Abstract

The neuropathological substrate of dementia in patients with Parkinson’s disease is still under debate, particularly in patients with insufficient alternate neuropathology for other degenerative dementias. In patients with pure Lewy body Parkinson’s disease, previous post-mortem studies have shown that dopaminergic and cholinergic regulatory projection systems degenerate, but the exact pathways that may explain the development of dementia in patients with Parkinson’s disease remain unclear. Studies in rodents suggest that both the mesocorticolimbic dopaminergic and septohippocampal cholinergic pathways may functionally interact to regulate certain aspects of cognition, however, whether such an interaction occurs in humans is still poorly understood. In this study, we performed stereological analyses of the A9 and A10 dopaminergic neurons and Ch1, Ch2 and Ch4 cholinergic neurons located in the basal forebrain, along with an assessment of α-synuclein pathology in these regions and in the hippocampus of six demented and five non-demented patients with Parkinson’s disease and five age-matched control individuals with no signs of neurological disease. Moreover, we measured choline acetyltransferase activity in the hippocampus and frontal cortex of eight demented and eight non-demented patients with Parkinson’s disease, as well as in the same areas of eight age-matched controls. All patients with Parkinson’s disease exhibited a similar 80–85% loss of pigmented A9 dopaminergic neurons, whereas patients with Parkinson’s disease dementia presented an additional loss in the lateral part of A10 dopaminergic neurons as well as Ch4 nucleus basalis neurons. In contrast, medial A10 dopaminergic neurons and Ch1 and Ch2 cholinergic septal neurons were largely spared. Despite variable Ch4 cell loss, cortical but not hippocampal cholinergic activity was consistently reduced in all patients with Parkinson’s disease, suggesting significant dysfunction in cortical cholinergic pathways before frank neuronal degeneration. Patients with Parkinson’s disease dementia were differentiated by a significant reduction in hippocampal cholinergic activity, by a significant loss of non-pigmented lateral A10 dopaminergic neurons and Ch4 cholinergic neurons (30 and 55% cell loss, respectively, compared with neuronal preservation in control subjects), and by an increase in the severity of α-synuclein pathology in the basal forebrain and hippocampus. Overall, these results point to increasing α-synuclein deposition and hippocampal dysfunction in a setting of more widespread degeneration of cortical dopaminergic and cholinergic pathways as contributing to the dementia occurring in patients with pure Parkinson’s disease. Furthermore, our findings support the concept that α-synuclein deposition is associated with significant neuronal dysfunction in the absence of frank neuronal loss in Parkinson’s disease.

Introduction

In addition to classic motor symptoms, the existence of significant non-motor symptoms in Parkinson’s disease is now increasingly acknowledged as part of the core features of the disorder. In particular, patients with end-stage Parkinson’s disease often have frank dementia (Aarsland and Kurz, 2010). Although loss of A9 dopaminergic cells located in the substantia nigra and subsequent reduction in striatal dopamine levels are responsible for the development of motor symptoms in Parkinson’s disease (Fahn et al., 1971; Hirsch et al., 1988), the exact pathophysiology underlying the development of dementia in Parkinson’s disease is less clear. In fact, other systems are known to be affected in Parkinson’s disease (for review see Jellinger, 1991; Irwin et al., 2013) and might be responsible for some of the non-motor symptoms of the disease. In particular, loss of neurons located in the A10 dopaminergic cell region has been reported (McRitchie et al., 1997), and studies have shown that alteration of the mesocortical dopaminergic pathway, leading to a direct depletion of dopamine in the prefrontal cortex, may cause deficits in certain memory tasks (Mattay et al., 2002).

Some aspects of cognition perturbed in Parkinson’s disease seem to respond to levodopa therapy (Beato et al., 2008) whereas other aspects do not, or may even be worsened by levodopa (for review, see Kehagia et al., 2013). Thus, it is now increasingly believed that dementia in Parkinson’s disease is likely to be secondary to a wider neurodegeneration extending beyond that affecting the dopaminergic systems (Kehagia et al., 2010, 2013) with co-morbid Alzheimer-type pathology present in up to 50% of demented patients with Parkinson’s disease (Irwin et al., 2013). Interestingly, studies also suggest that in up to 55% of cases, dementia in Parkinson’s disease cannot be attributed to a particular pathological finding, emphasizing the heterogeneous nature of this disorder (Hughes et al., 1993). Of note, anticholinergic agents are known to worsen cognitive dysfunction in patients with Parkinson’s disease (Bedard et al., 1999). In fact, a severe loss of cholinergic neurons in the basal forebrain neurons is reported in Parkinson’s disease, and overall the cholinergic system is more severely affected in demented, as opposed to non-demented patients (Dubois et al., 1983; Nakano and Hirano, 1984; Perry et al., 1985; Mufson et al., 1991). Studies in rodents support this notion by showing that the integrity of the mesocorticolimbic dopaminergic and cholinergic basal forebrain system is critical to regulate aspects of cognitive function, particularly those related to working memory (Wisman et al., 2008; Hall et al., 2013). Largely due to technical difficulties, assessment of structural degeneration in both the dopaminergic and cholinergic systems involved in cognition has not been performed in human post-mortem material although it may be a significant factor in the dementia observed in many patients with Parkinson’s disease.

The aim of this study was to investigate whether concurrent structural and/or functional deficits in both dopaminergic and cholinergic systems could be associated with the development of dementia in Parkinson’s disease in cases without significant occurrence of Alzheimer-type pathology. For this purpose, we selected cases not reaching current neuropathological criteria for any neurodegenerative condition other than Parkinson’s disease (Hyman et al., 2012; Montine et al., 2012) and analysed in detail selected brain regions in post-mortem material obtained from longitudinally followed demented and non-demented patients with Parkinson’s disease, as well as from age-matched individuals without any neurological disease. The brain areas selected for detailed immunohistochemical analyses included the A9 and A10 midbrain dopaminergic regions and the Ch1, Ch2 and Ch4 basal forebrain cholinergic regions. To perform quantitative analyses using stereological techniques in brain regions located at or close to the midline, we chose cases where the entire basal ganglia and brainstem were available for analysis and processed these bilaterally preserved human post-mortem tissue blocks. Thus the case types used here for inclusion in the immunohistological study are rare due to the following reasons: (i) they are screened to exclude anything but age-related pathologies and Parkinson’s disease; (ii) they have their entire bilateral fixed tissue blocks from the midbrain, basal ganglia and basal forebrain available for study; and (iii) they were longitudinally studied using the same protocol with clinical data on dementia status and severity available.

Materials and methods

Cases

Human brain tissues were obtained from the Sydney Brain Bank at Neuroscience Research Australia. Participants were studied longitudinally using standardized assessments, and informed consent was obtained from all individuals before donation of their brain. Standardized clinicopathological criteria were used for diagnosis (Halliday et al., 2002; Hyman et al., 2012; Montine et al., 2012) with dementia status and severity determined using the Clinical Dementia Rating (CDR) scale (Morris, 1997). Controls had no significant neuropathology and no evidence of neurological or psychiatric disease. Parkinson’s disease cases fulfilled the UK Parkinson’s Disease Society Brain Bank Diagnostic Criteria (Hughes et al., 1992); they were all levodopa responsive, medicated throughout the disease course, and had no other significant neuropathology. Ethics approval for this study was obtained from the Human Research Ethics Committee at the University of New South Wales. Brains were obtained from two separate groups of participants (each including demented and non-demented patients with Parkinson’s disease as well as control participants) to comply with the tissue requirement of each analysis, where histopathological and biochemical analyses require fixed and unfixed tissue, respectively. Tables 1 and 2 summarize the clinical, demographic and age-related pathological details of each case included in this study.

Table 1

Clinical and demographic characteristics of cases included in the histological analyses

Case number Age Onset of PD Duration of PD Duration of dementia HY stage (/5) CDR (/3) Plaque (/3) NFT (/6) Cause of death PMI (h) Fixed brain weight (g) 
Ctrl1 85 Multiple myeloma 943 
Ctrl2 89 Renal failure 1079 
Ctrl3 87 Sepsis 39.5 1099 
Ctrl4 84 Acute myocardial infarction 21 991 
Ctrl5 89 Septicaemia 78 1211 
Median (IQR) 87 (4)    0 0 1 (1) 3 (3)  21 (33) 1079 (108) 
PD1 76 48 28 Parkinson’s disease 28 years 18 1102 
PD2 72 56 16 Adenocarcinoma of lung with metastasis 1200 
PD3 84 74 10 Ruptured aortic aneurysm 1228 
PD4 91 79 12 Parkinson’s disease 20 1092 
PD5 74 45 29 Cardiopulmonary arrest 18 1066 
Median (IQR) 76 (10) 56 (26) 16 (16)  5 (0) 0 1 (1) 2 (2)  18 (13) 1102 (108) 
PDD1 88 76 12 Parkinson’s disease dementia 22 1253 
PDD2 76 47 29 Hypostatic chest infection 29 1370 
PDD3 83 70 13 Aspiration pneumonia 10 1208 
PDD4 80 63 17 Cerebrovascular accient (24 h); hypertension 12 months 9.5 1140 
PDD5 87 54 33 Cerebrovascular accident 20 1127 
PDD5 61 51 10 Hypostatic pneumonia 39 1795 
Median (IQR) 82 (9) 59 (16.5) 15 (14) 5 (0) 5 (0) 3 (1) 1 (1) 0 (1.5)  21 (15) 1231 (183) 
Case number Age Onset of PD Duration of PD Duration of dementia HY stage (/5) CDR (/3) Plaque (/3) NFT (/6) Cause of death PMI (h) Fixed brain weight (g) 
Ctrl1 85 Multiple myeloma 943 
Ctrl2 89 Renal failure 1079 
Ctrl3 87 Sepsis 39.5 1099 
Ctrl4 84 Acute myocardial infarction 21 991 
Ctrl5 89 Septicaemia 78 1211 
Median (IQR) 87 (4)    0 0 1 (1) 3 (3)  21 (33) 1079 (108) 
PD1 76 48 28 Parkinson’s disease 28 years 18 1102 
PD2 72 56 16 Adenocarcinoma of lung with metastasis 1200 
PD3 84 74 10 Ruptured aortic aneurysm 1228 
PD4 91 79 12 Parkinson’s disease 20 1092 
PD5 74 45 29 Cardiopulmonary arrest 18 1066 
Median (IQR) 76 (10) 56 (26) 16 (16)  5 (0) 0 1 (1) 2 (2)  18 (13) 1102 (108) 
PDD1 88 76 12 Parkinson’s disease dementia 22 1253 
PDD2 76 47 29 Hypostatic chest infection 29 1370 
PDD3 83 70 13 Aspiration pneumonia 10 1208 
PDD4 80 63 17 Cerebrovascular accient (24 h); hypertension 12 months 9.5 1140 
PDD5 87 54 33 Cerebrovascular accident 20 1127 
PDD5 61 51 10 Hypostatic pneumonia 39 1795 
Median (IQR) 82 (9) 59 (16.5) 15 (14) 5 (0) 5 (0) 3 (1) 1 (1) 0 (1.5)  21 (15) 1231 (183) 

Age and duration of Parkinson’s disease and dementia are indicated in years. Medians are presented with (interquartile range, IQR).

CDR = clinical dementia rating; ctrl = control; HY = Hoehn and Yahr; NFT = neurofibrillary tangle; PD = Parkinson's disease; PDD = Parkinson's disease with dementia; PMI = post-mortem interval.

Table 2

Clinical and demographic characteristics of cases included in the biochemical analyses

Case number Age Onset of PD Duration of PD Duration of dementia HY stage (/5) CDR (/3) Plaque (/3) NFT (/6) Cause of death PMI (h) Fixed brain weight (g) 
Ctrl1 84 Respiratory arrest 550 
Ctrl2 85 Pneumonia 23 495 
Ctrl3 87 Acute peritonitis 24 466 
Ctrl4 89 0.5 Metastatic adenocarcinoma 23 506 
Ctrl5 90 Acute myeloid leukaemia 26 545 
Ctrl6 92     Pancytopaenia 491 
Ctrl7 93     Congestive cardiac failure 15 516 
Ctrl8 94     Cardiorespiratory arrest 24 555 
Median (IQR) 90 (6)    0 0 (0.1) 0.5 (1) 3 (0.8)  23 (11) 501 (39) 
PD1 71 55 16 Cardiorespiratory failure 15 630 
PD2 72 63 Acute myocardial infarction 710 
PD3 75 57 18 Cardiorespiratory failure 24 690 
PD4 80 64 16 Acute myocardial infarction 23 530 
PD5 84 37 47 Cardiorespiratory failure 21 520 
PD6 84 72 12  Cholangitis, cholangiocarcinoma 666 
PD7 87 67 20  Cardiorespiratory failure 24 550 
PD8 88 65 23  Cardiorespiratory failure 498 
Median (IQR) 82 (10.5) 64 (9) 17 (6)  5 (0.3) 0 0 (1) 1 (1.3)  18 (18) 590 (144) 
PDD1 72 65 Cardiorespiratory failure 29 668 
PDD2 72 59 13 Multi-organ failure 555 
PDD3 78 57 21 Cardiorespiratory arrest, advanced leukaemia 22 584 
PDD4 79 62 17 Septicaemia 32 649 
PDD5 83 69 14 Cardiorespiratory failure 32 469 
PDD6 85 69 14 Pneumonia 624 
PDD7 83 68 17 Cerebrovascular accident 26 491 
PDD8 90 75 15 Cardiac failure, ischaemic heart disease 584 
Median (IQR) 81 (7) 67 (8) 15 (3) 3.5 (2) 5 (1) 1 (1.3) 1 (1.3) 0 (2.3)  24 (23) 584 (92) 
Case number Age Onset of PD Duration of PD Duration of dementia HY stage (/5) CDR (/3) Plaque (/3) NFT (/6) Cause of death PMI (h) Fixed brain weight (g) 
Ctrl1 84 Respiratory arrest 550 
Ctrl2 85 Pneumonia 23 495 
Ctrl3 87 Acute peritonitis 24 466 
Ctrl4 89 0.5 Metastatic adenocarcinoma 23 506 
Ctrl5 90 Acute myeloid leukaemia 26 545 
Ctrl6 92     Pancytopaenia 491 
Ctrl7 93     Congestive cardiac failure 15 516 
Ctrl8 94     Cardiorespiratory arrest 24 555 
Median (IQR) 90 (6)    0 0 (0.1) 0.5 (1) 3 (0.8)  23 (11) 501 (39) 
PD1 71 55 16 Cardiorespiratory failure 15 630 
PD2 72 63 Acute myocardial infarction 710 
PD3 75 57 18 Cardiorespiratory failure 24 690 
PD4 80 64 16 Acute myocardial infarction 23 530 
PD5 84 37 47 Cardiorespiratory failure 21 520 
PD6 84 72 12  Cholangitis, cholangiocarcinoma 666 
PD7 87 67 20  Cardiorespiratory failure 24 550 
PD8 88 65 23  Cardiorespiratory failure 498 
Median (IQR) 82 (10.5) 64 (9) 17 (6)  5 (0.3) 0 0 (1) 1 (1.3)  18 (18) 590 (144) 
PDD1 72 65 Cardiorespiratory failure 29 668 
PDD2 72 59 13 Multi-organ failure 555 
PDD3 78 57 21 Cardiorespiratory arrest, advanced leukaemia 22 584 
PDD4 79 62 17 Septicaemia 32 649 
PDD5 83 69 14 Cardiorespiratory failure 32 469 
PDD6 85 69 14 Pneumonia 624 
PDD7 83 68 17 Cerebrovascular accident 26 491 
PDD8 90 75 15 Cardiac failure, ischaemic heart disease 584 
Median (IQR) 81 (7) 67 (8) 15 (3) 3.5 (2) 5 (1) 1 (1.3) 1 (1.3) 0 (2.3)  24 (23) 584 (92) 

Age and duration of Parkinson’s disease and dementia are indicated in years. Medians are presented with (interquartile range, IQR).

CDR = clinical dementia rating; ctrl = control; HY = Hoehn and Yahr; NFT = neurofibrillary tangle; PD = Parkinson's disease; PDD = Parkinson's disease with dementia; PMI = post-mortem interval.

Tissue preparation for histological analysis

Whole brains were fixed in 15% neutral buffered formalin, weighed and the volume determined by fluid displacement. The cerebellum and brainstem were separated from the cerebrum, the weight and volume of the cerebrum determined and the length of each hemisphere measured. Each cerebrum was embedded in 3% agar, sectioned in 3-mm thick coronal slices and photographed. The brainstem was also embedded in agar and sectioned transversely at 3 mm intervals and the posterior surface photographed. Large tissue blocks containing the entire basal forebrain and any associated basal ganglia structures in the same coronal slices were taken for sectioning, and all transverse tissue blocks of the entire midbrain were taken for sectioning. In addition, a 3 mm coronal slice of the hippocampus at a level just posterior to the lateral geniculate nucleus was taken for sectioning.

Tissue blocks from the basal forebrain, midbrain and hippocampus were cryoprotected in 30% sucrose in 0.1 M Tris buffer (pH 7.4), sectioned on a freezing microtome (Microm HM 450) into coronal (forebrain and hippocampus) or transverse (midbrain) 50-µm thick sections and collected into 12 or 16 spaced series of sections. The sections were stored in antifreeze solution (0.5 M sodium phosphate buffer, 30% glycerol and 30% ethylene glycol) at −20°C until processed for histology. The first series of sections was mounted on chromatin gelatin-coated slides and processed for Nissl (0.5% cresyl violet, containing 0.1% acetic acid) for identification of cyto-architectural boundaries. Alternate series of sections were used for immunohistochemistry, as described below.

Tissue preparation for biochemical analysis

Frozen tissues from mid-level hippocampal dentate gyrus and surrounds (sampled just posterior to the coronal level of the lateral geniculate nucleus) and from the upper cortical layers of the superior frontal gyrus (sampled at the level of or just posterior to the genu of the corpus callosum) were used to determine the activity of choline acetyltransferase (ChAT), the enzyme catalysing the acetylation of choline to yield acetylcholine. Tissue samples were homogenized 200 mg/ml in 20 mM Tris-acetate buffer pH 6.1 using stainless steel beads and a Tissuelyser II system (Qiagen) at 20 Hz for 40 s. The homogenate samples were then centrifuged at 14 000rpm for 10 min at 4°C and the supernatant collected. Protein concentration in the supernatant was determined using the Protein A280 module in a NanoDrop™ 2000c spectrophotometer (Thermo Fisher Scientific).

Immunohistochemistry

Immunohistochemistry was performed on free-floating sections using the following antibodies: goat anti-ChAT (diluted 1:100; AB144P, Millipore), rabbit anti-tyrosine hydroxylase (a marker of dopaminergic neurons; diluted 1:1000; P40101-0, Pelfreeze) and mouse anti α-synuclein (LB509; diluted 1:500, ab27766, Abcam). To optimize staining properties of ChAT and α-synuclein antibodies, antigen retrieval was performed before the protocol described below by pre-incubating the sections at 80°C overnight in Tris EDTA buffer (pH adjusted to 9.0).

The sections were first rinsed in TBS solution (0.5 M Trisma base, 0.15 M NaCl, pH 7.6) followed by quenching of the endogenous peroxidase activity by incubation in 3% H2O2 and 10% methanol in TBS for 30 min and then rinsed in 0.05% Triton™ X-100 TBS. To eliminate non-specific secondary antibody binding to the tissue, sections were incubated for 1 h with 0.05% Triton™ X-100 TBS containing 5% normal serum matching the species used to raise the corresponding secondary antibodies. Primary antibodies were prepared in 0.05% Triton™ X-100 TBS containing 1% bovine serum albumin. For tyrosine hydroxylase and α-synuclein immunostaining, incubation with primary antibody was performed at room temperature overnight, whereas incubation with anti-ChAT antibody was carried out at 4°C for 48 h. On the following day, the sections were rinsed in 0.05% Triton™ X-100 TBS and incubated for 1 h at room temperature in 1:200 dilution of the appropriate biotinylated secondary antibody solution (Vector laboratories), followed by a 1-h incubation in avidin-biotin-peroxidase solution (ABC kit, Vector Laboratories). Antibody staining was visualized using the 3,3’-diaminobenzidine (DAB) chromogen and 0.01% H2O2. Midbrain tyrosine hydroxylase and α-synuclein immunostainings were visualized using 0.02% nickel 0.025% cobalt DAB, which yields a blue-black colour permitting differentiation of the immunostaining from the brown melanin pigment in these cells. The sections were then mounted on chromatin-gelatin coated slides, dehydrated in ascending alcohol concentrations, cleared in xylene and coverslipped using DPX (Sigma).

Stereological analysis

The total number of ChAT-immunoreactive neurons in the basal forebrain and tyrosine hydroxylase-immunoreactive and/or melanin-containing neurons in the ventral midbrain were estimated using an unbiased stereological quantification method according to the optical fractionator principle (West, 1999; Schmitz and Hof, 2005). A low magnification objective (×2) was used to draw the boundaries of the regions of interest. The basal forebrain region was delineated into medial septum/vertical limb of the diagonal Band of Broca (DBB), which includes the cholinergic Ch1 and Ch2 nuclei, and the nucleus basalis of Meynert (NBM), which contains the cholinergic Ch4 nucleus. The medial septum/vertical DBB extends from the genu of the corpus callosum, rostrally, to the decussation of the anterior commissure, caudally, its lateral edge occurring near the medial border of the olfactory tubercle. The anterior part of the NBM occurs lateral to the vertical DBB and ventral to the anterior commissure. More posteriorly, it extends between the lateral edge of the optic tract and the emergence of the anterior commissure from the temporal lobe and remains ventral to the globus pallidus, which defines its dorsolateral and ventromedial boundaries. Figure 1 shows examples of the delineation at four different anteroposterior levels of the basal forebrain.

Figure 1

Photomicrographs of four representative coronal cresyl violet-stained sections through the basal forebrain showing inclusion boundaries for Ch1 and Ch2 regions (depicted as CH1/2) and Ch4 cholinergic nuclei as defined for stereological analyses. The Ch1/2 nuclei (A and B) represent the cholinergic cells associated to the medial septum/vertical DBB and the Ch4 nuclei (C and D) represent the cholinergic cells located in the NBM. Scale bar = 2 mm. Solid lines show boundaries used for quantification, and dashed lines indicate borders of the other designated structures identified for orientation on these coronal sections. ac = anterior commissure; cd = caudate nucleus; fx = fornix; gp = globus pallidus; gpe = external globus pallidus; gpi = internal globus pallidus; ic = internal capsule; lv = lateral ventricule; nacc = nucleus accumbens; ox = optic chiasm; put = putamen.

Figure 1

Photomicrographs of four representative coronal cresyl violet-stained sections through the basal forebrain showing inclusion boundaries for Ch1 and Ch2 regions (depicted as CH1/2) and Ch4 cholinergic nuclei as defined for stereological analyses. The Ch1/2 nuclei (A and B) represent the cholinergic cells associated to the medial septum/vertical DBB and the Ch4 nuclei (C and D) represent the cholinergic cells located in the NBM. Scale bar = 2 mm. Solid lines show boundaries used for quantification, and dashed lines indicate borders of the other designated structures identified for orientation on these coronal sections. ac = anterior commissure; cd = caudate nucleus; fx = fornix; gp = globus pallidus; gpe = external globus pallidus; gpi = internal globus pallidus; ic = internal capsule; lv = lateral ventricule; nacc = nucleus accumbens; ox = optic chiasm; put = putamen.

In the midbrain, cell groups in the substantia nigra, and two parts of the ventral tegmental area (VTA), namely the parabrachial pigmented nucleus (known as the lateral VTA) and the medial ventromedial mesencephalic tegmentum (known as the medial VTA) were delineated. The medial VTA cell groups included the VTA nucleus, rostral linear nucleus, central linear nucleus, paranigral nucleus, parapeduncular nucleus and interfascicular nucleus, which with the parabrachial pigmented nucleus will be collectively referred to as the ‘VTA region’ in this paper. The VTA region is situated medial to the red nucleus and remains dorsal to the substantia nigra. When present, the oculomotor nerve fibres define the border between the medial and lateral VTA. In the caudal portion, the medial VTA extends dorsomedial to the substantia nigra and dorsal to the interpeduncular nucleus. The substantia nigra is found dorsomedial to the cerebral peduncle, lateral to the medial VTA and ventral to the lateral VTA, the latter separating the substantia nigra from the red nucleus. Figure 2 shows delineation of the regions at three different rostrocaudal midbrain levels.

Figure 2

Photomicrographs of three representative transverse tyrosine hydroxylase-stained sections through the midbrain showing inclusion boundaries for substantia nigra (SN) and lateral and medial VTA at rostral (A), intermediate (B) and caudal (C) levels. Scale bar = 1 mm. cp = cerebral peduncle; rn = red nucleus; rrf = retrorubral field; 3n = exiting third nerve.

Figure 2

Photomicrographs of three representative transverse tyrosine hydroxylase-stained sections through the midbrain showing inclusion boundaries for substantia nigra (SN) and lateral and medial VTA at rostral (A), intermediate (B) and caudal (C) levels. Scale bar = 1 mm. cp = cerebral peduncle; rn = red nucleus; rrf = retrorubral field; 3n = exiting third nerve.

Sampling was done using the CAST module in the VIS software (version 4.4.4.0, Visiopharm) by an investigator blinded to case group identity. The actual neuronal counts were obtained using a ×60 Plan-Apo oil-immersion objective with a high numerical aperture (NA = 1.4) on a Nikon Eclipse 80i microscope equipped with an x–y motorized stage and a z-axis motor (Prior). Precise position in the z-axis was measured with a microcator (Heidenhain). The counting frame was placed on the first counting position identified randomly, and then moved through all the counting positions until the entire delineated region was systematically sampled. The x-y step length was adjusted so that 150–250 cells were counted on each side of the brain for each structure. Estimates of total number of cells were obtained according to the optical fractionator formula. Coefficient of error (CE) attributable to the sampling was calculated according to Gundersen and Jensen (1987) and values ≤0.1 were accepted. Total volumes of each delineated area were estimated using the Cavalieri method in the same software platform.

Quantitative assessment of striatal cholinergic interneurons

Quantitative assessment of ChAT-immunoreactive interneurons in the putamen was carried out in a region extending across the decussation of the anterior commissure, 5 mm in the anteroposterior length of the putamen. The actual neuronal counts were performed on a total of five sections spaced 1.2 mm anteroposteriorly from each other, encompassing both the pre- and post-commissural putamen. The boundaries of the regions of interest were drawn at a low magnification objective (×2) and cholinergic interneurons were counted at ×10 magnification, using a Nikon Eclipse 80i microscope. A systematic sample principle, as described above, was applied and 100–150 cells were counted on each side of the putamen.

Assessment of α-synuclein immunoreactive Lewy bodies and Lewy neurites

Semi-quantitative assessment of Lewy bodies was performed in the basal forebrain (separately in the medial septum/vertical DBB and NBM) and midbrain regions at two representative levels for each region. An investigator blinded to the case group identity of the section counted the number of Lewy bodies per ×20 microscopic field. Lewy neurites were rated semi-quantitatively (0 = absent, 1 = sparse, 2 = moderate, 3 = severe) in the same regions as well as in the hippocampus. Lewy body and Lewy neurite counts were normalized into z-scores to create unified variables reflecting the Lewy pathology in the medial septum/vertical DBB and NBM. An additional combined variable was calculated by adding the medial septum/vertical DBB and NBM z-scores to reflect overall Lewy pathology in the basal forebrain.

Choline acetyltransferase activity assay

ChAT activity was measured using a modified version of the radiometric assay of Fonnum (1975). Briefly, tissue supernatants (prepared as detailed above) were mixed 1:1 with 100 mM sodium phosphate buffer pH 7.4 containing 1% Triton™ X-100 and 20 mM EDTA. Diluted sample (10 µl) was mixed with 20 µl of reaction buffer containing 50 mM sodium phosphate buffer pH 7.4, 300 mM NaCl, 8 mM choline chloride, 0.2 mM physostigmine and 0.2 mM acetyl-coenzyme A (ACoA; with a 1:5 ratio of [1-14C]-ACoA) and immediately incubated at 37°C for 15 min. All samples were run in duplicates and blanks were run substituting homogenate with water. The reaction was stopped by putting the tubes on ice and adding 4 ml of ice-cold 10 mM sodium phosphate buffer pH 7.4, followed by addition of 2 ml sodium tetraphenylborate (Sigma Aldrich; 5 mg/ml in 15:85 acetonitrile:toluene). The organic and inorganic phases were separated by centrifugation at 4500 rpm for 2 min. The organic phase (1 ml) was mixed with 5 ml of liquid scintillation cocktail (Ultima Gold Plus™, Perkin Elmer Inc) and the radioactivity was measured using a liquid scintillation counter (Beckman LS 6500, Beckman Inc.) and quantified by an external standard calibration curve. ChAT was expressed as nmol ACh/mg protein/h.

Statistical analysis

Data are presented as median in box plots, unless stated otherwise. All statistical analyses were conducted using the Statistical Package for the Social Sciences version 19 (SPSS Inc). As the data analysed in this study did not fulfil criteria for normal distribution, non-parametric tests were performed for all statistical comparisons, unless stated otherwise. Kruskal Wallis (three group comparison) or Mann Whitney (two group comparison) tests were used to assess if the demographics variables of the groups differed from each other. Left and right cell counts assessed by stereology in each structure were compared using a Wilcoxon signed rank test. Kruskal Wallis test with uncorrected Mann Whitney post hoc test was performed on all stereological and Lewy body counts and Lewy neurite scores, as well as ChAT activity measurements. In cases where cell counts were calculated as percentage of total or percentage of control, repeated measures ANOVA (with Greenhouse-Geisser correction when assumption of sphericity was violated) were performed using the general linear model, followed by Tukey HSD post hoc test with Bonferroni correction. Two-tailed analyses of correlation were performed using Spearman’s rank order correlation or Kendall’s tau correlation where appropriate. Statistical significance was set at 0.05. Data in Figs 3, 5 and 7 are shown as box plots where boundaries of the box represent the 25th and 75th percentiles, while the median is indicated by the horizontal lines. Whiskers represent the lowest and highest data point within 1.5 interquartile ranges (outliers are values that lie between 1.5 and 3 interquartile range and extreme values are values outside 3 interquartile ranges).

Figure 3

Stereological analysis of dopaminergic neurons in the midbrain. Dopaminergic neurons were quantified in the A9 substantia nigra (A), lateral A10 VTA (E) and medial A10 VTA (I) using stereology. Photomicrographs illustrate tyrosine hydroxylase-immunostained sections from a representative case in each group at the level of the A9 substantia nigra (B–D), lateral A10 VTA (F–H) and medial A10 VTA (J–L). Scale bar = 100 µm. Total numbers of dopaminergic neurons are calculated as the sum of tyrosine hydroxylase-immunoreactive and melanin-containing (TH+M+), TH+M− and TH−M+ cells and shown as box plots. Open circle indicates outlier value and filled circles indicate extreme values. *P < 0.05; **P < 0.01 different from control. (A) Kruskal Wallis, χ2(2) = 9.216, P = 0.0004 followed by uncorrected Mann Whitney post hoc test. (E) Kruskal Wallis, χ2(2) = 5.941, P = 0.009 followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease dementia; SN = substantia nigra.

Figure 3

Stereological analysis of dopaminergic neurons in the midbrain. Dopaminergic neurons were quantified in the A9 substantia nigra (A), lateral A10 VTA (E) and medial A10 VTA (I) using stereology. Photomicrographs illustrate tyrosine hydroxylase-immunostained sections from a representative case in each group at the level of the A9 substantia nigra (B–D), lateral A10 VTA (F–H) and medial A10 VTA (J–L). Scale bar = 100 µm. Total numbers of dopaminergic neurons are calculated as the sum of tyrosine hydroxylase-immunoreactive and melanin-containing (TH+M+), TH+M− and TH−M+ cells and shown as box plots. Open circle indicates outlier value and filled circles indicate extreme values. *P < 0.05; **P < 0.01 different from control. (A) Kruskal Wallis, χ2(2) = 9.216, P = 0.0004 followed by uncorrected Mann Whitney post hoc test. (E) Kruskal Wallis, χ2(2) = 5.941, P = 0.009 followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease dementia; SN = substantia nigra.

Results

A total of 16 cases (five with Parkinson’s disease, six with Parkinson’s disease dementia, five age-matched control subjects) were included in the immunohistochemical analyses, while the biochemical analysis was performed on a total of 24 cases (eight with Parkinson’s disease, eight with Parkinson’s disease dementia, eight age-matched control subjects). The demographic and clinical details of each case are presented in Tables 1 and 2. For each analysis, there was no significant difference in age, post-mortem interval, amyloid-β plaque score or neurofibrillary tangles score between the three groups. In addition, there was no significant difference in onset or duration of disease and Hoehn and Yahr stage between the Parkinson’s disease and Parkinson’s disease dementia groups. The diagnosis of dementia was established prospectively for each brain donor during follow-ups using the CDR scale, where a score of 0 indicates no dementia, and scores of 1-2-3 indicate mild-moderate-severe dementia, respectively (Morris, 1997). For each of the stereological analyses, we conducted a Wilcoxon signed rank test to exclude pairwise differences between left and right sides of the brain for each region of interest. There was no significant difference between the two sides in any of the regions analysed and therefore we report the total bilateral counts (substantia nigra, Z = −0465, P = 0.669; lateral VTA, Z = −0.465, P = 0.669; medial VTA, Z = −0.931, P = 0.375, medial septum/vertical DBB, Z = −0.621, P = 0.562; NBM, Z = −0.511, 0.639).

Loss of A9 dopaminergic neurons in Parkinson’s disease: selective loss of lateral ventral tegmental area A10 neurons in Parkinson’s disease dementia

The estimated numbers and the appearance of tyrosine hydroxylase-immunoreactive and melanin-containing A9 and A10 neurons in three defined regions of the midbrain (substantia nigra, lateral and medial VTA, as outlined above; Fig. 2) are shown in Fig. 3. As expected, a severe 80–85% reduction in the number of nigral A9 dopaminergic neurons was detected in both Parkinson’s disease patient groups compared to the controls (Fig. 3A–D). In contrast, there was a differential loss of A10 dopaminergic neurons in the lateral VTA that was significant in the cases with Parkinson’s disease dementia (with cell number ranging from 26 391 to 66 461, compared to controls ranging from 60 270 to 138 160), whereas in non-demented Parkinson’s disease cases, the trend for lower cell numbers in this structure (with cell number ranging from 43 901 to 123 723) did not reach significance (Fig. 3E–H). In addition, no statistically significant difference was observed in the number of A10 dopaminergic neurons located in the medial VTA between the three groups. It should be noted that the variability observed in the medial VTA counts was higher than in the substantia nigra and lateral VTA counts (one extreme value in the control group, one outlier and one extreme value in the Parkinson’s disease dementia group; Fig. 3I–L).

Melanin-containing dopaminergic neurons are targeted in Parkinson’s disease, whereas in Parkinson’s disease dementia non-melanized dopaminergic neurons are also reduced

The immunohistochemical method allowed the distinction of staining attributable to tyrosine hydroxylase by its dark-blue colour located in the cell soma and proximal processes from the brown granular cytoplasm-restricted endogenous melanin pigment. Neurons containing each of these markers of dopaminergic neurons were evaluated, identifying three types of dopaminergic neurons (see also Milber et al., 2012): tyrosine hydroxylase-immunoreactive and melanin-containing neurons, tyrosine hydroxylase-negative melanin-containing neurons and tyrosine hydroxylase-immunoreactive melanin-negative neurons. In the three regions of interest we estimated the proportion of each neuron type expressed as a percentage of total dopaminergic neurons in each case. To determine if there was a disproportionate loss of any cell subtype, a two-way repeated measures ANOVA of the percentage of total dopaminergic neurons within each case was performed. This analysis revealed a significant main effect of the cell subtype in the A9 substantia nigra, F(1.383,17.976) = 69.4, P < 0.001 (Greenhouse Geisser correction), as well as in the lateral A10 VTA, F(2,26) = 58.9, P < 0.001, and in the medial A10 VTA, F(1.089,14.159) = 38.2, P < 0.001 (Greenhouse Geisser correction), showing that the dopaminergic cells most abundant in all three regions in all cases are tyrosine hydroxylase-immunoreactive and melanin-containing (substantia nigra, 65 ± 3%; lateral VTA, 62 ± 4%; VTA, 60 ± 4%). In the A9 substantia nigra, the second largest subtype was the tyrosine hydroxylase-negative melanin-containing neurons (23 ± 3%), followed by the tyrosine hydroxylase-immunoreactive melanin-negative neurons (12 ± 2%). Tyrosine hydroxylase-immunoreactive melanin-negative neurons were more common than tyrosine hydroxylase-negative melanin-containing neurons in both the lateral and medial A10 VTA (23 ± 3% versus 14 ± 2% and 34 ± 4% versus 7 ± 1%, respectively) (Fig. 4A). There was no significant interaction between cell subtype and group in the A9 substantia nigra and medial A10 VTA, with melanin-containing neurons significantly reduced in the A9 substantia nigra but spared in the medial A10 VTA (Fig. 4A, C and E). Albeit weak, there was an interaction between cell subtype and group in the lateral A10 VTA, F(4,26) = 2.976, P = 0.038, with tyrosine hydroxylase-negative melanin-containing neurons reduced in all patients with Parkinson’s disease, whereas tyrosine hydroxylase-immunoreactive melanin-negative neurons were selectively affected in patients with dementia (Fig. 4A and D).

Figure 4

Proportion of tyrosine hydroxylase-immunoreactive and melanin-containing cells in dopaminergic neurons of the midbrain. The proportion of TH+M−, TH+M+ and TH−M+ cells, quantified using stereology (as shown in Fig. 2), is represented as percentage of total cells in each group (A), and for the Parkinson’s disease and Parkinson’s disease with dementia groups as percentage of corresponding cell subtypes in the control group (C–E) in A9 substantia nigra, lateral and medial A10 VTA regions. A representative section at the level of the substantia nigra shows tyrosine hydroxylase-immunoreactive neurons in blue and melanin pigment as a brown granular deposit (B). Insets show examples of TH+M+, TH+M− and TH−M+ cells. Scale bar = 100 µm (large panel) and 20 µm (insets). Data are presented as mean and plotted in stack bar graphs (A), or as mean + SEM with superimposed individual data points (CE). Black line indicates mean of control group ± SEM. PD = Parkinson’s disease; PDD = Parkinson’s disease with dementia; SN = substantia nigra; TH = tyrosine hydroxylase; M = melanin.

Figure 4

Proportion of tyrosine hydroxylase-immunoreactive and melanin-containing cells in dopaminergic neurons of the midbrain. The proportion of TH+M−, TH+M+ and TH−M+ cells, quantified using stereology (as shown in Fig. 2), is represented as percentage of total cells in each group (A), and for the Parkinson’s disease and Parkinson’s disease with dementia groups as percentage of corresponding cell subtypes in the control group (C–E) in A9 substantia nigra, lateral and medial A10 VTA regions. A representative section at the level of the substantia nigra shows tyrosine hydroxylase-immunoreactive neurons in blue and melanin pigment as a brown granular deposit (B). Insets show examples of TH+M+, TH+M− and TH−M+ cells. Scale bar = 100 µm (large panel) and 20 µm (insets). Data are presented as mean and plotted in stack bar graphs (A), or as mean + SEM with superimposed individual data points (CE). Black line indicates mean of control group ± SEM. PD = Parkinson’s disease; PDD = Parkinson’s disease with dementia; SN = substantia nigra; TH = tyrosine hydroxylase; M = melanin.

Preservation of cholinergic neurons in the basal forebrain in Parkinson’s disease: selective loss of Ch4 cholinergic neurons in Parkinson’s disease dementia

To quantify cholinergic neurons in the medial septum/vertical DBB and NBM, we performed stereological analyses on ChAT-immunostained sections of the basal forebrain. There was no significant difference in the number of Ch1 and Ch2 ChAT-immunoreactive neurons in the medial septum/vertical DBB (Fig. 5A–D). However, the number of Ch4 cholinergic cells was reduced by 54% in the NBM in patients with Parkinson’s disease dementia compared to control subjects, whereas the 30% reduction estimated in the non-demented Parkinson’s disease group did not reach significance (Fig. 5E–H). We estimated the density of Ch4 ChAT-immunoreactive neurons in that region by calculating the volume of the NBM. Interestingly, we found that, although there was no significant difference in the volume of the NBM in cases with Parkinson’s disease and Parkinson’s disease dementia compared with controls, the density of cholinergic cells in the Parkinson’s disease dementia group was significantly reduced by 60–65% compared with the control and non-demented Parkinson’s disease groups [Kruskal Wallis, χ2(2) = 9.912, P < 0.001, followed by uncorrected Mann Whitney post hoc test]. Loss of Ch4 ChAT-immunoreactive neurons in the NBM appeared to be specific, as there was also no significant difference in the number of cholinergic striatal neurons between the diagnostic groups (Fig. 5I–L).

Figure 5

Stereological analysis of cholinergic neurons in the basal forebrain. Cholinergic neurons were quantified in the Ch1/2 medial septum/vertical DBB (MS/vDBB; A), Ch4 NBM (E) and putamen (I) using stereology and shown as box plots where filled circles indicate extreme values. Photomicrographs illustrate ChAT-immunostained sections from a representative case in each group at the level of the medial septum/vertical DBB (B, C and D), NBM (F–H) and putamen (J–L). Scale bar = 100 µm. (E) *P < 0.05 different from controls. Kruskal Wallis, χ2(2) = 6.318, P = 0.009, followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease with dementia.

Figure 5

Stereological analysis of cholinergic neurons in the basal forebrain. Cholinergic neurons were quantified in the Ch1/2 medial septum/vertical DBB (MS/vDBB; A), Ch4 NBM (E) and putamen (I) using stereology and shown as box plots where filled circles indicate extreme values. Photomicrographs illustrate ChAT-immunostained sections from a representative case in each group at the level of the medial septum/vertical DBB (B, C and D), NBM (F–H) and putamen (J–L). Scale bar = 100 µm. (E) *P < 0.05 different from controls. Kruskal Wallis, χ2(2) = 6.318, P = 0.009, followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease with dementia.

Close inspection of the distribution of individual CDR scores in the Parkinson’s disease dementia group revealed that the highest NBM count was associated with the lowest CDR score (score = 1) and that the lowest NBM count was associated with the highest CDR score (score = 3; representing the more severely demented case). For the Ch1/2 cholinergic neurons in the medial septum/vertical DBB, the two cases with the lowest cell counts had the longest durations of dementia. However, there was no significant correlation between dementia severity or duration and cholinergic cell numbers in the NBM or medial septum/vertical DBB as assessed by a Kendall’s τ b correlation analysis, nor was there a correlation between dopaminergic and cholinergic cell counts, perhaps due in part to the small number of cases in each diagnostic group available for these extensive analyses.

Reduced choline acetyltransferase activity in frontal cortex in Parkinson’s disease: selective loss of hippocampal choline acetyltransferase activity in Parkinson’s disease dementia

We processed a second set of specimens (fresh frozen brain material) to determine if patients with Parkinson’s disease dementia have neocortical and/or hippocampal cholinergic deficits compared to non-demented patients with Parkinson’s disease using a robust ChAT activity radio-enzymatic assay. We found that ChAT activity was significantly decreased in the hippocampus in patients with Parkinson’s disease dementia, compared with controls and non-demented patients with Parkinson’s disease (Fig. 6A). Interestingly, ChAT activity was reduced in the frontal cortex in patients with Parkinson’s disease with and without dementia, compared to control subjects (Fig. 6B). However, there was no correlation between dementia severity and ChAT activity as assessed by a Kendall’s tau b correlation analysis.

Figure 6

Measurement of ChAT activity in the hippocampus and frontal cortex. ChAT activity is expressed as nmol acetylcholine/mg protein/h. Data are shown as box plots. Filled circle indicates an extreme value. *P < 0.05; **P < 0.01 different from controls, #P < 0 .05 different from Parkinson’s disease without dementia. Hippocampus: Kruskal Wallis, χ2(2) = 6.365, P = 0.041 followed by uncorrected Mann Whitney post hoc test. Frontal cortex: Kruskal Wallis, χ2(2) = 12.065, P = 0.002 followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease dementia.

Figure 6

Measurement of ChAT activity in the hippocampus and frontal cortex. ChAT activity is expressed as nmol acetylcholine/mg protein/h. Data are shown as box plots. Filled circle indicates an extreme value. *P < 0.05; **P < 0.01 different from controls, #P < 0 .05 different from Parkinson’s disease without dementia. Hippocampus: Kruskal Wallis, χ2(2) = 6.365, P = 0.041 followed by uncorrected Mann Whitney post hoc test. Frontal cortex: Kruskal Wallis, χ2(2) = 12.065, P = 0.002 followed by uncorrected Mann Whitney post hoc test. PD = Parkinson’s disease; PDD = Parkinson’s disease dementia.

More α-synuclein immunoreactive Lewy bodies and neurites occur in the basal forebrain and hippocampus of patients with Parkinson’s disease dementia

Immunohistochemistry for α-synuclein was performed in order to visualize Lewy body inclusions and Lewy neurites in the midbrain, basal forebrain and hippocampus. Semi-quantitative assessment of Lewy pathology in these regions revealed that patients with Parkinson’s disease and Parkinson’s disease dementia presented with numerous and scattered Lewy bodies and Lewy neurites compared with control subjects (Fig. 7A and E). In addition, α-synuclein pathology in the basal forebrain of patients with Parkinson’s disease dementia was more severe compared with non-demented patients with Parkinson’s disease, as seen by an increase in the overall unified Lewy z-score reflecting both Lewy body and Lewy neurite counts (Fig. 6E–H), highlighting the possible role of α-synuclein aggregation in the development of cortical and hippocampal cholinergic dysfunction. Of note, separate z-scores of Lewy pathology in the medial septum/vertical DBB and NBM showed similar results, i.e. an increased pathology in both regions in patients with Parkinson’s disease dementia compared with non-demented patients [medial septum/vertical DBB, χ2(2) = 11.740, P = 0.003; NBM, χ2(2) = 12.959, P = 0.023]. The presence of Lewy neurites was detected in the midbrain (Fig. 6C and D), basal forebrain (Fig. 6G and H) and hippocampus (Fig. 6L) of diseased brains. A few Lewy neurites were observed in the basal forebrain of two control cases (Fig. 6E and F). There were no Lewy neurites in the hippocampus in control cases. In the non-demented Parkinson’s disease group, the Lewy neurite distribution was sparse overall and moderate to severe in the Parkinson’s disease dementia group (Fig. 6I). In the non-demented Parkinson’s disease group, Lewy neurites were exclusively restricted to the CA2 region, while in the Parkinson’s disease dementia group, the Lewy neurite pathology extended outside the CA2 region in two cases. The Lewy neurite score was significantly higher in the Parkinson’s disease dementia group compared with control and non-demented Parkinson’s disease groups (Fig. 6I; and cf. L with J and K).

Figure 7

Lewy bodies and Lewy neurites in the midbrain, basal forebrain and hippocampus. The number of Lewy bodies in the midbrain and basal forebrain (A and E) were counted in a ×20 microscopic field of view on two representative levels of each region and Lewy neurites rated semi-quantitatively in the basal forebrain (E) and hippocampus (I). A unified z-score representing the overall Lewy pathology (Lewy bodies and Lewy neurites) in the medial septum/vertical DBB and NBM was calculated to bring both measurements into the same numerical space, thereby providing a single measure of Lewy pathology in the basal forebrain (E). High-power photomicrographs show α-synuclein-immunoreactive Lewy bodies (arrows) and dystrophic Lewy neurites (arrowheads) in the midbrain (C and D) and in the basal forebrain (F–H), whereas the hippocampus contains Lewy neurites only (L). Lewy bodies in the midbrain appeared blue-black after nickel-cobalt intensification of DAB staining (C and D), whereas melanin pigment is seen as brown deposits (B–D). Scale bar = 100 µm. Data (A, E and I) are shown as individual data points, while the horizontal line indicates the median value. *P < 0.05; **P < 0.01 different from control, #P < 0.05; ##P < 0.01 different from Parkinson’s disease without dementia. (A) Kruskal Wallis, χ2(2) = 10.405, P = 0.0004, followed by uncorrected Mann Whitney post hoc test. (E) Kruskal Wallis, χ2(2) = 12.972, P = 0.002, followed by uncorrected Mann Whitney post hoc test. (I) Kruskal Wallis, χ2(2) = 12.378, P = 0.00003 followed by uncorrected Mann Whitney post hoc test. LB = Lewy bodies; LN = Lewy neurites; PD = Parkinson’s disease; PDD = Parkinson’s disease dementia.

Figure 7

Lewy bodies and Lewy neurites in the midbrain, basal forebrain and hippocampus. The number of Lewy bodies in the midbrain and basal forebrain (A and E) were counted in a ×20 microscopic field of view on two representative levels of each region and Lewy neurites rated semi-quantitatively in the basal forebrain (E) and hippocampus (I). A unified z-score representing the overall Lewy pathology (Lewy bodies and Lewy neurites) in the medial septum/vertical DBB and NBM was calculated to bring both measurements into the same numerical space, thereby providing a single measure of Lewy pathology in the basal forebrain (E). High-power photomicrographs show α-synuclein-immunoreactive Lewy bodies (arrows) and dystrophic Lewy neurites (arrowheads) in the midbrain (C and D) and in the basal forebrain (F–H), whereas the hippocampus contains Lewy neurites only (L). Lewy bodies in the midbrain appeared blue-black after nickel-cobalt intensification of DAB staining (C and D), whereas melanin pigment is seen as brown deposits (B–D). Scale bar = 100 µm. Data (A, E and I) are shown as individual data points, while the horizontal line indicates the median value. *P < 0.05; **P < 0.01 different from control, #P < 0.05; ##P < 0.01 different from Parkinson’s disease without dementia. (A) Kruskal Wallis, χ2(2) = 10.405, P = 0.0004, followed by uncorrected Mann Whitney post hoc test. (E) Kruskal Wallis, χ2(2) = 12.972, P = 0.002, followed by uncorrected Mann Whitney post hoc test. (I) Kruskal Wallis, χ2(2) = 12.378, P = 0.00003 followed by uncorrected Mann Whitney post hoc test. LB = Lewy bodies; LN = Lewy neurites; PD = Parkinson’s disease; PDD = Parkinson’s disease dementia.

Discussion

In an effort to understand how dementia emerges in Parkinson’s disease, the present study investigated whether co-existing structural and functional deficits in dopaminergic and cholinergic hippocampal and/or neocortical projecting neurons could discriminate demented from non-demented patients with Parkinson’s disease and age-matched control subjects—a hypothesis tested in animal models (Wisman et al., 2008) but not yet directly addressed in human post-mortem brains. Indeed, although a few studies have investigated changes in dopaminergic and cholinergic systems in patients with Parkinson’s disease with dementia, they have focused on functional or structural changes in selective parts of the dopaminergic and cholinergic system, namely the nigrostriatal dopaminergic and basocortical cholinergic projections. Furthermore, by excluding the mesocorticolimbic dopaminergic and septohippocampal cholinergic pathways, these studies seem to provide incomplete evidence pertaining to the nature of the neurotransmitter deficits seen in these patients (Gaspar and Gray, 1984; Hilker et al., 2005; Klein et al., 2010). In contrast, we analysed the full extent of the basal forebrain cholinergic and midbrain dopaminergic nuclei and quantified specific dopaminergic and cholinergic nuclei in brain tissue obtained from characterized patients with Parkinson’s disease with and without dementia. This task is normally complicated by the rare availability of complete sets of post-mortem tissue from such patients whose cognitive functions have been prospectively assessed, and by the presence of overlapping pathologies in many patients with Parkinson’s disease with dementia. For the present study, all cases were screened to exclude any significant degree of Alzheimer-type or other neurodegenerative and cerebrovascular pathologies, and were matched to include controls with a similar degree of age-related pathology, as population-based autopsy series show that 50% of people >50 years of age have hippocampal tangle formation and 50% of people >70 years-old have mild neocortical amyloid plaque deposition (Braak and Braak, 1995). By processing sections from rarely available bilaterally preserved tissue blocks, we were able to perform reliable and complete stereological analyses of midline cholinergic and dopaminergic structures in these well characterized cases without the confound of substantial additional Alzheimer-type pathology. The data obtained from this material show that in patients with Parkinson’s disease dementia there are additional dopaminergic deficits and more substantial cholinergic deficits compared to non-demented patients with Parkinson’s disease (Table 3).

Table 3

Pathological correlates of dementia in Parkinson’s disease

 Controls Parkinson’s disease Parkinson’s disease dementia 
Age related pathologies 
    Amyloid plaques 
    Hippocampal tangles 
Dopaminergic systems 
    Midbrain LB Reference group 
    A9 neuron loss ++ ++ 
    A10 lateral VTA loss None 
Cholinergic systems 
    Basal forebrain LB Reference group ++ 
    Hippocampal LN ++ 
    Neocortical ChAT activity loss 
    Hippocampal ChAT activity loss None 
    Ch4 neuron loss None 
 Controls Parkinson’s disease Parkinson’s disease dementia 
Age related pathologies 
    Amyloid plaques 
    Hippocampal tangles 
Dopaminergic systems 
    Midbrain LB Reference group 
    A9 neuron loss ++ ++ 
    A10 lateral VTA loss None 
Cholinergic systems 
    Basal forebrain LB Reference group ++ 
    Hippocampal LN ++ 
    Neocortical ChAT activity loss 
    Hippocampal ChAT activity loss None 
    Ch4 neuron loss None 

LB = Lewy bodies; LN = Lewy neurites.

+ indicates presence of pathology; ++ indicates more severe state of pathology.

Although the involvement of the dopaminergic and cholinergic systems has long been suspected in the pathogenesis of Parkinson’s disease dementia (Dubois et al., 1983; Gaspar and Gray, 1984; Perry et al., 1985; Ito et al., 2002), the exact contribution of each pathway is still subject to debate. Somewhat conflicting, PET and SPECT studies measuring striatal dopaminergic hypofunction suggest that the disruption of the nigrostriatal pathway may to some extent mediate the development of cognitive impairment in Parkinson’s disease (Jokinen et al., 2009; O'Brien et al., 2009), whereas other studies show that striatal dopamine uptake cannot discriminate patients with Parkinson disease dementia from non-demented patients (Hilker et al., 2005; Klein et al., 2010). Histopathological studies seem to corroborate that dementia in Parkinson’s disease is independent of the severity of dopamine loss in the nigrostriatal system (Gaspar and Gray, 1984). In-line with the latter studies, it is interesting to note that the severe loss of A9 neurons reported in our cases is similar across all Parkinson’s disease groups, regardless of the dementia diagnosis, excluding the substantia nigra neuronal loss as the main factor driving the development of dementia.

The A10 VTA cell groups are usually considered as a whole and thought to be primarily located medial to the substantia nigra. However, the A10 dopaminergic cell group is more heterogeneous, particularly in humans (Halliday and Tork, 1986; McRitchie et al., 1996), with a well-recognized dorsolateral VTA region where the dopaminergic neurons display different electrophysiological properties to their better-known medial VTA counterparts (Di Salvio et al., 2010; Zhang et al., 2010). Projections from these two VTA regions also differ significantly (Ikemoto, 2007; Ferreira et al., 2008), including their dopaminergic innervation of the cholinergic basal forebrain; the lateral VTA innervates the Ch4 NBM, while the medial VTA innervates cholinergic neurons in the diagonal band (Gaykema and Zaborszky, 1996). Lateral A10 VTA neurons also send projections to the nucleus accumbens, hippocampus, amygdala as well as to the prefrontal cortex (Swanson, 1982; Oades and Halliday, 1987; Gasbarri et al., 1994).

In the present study we found that A10 dopaminergic neurons located in the lateral VTA specifically degenerated in Parkinson’s disease with dementia. In addition, we found that only a certain type of lateral A10 dopaminergic neuron degenerated—those not containing melanin. Most previous research on Parkinson’s disease has suggested that the dopaminergic neurons most vulnerable to Parkinson’s disease are those that are melanized (Hirsch et al., 1988; Double and Halliday, 2006) and that such cells initially lose the tyrosine hydroxylase-immunoreactive prior to degeneration (Milber et al., 2012). Our results support the concept that tyrosine hydroxylase-negative melanin-positive neurons are targeted in Parkinson’s disease, but also suggest that a possibly independent mechanism occurs in Parkinson’s disease dementia to cause degeneration of a different dopaminergic cell type in a different midbrain region. Only one other study to date has assessed the degree of degeneration in different regions of the A10 VTA in patients with Parkinson’s disease, and five of the seven Parkinson’s disease cases evaluated in that study had dementia (McRitchie et al., 1997). The authors found a similar loss of lateral A10 VTA dopaminergic neurons and a similar preservation of most medial A10 VTA dopaminergic neurons in such patients (McRitchie et al., 1997). Triple labelling studies in owl monkeys show that the lateral A10 VTA region contains substantial cortical projections and that single neurons innervate different cortical motor regions within the frontal lobe (Gaspar et al., 1992). Post-mortem studies have shown a selective loss of all dopaminergic innervation within the upper layers of the cortex in patients with end-stage Parkinson’s disease (Gaspar et al., 1991), which may correspond to the loss of lateral A10 dopaminergic VTA neurons we have observed in our patients with Parkinson’s disease dementia. In support of this concept, in vivo imaging studies confirm that the dopaminergic mesocortical projections are affected in Parkinson’s disease dementia (Ito et al., 2002).

The loss of cortical cholinergic function in the absence of neuronal degeneration in the basal forebrain is well documented in patients with mild cognitive impairment and early stages of Alzheimer’s disease (Schliebs and Arendt, 2011). In the present study, non-demented patients with Parkinson’s disease had reduced neocortical ChAT activity without degeneration of neurons in the Ch4 NBM, indicating significant functional impairment in cortical cholinergic pathways that may remain structurally relatively intact. We believe that such dysfunction in the absence of neuronal loss can be attributed to the presence of α-synuclein deposition, as we observed α-synuclein-immunopositive inclusions in basal forebrain cholinergic neurons in matching patient brains. It has previously been shown that α-synuclein inclusions in basal forebrain cholinergic neurons sequester ChAT from the cytoplasm, with the suggestion that the abnormal accumulation of α-synuclein in such neurons would decrease their neurotransmitter production (Dugger and Dickson, 2010). Likewise, wild-type mice with long-term dopamine depletion have reduced ChAT expression in the absence of cholinergic cell loss (Szego et al., 2011). As we did not observe significant degeneration in any A10 VTA regions that innervate the cholinergic basal forebrain in non-demented patients with Parkinson’s disease, our data suggest that α-synuclein deposition in the basal forebrain is the main factor affecting the cholinergic system in such patients, and not a loss of the dopamine innervation to this region. It is of interest to note that reduced cortical ChAT in non-demented patients with Parkinson’s disease has recently been shown to relate to the presence of olfactory dysfunction (Bohnen et al., 2010) and rapid eye movement sleep behaviour disorder (Kotagal et al., 2012), features known to occur early and before the onset of Parkinson’s disease dementia (Postuma et al., 2012).

More substantial cholinergic deficits are known to occur in patients with Parkinson’s disease dementia (Bohnen and Albin, 2011). In particular, a severe loss of cholinergic basal forebrain neurons has been reported to differentiate Parkinson’s disease dementia from non-demented patients with Parkinson’s disease (Gaspar and Gray, 1984). In-line with this, our study shows frank neuronal loss of Ch4 neurons in demented patients with Parkinson’s disease only, whereas hippocampal ChAT deficits and α-synuclein deposition in the basal forebrain were more severe in demented than non-demented patients with Parkinson’s disease. Patients with Parkinson’s disease dementia also have significantly increased cortical α-synuclein deposition (Hurtig et al., 2000; Harding and Halliday, 2001), which may impact on the viability of the Ch4 neurons, as it has been shown that widespread overexpression of α-synuclein in dopamine depleted transgenic animals induces Ch4 cholinergic cell loss (Szego et al., 2011). As there was also a loss of the lateral A10 VTA dopamine neurons that innervate Ch4 neurons in patients with Parkinson’s disease dementia only, we suggest that the loss of these A10 dopamine neurons may be a retrograde degenerative process.

In addition to the Ch4 cholinergic system, we were able to study the more anteriorly located Ch1 and Ch2 cells in the septal region, cells that provide the cholinergic innervation to limbic regions including the hippocampus (Mesulam et al., 1983). We found that the cholinergic cell loss observed in Parkinson’s disease dementia appeared restricted to the cortically projecting Ch4 neurons, while the Ch1/2 neurons remained structurally intact. Closer inspection of the cholinergic cell counts in the Ch1/2 cell groups in both Parkinson’s disease with and without dementia groups revealed a variation not seen in the control group. However, no significant loss of cholinergic neurons was observed in either of the two patient groups. This increased variation in Ch1/2 neuronal numbers may relate to our findings of a reduction in hippocampal ChAT activity and increased α-synuclein deposition in the CA2 region of the hippocampus in Parkinson’s disease dementia. Reduced hippocampal ChAT activity (Mattila et al., 2001) and prominent α-synuclein pathology in the CA2 sector of the hippocampus (Irwin et al., 2012) have previously been associated with cognitive deficits in Parkinson’s disease. As discussed above, when cortical ChAT activity was reduced without degeneration of Ch4 neurons, the α-synuclein pathology in the basal forebrain was more severe in cases with Parkinson’s disease dementia compared with non-demented cases with Parkinson’s disease, confirming previous studies correlating the extent of α-synuclein pathology to dementia in Parkinson’s disease (Aarsland et al., 2005; Braak et al., 2005). It is therefore striking that the number of corresponding Ch1/2 cholinergic cells remained seemingly unaltered. In fact, the lack of frank cell loss in the Ch1/2 nuclei in the presence of severe pathology in that region parallels what is seen in the dopaminergic terminals projecting to the ventral striatum. In contrast with the dorsal striatum (caudate and putamen) in which most of the dopaminergic terminals have degenerated at end-stage, the ventral striatum is much less dopamine depleted (Goldstein et al., 1982; Nyberg et al., 1983), despite the presence of α-synuclein pathology. Our observations in the cases with Parkinson’s disease dementia are consistent with α-synuclein deposition inducing cholinergic dysfunction in the hippocampal Ch1/2 system in these patients. Such hippocampal dysfunction seems more related to Parkinson’s disease dementia than the loss of cortical cholinergic function. Studies in rodents have shown that increased α-synuclein levels could lead to behavioural deficits without inducing obvious cell loss and suggested that neuronal dysfunction, rather than frank neurodegeneration, could be a pivotal mechanism in the toxic effects of α-synuclein (Masliah et al., 2000; Giasson et al., 2002; Maingay et al., 2006). In contrast, a depletion of cholinergic neurons in the Ch1/2 system has been reported in dementia with Lewy bodies (Fujishiro et al., 2006), possibly due to the retrograde degeneration of septal neurons following the selective loss of hippocampal neurons that occurs in this Lewy body syndrome compared with Parkinson’s disease (Harding et al., 2002).

To our knowledge, this study constitutes the first attempt at quantifying simultaneously cholinergic as well as nigral and non-nigral dopaminergic neurons in the same subset of brains obtained from patients diagnosed with Parkinson’s disease and Parkinson’s disease dementia. Taken together, our results from combined structural and functional analysis of cholinergic and dopaminergic systems suggest that the neuropathological substrate of Parkinson’s disease dementia might lie in both the cholinergic and non-nigral dopaminergic systems. Interestingly, although the minor reductions in lateral A10 VTA dopaminergic neurons and Ch4 NBM cholinergic neurons in the non-demented Parkinson’s disease group were not significant, the variation observed in the data (with cases having near-normal cell numbers and others having low numbers closer to those observed in cases with Parkinson’s disease dementia, in one and/or the other cell population) suggest that insidious cholinergic and non-nigral dopaminergic deficiencies may precede dementia in Parkinson’s disease. It is plausible that non-demented patients with Parkinson’s disease might have presented with symptoms that were not severe enough to fulfil the dementia criteria but might have had mild cognitive impairment preceding the onset of dementia in Parkinson’s disease. Although this is an interesting hypothesis, the design of our study and the available data in this cohort are not suitable for us to definitively establish such a link.

Our study was limited by small numbers of cases, and therefore the results should ideally be replicated in another set of cases or with other methods of analysis, for example a multi-tracer PET imaging study in a prospectively followed cohort of cases. The particular design of our study required material in which midline brain structures were entirely preserved, which is difficult to obtain since most brain banks cut the brains in half to freeze one side and fix the other. Thus very few longitudinally studied cases with limited age-related pathologies are available where the whole hemisphere has been fixed so that midline structures can be quantified with accuracy. Despite these small numbers, our study shows that Ch4 cholinergic neurons in the NBM and non-nigral A10 dopaminergic neurons are affected in Parkinson’s disease dementia, whereas they remain intact in non-demented patients with Parkinson’s disease. These findings corroborate the hypothesis that both systems may contribute to the emergence of dementia in Parkinson’s disease. However, they cannot determine whether both systems are affected sequentially or whether they may have independent effects on discrete cognitive functions. Our data might also have implications for the treatment of dementia in Parkinson’s disease, for which there are currently very few therapeutic options. Although dopaminergic replacement therapies have shown beneficial effects on some of the cognitive symptoms in Parkinson’s disease (Lange et al., 1992; Beato et al., 2008), none of these studies have reported beneficial effects in fully demented patients with Parkinson’s disease. They confirm that a dopaminergic dysfunction is likely involved in the initial cognitive impairment observed prior to Parkinson’s disease dementia. Interestingly, the only drug approved for the symptomatic treatment of dementia in Parkinson’s disease is an acetylcholine esterase inhibitor, rivastigmine, which however shows only moderate efficacy (Emre et al., 2004; Reingold et al., 2007). Our data suggest that both dopaminergic and cholinergic postsynaptic medications may prove promising in those patients without concomitant Alzheimer-type pathology, pathology that can now be assessed using PET imaging. Finally, our data also highlight increasing α-synuclein aggregation as a possible pathological substrate for the dysfunction and cell loss giving rise to Parkinson’s disease dementia, and suggest that the septohippocampal pathway might be a key area associated with the final development of dementia in Parkinson’s disease.

Acknowledgements

Human brain tissue was received from the Sydney Brain Bank, which is supported by Neuroscience Research Australia, the University of New South Wales and the National Health and Medical Research Council of Australia. The authors wish to acknowledge Ulla Samuelsson for technical support.

Funding

This study was supported by grants from the European Research Council (TreatPD 242932), the Swedish Research Council (2008-3092, 2009-2318) and the Swedish Foundation for Strategic Research (Parkinson’s models for translational research). Glenda Halliday is a Senior Principal Research Fellow of the National Health and Medical Research Council of Australia. Lachlan Thompson holds a Career Development Fellowship and Chris Bye a Peter Doherty Fellowship – both through the National Health and Medical Research Council of Australia.

Abbreviations

    Abbreviations
  • ChAT

    choline acetyltransferase

  • DBB

    diagonal band of Broca complex

  • NBM

    nucleus basalis of Meynert

  • VTA

    ventral tegmental area

References

Aarsland
D
Kurz
MW
The epidemiology of dementia associated with Parkinson disease
J Neurol Sci
 , 
2010
, vol. 
289
 (pg. 
18
-
22
)
Aarsland
D
Perry
R
Brown
A
Larsen
JP
Ballard
C
Neuropathology of dementia in Parkinson's disease: a prospective, community-based study
Ann Neurol
 , 
2005
, vol. 
58
 (pg. 
773
-
6
)
Beato
R
Levy
R
Pillon
B
Vidal
C
du Montcel
ST
Deweer
B
, et al.  . 
Working memory in Parkinson's disease patients: clinical features and response to levodopa
Arq Neuropsiquiatr
 , 
2008
, vol. 
66
 (pg. 
147
-
51
)
Bedard
MA
Pillon
B
Dubois
B
Duchesne
N
Masson
H
Agid
Y
Acute and long-term administration of anticholinergics in Parkinson's disease: specific effects on the subcortico-frontal syndrome
Brain Cogn
 , 
1999
, vol. 
40
 (pg. 
289
-
313
)
Bohnen
NI
Albin
RL
The cholinergic system and Parkinson disease
Behav Brain Res
 , 
2011
, vol. 
221
 (pg. 
564
-
73
)
Bohnen
NI
Muller
ML
Kotagal
V
Koeppe
RA
Kilbourn
MA
Albin
RL
, et al.  . 
Olfactory dysfunction, central cholinergic integrity and cognitive impairment in Parkinson's disease
Brain
 , 
2010
, vol. 
133
 
Pt 6
(pg. 
1747
-
54
)
Braak
H
Braak
E
Staging of Alzheimer's disease-related neurofibrillary changes
Neurobiol Aging
 , 
1995
, vol. 
16
 (pg. 
271
-
8
discussion 278–84
Braak
H
Rub
U
Jansen Steur
EN
Del Tredici
K
de Vos
RA
Cognitive status correlates with neuropathologic stage in Parkinson disease
Neurology
 , 
2005
, vol. 
64
 (pg. 
1404
-
10
)
Di Salvio
M
Di Giovannantonio
LG
Omodei
D
Acampora
D
Simeone
A
Otx2 expression is restricted to dopaminergic neurons of the ventral tegmental area in the adult brain
Int J Dev Biol
 , 
2010
, vol. 
54
 (pg. 
939
-
45
)
Double
KL
Halliday
GM
New face of neuromelanin
J Neural Transm Suppl
 , 
2006
, vol. 
70
 (pg. 
119
-
23
)
Dubois
B
Ruberg
M
Javoy-Agid
F
Ploska
A
Agid
Y
A subcortico-cortical cholinergic system is affected in Parkinson's disease
Brain Res
 , 
1983
, vol. 
288
 (pg. 
213
-
8
)
Dugger
BN
Dickson
DW
Cell type specific sequestration of choline acetyltransferase and tyrosine hydroxylase within Lewy bodies
Acta Neuropathol
 , 
2010
, vol. 
120
 (pg. 
633
-
9
)
Emre
M
Aarsland
D
Albanese
A
Byrne
EJ
Deuschl
G
De Deyn
PP
, et al.  . 
Rivastigmine for dementia associated with Parkinson's disease
N Engl J Med
 , 
2004
, vol. 
351
 (pg. 
2509
-
18
)
Fahn
S
Libsch
LR
Cutler
RW
Monoamines in the human neostriatum: topographic distribution in normals and in Parkinson's disease and their role in akinesia, rigidity, chorea, and tremor
J Neurol Sci
 , 
1971
, vol. 
14
 (pg. 
427
-
55
)
Ferreira
JG
Del-Fava
F
Hasue
RH
Shammah-Lagnado
SJ
Organization of ventral tegmental area projections to the ventral tegmental area-nigral complex in the rat
Neuroscience
 , 
2008
, vol. 
153
 (pg. 
196
-
213
)
Fonnum
F
A rapid radiochemical method for the determination of choline acetyltransferase
J Neurochem
 , 
1975
, vol. 
24
 (pg. 
407
-
9
)
Fujishiro
H
Umegaki
H
Isojima
D
Akatsu
H
Iguchi
A
Kosaka
K
Depletion of cholinergic neurons in the nucleus of the medial septum and the vertical limb of the diagonal band in dementia with Lewy bodies
Acta Neuropathol
 , 
2006
, vol. 
111
 (pg. 
109
-
14
)
Gasbarri
A
Packard
MG
Campana
E
Pacitti
C
Anterograde and retrograde tracing of projections from the ventral tegmental area to the hippocampal formation in the rat
Brain Res Bull
 , 
1994
, vol. 
33
 (pg. 
445
-
52
)
Gaspar
P
Gray
F
Dementia in idiopathic Parkinson's disease. A neuropathological study of 32 cases
Acta Neuropathol
 , 
1984
, vol. 
64
 (pg. 
43
-
52
)
Gaspar
P
Stepniewska
I
Kaas
JH
Topography and collateralization of the dopaminergic projections to motor and lateral prefrontal cortex in owl monkeys
J Comp Neurol
 , 
1992
, vol. 
325
 (pg. 
1
-
21
)
Gaspar
P
Duyckaerts
C
Alvarez
C
Javoy-Agid
F
Berger
B
Alterations of dopaminergic and noradrenergic innervations in motor cortex in Parkinson's disease
Ann Neurol
 , 
1991
, vol. 
30
 (pg. 
365
-
74
)
Gaykema
RP
Zaborszky
L
Direct catecholaminergic-cholinergic interactions in the basal forebrain. II. Substantia nigra-ventral tegmental area projections to cholinergic neurons
J Comp Neurol
 , 
1996
, vol. 
374
 (pg. 
555
-
77
)
Giasson
BI
Duda
JE
Quinn
SM
Zhang
B
Trojanowski
JQ
Lee
VM
Neuronal alpha-synucleinopathy with severe movement disorder in mice expressing A53T human alpha-synuclein
Neuron
 , 
2002
, vol. 
34
 (pg. 
521
-
33
)
Goldstein
M
Lieberman
A
Pearson
J
Relatively high levels of dopamine in nucleus accumbens of levodopa treated patients with Parkinson's disease
J Neural Transm
 , 
1982
, vol. 
54
 (pg. 
129
-
34
)
Gundersen
HJ
Jensen
EB
The efficiency of systematic sampling in stereology and its prediction
J Microsc
 , 
1987
, vol. 
147
 
Pt 3
(pg. 
229
-
63
)
Hall
H
Jewett
M
Landeck
N
Nilsson
N
Schagerlof
U
Leanza
G
, et al.  . 
Characterization of cognitive deficits in rats overexpressing human alpha-synuclein in the ventral tegmental area and medial septum using recombinant adeno-associated viral vectors
PLoS One
 , 
2013
, vol. 
8
 pg. 
e64844
 
Halliday
GM
Tork
I
Comparative anatomy of the ventromedial mesencephalic tegmentum in the rat, cat, monkey and human
J Comp Neurol
 , 
1986
, vol. 
252
 (pg. 
423
-
45
)
Halliday
G
Ng
T
Rodriguez
M
Harding
A
Blumbergs
P
Evans
W
, et al.  . 
Consensus neuropathological diagnosis of common dementia syndromes: testing and standardising the use of multiple diagnostic criteria
Acta Neuropathol
 , 
2002
, vol. 
104
 (pg. 
72
-
8
)
Harding
AJ
Halliday
GM
Cortical Lewy body pathology in the diagnosis of dementia
Acta Neuropathol
 , 
2001
, vol. 
102
 (pg. 
355
-
63
)
Harding
AJ
Lakay
B
Halliday
GM
Selective hippocampal neuron loss in dementia with Lewy bodies
Ann Neurol
 , 
2002
, vol. 
51
 (pg. 
125
-
8
)
Hilker
R
Thomas
AV
Klein
JC
Weisenbach
S
Kalbe
E
Burghaus
L
, et al.  . 
Dementia in Parkinson disease: functional imaging of cholinergic and dopaminergic pathways
Neurology
 , 
2005
, vol. 
65
 (pg. 
1716
-
22
)
Hirsch
E
Graybiel
AM
Agid
YA
Melanized dopaminergic neurons are differentially susceptible to degeneration in Parkinson's disease
Nature
 , 
1988
, vol. 
334
 (pg. 
345
-
8
)
Hughes
AJ
Ben-Shlomo
Y
Daniel
SE
Lees
AJ
What features improve the accuracy of clinical diagnosis in Parkinson's disease: a clinicopathologic study
Neurology
 , 
1992
, vol. 
42
 (pg. 
1142
-
6
)
Hughes
AJ
Daniel
SE
Blankson
S
Lees
AJ
A clinicopathologic study of 100 cases of Parkinson's disease
Arch Neurol
 , 
1993
, vol. 
50
 (pg. 
140
-
8
)
Hurtig
HI
Trojanowski
JQ
Galvin
J
Ewbank
D
Schmidt
ML
Lee
VM
, et al.  . 
Alpha-synuclein cortical Lewy bodies correlate with dementia in Parkinson's disease
Neurology
 , 
2000
, vol. 
54
 (pg. 
1916
-
21
)
Hyman
BT
Phelps
CH
Beach
TG
Bigio
EH
Cairns
NJ
Carrillo
MC
, et al.  . 
National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease
Alzheimers Dement
 , 
2012
, vol. 
8
 (pg. 
1
-
13
)
Ikemoto
S
Dopamine reward circuitry: two projection systems from the ventral midbrain to the nucleus accumbens-olfactory tubercle complex
Brain Res Rev
 , 
2007
, vol. 
56
 (pg. 
27
-
78
)
Irwin
DJ
Lee
VM
Trojanowski
JQ
Parkinson's disease dementia: convergence of alpha-synuclein, tau and amyloid-beta pathologies
Nat Rev Neurosci
 , 
2013
, vol. 
14
 (pg. 
626
-
36
)
Irwin
DJ
White
MT
Toledo
JB
Xie
SX
Robinson
JL
Van Deerlin
V
, et al.  . 
Neuropathologic substrates of Parkinson disease dementia
Ann Neurol
 , 
2012
, vol. 
72
 (pg. 
587
-
98
)
Ito
K
Nagano-Saito
A
Kato
T
Arahata
Y
Nakamura
A
Kawasumi
Y
, et al.  . 
Striatal and extrastriatal dysfunction in Parkinson's disease with dementia: a 6-[18F]fluoro-L-dopa PET study
Brain
 , 
2002
, vol. 
125
 
Pt 6
(pg. 
1358
-
65
)
Jellinger
KA
Pathology of Parkinson's disease. Changes other than the nigrostriatal pathway
Mol Chem Neuropathol
 , 
1991
, vol. 
14
 (pg. 
153
-
97
)
Jokinen
P
Bruck
A
Aalto
S
Forsback
S
Parkkola
R
Rinne
JO
Impaired cognitive performance in Parkinson's disease is related to caudate dopaminergic hypofunction and hippocampal atrophy
Parkinsonism Relat Disord
 , 
2009
, vol. 
15
 (pg. 
88
-
93
)
Kehagia
AA
Barker
RA
Robbins
TW
Neuropsychological and clinical heterogeneity of cognitive impairment and dementia in patients with Parkinson's disease
Lancet Neurol
 , 
2010
, vol. 
9
 (pg. 
1200
-
13
)
Kehagia
AA
Barker
RA
Robbins
TW
Cognitive impairment in Parkinson's disease: the dual syndrome hypothesis
Neurodegener Dis
 , 
2013
, vol. 
11
 (pg. 
79
-
92
)
Klein
JC
Eggers
C
Kalbe
E
Weisenbach
S
Hohmann
C
Vollmar
S
, et al.  . 
Neurotransmitter changes in dementia with Lewy bodies and Parkinson disease dementia in vivo
Neurology
 , 
2010
, vol. 
74
 (pg. 
885
-
92
)
Kotagal
V
Albin
RL
Muller
ML
Koeppe
RA
Chervin
RD
Frey
KA
, et al.  . 
Symptoms of rapid eye movement sleep behavior disorder are associated with cholinergic denervation in Parkinson disease
Ann Neurol
 , 
2012
, vol. 
71
 (pg. 
560
-
8
)
Lange
KW
Robbins
TW
Marsden
CD
James
M
Owen
AM
Paul
GM
L-dopa withdrawal in Parkinson's disease selectively impairs cognitive performance in tests sensitive to frontal lobe dysfunction
Psychopharmacology
 , 
1992
, vol. 
107
 (pg. 
394
-
404
)
Maingay
M
Romero-Ramos
M
Carta
M
Kirik
D
Ventral tegmental area dopamine neurons are resistant to human mutant alpha-synuclein overexpression
Neurobiol Dis
 , 
2006
, vol. 
23
 (pg. 
522
-
32
)
Masliah
E
Rockenstein
E
Veinbergs
I
Mallory
M
Hashimoto
M
Takeda
A
, et al.  . 
Dopaminergic loss and inclusion body formation in alpha-synuclein mice: implications for neurodegenerative disorders
Science
 , 
2000
, vol. 
287
 (pg. 
1265
-
9
)
Mattay
VS
Tessitore
A
Callicott
JH
Bertolino
A
Goldberg
TE
Chase
TN
, et al.  . 
Dopaminergic modulation of cortical function in patients with Parkinson's disease
Ann Neurol
 , 
2002
, vol. 
51
 (pg. 
156
-
64
)
Mattila
PM
Roytta
M
Lonnberg
P
Marjamaki
P
Helenius
H
Rinne
JO
Choline acetytransferase activity and striatal dopamine receptors in Parkinson's disease in relation to cognitive impairment
Acta Neuropathol
 , 
2001
, vol. 
102
 (pg. 
160
-
6
)
McRitchie
DA
Hardman
CD
Halliday
GM
Cytoarchitectural distribution of calcium binding proteins in midbrain dopaminergic regions of rats and humans
J Comp Neurol
 , 
1996
, vol. 
364
 (pg. 
121
-
50
)
McRitchie
DA
Cartwright
HR
Halliday
GM
Specific A10 dopaminergic nuclei in the midbrain degenerate in Parkinson's disease
Exp Neurol
 , 
1997
, vol. 
144
 (pg. 
202
-
13
)
Mesulam
MM
Mufson
EJ
Levey
AI
Wainer
BH
Cholinergic innervation of cortex by the basal forebrain: cytochemistry and cortical connections of the septal area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey
J Comp Neurol
 , 
1983
, vol. 
214
 (pg. 
170
-
97
)
Milber
JM
Noorigian
JV
Morley
JF
Petrovitch
H
White
L
Ross
GW
Duda
JE
Lewy pathology is not the first sign of degeneration in vulnerable neurons in Parkinson disease
Neurology
 , 
2012
, vol. 
79
 (pg. 
2307
-
14
)
Montine
TJ
Phelps
CH
Beach
TG
Bigio
EH
Cairns
NJ
Dickson
DW
, et al.  . 
National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease: a practical approach
Acta Neuropathol
 , 
2012
, vol. 
123
 (pg. 
1
-
11
)
Morris
JC
Clinical dementia rating: a reliable and valid diagnostic and staging measure for dementia of the Alzheimer type
Int Psychogeriatr
 , 
1997
, vol. 
9
 
Suppl 1
(pg. 
173
-
6
discussion 177–8
Mufson
EJ
Presley
LN
Kordower
JH
Nerve growth factor receptor immunoreactivity within the nucleus basalis (Ch4) in Parkinson's disease: reduced cell numbers and co-localization with cholinergic neurons
Brain Res
 , 
1991
, vol. 
539
 (pg. 
19
-
30
)
Nakano
I
Hirano
A
Parkinson's disease: neuron loss in the nucleus basalis without concomitant Alzheimer's disease
Ann Neurol
 , 
1984
, vol. 
15
 (pg. 
415
-
8
)
Nyberg
P
Nordberg
A
Wester
P
Winblad
B
Dopaminergic deficiency is more pronounced in putamen than in nucleus caudatus in Parkinson’s disease
Neurochem Pathol
 , 
1983
, vol. 
1
 (pg. 
193
-
202
)
O'Brien
JT
McKeith
IG
Walker
Z
Tatsch
K
Booij
J
Darcourt
J
, et al.  . 
Diagnostic accuracy of 123I-FP-CIT SPECT in possible dementia with Lewy bodies
Br J Psychiatry
 , 
2009
, vol. 
194
 (pg. 
34
-
9
)
Oades
RD
Halliday
GM
Ventral tegmental (A10) system: neurobiology. 1. Anatomy and connectivity
Brain Res
 , 
1987
, vol. 
434
 (pg. 
117
-
65
)
Perry
EK
Curtis
M
Dick
DJ
Candy
JM
Atack
JR
Bloxham
CA
, et al.  . 
Cholinergic correlates of cognitive impairment in Parkinson's disease: comparisons with Alzheimer's disease
J Neurol Neurosurg Psychiatry
 , 
1985
, vol. 
48
 (pg. 
413
-
21
)
Postuma
RB
Aarsland
D
Barone
P
Burn
DJ
Hawkes
CH
Oertel
W
, et al.  . 
Identifying prodromal Parkinson's disease: pre-motor disorders in Parkinson's disease
Mov Disord
 , 
2012
, vol. 
27
 (pg. 
617
-
26
)
Reingold
JL
Morgan
JC
Sethi
KD
Rivastigmine for the treatment of dementia associated with Parkinson's disease
Neuropsychiatr Dis Treat
 , 
2007
, vol. 
3
 (pg. 
775
-
83
)
Schliebs
R
Arendt
T
The cholinergic system in aging and neuronal degeneration
Behav Brain Res
 , 
2011
, vol. 
221
 (pg. 
555
-
63
)
Schmitz
C
Hof
PR
Design-based stereology in neuroscience
Neuroscience
 , 
2005
, vol. 
130
 (pg. 
813
-
31
)
Swanson
LW
The projections of the ventral tegmental area and adjacent regions: a combined fluorescent retrograde tracer and immunofluorescence study in the rat
Brain Res Bull
 , 
1982
, vol. 
9
 (pg. 
321
-
53
)
Szego
EM
Gerhardt
E
Outeiro
TF
Kermer
P
Dopamine-depletion and increased alpha-synuclein load induce degeneration of cortical cholinergic fibers in mice
J Neurol Sci
 , 
2011
, vol. 
310
 (pg. 
90
-
5
)
West
MJ
Stereological methods for estimating the total number of neurons and synapses: issues of precision and bias
Trends Neurosci
 , 
1999
, vol. 
22
 (pg. 
51
-
61
)
Wisman
LA
Sahin
G
Maingay
M
Leanza
G
Kirik
D
Functional convergence of dopaminergic and cholinergic input is critical for hippocampus-dependent working memory
J Neurosci
 , 
2008
, vol. 
28
 (pg. 
7797
-
807
)
Zhang
TA
Placzek
AN
Dani
JA
In vitro identification and electrophysiological characterization of dopamine neurons in the ventral tegmental area
Neuropharmacology
 , 
2010
, vol. 
59
 (pg. 
431
-
6
)

Author notes

*These authors contributed equally to this work.