OUP user menu

Striatal dopaminergic markers in dementia with Lewy bodies, Alzheimer's and Parkinson's diseases: rostrocaudal distribution

M. A. Piggott, E. F. Marshall, N. Thomas, S. Lloyd, J. A. Court, E. Jaros, D. Burn, M. Johnson, R. H. Perry, I. G. McKeith, C. Ballard, E. K. Perry
DOI: http://dx.doi.org/10.1093/brain/122.8.1449 1449-1468 First published online: 1 August 1999

Summary

Dementia with Lewy bodies (DLB) is a neuropsychiatric disease associated with extrapyramidal features which differ from those of Parkinson's disease, including reduced effectiveness of L-dopa and severe sensitivity reactions to neuroleptic drugs. Distinguishing Alzheimer's disease from DLB is clinically relevant in terms of prognosis and appropriate treatment. Dopaminergic activities have been investigated at coronal levels along the rostrocaudal striatal axis from a post-mortem series of 25 DLB, 14 Parkinson's disease and 17 Alzheimer's disease patients and 20 elderly controls. [3H]Mazindol binding to the dopamine uptake site was significantly reduced in the caudal putamen in DLB compared with controls (57%), but not as extensively as in Parkinson's disease (75%), and was unchanged in Alzheimer's disease. Among three dopamine receptors measured (D1, D2 and D3), the most striking changes were apparent in relation to D2. In DLB, [3H]raclopride binding to D2 receptors was significantly reduced in the caudal putamen (17%) compared with controls, and was significantly lower than in Parkinson's disease at all levels. D2 binding was significantly elevated at all coronal levels in Parkinson's disease compared with controls, most extensively in the rostral putamen (71%). There was no change from the normal pattern of D2 binding in Alzheimer's disease. The only significant alteration in D1 binding ([3H]SCH23390) in the groups examined was an elevation (30%) in the caudal striatum in Parkinson's disease. There were no differences in D3 binding, measured using [3H]7-OH-DPAT, in DLB compared with controls. A slight, significant decrease in D3 binding in the caudal striatum of Parkinson's disease (13%) patients and an increase in Alzheimer's disease (20%) in the dorsal striatum at the level of the nucleus accumbens were found. The concentration and distribution of dopamine were disrupted in both DLB and Parkinson's disease, although in the caudate nucleus the loss of dopamine in DLB was uniform whereas in Parkinson's disease the loss was greater caudally. In the caudal putamen, dopamine was reduced by 72% in DLB and by 90% in Parkinson's disease. The homovanillic acid : dopamine ratio, a metabolic index, indicated compensatory increased turnover in Parkinson's disease, which was absent in DLB despite the loss of substantia nigra neurons (49%), dopamine and uptake sites. These differences between DLB, Parkinson's disease and Alzheimer's disease may explain some characteristics of the extrapyramidal features of DLB and its limited response to L-dopa and severe neuroleptic sensitivity. The distinct changes in the rostrocaudal pattern of expression of dopaminergic parameters are relevant to the interpretation of the in vivo imaging and diagnosis of DLB.

  • human striatum
  • dementia with Lewy bodies
  • Alzheimer's disease
  • Parkinson's disease
  • dopamine receptors
  • ANOVA = analysis of variance
  • DLB = dementia with Lewy bodies
  • HVA = homovanillic acid
  • 7-OH-DPAT = R(+)-7-hydroxy-dipropylaminotetralin-2-N,N-di[2,3(n)-propylamino]-7-hydroxy-1,2,3,4-tetrahydronaphthalene
  • PPAP = R(–)-N-(3-propyl-1-propyl)-1-phenyl-2-aminopropane hydrochloride
  • SCH23390 = R(+)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine hydrochloride

Introduction

Dementia with Lewy bodies (DLB) is a progressive, degenerative dementia which is the second commonest after Alzheimer's disease. Clinical features of DLB include disturbances of consciousness and recurrent visual hallucinations, with progressive cognitive decline which develops to severe dementia. Spontaneous extrapyramidal symptoms occur in the majority of cases, including masked face, stooped posture, slow gait and rigidity. Resting tremor also occurs, but probably less than in Parkinson's disease (Galasko et al., 1996; McKeith et al., 1996). These parkinsonian signs are associated with reduced concentration of striatal dopamine (Langlais et al., 1993; Marshall et al., 1994; Piggott and Marshall, 1996) and reduced neuron density in the substantia nigra pars compacta (Perry et al., 1990, 1993). Dopamine uptake site density in the striatum is also reduced but to a lesser extent than in Parkinson's disease (Piggott et al., 1998), and D2 receptors are not upregulated (Piggott et al., 1994, 1998). Extrapyramidal signs in DLB can be severely exacerbated, or appear for the first time, following administration of the neuroleptic drugs that are usually prescribed to control psychotic symptoms such as hallucinations and delusions. Severe neuroleptic reactions, which do not necessarily resolve upon drug withdrawal, include rigidity, reduced consciousness, pyrexia, falling, postural hypotension and collapse (McKeith et al., 1992, 1995; Ballard et al., 1998a).

In Parkinson's disease, a progressive extrapyramidal movement disorder (rigidity, bradykinesia and tremor) is accompanied by increasing loss of striatal dopamine and severely reduced substantia nigra neuron density as well as reduced dopamine uptake sites, particularly in the putamen (Antonini et al., 1995; Wilson et al., 1996). At early stages of the disease or without dopamine replacement therapy, increased dopamine D2 receptors have been demonstrated in the striatum, both in vitro (Guttman, 1987; Piggott and Marshall, 1996) and by in vivo imaging (Antonini et al., 1995, 1997b).

In Alzheimer's disease, extrapyramidal features may also emerge, particularly in the later stages (Stern et al., 1996; Lopez et al., 1997). Reports of nigrostriatal neurochemical abnormalities vary, with some reports of little or no loss in dopamine uptake in vivo in Alzheimer's disease (Tyrrell et al., 1990; Donnemiller et al., 1997) and other reports of reduced dopamine uptake sites in vitro (Sahlberg et al., 1998) and in vivo (Rinne et al., 1998). There are reports that substantia nigra neuron numbers and striatal dopamine concentration are unchanged (Perry et al., 1990, 1993; Love et al., 1996; Liu et al., 1997), but there is one report of reduced substantia nigra neuron density (Kazee et al., 1995). Extrapyramidal symptoms in Alzheimer's disease may be associated with neurofibrillary tangles in the substantia nigra (Liu et al., 1997). Unaltered or reduced D2 receptors have also been reported (Pizzolato et al., 1996). This variability of reported findings of dopaminergic activities in Alzheimer's disease may be due to heterogeneity within the cases selected and the symptoms displayed (Forstl et al., 1994; Victoroff et al., 1996) and also to the potential inclusion of patients with DLB (Ellis et al., 1996; Kalra et al., 1996).

We have recently reported increasing gradients of dopamine uptake sites and D2 receptors in the striatum of normal elderly individuals from the head of the caudate rostrally to the internal globus pallidus caudally (Piggott et al., 1999). The number of D1 receptors was higher in the rostral putamen, D3 receptors were mainly expressed in the ventral striatum, and dopamine and homovanillic acid concentrations were higher at the level of the nucleus accumbens and caudal to the anterior commissure (Piggott et al., 1999).

In the present paper we report the rostrocaudal striatal distribution of dopaminergic markers in a post-mortem series of DLB, Parkinson's disease and Alzheimer's disease cases defined clinically and pathologically. We have distinguished changes characteristic of disease-related movement disorder from normal, aged individuals, and established parameters which have the potential to differentiate DLB from Alzheimer's disease and Parkinson's disease by in vivo imaging.

Methods

Cases

The series, selected from the Newcastle Brain Tissue Bank, included 25 cases with DLB, 14 Parkinson's disease cases and 17 Alzheimer's disease cases. Controls were 20 normal elderly individuals with no history of neurological or psychiatric disease, no record of l-dopa or neuroleptic prescription and no evidence of significant age-related neurodegeneration. Permission for post-mortem and donation of brain tissue were obtained by prior consent from next of kin, in accordance with the rules of the Joint Ethics Committee, Newcastle and North Tyneside Health Authorities, and brain-banking procedures were in line with MRC guidelines. Brains were removed at autopsy. The right hemisphere and brainstem were fixed in formalin and examined histologically, while the left hemisphere was coronally sliced, snap-frozen and stored at –70°C. Table 1 shows demographic case variables and the coronal levels that were quantified for each parameter. Clinically, DLB cases fulfilled the criteria of McKeith and colleagues (McKeith et al., 1992, 1996) and Alzheimer's disease cases were diagnosed with regard to published criteria (McKhann et al., 1984). All Parkinson's disease cases presented with clinical signs of extrapyramidal movement disorder, including tremor, rigidity and akinesia (Perry et al., 1985). The presence of dementia in Parkinson's disease was assessed using previously described criteria (Perry et al., 1990); if dementia developed within a year of onset of motor symptoms the cases were assigned to the DLB group (McKeith et al., 1996). Pathologically, DLB and Parkinson's disease were distinguished from Alzheimer's disease by the presence of brainstem Lewy bodies, cortical Lewy bodies, Lewy neurites in the CA2/3/4 segments of the hippocampus, and low or moderate Alzheimer-type pathology with fewer tangles than found in Alzheimer's disease (Perry et al., 1990, 1996) (Table 2). Parkinson's disease cases showed greater substantia nigra neuron loss, less extensive Lewy body formation in the neocortex and less Alzheimer-type pathology than DLB cases. Pathologically, Alzheimer's disease cases showed tangles and neuritic plaques in the hippocampus and neocortex, which were present at sufficient densities to satisfy published criteria (Braak and Braak, 1991; Mirra et al., 1991) (Table 2).

View this table:
Table 1

Details of subjects

Controls
Case no.Age (years)SexPM delay (h)Cause of deathCoronal levels at which parameters were measured
DATD1D2D3DA and HVA
189F24Pulmonary embolism1414151414
278F17Cancer101010,1110
356M18Cancer11111011
494F74Renal and heart failure10,12,13,1410,11,13,1410,13,1510,11,13,1410,12,13,14
580F48Ischaemic heart disease11101111
677M48Cancer13131313,1414
772F14Cancer11101111
878M11Bronchopneumonia11
963M16Cancer + haemorrhage121111,1312
1082F15Emphysema1211111112
1178M42Heart disease + pneumonia1314131413
1282M24Cancer11111111
1375M12Ischaemic heart disease11131111
1484M76Cancer12,1311,13131312,13
1576F24Cancer1111111111
1690F48Ischaemic heart disease10,11111010,11
1782F12Bronchopneumonia13,141313,151413,14
1855M24Cancer10,111010,111010,11
1991F48Ischaemic heart disease11111111
2064F89Cancer12,13,1413,1413,151312,13,14
Mean77.334.2
SD10.923.8
Dementia with Lewy bodies
Case no.Age (years)SexPM delay (h)Cause of deathDisease duration (years)Neuroleptic medication*EPSCoronal levels at which parameters were measured
DATD1D2D3DA and HVA
2178F48Ischaemic heart disease 1Yes S+11111311
2271F10Bronchopneumonia 6NoNo1213131412
2374M45Duodenal ulcer haemorrhage 1No+1110111011
2475M24Bronchopneumonia 1.5YesNo?1111111111
2578M47Cancer 2No1313141413
2680M 5Pulmonary embolism ?Yes S+++1313111313
2790M52Bronchopneumonia 4Yes S+1213111313
2884M16? 1Yes S+1111111111
2983M72Ischaemic heart disease11Yes S+11,1311,1311,1310,1311,13
3066F10Pyelonephritis 3Yes S+11,1311,1311,1410,1411,13
3182F24Bronchopneumonia ?Yes+1111111111
3269M27Ischaemic heart disease 2NoNo11,1311,131110,1411,13
3385M46Bronchopneumonia ?No+1213141312
3478M24? ?Yes+1111111011
3577F32Pyelonephritis11Yes S+1111111111
3672M18Bronchopneumonia 6Yes11101110,11
3782M90Bronchopneumonia 1Yes S++1111111011
3890M24Bronchopneumonia 5Yes++11101111,1410
3983M61Pneumonitis 2Yes S++11101110
4078M14Bronchopneumonia 5No1111111011
4169M 7Cachexia 4No1213111112
4266F72Bronchopneumonia 3?+131313
4385F18Bronchopneumonia 5No+13131413
4484M51Cancer 2No+11111110,11
4574F12Heart disease 1Yes+1313131313
Mean78.1233.96
SD 6.823.5
Parkinson's disease
Case no.Age (years)SexPM delay (h)Cause of deathDisease duration (years)l-Dopa medicationNeuroleptic medicationCoronal levels at which parameters were measured
DATD1D2D3DA and HVA
4673M 22Peritonitis 8YesNo11111111
4765M 60Cancer20YesNo1010101010
4872F 35Bronchopneumonia 3BrieflyNo1111111110
4972F 80Myocardial infarction 7YesNo1413131113
5072M 7Bronchopneumonia11YesNo11131311
5165M 30Ischaemic heart disease 3YesNo13
5270M 35Pulmonary embolism 3NoNo10,11,1210,1110,11,1210,1110,11,13
5386M 48Ischaemic heart disease13YesNo1212101113
5484M 78Pulmonary embolism 9YesYes1213121113
5567M 24Renal failure 5YesNo1010101010
5678M 24Bronchopneumonia 4YesNo11,131311,1311,1311,13
5783F120Bronchopneumonia 8YesNo1312131313
5873F 8Cancer11YesYes10,111110,1110,11
5983F 48? ?YesNo1312131013
Mean74.5 44.21
SD 7.1 31.4
Alzheimer's disease
Case no.Age (years)SexPM delay (h)Cause of deathDisease duration (years)Neuroleptic medicationEPS†Coronal levels at which parameters were measured
DATD1D2D3DA and HVA
PM = post-mortem; *S = severe reaction; EPS = extra pyramidal symptoms: + to +++ = increasing severity of parkinsonism.
6089F37Bronchopneumonia 1YesNo910141010,13
6191M21Coronary occlusion 3NoNo11,1311111111
6266F39Bronchopneumonia 3YesYes9,11111110,11
6377M96Bronchopneumonia 6YesNo1110111010
6476F48??No1313141313
6586F53Bronchopneumonia?No1313141313
6684F21Cardiac failure?Yes1111111111
6774M36Pulmonary embolism13Yes1111111111
6883F24Peritoneal abscess 5YesNo1313141313
6989M28Myocardial infarction 6Yes, initially1111111111
7084M24Bronchopneumonia 7YesNo1313141313
7183F24Cancer 1NoNo11111111
7288F48Ischaemic heart disease10YesNo11111111
7383F90Cancer, pulmonary embolism 5YesYes1111111111
7469M24Bronchopneumonia 9NoYes9,11,1310,11111011
7572M10Bronchopneumonia 5YesYes, terminally9
7679F72? ??13131413
Mean80.7640.88
SD 7.424.7
View this table:
Table 2

Pathological findings in DLB, Parkinson's disease and Alzheimer's disease

GroupLewy bodiesSubstantia nigra neuron numbersPlaque count per mm2NFT count per mm2Braak stage
SNCortexMean ± SDRangeMean ± SDRangeMean ± SDRangeMean ± SDRange
AD = Alzheimer's disease; PD = Parkinson's disease; NFT = neurofibrillary tangles; SN = substantia nigra.
Control494 ± 142307–683 3.4 ± 4.70–14.10.04 ± 0.120–0.52.3 ± 1.30–4
DLB+++251 ± 119 79–51512.9 ± 7.90–26 1.2 ± 2.60–93.3 ± 1.40–5
PD++156 ± 82 49–356 4.1 ± 6.50–220.15 ± 0.20–0.62.2 ± 1.40–4
AD489 ± 197241–94823.5 ± 18.17.9–4712.9 ± 5.81.6–23.25.3 ± 0.74–6

Neuropathological methods

The neuropathological methods used to diagnose the Alzheimer's disease, Parkinson's disease and DLB cases and to quantify the plaques, neurofibrillary tangles and Lewy bodies have been described previously (Perry et al., 1990). Briefly, the fixed right cerebral hemisphere was sliced coronally and tissue was sampled from the midfrontal cortex, lateral parietal cortex, occipital cortex, temporal cortex, hippocampus, basal ganglia and brainstem. Paraffin sections 5 μm thick were cut and stained with haematoxylin and eosin for general histology, and used for the identification of Lewy bodies in the substantia nigra. In cortical areas, ubiquitin immunochemistry was used for the identification and quantification of Lewy bodies. Paraffin sections (20 μm) of the upper and lower midbrain were stained with cresyl fast violet and used for quantification of pigmented neurons in the substantia nigra. Paraffin sections (20 μm) of cortical areas were stained with a modification of Palmgren's silver technique for neurofibrillary tangle demonstration and quantification (Cross, 1982). The von Braunmühl silver impregnation technique was used to demonstrate plaques in 25-μm thick frozen sections cut from fixed tissue blocks adjacent to those taken for paraffin processing. For each of the four neocortical areas the mean tangle density was obtained by counting tangles in consecutive fields (0.61 mm2 area) through the full width of the cortical ribbon in five randomly marked positions around the gyri (two at the crest, two in the mid-sulcal zone and one at the base of the sulcus); the mean plaque density was calculated from counts in fields (area = 3.1 mm2), at five similarly marked points. In each case the tangle and plaque densities in the whole of the neocortex were both expressed as mean values per mm2.

Brain sampling

For autoradiography, frozen tissue blocks were subdissected at –20°C, and 20 μm cryostat sections were cut and dried onto glass slides previously coated with Vectabond. All dopaminergic parameters were measured along the rostrocaudal axis, but because of tissue availability it was not possible to measure all parameters at all levels. Coronal levels in the anterior–posterior commissure axis (Fig. 1) were 0.5 cm apart and were designated according to Perry (Perry, 1993). Level 9 (not shown) marks the head of the caudate; level 10 the head of the putamen; level 11 the nucleus accumbens; level 12 the first appearance of the external globus pallidus and anterior commissure; level 13 the rostral limit of the temporal/frontal lobe junction; level 14 the rostral fornix, internal globus pallidus, amygdala and temporal horn of the lateral ventricle; and level 15 displays the lentiform nucleus formation, the mammillary body and the anterior nucleus of the thalamus.

Fig. 1

Coronal levels of striatum, 0.5 cm apart; taken from the publication by R. Perry (Perry, 1993). AC = anterior commissure; Amyg = amygdala; Ant Nuc = anterior nucleus of thalamus; ant = anterior; Caud = caudate; Cerb Ped = cerebral peduncle; Chi = optic chiasma; cl = coronal level; Fx = fornix; GPe = external globus pallidus; GPi = internal globus pallidus; Int Caps = internal capsule; Lat Vent Temp = temporal horn of lateral ventricle; MB = mammillary body; NA = nucleus accumbens; O Tb = olfactory tubercle; post = posterior; Putm = putamen. Receptor densities were measured in the dorsal and ventral caudate and putamen, and in the nucleus accumbens.

Materials

[3H]7-Hydroxy-dipropylaminotetralin (7-OH-DPAT) was supplied by Amersham (Amersham, UK), and NEN (London, UK) supplied [3H]mazindol, [3H]7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine hydro-chloride (SCH23390) and [methoxy-3H]raclopride. Butaclamol, desipramine, dopamine, cis-flupenthixol, ketanserin, nomifensine and R(–)-N-(3-propyl-1-propyl)-1-phenyl-2-aminopropane hydrochloride (PPAP) were supplied by Sigma (Poole, UK) and Vectabond by Vector Laboratories (Peterborough, UK). Other laboratory reagents were analytical grade, supplied by Sigma (Poole, UK).

Autoradiography protocols

For all ligands except [3H]mazindol, cryostat sections were dried at room temperature for 2 h before storage at –70°C for up to 4 weeks prior to use. For mazindol binding, slides were dried for only 5 min prior to storage at –70°C overnight before use. At least four control cases were measured more than once at a particular coronal level to provide an additional control for consistency between each autoradiography experiment. Autoradiography protocols included preincubation in buffer to remove endogenous ligands and residual drugs, followed by incubation with radioactive ligand to determine total binding. Adjacent sections were incubated with the addition of a displacer to determine non-specific binding. Sections were given three washes in buffer, followed by a water dip to remove buffer salts, and dried under a stream of air prior to exposure to film. Triplicate determinations for both total and non-specific binding were made, and specific binding was calculated by subtracting the mean value of non-specific binding from the mean total binding.

Dopamine receptor autoradiography

Dopamine uptake sites

Dopamine uptake site binding was determined with 3 nM [3H]mazindol, a concentration about half the dissociation constant Kd(h) reported previously (Singer et al., 1991; Alexander et al., 1992). Sections were removed from overnight storage at –70°C for 1 min prior to preincubation in 50 mM Tris–HCl (pH 7.9) at 4°C for 5 min, and were incubated with 3 nM [3H]mazindol (specific activity = 24 Ci/mmol) in 50 mM Tris–HCl buffer (pH 7.9) containing 300 mM NaCl, 5 mM KCl and 100 nM desipramine (to block other monoamine uptake sites). Non-specific binding was defined in the presence of 100 μM nomifensine and was ~60% of total binding. Sections were given three 1-min washes at 4°C in buffer as used for the prewash.

Dopamine D1 receptors

For dopamine D1 receptor autoradiography, a modification of the method of Cortes (Cortes et al., 1989) was used. Preincubation for 20 min at 4°C in 50 mM Tris–HCl (pH 7.4) was followed by incubation in 50 mM Tris–HCl (pH 7.4) containing 120 mM NaCl, 5 mM KCl, 2.5 mM CaCl2, 1 mM MgCl2 and 1 nM [3H]SCH23390 (specific activity = 71.1 Ci/mmol) with 100 nM ketanserin (to block 5-HT2 binding sites) for 150 min at room temperature. A [3H]SCH23390 concentration of 1 nM is slightly less than the Kd (dissociation constant) previously reported (Waddington and O'Boyle, 1987; Hall et al., 1994). Non-specific binding was 30–40% of total binding, and was determined in the presence of 2 μM cis-flupenthixol. Slides were washed for a total of 9 min in 50 mM Tris–HCl (pH 7.4) buffer at 4°C.

Dopamine D2 receptors

For D2 receptors a protocol developed from the method of Kohler (Kohler and Radesater, 1986) was used. Sections were preincubated for 30 min in 50 mM Tris–HCl (pH 7.4) at room temperature and incubated with 3 nM [3H]raclopride (specific activity = 69.5 Ci/mmol) in buffer composed of 50 mM Tris–HCl (pH 7.4) containing 120 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2 and 0.001% (w/v) ascorbic acid for 60 min at room temperature. This concentration of [3H]raclopride is above the Kd reported by Hall (Hall et al., 1994) of 1.25 nM and within the range reported by Dean (Dean et al., 1997) of 2–10 nM, in man. Non-specific binding was assessed in the presence of 1 μM butaclamol and was ~25% of total binding. Slides were washed for a total of 9 min in 50 mM Tris–HCl (pH 7.4) buffer at 4°C. The method for dopamine D2 receptors also measures a proportion of the smaller population of D3 receptors (Malmberg et al., 1993), although Landwehrmeyer and colleagues (Landwehrmeyer et al., 1993) found that [3H]raclopride did not label D3 receptors in the islands of Calleja.

Dopamine D3 receptors

D3 autoradiography was carried out using modifications of the methods of Lévesque and colleagues (Lévesque et al., 1992) and Herroelen and colleagues (Herroelen et al., 1994), as previously described (Piggott et al., 1999). Sections were preincubated twice for 10 min in 50 mM Tris–HCl (pH 7.4) containing 120 mM NaCl at room temperature to remove endogenous dopamine. The sections were incubated for 1 h at room temperature in 50 mM HEPES/NaOH buffer (pH 7.5) containing 1 mM EDTA, 100 nM PPAP (to block sigma sites) and 1 nM [3H]7-OH-DPAT (specific activity = 152 Ci/mmol), a concentration similar to the Kd previously reported (Lévesque et al., 1992; Herroelen et al., 1994). Non-specific binding (10% of total binding) was in the presence of 10 μM dopamine. Sections were washed for a total of 9 min in 50 mM Tris–HCl containing 120 mM NaCl (pH 7.4) at 4°C.

Development and quantitative imaging of autoradiographs

The sections, together with autoradiographic standards (Amersham), were opposed to Hyperfilm 3H (Amersham). Exposure times were 3 weeks for [3H]mazindol, 1 month for [3H]raclopride, 7 weeks for [3H]SCH23390 and 3 months for [3H]7-OH-DPAT. The resulting films were developed then scanned and quantified using a Lynx densitometry system (Applied Imaging, Sunderland, UK). Average values, calculated as fmol ligand bound per mg tissue equivalent, were measured by comparison with tritium standards (Amersham) at each coronal level in the areas of the dorsal and ventral caudate and putamen and in the nucleus accumbens.

Dopamine and homovanillic acid measurements

Dopamine and homovanillic acid (HVA) were measured in tissue homogenates of both caudate and putamen, dissected from the frozen tissue blocks, using high-performance liquid chromatography with electrochemical detection (Marshall et al., 1994).

Statistical analysis

Statistical analysis was performed using Minitab for Windows, version 12. Comparison of binding between disease groups in the same striatal area at the same level was by one-way analysis of variance (ANOVA). Two-way ANOVA was used to compare binding in different areas at separate levels, and if no interaction between these factors was observed comparisons were then made by one-way ANOVA followed by Fisher's pairwise comparisons post hoc test, with the significance level set at P ≤ 0.05, as suggested by Kinnear and Gray (Kinnear and Gray, 1994). The caudate and putamen were compared at similar levels and between pairs of levels by two-tailed t test. Assessment of correlation between neurochemical measures and demographic variables was calculated using Pearson's product moment correlation coefficient. Comparison of gender measurements was by two-tailed t test.

Results

The dopamine uptake site and D1, D2 and D3 receptor values (fmol/mg tissue; mean ± standard deviation) are shown in the dorsal and ventral caudate and putamen at the coronal levels measured in controls and in DLB, Parkinson's disease and Alzheimer's disease patients in Table 3. The graphs (Figs 2, 3, 4 and 6) illustrate the rostrocaudal distributions of the mean values of these parameters. Dopamine and HVA concentrations (pmol per mg protein; mean ± standard deviation) and the HVA : dopamine ratio are displayed in the rostrocaudal dimension (Fig. 7) in the caudate and putamen from controls and DLB, Parkinson's disease and Alzheimer's disease patients.

View this table:
Table 3

Dopamine uptake site and D1, D2 and D3 receptor binding values (fmol per mg tissue, mean ± standard deviation) in dorsal and ventral caudate and putamen at each coronal level in controls, DLB, Parkinson's disease and Alzheimer's disease

Dopamine uptake siteCoronal levelDorsal caudateVentral caudateDorsal putamenVentral putamenNucleus accumbens
Significant differences between disease groups by ANOVA followed by Fisher's post hoc comparison of means with significance set at P < 0.05 are indicated. AD = Alzheimer's disease; PD = Parkinson's disease. *Significantly lower than controls; significantly lower than control and Alzheimer's disease cases; significantly lower than Alzheimer's disease cases; §significantly lower than control, DLB and Alzheimer's disease cases; significantly lower than control and DLB cases; #significantly higher than control, DLB and Alzheimer's disease cases; **significantly higher than control and DLB cases; ††significantly higher than control and Alzheimer's disease cases; ‡‡, ‡‡‡significantly lower than controls at P < 0.05 and P < 0.01, respectively (t test).
Controls10 62.1 ± 13.1 64.1± 29.748.7 ± 16.4 73.2 ± 38.7
11 88.5 ± 27.1 89.3 ± 21.679.9 ± 21.5 78.1 ± 22.967.1 ± 17.1
12 90.2 ± 27.6109.1 ± 18.790.0 ± 17.3 91.7 ± 13.1
13100.1 ± 25.1121.7 ± 24.494.2 ± 26.3 97.0 ± 26.5
14101.8 ± 41.6112.8 ± 51.682.7 ± 28.9 94.4 ± 28.2
DLB11 81.1 ± 41.4 96.8 ± 43.259.3 ± 26.5 62.8 ± 31.865.1 ± 36.7
12 51.6 ± 21.2‡‡ 87.4 ± 33.138.3 ± 16.3‡‡‡ 53.6 ± 32.5
13 57.4 ± 38.6 83.2 ± 42.738.4 ± 28.8 45.1 ± 26.9
PD10 39.1 ± 28.9‡‡ 53.2 ± 35.930.1 ± 25.6 39.3 ± 27.8
11 54.7 ± 46.4 78.2 ± 53.535.5 ± 28.4§ 42.9 ± 25.156.8 ± 39.8
12 28.1 ± 22.1‡‡ 56.5 ± 21.0‡‡18.8 ± 25.3‡‡ 27.2 ± 22.2‡‡
13 37.2 ± 18.3 86.4 ± 27.122.6 ± 17.3 25.1 ± 21.2
AD 9 76.0 ± 29.4 83.8 ± 34.2
11 76.2 ± 37.5 82.2 ± 37.468.5 ± 34.5 76.4 ± 38.664.0 ± 32.4
13105.1 ± 36.6122.2 ± 42.395.9 ± 39.8108.5 ± 38.5
D1Controls10 33.4 ± 7.1 32.6 ± 9.137.4 ± 7.1 32.8 ± 9.2
11 29.9 ± 5.7 29.2 ± 8.236.0 ± 7.2 29.4 ± 7.134.7 ± 9.7
13 36.4 ± 10.7 33.8 ± 9.131.2 ± 9.0 25.8 ± 6.8
14 37.4 ± 2.5 35.0 ± 2.428.6 ± 9.1 26.9 ± 12.0
DLB10 29.2 ± 11.7 33.2 ± 9.834.2 ± 10.5 36.5 ± 12.5
11 30.9 ± 8.7 30.1 ± 7.832.6 ± 7.2 31.7 ± 7.336.8 ± 8.2
13 29.0 ± 9.0 28.4 ± 10.631.6 ± 11.1 27.7 ± 10.6
PD10 32.6 ± 9.8 33.4 ± 10.632.3 ± 7.1 30.8 ± 17.4
11 30.3 ± 7.1 29.9 ± 9.036.4 ± 6.9 32.1 ± 9.332.0 ± 10.8
12 34.1 ± 17.6 32.2 ± 17.433.7 ± 18.2 30.6 ± 14.5
13 45.8 ± 9.2# 42.1 ± 1.5#40.5 ± 11.7# 38.0 ± 7.7#
AD10 35.2 ± 8.0 37.3 ± 12.025.9 ± 16.4 20.7 ± 21.0
11 31.6 ± 14.4 26.0 ± 14.636.0 ± 14.5 28.0 ± 13.332.0 ± 20.2
13 29.0 ± 12.4 27.2 ± 14.527.4 ± 10.2 22.8 ± 12.0
D2Controls10 19.4 ± 7.3 19.8 ± 8.322.3 ± 9.9 21.2 ± 9.9
11 19.3 ± 10.4 19.1 ± 12.123.1 ± 12.0 22.0 ± 11.824.8 ± 14.6
13 29.2 ± 9.1 24.5 ± 8.227.6 ± 10.4 25.7 ± 9.8
15 31.8 ± 14.1 32.1 ± 13.727.4 ± 11.7 28.0 ± 11.1
DLB11 18.1 ± 4.9 17.0 ± 5.620.5 ± 6.3 20.6 ± 7.021.5 ± 7.9
13 16.9 ± 11.7§ 18.0 ± 8.919.5 ± 12.9 24.0 ± 16.2
14 21.1 ± 13.0 18.4 ± 13.623.6 ± 14.0 22.8 ± 14.0
PD10 34.0 ± 4.6 30.7 ± 7.939.1 ± 4.3 36.9 ± 8.5
11 34.7 ± 7.5# 28.5 ± 5.4#41.4 ± 13.3# 36.1 ± 10.1#37.2 ± 13.2#
12 33.7 35.138.5 40.2
13 32.8 ± 10.4 28.2 ± 6.135.3 ± 11.1** 35.6 ± 9.5**
AD11 24.4 ± 11.4 21.3 ± 10.824.9 ± 12.5 23.9 ± 11.720.3 ± 13.1
14 31.4 ± 13.4 28.0 ± 12.729.1 ± 13.2 31.3 ± 15.7
D3Controls10 5.8 9.8
11 4.5 ± 1.0 6.0 ± 1.6 6.9 ± 1.2 10.9 ± 1.716.3 ± 2.7
12 4.3 ± 1.2 5.2 ± 0.9 6.7 ± 1.2 11.3 ± 0.9
14 3.4 ± 0.5 3.9 ± 0.5 4.1 ± 1.04 5.0 ± 1.4
DLB10 4.4 ± 1.7 9.6 ± 2.6 6.2 ± 1.9 7.3 ± 2.0
11 4.9 ± 2.3 7.7 ± 3.1 8.2 ± 3.4 12.0 ± 3.016.4 ± 3.8
13 4.6 ± 1.9 5.6 ± 2.0 8.2 ± 2.6 11.8 ± 2.7
14 3.5 ± 1.1 4.4 ± 1.2 5.1 ± 3.6 5.5 ± 3.9
PD10 4.8 ± 2.0 9.7 ± 3.7 6.7 ± 0.8 10.3 ± 2.1
11 4.2 ± 0.7 6.3 ± 0.7 6.3 ± 0.4 11.0 ± 0.813.0 ± 2.3
13 3.4 ± 1.5 4.6 ± 1.6 4.4 ± 2.4 6.7 ± 3.4
AD10 5.5 ± 3.1 8.9 ± 5.9 9.2 ± 0.6 12.3 ± 2.9
11 6.5 ± 2.0†† 8.9 ± 2.8 9.3 ± 3.5†† 13.5 ± 5.3††15.6 ± 5.4
13 5.2 ± 1.8 5.0 ± 1.1 6.8 ± 2.2 8.4 ± 1.8
Fig. 2

[3H]Mazindol binding to dopamine uptake sites (mean of n as indicated, fmol/mg tissue at each point) in the dorsal and ventral caudate and putamen in controls, DLB patients, Parkinson's disease (PD) patients and Alzheimer's disease (AD) patients. In Parkinson's disease, coronal level 13 represents the mean value of three cases at coronal level 13 and one case at coronal level 14. ▪, dorsal caudate; ▴, ventral caudate; ⧫, dorsal putamen; ▾, ventral putamen; •, nucleus accumbens.

Fig. 3

[3H]SCH23390 binding to dopamine D1 receptors (mean of n as indicated, fmol/mg tissue) in the dorsal and ventral caudate and putamen in controls, DLB patients, Parkinson's disease (PD) patients and Alzheimer's disease (AD) patients.

Fig. 4

[3H]Raclopride binding to dopamine D2 receptors (mean of n as indicated, fmol/mg tissue) in the dorsal and ventral caudate and putamen in controls, DLB patients, Parkinson's disease (PD) patients and Alzheimer's disease (AD) patients.

Dopamine uptake site (Fig. 2)

In control cases, binding to the dopamine uptake site was between 49 and 97 fmol/mg tissue in the putamen, and 62–122 fmol/mg in the caudate, with an increase in binding along the rostrocaudal axis in the caudate (especially ventrally), and higher binding in the ventral caudate than in the dorsal caudate and the putamen at coronal level 12 and coronal level 13, as previously described (Piggott et al., 1999). Binding values were similar to those in previous reports (Singer et al., 1991; Hurd and Herkenham, 1993) at equivalent ligand concentrations.

Comparison of disease groups (Fig. 2) in the dorsal putamen at coronal level 11 [F(3,75) = 6.17, P = 0.001] showed dopamine uptake site binding in Parkinson's disease cases to be reduced compared with Alzheimer's disease, DLB and controls, and binding in DLB cases to be slightly reduced compared with controls, with no significant differences between diseases in the coronal level 11 caudate. In the ventral putamen at coronal level 11, binding in Parkinson's disease cases was lower than in controls and Alzheimer's disease cases [F(3,79) = 3.39, P = 0.011]. There were no significant differences between disease groups in the nucleus accumbens. At coronal level 13, the Parkinson's disease and DLB groups had similarly low levels of dopamine uptake sites compared with control and Alzheimer's disease cases in the dorsal caudate [F(3,45) = 10.98, P < 0.001] and ventral caudate [F(3,42) = 4.13, P = 0.012), dorsal putamen [F(3,48) = 18.36, P < 0.001] and in the ventral putamen [F(3,46) = 20.99, P < 0.001].

In DLB at coronal level 12, dopamine uptake site binding was significantly reduced compared with controls in the dorsal caudate (t test, P = 0.026) and in the dorsal putamen (P = 0.001). There was no increasing rostrocaudal gradient of dopamine uptake site binding in DLB cases as in the controls; in the dorsal caudate there was a trend to a decreasing rostrocaudal gradient, which approached significance (correlation coefficient r = –0.35, 0.1 > P > 0.05).

In Parkinson's disease, dopamine uptake site binding was reduced at all levels, being significantly lower than in controls in the dorsal caudate at coronal level 10 (t test, P = 0.03) and in the caudate and putamen at coronal level 12 (P = 0.019 and 0.02, respectively). The greatest losses of dopamine uptake sites in Parkinson's disease were at caudal coronal levels (75% reduction in the putamen at coronal level 13) and the least reduction was in the ventral caudate; there was no significant gradient of binding in the rostrocaudal dimension in any area.

In Alzheimer's disease there was no significant deviation from control dopamine uptake site binding at any level, with an increasing rostrocaudal gradient which was almost significant in the ventral caudate (r = 0.375, P ≈ 0.05).

There were no significant age or post-mortem delay differences between disease groups [F(3,72) = 1.5, P = 0.22 and F(3,72) = 0.84, P = 0.48, respectively]. Significant changes with demographic variables within groups were a decline in dopamine uptake site binding in DLB cases with age, which was most significant in the ventral caudate at posterior levels (r = –0.56, n = 11, P < 0.05). In Parkinson's disease there was a tendency to a reduced concentration of dopamine uptake sites with disease duration, which reached significance in the ventral caudate at posterior levels (r = –0.77, n = 7, P < 0.02). In the Alzheimer's disease group there was a significant negative correlation with increasing age in all Alzheimer's disease cases, especially in the caudate nucleus and also when the analysis was restricted to coronal level 11 (dorsal caudate, r = –0.758, n = 10, P < 0.001; ventral caudate, r = –0.697, n = 10, P < 0.01). There was also a significant decline in dopamine uptake sites with increasing age of onset of Alzheimer's disease, particularly in the caudate (r = –0.53, n = 14, P < 0.05, coronal level 11 only).

D1 receptor (Fig. 3)

In control cases, D1 receptor binding was ~30–35 fmol/mg tissue, comparable to other published values (Cortes et al., 1989; De Keyser et al., 1990); it was higher in the caudate compared with the putamen at posterior levels and displayed a rostrocaudal decline in the putamen which was significant dorsally, as previously reported (Piggott et al., 1999). The only difference in D1 binding in disease groups was between Parkinson's disease and all other groups at coronal level 13. A two-way ANOVA at coronal level 13 revealed that there was no significant difference between areas but a significant difference between groups [F(3,96) = 9.69, P < 0.001], with no group × area interaction. Subsequent one-way ANOVA showed a significant group difference [F(3,227) = 14.92, P < 0.001] and post hoc analysis (with significance level set at P < 0.05) showed the Parkinson's disease group to have higher D1 binding than all other groups at coronal level 13. This was particularly so in the ventral putamen (t test, P = 0.014 compared with controls).

With respect to demographic variables, in the control group there was a significant negative correlation of D1 binding with age in the putamen at coronal level 11 (r = –0.83, n = 8, P < 0.01). In DLB cases there was a significant negative correlation with age at coronal levels 10 and 11 in the ventral caudate and putamen (r = –0.5, n = 15, P < 0.05) and in the dorsomedial caudate (r = –0.62, n = 15, P < 0.01). There was a significant decline in D1 binding with age of onset of DLB in the caudate (r = –0.53, n = 14, P < 0.05). D1 binding declined with disease duration in Parkinson's disease; the decline was significant in the posterior caudate (r = –0.75, n = 6, P < 0.05).

D2 receptor (Fig. 4)

D2 binding in controls of ~20–30 fmol/mg tissue was in a similar range to other published values, allowing for different ligand concentrations (Hall et al., 1994; Dean et al., 1997). D2 receptor binding in controls was relatively uniform in the dorsoventral and lateromedial dimensions, with more variation in the increasing gradient rostrocaudally (Fig. 4). At coronal levels 13 and 15, D2 binding was higher than at coronal levels 10 and 11, with no difference between the caudate and putamen except at coronal level 11, where D2 binding in the putamen was higher, as previously reported (Piggott et al., 1999).

D2 binding was significantly different between disease groups [F(3,744) = 49.58, P < 0.001] and between areas [F(9,744) = 3.05, P = 0.001], with no significant interaction term. At coronal level 11, D2 binding was elevated in Parkinson's disease above control, DLB and Alzheimer's disease values in the dorsal caudate [F(3,80) = 10.93, P < 0.001], ventral caudate [F(3,88) = 4.99, P= 0.003), dorsal putamen [F(3,84) = 11.01, P < 0.001], ventral putamen [F(3,90) = 6.97, P < 0.001] and nucleus accumbens [F(3,72) = 4.98, P = 0.003); in the dorsal caudate at coronal level 11, D2 binding was significantly lower in DLB than in Alzheimer's disease [F(3,80) = 10.93, P < 0.001]. At caudal levels (coronal levels 12 and 13), D2 binding in Parkinson's disease was elevated above control and DLB values in the dorsal putamen [F(3,68) = 3.62, P = 0.017] and the ventral putamen [F(3,66) = 3.15, P = 0.03]. In addition, in DLB patients the level of D2 binding in the dorsal caudate at coronal levels 13 and 14 was lower than in controls and in Alzheimer's disease and Parkinson's disease patients [F(3,62) = 4.19, P = 0.009]. Figure 5 shows that D2 binding at coronal level 11 was lower in DLB and higher in Parkinson's disease than in controls. There was no difference in D2 binding between controls and Alzheimer's disease at rostral or caudal levels, and in Alzheimer's disease a rostrocaudal and dorsoventral distribution similar to that of controls was found. There was much less rostrocaudal variation in DLB and Parkinson's disease.

Fig. 5

Autoradiographs of [3H]raclopride binding to D2 receptors in the striatum in a control case, a DLB case and a Parkinson's disease case at coronal level 11, showing higher binding in Parkinson's disease and lower binding in DLB patients compared with the control group.

Significant changes with demographic variables included an effect of gender in DLB, with lower D2 binding in males in the ventral caudate and ventral putamen (t test, P = 0.01 and 0.022, respectively). There was a trend for decreased D2 binding with longer duration of DLB in all areas (r = –0.38, n = 23, 0.1 > P > 0.05 in the dorsal caudate). A reduction in D2 binding with age in Parkinson's disease was most significant in the rostral ventral putamen (r = –0.58, n = 10, P < 0.05). There was a trend in all striatal areas to a lower level of D2 binding with older age of onset in Parkinson's disease, which reached significance in the ventral putamen (r = –0.56, n = 10, P < 0.05). Disease duration in Alzheimer's disease was associated with a significant increase in D2 receptors in the rostral putamen (r = 0.64, n = 10, P < 0.05).

D3 receptors (Fig. 6)

Fig. 6

[3H]7-OH-DPAT binding to dopamine D3 receptors (mean of n as indicated, fmol/mg tissue) in the dorsal and ventral caudate and putamen in controls, DLB patients, Parkinson's disease (PD) patients and Alzheimer's disease (AD) patients.

D3 binding was 5–20 fmol/mg tissue, similar to or higher than that reported in other studies, none of which used a method identical to that presently described (Herroelen et al., 1994; Gurevich et al., 1997). D3 binding was distributed unevenly, with the highest density binding ventrally, as previously described (Piggott et al., 1999). Analysis of all groups at coronal level 11 showed significantly different D3 binding between disease groups [F(3,95) = 4.78, P = 0.004] and in striatal areas [F(4,95) = 53.38, P < 0.001], but with no significant interaction. There was no significant difference in D3 binding in DLB compared with controls in any area. D3 binding in the dorsal caudate and putamen in Alzheimer's disease at coronal level 11 was elevated compared with controls and Parkinson's disease cases [F(3,65) = 4.42, P = 0.007]. Comparing the nucleus accumbens only, there was no significant difference between groups. Comparison of groups at coronal level 13 showed a significant group [F(3,58) = 6.41, P = 0.001] and area difference [F(3,58) = 21.0, P < 0.001] and no significant interaction term. At coronal level 13, D3 binding in the caudate and putamen combined showed significant disease differences [F(3,70) = 2.81, P = 0.046], post hoc analysis (P < 0.05) showing significantly reduced D3 binding in Parkinson's disease compared with controls and DLB cases, particularly in the ventral putamen [F(3,13) = 4.04, P = 0.03].

Significant changes with demographic variables were that in DLB with increasing disease duration in the posterior caudate, D3 binding declined (dorsal caudate, r = –0.72, n = 7, P < 0.05; ventral caudate, r = –0.78, n = 7, P < 0.02). In Parkinson's disease there was a tendency for D3 binding to decline with age in the anterior dorsal putamen (r = –0.7, n = 8, 0.1 > P > 0.05).

Dopamine concentration and metabolism (Fig. 7)

Fig. 7

Rostrocaudal distribution of dopamine and HVA (mean ± standard deviation of n as indicated, pmol/mg protein), and HVA : dopamine (DA) ratio in controls, DLB patients, Parkinson's disease (PD) patients and Alzheimer's disease (AD) patients. Significant differences between disease groups by ANOVA and Fisher's post hoc comparison of means with significance set at P < 0.05 are indicated by Embedded Image, significantly lower than controls and Alzheimer's disease cases; *, significantly lower than controls; Embedded Image, significantly higher than controls, DLB and Alzheimer's disease cases. ▪, Caudate; □, putamen.

The dopamine concentration in control cases was greatest at coronal levels 11–13 in the caudate and putamen while the concentration of HVA displayed no gradient in the caudate but peaked in the putamen at coronal level 12. The HVA : dopamine ratio in the posterior striatum was higher in the putamen than in the caudate, as previously reported (Piggott et al., 1999).

Comparing disease groups, dopamine was significantly reduced in the caudate in DLB and Parkinson's disease relative to controls and Alzheimer's disease cases [F(3,91) = 12.02, P < 0.001]. The loss of dopamine in the caudate in DLB was uniform along the rostrocaudal axis (42–46%), but in Parkinson's disease the dopamine reduction in the posterior caudate was much more extensive (86%, compared with 60% rostrally). In the putamen, the dopamine concentration was significantly reduced in both DLB and Parkinson's disease compared with controls and Alzheimer's disease cases [F(3,80) = 18.38, P < 0.001]. In the putamen in both DLB and Parkinson's disease, dopamine loss was more extensive caudally. In the putamen in DLB, dopamine was reduced by 35% rostrally and 72% caudally, while in the putamen in Parkinson's disease dopamine was reduced by 79% rostrally and 90% caudally.

In DLB, HVA was extensively reduced, to an extent similar to that in Parkinson's disease, in both the caudate and the putamen [group comparisons: F(3,90) = 5.89, P = 0.001 and F(3,80) = 5.31, P = 0.002, respectively]. HVA concentrations in Parkinson's disease and DLB were significantly lower than in controls and Alzheimer's disease cases in the putamen, and significantly lower than in controls in the caudate. The HVA : dopamine ratio was not significantly different in DLB compared with controls, although it tended to be raised in the caudal putamen, while the HVA : dopamine ratio in Parkinson's disease was highly elevated in all areas, especially in the putamen, where it was 6–13 times higher than in controls [F(3,80) = 8.82, P < 0.001].

In Alzheimer's disease the dopamine content in the putamen was significantly higher than in controls [F(3,80) = 18.38, P < 0.001] both rostrally and caudally, while the HVA concentration was slightly less than in controls in the caudate [F(3,90) = 5.89, P = 0.001]. The HVA : dopamine ratios in Alzheimer's disease were not significantly different from those in controls in the caudate and putamen, but the standard deviation of the ratio was larger in the caudal putamen.

Significant demographic influences were that in DLB there was a loss of dopamine with age in the caudate (r = –0.46, n = 25, P < 0.02), longer disease duration tended to result in an increased HVA : dopamine ratio in the putamen in DLB, and increased age of onset of DLB was positively correlated with HVA concentration in the caudate (r = 0.51, n = 15, P < 0.05) and putamen (r = 0.8, n = 15, P < 0.001) at coronal levels 11–13. Age in Parkinson's disease was inversely correlated with dopamine and HVA concentrations in the caudate (r = –0.51, n =18, P < 0.01 and r = –0.7, n = 12, P < 0.01, respectively, at coronal levels 11–13). There was also a significant decline in HVA concentration with increasing Parkinson's disease duration in the caudate and putamen (r = –0.69, n = 12, P < 0.01 and r = –0.73, n = 12, P < 0.01, respectively, at coronal levels 11–13). In Alzheimer's disease there was a significant decline in dopamine concentration with increasing age in the caudate and putamen (r = –0.73, n = 15, P < 0.01 and r = –0.52, n = 13, P < 0.05, respectively, at coronal levels 11–13). In Alzheimer's disease, dopamine and HVA concentrations declined with increasing disease duration in the putamen (r = –0.78, n = 8, P < 0.01 and r = –0.62, n = 8, 0.1 > P > 0.05, respectively, at coronal level 11).

Discussion

The principal findings of the present study of clinically and pathologically assessed cases of DLB, Parkinson's disease and Alzheimer's disease were of a differential loss of dopamine uptake sites in DLB, which was restricted to levels of the striatum caudal to the anterior commissure, with reduction at rostral levels limited to the dorsal putamen. D2 receptor expression was also distinctive in DLB, being much lower than in Parkinson's disease and reduced caudally compared with controls and Alzheimer's disease. These differences in the neurochemical pathology of DLB may explain some clinical features of the disease, including the spontaneous extrapyramidal symptoms which respond poorly to l-dopa medication, and marked neuroleptic sensitivity. The distinct neurochemical profile may be of diagnostic value using in vivo imaging.

Dopamine uptake sites

In DLB, reduced dopamine uptake sites were restricted to levels posterior to the anterior commissure, affecting the putamen and dorsal caudate rather than the ventral caudate. This may be similar to the pattern of loss in very early Parkinson's disease (Wilson et al., 1996; Booij et al., 1997). In contrast, dopamine uptake sites were found to be reduced at all coronal levels of the putamen in Parkinson's disease and in all of the caudate posterior to the anterior commissure, consistent with many previous reports both in vitro (Guttman, 1987; Wilson et al., 1996) and in vivo (Antonini et al., 1995; Brucke et al., 1997; Tatsch et al., 1997; Tissingh et al., 1998). In early drug-naïve Parkinson's disease, dopamine uptake site losses measured by single photon emission computerized tomography (SPECT) imaging are bilateral even in unilateral disease (Tissingh et al., 1998). Reductions in markers of dopaminergic terminals in Parkinson's disease are well reported, with greater loss in the caudal putamen initially and later reductions spreading rostrally and involving the caudate (Kish et al., 1988; Murray et al., 1995; Wilson et al., 1996), similar to the pattern seen in the MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) monkey (Moratalla et al., 1992). The threshold of reduction in dopamine uptake sites for the appearance of mild parkinsonism may be quite low—below 20% in the putamen overall and ~50% in the posterior putamen, as measured in vivo (Eising et al., 1997; Guttman et al., 1997)—suggesting that the loss of dopamine uptake sites in DLB is sufficient to evoke extrapyramidal symptoms.

In Alzheimer's disease, mazindol binding to the dopamine uptake site was not reduced at all, in agreement with previous reports showing little reduction in vitro (Murray et al., 1995) or in vivo (Tyrrell et al., 1990; Donnemiller et al., 1997), but in contrast to reports of reduced dopamine uptake sites (Rinne et al., 1998; Sahlberg et al., 1998) that were correlated with the severity of extrapyramidal symptoms. In the present study, older age of onset of Alzheimer's disease was associated with lower dopamine uptake site density, perhaps reflecting a decline in binding with age or more extrapyramidal symptoms with older age of Alzheimer's disease onset. The pathological correlates of extrapyramidal features in Alzheimer's disease may also be in other parts of the basal ganglia–thalamocortical circuit or, more controversially, in neurofibrillary tangles in the substantia nigra (Liu et al., 1997).

The decline in dopamine uptake sites with disease duration in Parkinson's disease and Alzheimer's disease was not seen significantly in DLB, which may parallel the observation that extrapyramidal symptoms progress only slowly during the course of the disease (Ballard et al., 1998b). Imaging of dopamine uptake has been shown to correlate with disease duration and increasing movement disability (Tatsch et al., 1997). Dopamine uptake sites tended to be reduced with age, especially in the caudate in DLB and Alzheimer's disease, consistent with previous reports over more extended age ranges (De Keyser et al., 1990; Rinne et al., 1998; Volkow et al., 1998). A proportion of the cases received l-dopa therapy or neuroleptic treatment, and the possible effects of these must be considered. Longer duration of l-dopa therapy may increase the rate of loss of dopamine uptake sites as substantia nigra neurons are lost, due to the metabolic cost of processing l-dopa (Kopin, 1993; Watts, 1997), which may contribute to the reduction in uptake site binding in Parkinson's disease with increased disease duration. Conversely, however, l-dopa treatment of MPTP monkeys has been reported to increase mazindol binding (Rioux et al., 1997). Neuroleptic treatment is reported not to affect dopamine uptake sites in long-term therapy in schizophrenia (Czudek and Reynolds, 1989) and in rats, but may tend to reduce presynaptic markers in older groups of DLB and Alzheimer's disease patients, particularly in the caudate (Perry et al., 1998).

D2 receptor binding

D2 receptor binding was elevated in Parkinson's disease, especially rostrally, while in DLB values were lower than in Parkinson's disease at all coronal levels, and lower than in controls caudally. In Alzheimer's disease, D2 receptor densities in the striatum were unchanged. Consequently, while D2 binding was elevated in Parkinson's disease at all levels, the elevation was relatively greater rostrally; conversely, in DLB the deficit in D2 binding was relatively greater caudally, compared with controls. Several reports have shown raised D2 binding in early Parkinson's disease, as measured by SPET or PET (Giobbe et al., 1993; Antonini et al., 1994, 1997a; Reiche et al., 1995), and it has been suggested that in vivo imaging may help to differentiate idiopathic Parkinson's disease from other parkinsonism-plus syndromes like multiple system atrophy and progressive supranuclear palsy (Tissingh et al., 1997), in which, as in DLB, reduced D2 density occurs, and l-dopa therapy is less effective. In addition, it has been reported that patients with multiple system atrophy are at risk of developing a neuroleptic malignant syndrome-like condition (Konagaya et al., 1997). How early in the Parkinson's disease process D2 elevation occurs or whether a threshold of dopamine loss needs to be reached is not clear. Antonini and colleagues (Antonini et al., 1995) found in a PET study that D2 elevation correlated with reduced [18F]fluoro-l-dopa uptake in early Parkinson's disease, and they did not have any patients so early in disease that they did not show D2 upregulation. Rinne and colleagues (Rinne et al., 1993) found D2 upregulation to be present at the time of diagnosis. Striatal dopamine concentrations are required to be significantly reduced before clinical symptoms become apparent, and the tendency for D2 receptors to upregulate as a compensatory measure will mask clinical symptoms in the earliest phase. l-Dopa therapy and longer Parkinson's disease duration (over years rather than months) have been found to lower previously elevated D2 receptors in vivo (Antonini et al., 1997b), although levels have been reported to remain high for >36 months with dopaminergic therapy (Hierholzer et al., 1998). In DLB, D2 binding was reduced by 17% in the caudal putamen compared with controls, in an area where dopamine uptake sites were reduced to 52% of the control value, and therefore dopamine terminals are probably depleted sufficiently for D2 upregulation to be expected, as in Parkinson's disease. A recent report has suggested that D2 upregulation is less persistent in Parkinson's disease that presents bilaterally (Wenning et al., 1998), and it is interesting that extrapyramidal symptoms in DLB have been noted to be more symmetrical at presentation than in Parkinson's disease (Gnanalingham et al., 1997). The low levels of D2 receptors in DLB are likely to be responsible for the adverse neuroleptic reactions seen in these patients, and may also explain the reduced benefit from l-dopa compared with Parkinson's disease. Lower numbers of D2 receptors are more likely to be completely blocked, producing parkinsonian signs, by neuroleptic administration, and DLB patients can show severe sensitivity to typical as well as recently developed neuroleptics, which have reduced propensity to induce extrapyramidal symptoms in other patient groups (McKeith et al., 1995).

In Alzheimer's disease there was no change in D2 receptor binding or distribution compared with controls, consistent with the normal level of dopamine uptake site binding in the present report. This is in spite of neuroleptic medication in the majority of patients, which has been reported to increase D2 density in Alzheimer's disease and also in schizophrenia and in animal models (Seeman et al., 1987). Alzheimer's disease patients seldom show sensitivity to neuroleptics (McKeith et al., 1992).

D2 binding was found to be reduced with age in Parkinson's disease, with non-significant trends in other groups, consistent with the reported loss of D2 binding with increasing age previously reported in normal individuals over a wider age range in vivo (Volkow et al., 1998), but in contrast to the lack of a decline with age in vitro (De Keyser et al., 1990). Loss of D2 with age in normal cases may be slight and detectable only over a wide age range. Only in the DLB group was a gender difference apparent, males having lower D2 binding in the ventral striatum. This may relate to greater severity of extrapyramidal symptoms or to susceptibility to adverse neuroleptic reaction in males, and although this was not borne out in the present study it will be investigated in larger studies in DLB in future.

D1 receptor binding

There was no difference between disease groups in the density or pattern of binding, except for an increase in D1 binding in the caudal striatum in Parkinson's disease. Previously, D1 receptors have been found to be unaltered in Parkinson's disease (Mizukawa et al., 1993; Piggott and Marshall, 1996), although there are some reports of a raised level of D1 receptors in Parkinson's disease untreated with l-dopa (Seeman et al., 1987; Pearce et al., 1990), and an elevation has been reported in one case of postencephalitic parkinsonism (Piggott and Marshall, 1996) and in narcolepsy (Aldrich et al., 1992). D1 receptors are also elevated in MPTP experimental models of Parkinson's disease in cats and primates (Gnanalingham et al., 1993; Graham et al., 1993; Frohna et al., 1995; Rioux et al., 1997). The variation in reports of D1 binding density in Parkinson's disease may be due to the different coronal levels studied, since in the present work the elevation was only apparent caudally. l-Dopa therapy has been reported to increase D1 receptors in parkinsonian monkeys in the caudal striatum (Graham et al., 1993; Rioux et al., 1997), and raised D1 density might relate to the propensity for l-dopa medication to produce dyskinesias in long-term treated Parkinson's disease, in postencephalitic parkinsonism and in MPTP models (Graham et al., 1993). Conversely, l-dopa therapy was reported to tend to decrease D1 binding in human in vivo imaging of Parkinson's disease (Turjanski et al., 1997), which is consistent with the present finding of declining D1 receptors with disease duration in Parkinson's disease.

The present study confirms previous findings of no change in D1 binding in Alzheimer's disease (Cross et al., 1984; Seeman et al., 1987) or DLB (Piggott and Marshall, 1996). Reports of declining D1 receptor level with ageing have been conflicting. In the present study, D1 binding declined with age in the control and DLB groups, consistent with the decline previously reported (Rinne et al., 1990) but in contrast to a report of no decline in D1 receptors with age (De Keyser et al., 1990). It is probable that the age-related decline in D1 receptors is slight and reliably detected only over a long age range.

D3 receptor binding

The pattern of D3 binding is localized to limbic areas, particularly the nucleus accumbens, rather than motor areas of the striatum, with significant alterations in disease groups restricted to slightly elevated D3 binding in Alzheimer's disease in the dorsal striatum and somewhat reduced D3 binding in Parkinson's disease cases, with no deviation from the pattern of expression in DLB from controls. D3 receptors, while being structurally related to the D2-like dopamine receptor group, are localized to striosomes, like D1 receptors (Murray et al., 1994), and may function in an inverse manner to D2 receptors (Levant, 1997). A previous report has shown no change in D3 binding in Parkinson's disease (Hurley et al., 1996), while a recent investigation found reduced D3 receptors, particularly in the ventral striatum, in Parkinson's disease of more than 10 years' duration (Ryoo et al., 1998). This latter finding is consistent with the lower D3 binding in Parkinson's disease in the present report, where there was an apparent 20% reduction in binding (although not statistically significant) in the nucleus accumbens, as well as the significant reduction in the caudal striatum. 6-Hydroxydopamine lesion decreased D3 receptors in rats (Lévesque et al., 1995) and in MPTP monkeys (Morisette et al., 1998). The present study compares D2 and D3 binding in the same cohort of patients (where only four cases have Parkinson's disease duration of more than 10 years) showing no upregulation of D3 receptors where D2 upregulation is demonstrated. Dopamine replacement by a dopamine agonist with D1 activity (but not by a D2-like agonist) has been reported to reverse experimentally reduced D3 receptors in the MPTP monkey (Morissette et al., 1998). l-Dopa therapy, with presumed D1, D2 and D3 activity, might be expected to tend to reverse D3 receptor downregulation. Many typical neuroleptics are also D3 antagonists, and D3 binding or mRNA (in the olfactory tubercle) has been reported to be elevated in rats with neuroleptic administration (Wang et al., 1996), but also not to have any effect (Lévesque et al., 1995). The present study found slightly elevated D3 binding in Alzheimer's disease in the dorsal striatum.

Dopamine and homovanillic acid concentration

In the rostral striatum in DLB, greater loss occurred in the caudate compared with the putamen, and in the caudate the loss was uniform rostrocaudally. This implies different vulnerability of cell groups in the substantia nigra, and also suggests that there may be an influence of dopamine loss on non-motor characteristics of DLB, such as depression and reduced cognitive abilities (Gnanalingham et al., 1997). It has been suggested that dopamine reductions of ~50–75% are required for the manifestation of extrapyramidal symptoms in Parkinson's disease (Bernheimer et al., 1973; Hornykiewicz, 1973; Kish et al., 1985), which is the degree of loss just reached in the DLB cases. Other reports of dopamine concentration in DLB have shown reductions at least as great as here (Langlais et al., 1993; Ohara et al., 1998). This should be sufficient explanation for the mild extrapyramidal symptoms in DLB. There would also seem to be sufficient reduction in dopamine concentration in the caudal putamen to have the potential to induce D2 upregulation in DLB as observed in Parkinson's disease. The HVA : dopamine ratio was increased greatly in Parkinson's disease, but in DLB the tendency to a higher ratio was not significantly raised over controls, even in the caudal putamen. DLB patients do not, therefore, appear to have the same capacity for presynaptic compensation for reduced dopamine as the remaining substantia nigra neurons do in Parkinson's disease and also in progressive supranuclear palsy (Bokobza et al., 1984). This, combined with low level of postsynaptic D2 receptors, will probably result in greater motor deficits in DLB for the same amount of dopamine loss than in Parkinson's disease, which has both pre- and postsynaptic compensatory mechanisms.

The elevated dopamine concentration in the putamen in Alzheimer's disease may be a consequence of neuroleptic medication, or is perhaps some artefact of case selection. Previous reports have usually shown unchanged dopamine and HVA in Alzheimer's disease (Arai et al., 1984; Langlais et al., 1993), but decreased dopamine in the basal ganglia (Storga et al., 1996; Ohara et al., 1998) has also been found. There may be other factors contributing to movement abnormalities in Alzheimer's disease, involving nuclei beyond the nigrostriatal pathway, or synthesis and storage of dopamine rather than release (Marshall et al., 1994). There was a greater range of HVA : dopamine ratio values in Alzheimer's disease than in DLB patients, especially caudally in the caudate, which may point to greater heterogeneity among the Alzheimer's disease patients. Neuroleptic administration tends to raise dopamine turnover in rats (Clow et al., 1980; See et al., 1992), and the raised HVA : dopamine ratio in Alzheimer's disease and DLB may arise partly in response to drug treatment.

Conclusion

Variations in the dopaminergic features which may underlie extrapyramidal symptoms in DLB, Parkinson's disease and Alzheimer's disease were revealed, with differences between DLB and Parkinson's disease presynaptically in dopamine loss and turnover, and dopamine uptake sites, as well as postsynaptically in D2 receptor density. DLB and Parkinson's disease both showed patterns of dopamine deficits which differed from those of controls and Alzheimer's disease patients. These findings are relevant to clinical practice since it is important to be able to distinguish Alzheimer's disease and DLB because prognosis and treatment strategies with neuroleptics, acetylcholinesterase inhibitors and l-dopa may need to vary. The caudal loss in dopamine uptake sites may be an effective way of separating DLB from Alzheimer's disease and Parkinson's disease in vivo with a suitable ligand, e.g. [123I]β-CIT (β-carboxymethyl-iodophenyl-tropane), in combination with clinical diagnostic criteria, as reported by Donnemiller and colleagues (Donnemiller et al., 1997). The lower expression caudally of D2 receptors in DLB compared with Alzheimer's disease, and especially the low D2 binding rostrally in DLB compared with Parkinson's disease, may also provide assistance in diagnosis by in vivo imaging of the D2 receptor in a way similar to that suggested for separating Parkinson's disease from parkinsonism-plus syndromes. Sequential imaging of dopamine uptake sites and D2 binding may also be valuable in monitoring the emergence of reduced l-dopa responsiveness and evaluating the effectiveness of therapeutic drugs in DLB.

References

View Abstract