A role for the substantia nigra pars reticulata in the gaze palsy of progressive supranuclear palsy.

We examined the topography and degree of cell loss within basal ganglia structures commonly involved in progressive supranuclear palsy in order to identify any relationship between degeneration in these nuclei and gaze palsy. Serial section analyses and unbiased quantitative techniques were applied to brain tissue from six cases with progressive supranuclear palsy (four with gaze palsy and two without) and six controls with no neurological or neuropathological abnormalities. The total number of nucleolated neurons within the substantia nigra pars compacta (SNc) and reticulata (SNr), the subthalamic nucleus, and the internal and external segments of the globus pallidus was determined for all subjects and the data expressed as percentages of control values to compare degeneration across these basal ganglia structures. The density of neurofibrillary tangles was also evaluated within these structures. Despite significant subcortical neurofibrillary tangle formation in all cases, there was considerable variability in the degree of neuronal cell loss in all basal ganglia regions, except the SNc which was consistently affected. There was no correlation between the ranked density of neurofibrillary tangles and the degree of neuronal cell loss in any basal ganglia region. Comparisons between cases with and without gaze palsy revealed a 40% greater decrease in the number of SNr neurons in cases with gaze palsy (75 +/- 8% loss) compared with those without (35 +/- 14% loss). This was the largest difference between these cases. As the SNr projects to the superior colliculus, degeneration of this basal ganglia structure may disrupt eye movements in progressive supranuclear palsy.

The external globus pallidus in patients with Parkinson's disease and progressive supranuclear palsy.

Underactivity of the external segment of the globus pallidus is thought to contribute to the generation of parkinsonian hypokinetic symptoms in association with striatal dopaminergic dysfunction and overactivity of the subthalamus. These symptoms can be corrected by neurosurgical techniques aimed at normalizing subthalamic overactivity. The aim of the present study was to compare the amount of neurodegeneration and changes in the calcium-binding protein parvalbumin in the external segment of the globus pallidus in parkinsonian disorders. Cases with progressive supranuclear palsy were compared with cases with Parkinson's disease and control subjects. The number of neurones and neurofibrillary tangles was estimated using unbiased stereologic techniques. The external segment of the globus pallidus in Parkinson's disease was not significantly different from that in control subjects. In contrast, most patients with progressive supranuclear palsy had significant neurodegeneration of the external pallidum, particularly patients with significant degeneration of both the subthalamus and substantia nigra. These results suggest that the parkinsonian symptoms in progressive supranuclear palsy are caused by the degeneration of the external segment of the globus pallidus because such degeneration would increase thalamic inhibition through the basal ganglia output nuclei, particularly in patients with a loss of excitatory drive from the subthalamus.

The internal globus pallidus is affected in progressive supranuclear palsy and Parkinson's disease.

Our structural studies of the substantia nigra in parkinsonian patients identified previously unsuspected changes in the pars reticulata, suggesting significant dysfunction in this basal ganglia output. There have been few similar structural studies of the other major basal ganglia output, the internal segment of the globus pallidus. This is despite significant evidence that this basal ganglia region is crucially important for generating parkinsonian symptoms. In fact current surgical interventions target this region in Parkinson's disease. The cellular anatomy of the internal globus pallidus was compared among five controls, six patients with Parkinson's disease, and five patients with progressive supranuclear palsy. Neurons and pathological structures were quantified using the unbiased fractionator method. Only cases with progressive supranuclear palsy had detectable pathology within the internal globus pallidus in the form of tau-positive neuronal and glial tangles and substantial neurodegeneration. Cases with Parkinson's disease had a significant reduction in the proportion of neurons containing parvalbumin but were without significant neurodegeneration, consistent with dysfunction of both basal ganglia output nuclei in advanced parkinsonism. Surgical ablation of the internal globus pallidus for Parkinson's disease appears at odds with the significant neurodegeneration in the similarly akinetic and rigid patients with progressive supranuclear palsy. The results are discussed in association with current hypotheses of basal ganglia function and recent experimentation in patients undergoing pallidotomy for hyperkinetic disorders.

Serotonergic neurons in the brainstem of the wallaby, Macropus eugenii.

The organisation and cytoarchitecture of the serotonergic neurons in a diprotodont marsupial were examined by using serial sections of the brainstem processed for serotonin immunohistochemistry and routine histology. The topographic distribution of serotonergic neurons in the brainstem of the adult wallaby (Macropus eugenii) was similar to that of eutherian mammals. Serotonergic neurons were divided into rostral and caudal groups, separated by an oblique boundary through the pontomedullary junction. Approximately 52% of the serotonergic neurons in the wallaby brainstem were located in the rostral midline nuclei (caudal linear nucleus, dorsal, median, and pontine raphe nuclei and the interpeduncular nucleus), whereas 21% were found in the caudal midline region (nuclei raphe magnus, obscurus, and pallidus). The remaining serotonergic neurons (27%) were located in more lateral regions such as the pedunculopontine tegmental nuclei, the supralemniscal nuclei (B9 group), and the ventrolateral medulla. The largest serotonergic group, the dorsal raphe, contained one-third of the brainstem serotonergic neurons and showed five subdivisions, similar to that described in other species. In contrast, the median raphe did not show clear subdivisions. The internal complexity of the raphe nuclei and the degree of lateralisation of serotonergic neurons suggest that the wallaby serotonergic system is similar in organisation to that described for the cat and rabbit. This study supports the suggestion that the serotonergic system is evolutionally well conserved and provides baseline data for a quantitative study of serotonergic innervation of the developing cortex in the wallaby.

Progressive supranuclear palsy affects both the substantia nigra pars compacta and reticulata.

We have analyzed the neuropathology of the substantia nigra in four cases of progressive supranuclear palsy compared with age-matched controls and patients with Parkinson's disease. Although there are many reports of severe dopaminergic cell loss in progressive supranuclear palsy, the fate of the GABAergic pars reticulata neurones remains unclear. Serial section analysis and fractional counts of pars compacta neurones (identified by their neuromelanin pigment) and pars reticulata neurones (identified using parvalbumin immunohistochemistry) were performed, and the type and distribution of neuropathology were described. Severe neurodegeneration within the dopaminergic pars compacta was seen in all cases of progressive supranuclear palsy and all cases of Parkinson's disease compared with controls. Lewy body pathology was found only in cases of Parkinson's disease, while neurofibrillary tangles were seen only in cases of progressive supranuclear palsy. Tau-positive astrocytes and neuropil threads were occasionally seen in controls and cases of Parkinson's disease (particularly those of advanced age) but were extremely numerous in all cases of progressive supranuclear palsy. There was a similar decrease in parvalbumin immunoreactivity within the pars reticulata in both progressive supranuclear palsy and Parkinson's disease. However, there was a striking 70% reduction in the number of pars reticulata neurones in progressive supranuclear palsy, with no cell loss observed in Parkinson's disease compared with controls. Our results show that both the dopaminergic pars compacta and the GABAergic pars reticulata are significantly damaged in cases of progressive supranuclear palsy. The distribution of neurodegeneration in patients with Parkinson's disease and progressive supranuclear palsy is discussed with respect to the current theories on pathophysiology in basal ganglia circuitry.

The subthalamic nucleus in Parkinson's disease and progressive supranuclear palsy.

The subthalamus has become a promising target for the neurosurgical treatment of parkinsonian symptoms. We have used unbiased counting techniques to quantify the neuronal populations of the subthalamic nucleus in patients with idiopathic Parkinson's disease and progressive supranuclear palsy. In addition, the type of calcium binding proteins contained within these subthalamic neurons was established using immunohistochemistry. Most of the 550,000 subthalamic neurons contain either parvalbumin or calretinin calcium binding proteins, and patients with idiopathic Parkinson's disease sustained no damage to this nucleus. This is consistent with current theories of basal ganglia circuitry, which postulate that overstimulation of this excitatory nucleus contributes to the inhibition of the motor thalamus via the activation of inhibitory relays. In contrast, we found that there was substantial cell loss in the subthalamus in progressive supranuclear palsy (45 to 85% neuronal reduction) and that both cell types were equally affected. Extracellular neurofibrillary tangles as well as tau-positive glia were observed in the subthalamus of these cases. As the patients with Parkinson's disease and progressive supranuclear palsy all had overlapping parkinsonian symptoms, the loss of subthalamic stimulation within the basal ganglia of progressive supranuclear palsy cases is puzzling, unless their parkinsonian symptoms were generated by an alternate mechanism.

Substantia nigra pars reticulata neurons in Parkinson's disease.

We have examined by immunohistochemistry the parvalbumin-containing neurons of the substantia nigra in patients with Parkinson's disease and in age-matched controls. Parvalbumin, a calcium binding protein, is involved in buffering intracellular calcium and in this study was localized within the majority of non-pigmented neurons of the human pars reticulata. Previous studies have shown that the parvalbumin-immunoreactive pars reticulata neurons are GABAergic and project to the motor thalamus and tectum. Their increased output, due to the loss of dopaminergic inhibition in Parkinson's disease, decreases cortical activation via thalamic pathways, causing parkinsonian symptoms. In Parkinson's disease there was a significant loss of parvalbumin-immunoreactivity from these neurons, though there was no evidence of actual cell loss. This loss of parvalbumin-immunoreactivity was detected only in those cases with end-stage Parkinson's disease.

Anatomical and immunohistochemical identification of catecholaminergic neurones in brain slice preparations used in electrophysiology.

The physiological characteristics of central neural populations are being increasingly explored in slice preparations. A major challenge of this approach is to correlate the physiological properties of individual neurones or groups of neurones with their anatomical and chemical properties in order to gain key insights into their functional identities. The present study describes a method for determining the precise topographical position and the immunohistochemical characteristics of neurones in brain slice preparations that are used frequently in electrophysiological investigations. Thick horizontal slices of rat brainstem were re-cut using a method that provided thin sections that were always in the same plane as the parent slice and that were of suitable thickness for immunohistochemistry. Catecholaminergic neurones in these co-planar (horizontal) sections were stained using antisera to tyrosine hydroxylase, the rate-limiting enzyme for catecholamine synthesis. To identify individual catecholamine neurones in the co-planar sections, we constructed a reference atlas of the distribution of catecholamine neurones in the horizontal plane of the rat brain. The combined use of the horizontal atlas and of immunohistochemical techniques in co-planar sections of horizontal slices enables the determination of several key properties: (1) whether a neurone is TH-positive, (2) its precise topographical position and (3) its content of neuropeptides and other immunohistochemical markers. Thus our study offers a readily feasible method for correlative anatomy and immunohistochemistry of physiologically identified catecholaminergic neurones in brain slices.

Cytoarchitectural distribution of calcium binding proteins in midbrain dopaminergic regions of rats and humans.

The present study compares the distribution of three calcium binding proteins, calbindin-D28k, calretinin, and parvalbumin, in the midbrain tegmentum of rats and humans. In order to compare the distributions of these proteins directly, the cytoarchitecture of this region was evaluated by using immunohistochemistry for tyrosine hydroxylase and substance P in serial sections in both transverse and horizontal planes. There was a high degree of homology in the cytoarchitecture of the three main dopaminergic regions identified. The A8 group was localised in the retrorubral fields, which extended rostrally into the midbrain reticular fields in the human. The A9 group corresponded to the substantia nigra, which was delimited by its dense substance P innervation. The heterogeneous A10 group, situated along the dorsal border as well as medial to the A9 group, comprised multiple nuclei. The distribution of calcium binding proteins was similar in both species, although a larger proportion of neurons contained these proteins in the rat. Calbindin-D28k was localised in neurons within A8 and A10 nuclei and within the caudomedial A9 region (and rostrolateral A9 in the rat only). Calretinin was localised in similar regions. In contrast, neurons containing parvalbumin were concentrated in the substantia nigra pars reticulata. The results suggest that few dopaminergic neurons receiving striatal input in the substantia nigra contain calcium binding proteins; rather, the nondopaminergic nigral neurons contain parvalbumin. Interestingly, dopaminergic neurons are more numerous in humans, whereas nondopaminergic neurons predominate in rats, which suggests that functional differences may exist between rats and humans.