynaptic dopamine nerve terminals. [11C]Raclopride (RAC) PET is an in vivo marker of dopamine D2/D3 receptor binding (Garnett et al., 1983; Farde et al., 1986; Firnau et al., 1987; Martin et al., 1989; Brooks et al., 1990). To date, combined studies of presynaptic dopaminergic function and postsynaptic D2 receptor binding have only been reported for single parkin carriers and two parkin kindreds and just at a striatal level (Broussolle et al., 2000; Hilker et al., 2001; Portman et al., 2001; Pramstaller et al., 2002; Thobois et al., 2003). As a consequence, possible cortical dopaminergic changes in these patients have not been investigated.

The purpose of the present study was to: (i) examine patterns of disruption of pre‐ and postsynaptic dopaminergic function across the whole brain in a group of unrelated parkin patients using [18F]dopa and RAC PET; and (ii) determine whether differences in dopaminergic status were evident between parkin patients and young‐onset Parkinson’s disease (YOPD) patients negative for parkin mutations. [18F]Dopa and RAC PET findings were compared between young‐onset parkin‐positive and parkin‐negative patients and an older group of IPD patients all matched for clinical disease severity.



Thirteen patients homozygous or compound heterozygous for parkin gene mutations have been identified from a cohort of unrelated sporadic Parkinson’s disease patients with disease onset below age 40 years (YOPD; Table 1). Eleven of these 13 parkin‐positive and 10 parkin‐negative YOPD patients, matched for age, disease duration and severity, underwent [18F]dopa PET (Table 2). The PET findings for the YOPD patients were also compared with those of a group of 11 older IPD patients with similar disease severity but significantly shorter disease duration.

View this table:
Table 1

Parkin mutations identified in 13 cases with young‐onset Parkinson’s disease (YOPD)

Patientparkin mutationPET scans
Allele 1Allele 2
1Exon 8 deletionIntron 5 (+2 T>A)[18F]Dopa/RAC
2Exon 2 202–203 AG deletionExon 2 202–203 AG deletion[18F]Dopa/RAC
3Exon 7 924c>t Arg275TrpExon 12 1390g>a Gly430As[18F]Dopa/RAC
4Exon 7 924c>t Arg275TrpExon 5/6 deletion[18F]Dopa/RAC
5Exon 6/7 deletionUncharacterized[18F]Dopa/RAC
6Exon 3 40 bp deletion>140stopExon 11 1281g>a Asp394Asn[18F]Dopa/RAC
8Exon 12 1390g>a Gly430AspUncharacterized[18F]Dopa
9Exon 7 905t>a Cys268stopExon 2 202–203 AG deletion[18F]Dopa
10Exon 5 deletionUncharacterized[18F]Dopa
11Exon 5 deletionUncharacterized[18F]Dopa
12Exon 9 1101c>t Arg334CysExon 7 939g>a Asp280AsnRAC
13Exon 8 deletionUncharacterizedRAC
View this table:
Table 2

Demographic and clinical characteristics of parkin‐positive, parkin‐negative YOPD, IPD and control subjects

[18F]Dopa PET[11C]RAC PET
Age (years)Sex (M/F)Disease duration (years)HY stageUPDRS IIIAge (years)Sex (M/F)Disease duration (years)HY stageUPDRS III
Parkin‐positive (n = 11)Parkin‐positive (n = 8)
Mean ± SD44.8 ± 10.27/417.9 ± 7.42.9 ± 0.640.1 ± 13.3Mean ± SD52.4 ± 6.15/323.3 ± 4.93.1 ± .644.4 ± 11.3
Parkin‐negative (n = 10)Parkin‐negative (n = 9)
Mean ± SD52.7 ± 8.88/218 ± 8.72.9 ± 0.736.3 ± 10.3Mean ± SD.54 ± 8.38/119.7 ± 7.32.9 ± 0.736.4 ± 10.9
IPD (n = 11)IPD (n = 7)
Mean ± SD61.2 ± 6.9*6/5.8.6 ± 3.2*3.0 ± 0.734.1 ± ±9.5Mean ± SD.53 ± 4.74/37.9 ± 3.6**2.7 ± 0.535.4 ± 9.5
Controls (n = 16)Controls (n = 10)
Mean ± SD47.1 ± 14.416/0Mean ± SD45.8 ± 10.310/0

Values represent the means (± 1 SD). Intergroup differences were calculated by one‐way ANOVA with post hoc LSD correction (least statistical significance). *P < 0.01 and **P < 0.001 versus the parkin‐positive and parkin‐negative group. HY = Hoehn and Yahr.

Eight out of the 13 parkin‐positive YOPD patients underwent RAC PET, and findings were compared with those for nine of the 10 YOPD patients negative for parkin mutations who were also scanned with [18F]dopa PET. Additionally, a group of seven older IPD patients with similar disease severity but significantly shorter disease duration had RAC PET. [18F]Dopa and RAC PET findings for parkinsonian cases were also compared with those for separate groups of 16 and 10 normal control subjects, respectively, age‐matched to YOPD.

All parkin‐positive and parkin‐negative patients fulfilled the UK Parkinson’s Disease Society Brain Bank diagnostic criteria for IPD (Gibb and Lees, 1988). In all patients, the disease severity was measured using the Hoehn and Yahr scale (Hoehn and Yahr, 1967) and the Unified Parkinson’s Disease Rating Scale (UPDRS) in off‐drug states (Fahn and Elton, 1987). All patients were receiving regular levodopa therapy (minimum 450 mg daily; maximum 1200 mg daily) except one patient, who was on an apomorphine pump. In addition, five of the parkin‐positive patients and four of the parkin‐negative patients were taking adjunct dopamine agonist medication. One patient of the parkin‐positive group and two patients of the parkin‐negative group were taking amantadine.

Our patient subgroups comprised a higher number of males than females. Female striatal [18F]dopa and RAC uptake values lay within 1 SD of corresponding mean male values. Due to ethical issues, control subjects were males, since administering radioisotopes to healthy females could affect female germ cells more severely as they are not renewed, in contrast to male germ cells. This study was approved by the Ethics Committee of the Hammersmith Hospitals Trust, London, UK and by the Administration of Radioactive Substances Advisory Committee (ARSAC), UK. Subjects’ consent was obtained according to the Declaration of Helsinki.

Molecular analysis

PARK2 PCR amplification and sequence analysis

The 12 coding exons of the parkin gene were amplified from genomic DNA by polymerase chain reaction (PCR) using primers previously described, except for the primer for exon 3, for which exonic primers Ex3iFor and Ex3iRev were used (Kitada et al., 1998; Abbas et al., 1999). The same primers were used for the sequencing of the PCR products of the 12 exons on both strands using the Big Dye Terminator Cycle Sequencing Ready Reaction DNA Sequencing kit (Applied Biosystems, Foster City, CA), on an ABI 3100 automated sequencer with the Sequence Analysis v.3.4.1 (Applied Biosystems) software.

PARK2 semi‐quantitative PCR

Exon deletions/duplications were screened using a semi‐quantitative PCR protocol as previously described (Lucking et al., 2000). The PCR products were analysed on an ABI 377/3100 automated sequencer and Genescan v.3.1.2 and Genotype v.2.5.1 software (Applied Biosystems). Deletions and insertions of bases were deduced from the size of the PCR products.

Scanning protocol

[18F]Dopa and RAC PET were performed in patients and control subjects with a CTI‐Siemens ECAT EXACT HR++/966 positron emission tomograph (CTI, Knoxville, TN) which has an axial field of view of 23.4 cm. Emission scan data were acquired in a high‐sensitivity, three‐dimensional mode giving mean reconstructed full‐width at half‐maximum (FWHM) transaxial and axial spatial resolutions of 4.8 ± 0.2 and 5.6 ± 0.5 mm, respectively (Spinks et al., 2000). A correction for tissue attenuation of 511 KeV γ‐radiation was measured with a 5 min 3D transmission scan performed prior to tracer injection and acquired using a retractable 137Cs source. Analysis of data was performed on a SUN Ultra 10 workstation (Sun Microsystems, Silicon Valley, CA).

To minimize artefacts arising from head motion, subjects were positioned within a moulded head holder such that their orbitomeatal line was parallel to the transaxial plane of the tomograph. No obvious linear/streak artefacts from movement were evident in our statistical parametric mapping (SPM) series. Head position was carefully monitored with a video camera and by direct observation throughout. Each patient had his anti‐parkinsonian medication stopped at least 12 h before PET. Patients were pre‐treated with 150 mg of carbidopa to block peripheral dopa‐decarboxylase activity and hence prolong the availability of [18F]dopa within the blood system. Entacapone was not given to improve [18F]dopa bioavailability further as this would have precluded statistical comparison of [18F]dopa data of previously scanned IPD patients.

[18F]Dopa PET

Following a transmission scan, [18F]dopa (100–111 MBq in 10 ml of normal saline solution) was infused intravenously as a bolus over 30 s. Scanning began 30 s prior to tracer injection with a protocol of 26 time frames over 94 min 30 s (1 × 30 s; 4 × 1 min; 3 × 2 min; 3 × 3 min; and 15 × 5 min). Following correction for decay, parametric images of [18F]dopa influx rate constants (Ki0) were generated using the multiple time graphical analysis approach with an occipital reference tissue input function. [18F]Dopa Ki0 values, representing the rate of [18F]dopa uptake and storage as [18F]dopamine, were calculated for each voxel with in‐house Kronos software (Dale Bailey) written in IDL (Research Systems, Inc., Boulder, CO) (Patlak et al., 1983).


Following the intravenous injection of a bolus of 110–129 MBq of RAC, a dynamic time series of 26 frames over a period of 90 min 30 s (1 × 30 s; 1 × 15 s; 1 × 5 s; 1 × 10 s; 1 × 30 s; 4 × 1 min; and 17 × 5 min) was acquired starting 30 s prior to injection. Parametric images of RAC‐binding potential (RAC‐BP) were generated at a voxel level from the dynamic RAC scan time series using a basis function implementation of the simplified reference region compartmental model with the cerebellum as the reference tissue (Gunn et al., 1997).

Data analysis

Two methods of analysis were employed: (i) a region of interest (ROI) approach; and (ii) SPM, allowing exploratory voxel by voxel group comparisons throughout the entire brain volume without requiring an a priori hypothesis.

Four ROIs were outlined by inspection of integrated transversal RAC and [18F]dopa PET images using Analyze software (version 7.5, BRU, Mayo Foundation, Rochester, MN). These included the head of the caudate nucleus (circle 10 × 10 mm), the anterior part of the putamen (circle 10 × 10 mm), the posterior part of the putamen (circle 10 × 10 mm) and the entire putamen (ellipsis 10 × 24 mm).

Caudate and putaminal asymmetry indices (AIs) were calculated, reflecting the percentage difference between the PET signal in the respective region of higher tracer uptake compared with the contralateral region.

AI = (higher PET signal – lower PET signal/higher PET signal) × 100

Additionally, the hemispheric ratio between the caudate and putaminal PET signal was measured. Right and left caudate and putaminal [18F]dopa Ki0 and RAC‐BP were averaged for each subject to allow for one‐way analysis of variance (ANOVA) and post hoc LSD (least significant difference). For multiple comparisons between groups, a Bonferroni correction was applied. Statistical analyses were carried out using a commercial software package (SPSS for Windows 8.0, Surrey, UK).

SPM was performed using the software package SPM99 (Wellcome Department of Cognitive Neuroscience, Institute of Neurology, Queen Square, London, UK) implemented in Matlab 5.3 (Mathworks Inc., Sherborn, MA) (Friston et al., 1995). In order to localize significant changes in [18F]dopa Ki0 and RAC‐BP at a voxel level, between‐groups individual images were normalized into standard stereotaxic MNI (Montreal Neurological Institute) space with SPM. Spatial normalization of parametric maps was achieved by generating individual integrated ‘add images’ (combined time frames 1–26 of the dynamic data set) and normalizing those onto in‐house made [18F]dopa and RAC templates. Subsequently, the resulting transformation parameters were applied to the patient’s corresponding parametric [18F]dopa Ki0 and RAC‐BP images. A Gaussian kernel (6 × 6 × 6 mm) was then convolved with the spatially normalized parametric images to smooth them in order to accommodate inter‐individual anatomic variability and to improve signal to noise for the statistical analysis. Since dopamine uptake and D2 receptor density are known to be low in the occipital lobe and the cerebellum, a brain mask for those areas was created in order to minimize voxels of no interest for multicomparison corrections. A total of 61 774 voxels were analysed. The obtained data sets allowed for categorical comparisons of mean [18F]dopa Ki0 values and RAC‐BP among the parkin‐positive, parkin‐negative and control groups.


ROI analysis of [18F]dopa

Regional mean [18F]dopa Ki0 and RAC‐BP of the study groups and the intergroup statistics are detailed in Table 3. One‐way ANOVA showed a significant decrease of caudate (P < 0.001) and putamen (P < 0.001) [18F]dopa uptake in parkin‐positive patients, parkin‐negative YOPD patients, and in IPD versus normal control subjects. Parkin‐positive and parkin‐negative YOPD patients and IPD cases had similar levels and patterns of striatal [18F]dopa uptake. Comparing parkinsonian patient groups with the normal control group, the asymmetry index of [18F]dopa uptake was significantly increased in the caudate (parkin‐positive group, < 0.01; parkin‐negative YOPD group, P < 0.01; IPD group, P < 0.05) and putamen (parkin‐positive group, P < 0.05; parkin‐negative YOPD group, P < 0.001; IPD group, P < 0.05). No significant difference in striatal [18F]dopa asymmetry was found between parkin‐positive and parkin‐negative YOPD patients and IPD patients. The caudate to putamen ratio of [18F]dopa Ki0 was significantly increased in parkin‐positive (P < 0.001), parkin‐negative (P < 0.001) and IPD patients (P < 0.001) versus control subjects. There was no significant difference in caudate and putamen [18F]dopa uptake, the AI and the ratio between caudate and putamen comparing parkin‐negative YOPD patients and the IPD group.

View this table:
Table 3

Mean regional putamen (anterior, posterior) and caudate [18F]dopa Ki0 and [11C]RAC‐BP values in parkin‐positive YOPD, parkin‐negative YOPD, IPD and control subjects

Caudate+Putamen+Putamen ant.+Putamen pos. +AI (%) caudate++AI (%) putamen++Ratio caudate/putamen++
[18F]Dopa Ki0
Parkin‐positive (n = 11)
Mean ± SD0.0099 ± 0.0025***0.0056 ± 0.001***0.0072 ± 0.0015***0.0049 ± 0.0011***21 ± 12**11 ± 10*1.76 ± 0.3***
Parkin‐negative (n = 10)
Mean ± SD0.0094 ± 0.0023***0.0054 ± 0.0011***0.0067 ± 0.0019***0.0043 ± 0.0006***19 ± 14**17 ± 12***1.77 ± 0.35***
IPD (n = 11)
Mean ± SD0.0099 ± 0.0019***0.0051 ± 0.0015***0.0065 ± 0.0021***0.0042 ± 0.001***16 ± 14*11 ± 11*2.02 ± 0.4***
Controls (n = 16)
Mean ± SD0.0165 ± 0.00240.0168 ± 0.00280.0179 ± 0.00290.0179 ± 0.00376 ± 44 ± 20.99 ± 0.07
Parkin‐positive (n = 8)
Mean ± SD2.21 ± 0.35***2.84 ± 0.34***2.9 ± 0.31***2.85 ± 0.41**†6 ± 56 ± 50.78 ± 0.09**
Parkin‐negative (n = 9)
Mean ± SD2.6 ± 0.5**3.37 ± 0.543.31 ± 0.54*3.39 ± 0.596 ± 36 ± 40.77 ± 0.06**
IPD (n = 7)
Mean ± SD2.6 ± 0.46**3.38 ± 0.693.37 ± .593.51 ± 0.835 ± 35 ± 50.78 ± 0.07**
Controls (n = 10)
Mean ± SD3.16 ± 0.293.58 ± 0.33.72 ± 0.213.55 ± 0.354 ± 33 ± 30.88 ± 0.06

Values represent the means (± 1 SD). Intergroup differences were calculated by one‐way ANOVA with post hoc LSD correction (least statistical significance). *P < 0.05; **P < 0.01; ***P < 0.001versus normal controls.P < 0.05 versus parkin‐negative patients. +P <0.001; ++P < 0.01 from overall ANOVA.

ROI analysis of RAC PET

One‐way ANOVA revealed significant reductions in caudate and putamen RAC‐BP in parkin‐positive patients compared with parkin‐negative YOPD patients (caudate P < 0.05 and putamen P < 0.05) and control subjects (caudate and putamen P < 0.001). Parkin‐negative YOPD patients showed significant decreases of caudate (P < 0.01) and anterior putamen BP (P < 0.05) when compared with healthy controls. RAC‐BP was significantly decreased in the caudate of the IPD group versus control subjects (P < 0.01). The asymmetry indices were not significantly different among the four groups; however, we found significant decreases of the caudate to putamen ratio in the three patient groups when compared with the control group (parkin‐positive patients, P < 0.01; parkin‐negative patients, P < 0.01, IPD group, P < 0.01). There was no significant difference in caudate and putamen RAC‐BP and AI and the ratio between caudate and putamen comparing parkin‐negative YOPD patients and the IPD group.

[18F]Dopa PET SPM findings (Table 4A)

SPM of parametric [18F]dopa Ki0 images confirmed the results of the putaminal ROI analysis. However, significant relative decreases of Ki0 values were also localized bilaterally in the caudate of parkin‐positive compared with parkin‐negative YOPD patients (Fig. 1, Table 4AA). Furthermore, we found significant decreases in [18F]dopa Ki0 values in the ventral and dorsal midbrain in the parkin‐positive patients (areas of the red nucleus, substantia nigra and raphe nuclei) compared with both parkin‐negative YOPD patients and controls. No significant increases in [18F]dopa Ki0 values were detected in the parkin‐positive group versus the parkin‐negative YOPD and control groups. Comparing the parkin‐negative YOPD with the IPD group, no significant differences were apparent (data not shown). When compared with control subjects, significant reductions of [18F]dopa signal were localized in the caudate and putamen in all three parkinsonian patient groups, whereas significant decreases in the dorsal and ventral midbrain [18F]dopa Ki0 values were observed only in the parkin‐positive group.

Fig. 1 SPM{Z} transverse and sagittal maximum intensity projection maps rendered on to a stereotactically normalized MRI scan, showing areas of significant decreases in midbrain (A, axial slice; C, sagittal slice) and caudate (B, axial slice) [18F]dopa Ki0 uptake in parkin‐linked parkinsonism compared with parkin‐negative YOPD. Numbers in A and B correspond to the z coordinate; the number in C corresponds to the x coordinate in Talairach space.

Fig. 2 SPM{Z} transverse maximum intensity projection maps rendered on to a stereotactically normalized MRI scan, showing areas of significant decreases in temporal (A), striatal and temporal (B), frontal, striatal and parietal (C), and frontal and parietal (D) RAC‐BP in parkin‐linked parkinsonism compared with parkin‐negative YOPD. Numbers correspond to the z coordinate in Talairach space.

View this table:
Table 4A

Between‐group SPM findings showing the locations of significant decreases of [18F]dopa Ki0 in parkin‐positive, parkin‐negative YOPD and control subjects

[18F]Dopa Ki0Talairach coordinatesZ scoreP values (corrected)Height threshold
BA x y z
Decreases in parkin‐positive patients compared with parkin‐negative patients
 Left ventral midbrain*–14–8–84.030.00010.01
 Right ventral midbrain*)10–20–23.560.0001
 Right dorsal midbrain8–36–153.230.0001
 Left caudate–160243.730.00010.05
 Right caudate2211203.320.0001
Decreases in parkin‐positive patients compared with control subjects
 Left putamen–3000Inf0.00010.001
 Right putamen30–227.340.0001
 Right caudate201686.290.0001
 Left caudate–141606.050.0001
 Right ventral midbrain6–20–115.350.0001
 Left ventral midbrain–8–16–94.910.0001
 Dorsal midbrain0–31–154.410.0001
Decreases in parkin‐negative patients compared with control subjects
 Left putamen –3000Inf0.00010.001
 Right putamen 28247.270.0001
 Left caudate –16896.440.00010.05
 Right caudate 168115.970.0001

*Red nucleus, substantia nigra and raphe nuclei.

RAC PET SPM findings (Table 4B)

The categorical comparison of striatal RAC‐BP between parkin‐positive and parkin‐negative YOPD patients reflected the results obtained from the ROI analysis, showing significant relative reduction of RAC‐BP in the caudate and putamen in the parkin‐positive group. Additionally, significant relative reductions in RAC‐BP were localized in the parietal cortex, temporal cortex and bilaterally in the prefrontal cortex. When compared with controls, parkin‐positive patients showed widespread decreases in RAC‐BP over cortical [Brodmann area (BA) 10; 21; 22; 40 (bilateral); 37 (right)], striatal and thalamic (medial dorsal nucleus, bilateral) areas. Significant reductions of RAC‐BP in parkin‐negative YOPD patients versus controls were localized in the left and right caudate and anterior putamen, in the thalamus and parieto‐temporal cortical areas (Table 4BB). No significant differences in RAC‐BP were observed in the group of parkin‐negative YOPD patients when compared with IPD patients.

View this table:
Table 4B

Between‐group SPM findings showing the locations of significant decreases of [11C]RAC‐BP in parkin‐positive, parkin‐negative YOPD and control subjects.

[11C]RAC‐ BPTalairach coordinatesZ scoreP values (corrected)Height threshold
BA x y z
Decreases in parkin‐positive patients compared with parkin‐negative patients
 Left area 40–39–48454.610.0380.01
 Left putamen (ant.)–26694.30.0001
 Right putamen29–5134.130.0001
 Right area 3755–49–113.980.0001
 Right area 2161–24–93.730.0001
 Left area 22–50–53213.750.01
 Right area 2253–34163.490.002
 Right caudate169203.320.0001
 Right area 4057–27363.280.002
 Left caudate–1614123.070.02
 Right area 102648203.450.00010.05
 Left area 10–1058–83.370.048
Decreases in parkin‐positive patients compared with control subjects
 Right putamen28836.040.00010.001
 Left caudate–1619–35.750.0001
 Right caudate–2612–15.110.0001
 Left putamen1219–15.640.0001
 Left area 40–40–50474.860.029
 Right area 102851204.610.002
 Right area 3746–41–134.480.0001
 Left thalamus (dm)–4–9124,360.001
 Right area 4046–33464.300.006
 Right area 21/2257–3763.880.046
 Right thalamus (dm)6–2283.740.0001
 Left area 21–50–12–103.620.022
 Left area 22–59–3173.580.025
Decreases in parkin‐negative patients compared with control subjects.
 Right anterior putamen261614.740.00010.001
 Left caudate–161634.630.0001
 Right caudate1019–13.880.0001
 Left anterior putamen–186–53.820.0001
 Right thalamus (dm)4–1043.930.00010.01
 Left thalamus (dm)–4–1783.570.0001

Dm = dorsomedial; ant = anterior.


This is the first [18F]dopa and RAC‐PET study to characterize, within the entire brain volume, abnormalities in pre‐ and postsynaptic dopaminergic function in a group of sporadic patients with parkin‐linked parkinsonism.

ROI analysis revealed a severe reduction of caudate and putaminal [18F]dopa uptake in parkin‐positive and parkin‐negative YOPD patients and in IPD when compared with controls. A similar striatal pattern of reduced [18F]dopa uptake in parkin‐positive and parkin‐negative YOPD patients was evident with both groups showing the typical IPD posterior–anterior gradient with most prominent loss in the posterior part of the putamen. These findings are in line with previously published case reports (Broussolle et al., 2000; Hilker et al., 2001; Portman et al., 2001; Khan et al., 2002a; Thobois et al., 2003).

SPM interrogation of [18F]dopa uptake, comparing parkin‐positive and parkin‐negative YOPD patients, localized significant symmetrical relative reductions in the caudate and midbrain areas of parkin‐positive patients. A possible explanation for the failure of the ROI approach to detect these caudate decreases in the parkin‐positive cases may be that the ROIs sampled the dorsal head of caudate, whereas SPM localized relative reductions in the body of the nucleus. Furthermore, ROI analysis is based on sampling brain volumes with a priori categorical assumptions as to their size and shape, whereas no a priori hypothesis regarding the localization of PET signal change is required by SPM. Monoaminergic neurons in striatal regions are predominantly dopaminergic and so [18F]dopa PET provides a marker of functional integrity of nigrostriatal dopaminergic neurons. However, in midbrain areas, [18F]dopa is taken up not only by dopaminergic neurons but also by serotoninergic and noradrenergic neurons expressing AADC (Tison et al., 1991). Consequently, reduced [18F]dopa Ki0 in the dorsal and ventral midbrain of the parkin‐positive group may imply a more widespread dysfunction of catecholaminergic and serotoninergic neurotransmitter systems. This would be in line with the recently reported neuropathological findings of loss of pigmented neurons and gliosis in both the substantia nigra pars compacta and locus coeruleus in patients with parkin mutations in the absence of Lewy bodies (Hayashi et al., 2000). Decreases in midbrain [18F]dopa uptake have also been reported in IPD patients with advanced disease (Rakshi et al., 1999).

The discordance between moderate disability and severe nigrostriatal dopaminergic dysfunction in parkin patients parallels findings reported for patients with parkinsonism due to PARK6 (Khan et al., 2002b). It suggests that significant dopamine cell loss occurs early, but then progresses slowly enough to allow compensatory mechanisms to develop. No significant [18F]dopa Ki0 reductions were found in cortical regions on comparing parkin‐positive and parkin‐negative YOPD patients with normal controls. In agreement with our findings, Rakshi et al. (1999) using SPM also detected no significant reductions in frontal [18F]dopa Ki0 in advanced IPD, suggesting that AADC activity is maintained in this area. It has also been observed recently that, even in advanced stages of the disease, the release of endogenous dopamine in prefrontal cortical areas after amphetamine is similar to that of normal subjects (Piccini et al., 2003).

When comparing parkin‐negative patients with older IPD patients matched for disease severity, no significant differences in mean regional cerebral [18F]dopa Ki0 values were found. The effect of age on [18F]dopa Ki0 values in healthy subjects has remained a controversial issue. Two studies found a mild inverse relationship between age and striatal [18F]dopa Ki0 values (Martin et al., 1989; Laasko et al., 2002), whereas others found no correlation (Sawle et al., 1990; Eidelberg et al., 1993; Ishikawa et al., 1996). If a decline of striatal [18F]dopa Ki0 is associated with age, our data would indicate that [18F]dopa uptake in YOPD patients is either similarly or more affected than in IPD patients with similar disease severity.

Both ROI analysis and SPM detected significant and uniform decreases of RAC‐BP in the caudate and putamen of parkin‐positive patients compared with parkin‐negative cases. This finding in sporadic parkin patients is in line with the study of Hilker et al. (2001) who reported decreases in caudate and putaminal RAC‐BP in five members of the same parkin kindred. Compared with controls, RAC‐BP was significantly reduced in the caudate but not in the putamen of parkin‐negative YOPD and IPD patients. These findings are in keeping with other RAC‐PET and neuropathological data indicating that putamen dopamine D2 receptor binding remains normal in treated patients with IPD (Guttman et al., 1985; Ahlskog et al., 1991; Brooks et al., 1992; Griffiths et al., 1994; Antonini et al., 1997). Over the 31–74 year age range of normal control subjects, a mild but insignificant decline in RAC‐BP has been reported, while a steep decline in striatal D2 receptor densities has been noted over the first three decades of life (Brooks et al., 1992; Antonini et al., 1993). There appears to be little further loss of D2 sites in later life. In our study no statistical difference with regard to mean age was apparent when comparing patient groups versus controls, although patients were on average 8 years older than control subjects. Assuming that putaminal RAC‐BP declines by ∼0.6% per year (Antonini et al., 1993), one would expect at most a 5% decline of RAC‐BP in our patient groups compared with normal subjects, whereas there was an observed 18% reduction of putaminal RAC binding.

While nigral degeneration of dopaminergic neurons results in an initial 10–20% upregulation of D2 receptor binding in early IPD, adaptive postsynaptic mechanisms and treatment exposure normalize this as the condition advances (Brooks et al., 1992; Ahlskog et al., 1991; Antonini et al., 1997). This viewpoint is supported by animal studies: RAC‐BP was normal in four parkinsonian rhesus monkeys that had been treated with 1‐methyl‐4‐phenyl‐1,2,3,6‐tetrahydropyridine 6–7 years before PET and never received antiparkinsonian treatment (Doudet et al., 2000). As initial putamen D2 receptor upregulation normalizes in chronically medicated patients with IPD, the reduced D2 availability found in parkin‐linked parkinsonism suggests that the genetic defect itself may be responsible rather than loss of dopamine projections or exposure to medication. Longitudinal studies on striatal D2 site density in cohorts of untreated and treated parkin patients are required to clarify the issue.

Cortical D2 receptor binding

Autoradiography and PET studies with dopamine D2/D3 receptor ligands such as [3H]RAC, [125I]epidepride and [11C]FLB 457 have shown that extrastriatal D2 density is highest in the thalamus, lower in prefrontal, temporal and parietal cortex, and lowest in the occipital cortex (Lidow et al., 1989; Kessler et al., 1993; Hall et al., 1996). Although the D2/D3 receptor affinity for [11C]FLB is ∼100 times higher than it is for RAC (Halldin et al., 1995), recent improvement of PET scanner sensitivity and use of SPM have revealed significant reductions in RAC‐BP in prefrontal, temporal, parietal and thalamic areas when normal subjects receive D2 receptor antagonist neuroleptics (unpublished data) and in patients with Huntington’s disease as a consequence of disease progression (Pavese et al., 2003). The location of the extrastriatal loss of D2 function in our parkin‐linked parkinsonism is consistent with the distribution of D2 receptor reported in post‐mortem studies of the human brain (Kessler et al., 1993; Hall et al., 1996; Joyce et al., 1998).

Significant reductions in extrastriatal RAC‐BP were observed in thalamic, temporal, parietal and prefrontal brain areas of parkin patients when compared with controls and patients with parkin‐negative YOPD. RAC‐BP was also significantly decreased in thalamic, temporal, parietal and frontal brain areas of patients with parkin‐negative YOPD versus control subjects. The latter finding is in line with a recently published study investigating [11C]FLB distribution in medicated patients with IPD (Kaasinen et al., 2000). To date, there are no studies correlating cortical D2 receptor status and clinical symptoms. It has been reported that patients with parkin mutations have an increased prevalence of behavioural and neuropsychiatric disorders in comparison with patients with IPD (Khan et al., 2002a). A retrospective analysis of the clinical notes revealed symptoms such as depression (two patients), panic attacks (three patients) and paranoia (one patient) in our patients with parkin‐linked parkinsonism. These behavioural disorders might be associated in part with the disruption of dopaminergic pathways interconnecting the prefrontal cortex and ventral striatum (Ribeyre et al., 1994; Farde et al., 1997; Joyce et al., 1997).


In contrast to parkin‐negative YOPD and IPD patients, parkin‐linked parkinsonism is associated with more widespread and severe disruption of striatal dopaminergic and midbrain catecholaminergic and serotoninergic pathways. Reductions of postsynaptic D2 receptors in parkin‐linked parkinsonism have been localized in striatal, thalamic and cortical areas. This may be a direct consequence of the genetic defect per se or a higher susceptibility to dopaminergic treatment.


We wish to thank our colleagues at the MRC Clinical Sciences Centre, Radiochemistry and Methods Section whose expertise made these studies possible, Hope McDevitt, Stella Ahier and Andrew Blyth for their expert help with scanning, and colleagues Mary Sweeney and Peter Dixon in the Neurogenetics Unit, Institute of Neurology, for helpful discussions. This study was supported by the Parkinson’s Disease Society, UK.


View Abstract