Introduction

The Kii peninsula of Japan and the island of Guam are focal points of a high prevalence of amyotrophic lateral sclerosis (ALS) and parkinsonism–dementia complex (PDC) in the western Pacific [5, 14]. Clinically, western Pacific ALS in Kii and Guam is indistinguishable from sporadic ALS in other areas of the world, while PDC is unique in these endemic areas and is characterized by insidious progression of mental slowness and extrapyramidal symptoms of bradykinesia, rigidity, and occasional tremor [2]. Both the ALS and PDC features overlap in some patients presenting with “PDC plus ALS” where PDC features usually precede ALS features.

We previously reported that Kii PDC cases are neuropathologically characterized by nerve cell loss and widespread neurofibrillary tangles (NFTs) of severe degree without senile plaques in the cerebral cortex, basal ganglia, or brainstem, and with upper and lower motor neuron degeneration of varying degrees, while Kii ALS cases are characterized by classical ALS pathology and the additional findings of nerve cell loss and NFTs of varying degree in various brain areas [9, 10]. “PDC plus ALS” cases are neuropathologically characterized by superimposition of PDC pathology and ALS pathology. These findings may suggest that ALS, PDC and “PDC plus ALS” in the Kii peninsula share a common neuropathology of a combination of motor neuron disease and widespread neurofibrillary degeneration, and that they are neuropathologically on a spectrum of a single disease entity despite the difference in clinical manifestations [7, 9, 10].

However, there are no quantitative or semiquantitative data on the regional distribution of NFTs in Kii ALS/PDC. We have, therefore, investigated topography and semiquantitative density of NFTs in Kii ALS, PDC, and “PDC plus ALS” cases in order to clarify whether or not Kii ALS and PDC are on a spectrum of a single entity from the viewpoint of NFT pathology.

Materials and methods

Seven autopsied brains and five spinal cords from seven Kii ALS/PDC patients who were born and raised in the Kii peninsula ALS focal point were submitted for the present study. The clinical features are shown in Table 1. Clinical diagnosis was based on the clinical features: “ALS” refers to pure motor neuron disease features throughout the course of the disease and included two patients. “PDC” refers to progressive parkinsonism and dementia with modest, if present, motor neuron signs, and included three patients. “PDC plus ALS” refers to an overlap of PDC and ALS, where PDC features preceded ALS features and included two patients.

Table 1 Summary of clinical features of Kii ALS/PDC cases
Full size table

The brains and spinal cords were fixed in formalin solution. The brains were sliced coronally, and the spinal cords were sliced axially, and the slices were embedded in paraffin. The paraffin-embedded specimens were cut into 9-μm thick sections and prepared for hematoxylin-eosin (HE), Klüver-Barrera (KB), Bodian, and Gallyas-Braak staining. For the detection of NFTs, the sections were stained with anti-human tau protein (AT8, monoclonal mouse, dilution at 1:2,000, Innogenetics, Zwijndrecht, Belgium) using the dextranpolymer method with an ENVISION kit (Dako, Glostrup, Denmark). Peroxidase-conjugated streptavidin was visualized with 3,3′-diaminobenzidine (DAB; Wako, Osaka, Japan) as the final chromogen and counterstained lightly with Mayer’s hematoxylin.

Gallyas-Braak stained sections of the frontal cortex (motor area), frontal pole, temporal pole, amygdaloid nucleus, hippocampus, basal ganglia, Meynert’s nucleus, midbrain, upper and middle pons, medulla oblongata, cerebellum including the dentate nucleus, and spinal cord were used for topographical mapping and counting of NFTs. NFTs were counted and plotted on brain maps under ×100 magnification using an Olympus micrometer (Olympus, Tokyo, Japan).

Neurofibrillary tangles were counted in five different areas of 1-mm2 in the temporal cortex, amygdaloid nucleus, substantia nigra, and in the entire area of the unilateral locus ceruleus.

Results

Neuropathological findings

The neuropathological findings of ALS, PDC, and “PDC plus ALS” cases were essentially similar to those previously reported [4, 9] and are summarized in Table 2. The brain maps of NFT topography of representative cases are shown, to the same scale, in Figs. 1 and 2.

Table 2 Summary of neuropathological findings of Kii ALS/PDC cases
Full size table
Fig. 1
figure 1

Distribution of NFTs in representative cases. a ALS case (case 1) with abundant NFTs. b PDC case (case 5). c “PDC plus ALS” case (case 6). One dot denotes presence of one NFT, and one square denotes 10 NFTs. Scale bar = 1 cm

Full size image
Fig. 2
figure 2

Distribution of NFTs in the spinal cord of representative cases. a ALS cases (Thoracic and lumbar cord: case 1, sacral cord: case 2). b PDC case (case 5). c “PDC plus ALS” case (case 6). One dot denotes presence of one NFT. Scale bar = 1 cm

Full size image

Macroscopically, the brains of the ALS cases were unremarkable, whereas those of PDC and “PDC plus ALS” cases showed moderate to marked atrophy of the frontal and temporal lobes as shown in Fig. 1, with a decrease in brain weight (Table 2). The substantia nigra and locus ceruleus were depigmented in not only the PDC and “PDC plus ALS” cases, but also in the ALS cases.

Microscopically, the loss of pigmented nerve cells of the substantia nigra and locus ceruleus was marked in the PDC and “PDC plus ALS” cases, while it was mild to moderate in the ALS cases. Nerve cell loss and NFTs of the cerebral cortex and brainstem were severe in the PDC and “PDC plus ALS” cases, while they were mild in one ALS case and severe in the other. Typical senile plaques and amyloid angiopathy were absent, though a few diffuse plaques were found in the Ammon’s horn of a single PDC case (case 3, Table 2). Neuropil threads were observed in the parahippocampal gyrus to a mild degree in all cases.

Topography and density of NFTs

In all cases, NFTs were distributed in a similar topography with a predilection for the limbic system; NFTs were numerous in the hippocampus, particularly in the CA1 field, parahippocampal gyrus, temporal pole, insular cortex, and amygdaloid nucleus; they were moderate in number in the Meynert nucleus, hypothalamus, periaqueductal gray matter, substantia nigra, locus ceruleus, and brainstem reticular formation; and there were few in the pontine nuclei, basal ganglia, thalamus, dentate nucleus, other cerebral cortices including the primary motor cortex and spinal cord (Figs. 1, 2). The regional densities of NFTs including ghost tangles were high in all PDC and “PDC plus ALS” cases and one ALS case, while it was low in the other ALS case (Fig. 3, Table 3). The density of NFTs did not correlate with age or the duration of the disease in ALS, PDC, or “PDC plus ALS” cases.

Fig. 3
figure 3

Distribution of NFTs in the parahippocampal gyrus (a–c), substantia nigra (d–f), and primary motor cortex (g, h) of Kii ALS and PDC cases. a, d, g; case 1 (ALS with abundant NFTs), b, e; case 4 (PDC), c, f, h; case 6 (“PDC plus ALS” case). NFTs were distributed in dense layers II and III. The topography of NFTs was similar in all cases. Modified Gallyas stain with faint Nissl and Luxol-fast blue stains. a–f scale bar = 200 μm. g and h scale bar = 100 μm

Full size image
Table 3 Density of NFTs/mm2 area in the cerebral cortex, amygdala and brain stem of Kii ALS/PDC cases
Full size table

Discussion

Amyotrophic lateral sclerosis and PDC of the Kii peninsula occur in the same geographic area, in members of the same family, and frequently in the same individual simultaneously, which suggests that they are different clinical phenotypes of a single disease entity. Neuropathological findings of NFTs and motor neuron involvement were also common to both ALS and PDC [7, 9, 10].

The present study has demonstrated a similar topographic distribution of NFTs in cases of clinical ALS, PDC, and “PDC plus ALS” phenotypes. In all cases, hippocampus, especially the CA1 field, parahippocampal gyrus, temporal pole, insular cortex, and amygdaloid nucleus were most severely affected. NFTs of the neocortex were denser in the superficial layers II–III than in the layers V–VI, as we reported previously in Kii ALS/PDC [4, 9]. This pattern is different from that of Alzheimer’s disease cases, where NFTs are densest in layers V–VI. Only a few NFTs were observed in the five spinal cord cases that were investigated. They were diffusely distributed, with lamina VII that was described by Rexed [13], which is known as the intermediate gray, showing higher NFT density than other parts of the spinal cord.

In 1979, Anderson et al. [1] reported that the presence of many NFTs was not unique to the brains of Guamanian ALS or PDC patients, but was common to the brains of neurologically normal Guamanian Chamorros. Similar findings were confirmed by several studies [2, 11, 12], and some researchers have suggested that the presence of NFTs in Guamanian brains represents a non-pathological background of Chamorros rather than a pathological hallmark.

Oyanagi et al. [11] reported that the numbers of NFTs were almost similar between normal Guamanian Chamorros and Guamanian ALS patients, and have proposed that “Guam ALS” is not a particular ALS but a classical ALS in Guamanian Chamorros, and that Guamanian PDC is nosologically different from Guamanian ALS.

Recently, Hof et al. [3] reported that the primary motor cortex in Guamanian cases contained high numbers of NFTs, contrasting sharply with the situation in AD and in non-Chamorro cases of ALS. They have suggested that Guamanian ALS/PDC may represent a single disorder with a broad spectrum of neurodegeneration, ranging from relatively pure ALS to PDC.

The present study has shown that the topography of NFTs is similar among ALS, PDC, and “PDC plus ALS” in Kii cases. These findings suggest that ALS and PDC in the Kii peninsula are on a spectrum of a single disease, and abundant NFTs are pathognomonic to Kii ALS/PDC.

In Kii and Guam, the distribution of NFTs of ALS/PDC cases was similar. However, the densities of NFTs were different between Kii and Guam cases. In Guam, pure ALS cases show considerably lower densities of NFTs in the neocortex and hippocampus, excluding the entorhinal cortex, than the PDC cases [2, 3]. Interestingly, in our “ALS” cases, the density of NFTs was abundant in one but only a few in the other. We could not find “pure ALS” cases with a high density of NFTs in Guam or Kii in previous reports. Although more cases are necessary, the variety of the NFT densities might be a peculiarity in the Kii cases.

The duration of the disease was longer in PDC and “PDC plus ALS” cases (6–8 years) than in the ALS cases (3–5 years). Though our data failed to show a correlation between the density of the NFTs and the duration of the disease, the longer survival of PDC and “PDC plus ALS” cases than ALS cases may be one of the causes of denser NFTs and more marked brain atrophy in the former.

Our previous investigation [7] has revealed that the frequency of NFTs in the brains of people without neurodegenerative disease in the ALS focal point of the Kii peninsula was as low as that of the general Japanese population. In addition, a neuropathological study on NFTs in ALS patients from the whole Mie prefectural area including the high incidence focus [6] has shown that the presence of abundant NFTs in the CNS was unique to ALS cases from the high incidence ALS focus area and never occurred in ALS cases outside the focus area. More than 70% of the patients had a positive family history of ALS/PDC, and the rates of incidence and prevalence remain continuously very high despite the dramatic changes in environmental factors, including food and water supply and by westernization of daily life [8]. These findings suggest that genetic factors may play a major role, although further studies on more cases are necessary to clarify the etiopathogenesis of Kii ALS/PDC.