Brain, Vol. 126, No. 4, 827-840,
April 2003
© 2003 Guarantors of Brain
doi: 10.1093/brain/awg085
Severity of gliosis in Pick’s disease and frontotemporal lobar degeneration: tau-positive glia differentiate these disorders
1 Prince of Wales Medical Research Institute and University of New South Wales, 2 Centre for Education and Research on Ageing, University of Sydney, Concord Hospital and 3 University of Western Sydney, Sydney, Australia
Correspondence to: A/Prof. Glenda Halliday, Prince of Wales Medical Research Institute, Barker Street, Randwick, Sydney 2031, NSW, Australia E-mail: g.halliday{at}unsw.edu.au
Received May 16, 2002. Revised July 25, 2002. Second revision September 13, 2002. Accepted September 16, 2002.
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Frontotemporal dementia is a term used to characterize diverse neuropathological conditions that can present with the same clinical phenotype. Five different neuropathologies underlie this disorder. However, consistent frontal and/or temporal neuronal loss and gliosis characterize all cases, the majority having no obvious pathological inclusions. Because neuronal loss and gliosis are consistent features across all cases, the present study aimed to determine the relationship between neuronal loss, gliosis and, for cases with abnormal tau inclusions, intracellular tau deposition. Formalin-fixed brain specimens from sporadic cases with frontotemporal dementia (eight with tau-positive Pick bodies, five with frontotemporal lobar degeneration without inclusions) were compared with those from non-diseased controls (n = 5). Brain specimens were cut into 3 mm coronal slices for evaluation and tissue samples from the superior frontal gyrus were taken for microscopic analysis. Immuno histochemistry for glia-specific proteins (astrocytic glial fibrillary acidic protein and microglial major histocompatibility complex II) and different tau epitopes was performed on 50 µm free-floating sections. Gross patterns of brain atrophy were analysed and upper and lower layer pyramidal neurons and glial cell numbers were quantified. A disease severity scheme was devised using the degree of gross macroscopic frontal and temporal atrophy to establish the relationship between the gliosis and neurodegeneration. In this small sample, the patterns of gross atrophy could be grouped reliably into four stages of severity. These stages were the same across disease groups and correlated with volume- corrected pyramidal neuron densities. In cases with Pick bodies, disease stage also correlated with duration, providing further evidence that these stages represent the progression of degeneration in this limited sample. Whereas there were, on average, many more reactive astrocytes in the cases with Pick bodies than in those with frontotemporal lobar atrophy, there was significant overlap between cases in the degree of astrocytosis. However, a large proportion of the astrocytes in Pick’s disease displayed phosphorylated tau immunoreactivity, whereas no tau-positive astrocytes were found in frontotemporal lobar degeneration. The pattern and degree of microglia activation were similar in all the dementia cases analysed, with considerably more activated microglia accumulating in white matter. In this small sample, the abundance of white matter microglia at early disease stages suggests a prominent role for this cell type in the neurodegenerative process. In frontotemporal lobar degeneration, a significant proportion of the activated white matter microglia were tau-2-immunoreactive, suggesting direct involvement in axonal degeneration, possibly via immune processes.
Keywords: astrocytes; frontotemporal dementia; microglia; Pick’s disease; tau
Abbreviations: FTD = frontotemporal dementia; FTLD = frontotemporal lobar degeneration; GFAP = glial fibrillary acidic protein; HLA-DR = major histocompatibility antigen II; MSR = macrophage scavenger receptor antibody; PiD = Pick’s disease
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Frontotemporal dementia (FTD) is an insidious dementia syndrome which usually presents with progressive changes in behaviour, but may also present with progressive language dysfunction (Hodges and Miller, 2001
; McKhann et al., 2001). The vast majority of cases have frontal and/or anterior temporal lobar atrophy consistent with the pattern of neuronal loss (Dickson, 1998; Mann, 1998; Neary et al., 1998; Spillantini et al., 1998; Hodges and Miller, 2001). At autopsy the degree of cerebral atrophy varies markedly in different patients (Chan et al., 2001), suggesting considerable variability in the disease process. This variability is confirmed by the identification of five basic underlying histopathologies (Hodges and Miller, 2001; McKhann et al., 2001): (i) cases with neuronal inclusions with a predominance of insoluble 3-repeat tau; the largest group have Pick’s disease (PiD) with tau-positive Pick bodies [usually sporadic (Morris et al., 2001; Rossor, 1999)]; (ii) cases with neuronal inclusions with a predominance of insoluble four-repeat tau; they have familial FTD with tau gene mutations (Morris et al., 2001) and sporadic corticobasal degeneration or progressive supranuclear palsy (Bergeron et al., 1998); (iii) cases with neuronal inclusions with insoluble three- and four-repeat tau; the largest group of these have familial FTD with tau gene mutations (Morris et al., 2001); (iv) cases lacking distinctive intraneuronal inclusions; these cases have frontotemporal lobar degeneration (FTLD), which can be sporadic or familial (Mann, 1998; Morris et al., 2001); one family is linked to an abnormality on chromosome 3 (Ashworth et al., 1999); (v) cases with ubiquitin-positive and tau-negative neuronal inclusions; these have FTLD with motor neuron disease or with inclusions typical of motor neuron disease, which can be sporadic or familial (Jackson et al., 1996; Morris et al., 2001); one family is linked to an abnormality on chromosome 9 (Hosler et al., 2000). As can be seen, significant advances have been made concerning the molecular mechanisms of cell death in FTD cases with intraneuronal inclusions. However, the pathophysiology of FTLD is still poorly understood.
The major pathological change in FTLD is the substantial gliosis associated with cell loss (Mann, 1998). Unlike other degenerative diseases affecting the cortex, FTLD is also characterized by substantial gliosis in the white matter (Cooper et al., 1996; Kitagaki et al., 1997; Nichol et al., 2001), as indicated by the descriptive diagnosis of progressive subcortical gliosis in a proportion of cases (Mann, 1998). Recent quantitation of the histological changes in FTLD cases reveals significantly more astrocytes and microglia in the frontal cortices compared with controls (Arnold et al., 2000; Martin et al., 2001). Arnold and colleagues suggest that the microglia could be involved mechanistically in the cell death as neuronal expression of microglia-specific antigens are observed in the frontal and temporal cortices of some cases (Hollister et al., 1997; Arnold et al., 2000). In corticobasal degeneration and progressive supranuclear palsy, significant microglia activation has been shown to correlate with the distribution and degree of neuronal tau deposition (Ishizawa and Dickson, 2001). Of course, in these disorders as well as in PiD, significant tau deposition also occurs in the cortical astroglia and oligodendroglia (Komori, 1999; Berry et al., 2001), with a suggestion that such tau accumulation provides some protection for these cell types (Ishizawa and Dickson, 2001). Overall, the data favour an important role for microglia in the pathogenesis of FTLD and PiD, whether in response to a neuronal immune stimulus (FTLD; Arnold et al., 2000) or in response to neuronal tau deposition (PiD; Komori, 1999). The present study evaluated the relationship between microglial activation, neuronal loss and tau deposition in FTLD compared with PiD. A role for microglia in the pathogenesis of the disease can only be inferred if activation is an early rather than a late event.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Cases
Five cases with FTLD and eight with PiD (Fig. 1) were compared with five control cases without neurological or neuropathological disease. Brains were collected at routine or brain-only autopsy through a regional brain donor programme. The programme includes prospective clinical evaluations of cases using consensus criteria, as recommended recently (McKhann et al., 2001
), and is approved by the Ethics Committees of the South Eastern and Central Sydney Areas Health Services and the Universities of Sydney and New South Wales. All subjects were examined by a neurologist or geriatrician and were assessed clinically within 12 months before death. Additional corroborative information for each patient was obtained from an informant interview to ascertain the pattern and type of deficits, to aid in the estimation of disease duration and obtain information on any family history of disease. Cases with infarction, head injury or any neurodegenerative disease other than FTLD or PiD were excluded. Longitudinal clinical information revealed that all FTLD and PiD cases had behavioural change at onset. Nine out of 13 cases had no family history of dementia or other neurological condition. Four of these cases tested negative for tau gene abnormalities (no genetic screening tests were found in the medical records of the remaining cases). Four out of 13 cases had family members with dementia. Two of these cases had a parent with dementia and were negative in screening tests for tau gene mutations. The other two cases had an elderly aunt or uncle with dementia and one had records of a negative tau gene screening test (the other case was not screened). We believe that none of the cases examined had known tau gene mutations. Control cases were without neurological or neuropathological disease and had no family history of disease. Case details, including diagnosis, age, post-mortem delay and dementia duration, are given in Table 1. Non-parametric Kruskal–Wallis tests were used to determine any significant group differences between these variables.
|
|
Tissue preparation
Each brain was weighed and the volume determined by fluid displacement. Following fixation for 14 days in neutral buffered formalin, the cerebellum and brainstem were separated from the cerebrum by sectioning through the cerebral peduncles. The weight and volume of the cerebrum were determined and the length of each hemisphere was measured. Prior to sectioning, the superior frontal gyrus was painted with coloured dyes as described previously (Harasty et al., 1999
). The cerebrum was embedded in 3% agarose and sliced coronally at 3 mm intervals. The posterior surface of each slice was photographed and printed in black and white at x1 magnification. For neuropathological diagnosis, blocks from the limbic (amygdala, hippocampus and anterior cingulate), frontal, parietal, temporal and occipital neocortex and from the basal ganglia, diencephalon, brainstem and cerebellum were sampled. Sections were cut and stained with haematoxylin and eosin, the modified Bielschowsky silver stain and immunohistochemistry for tau-2 (mouse anti-human tau; T5530 from Sigma, St Louis, MO, USA; diluted 1 : 10 000), 
-synuclein (mouse anti-human -synuclein; 18 0215 from Zymed Laboratories, South San Francisco, CA, USA; diluted 1 : 200), ubiquitin (rabbit anti-cow ubiquitin; Z0458 from Dako, Botany, Australia; diluted 1 : 200) and phosphorylated neurofilament (mouse anti-human phosphorylated neurofilament; MAS330 from Seralab, Leicester, UK; diluted 1 : 2000). Standard peroxidase visualization was used with diaminobenzidine (Sigma) as the chromogen.
For detailed analysis of the glia, tissue blocks were sampled from the superior frontal cortex and underlying white matter and placed in 30% sucrose solution overnight. Recent in vivo blood flow and imaging studies have identified the superior frontal gyrus as the region most affected in FTD (Sjogren et al., 2000). Serial 50 µm thick sections were cut on a cryostat and consecutive free-floating sections were stained immunohistochemically (Shepherd et al., 2000) for microglia using antibodies for the major histocompatibility antigen II (HLA-DR, mouse anti-human HLA-DR; M0775 from Dako; diluted 1 : 1000) and the anti-macrophage scavenger receptor (MSR; goat anti-macrophage scavenger receptor type 1 polyclonal antibody from Chemicon International, Temecula, CA; diluted 1 : 5000). Two microglial markers were required, as HLA-DR could not be used simultaneously with tau antibodies to double-label tau-positive microglia. Sections were also stained for astrocytes using an antibody for glial fibrillary acidic protein (GFAP; rabbit anti-cow GFAP; Z0334; Dako; diluted 1 : 750). Tau-containing pathologies were detected using the AT8 antibody (mouse anti-human paired helical filament tau; MN1020; Pierce Endogen, Chicago, IL, USA; diluted 1 : 10 000) and the tau-2 antibody (mouse anti-human tau-2; T5530; Sigma; diluted 1 : 20 000). AT8 antibody detects tau, which is phosphorylated at serines 202 and 205. In contrast, tau-2 antibody is phosphorylation-independent, detecting tau with the epitope proline 101 masked (Watanabe et al., 1992). Tau-2 localizes tau with residues 92–108 in a serine conformation (found in paired helical filament tau) and in an unfolded conformation (found in reactive glia) (Watanabe et al., 1992).
Antigen retrieval using 4% aluminium chloride (AlCl3) buffer was carried out on the tissue sections to maximize protein visualization, as described previously (Shepherd et al., 2000). Briefly, tissue sections were rinsed in 4% AlCl3 and microwaved (1000 W microwave) for 6 s on 80% power. Sections were then washed in 0.1 M Tris buffer (0.1 M Tris and 0.1% azide, pH 7.4) followed by 50% alcohol for 3 x 15 min. This was followed by a further wash in 50% alcohol with 3% H2O2 (20 min) before immunohistochemical staining. Sections were incubated with 10% blocking serum before incubation with the primary antibody overnight at 4°C. Sections were washed in 0.1 M Tris–Triton (3 x 15 min; 0.1% Triton). Biotinylated secondary antibody (biotinylated IgG; Vector Laboratories, Burlingame, CA, USA) was applied and incubated at room temperature for 1 h. After further washing for 3 x 15 min, sections were incubated in Vectastain Elite ABC (Vector Laboratories) for 1 h and washed for 3 x 15 min, and the peroxidase was visualized with diaminobenzidine. Sections were then mounted onto gelatinized slides, dehydrated through alcohol (70%, 90%, 2 x 100% for 5 min), cleared in xylene and coverslipped with Depex. The specificity of the immunohistochemical reactions were tested by omitting the primary antisera. No peroxidase reaction was observed in these test sections.
To determine the glial types containing tau, double-labelling immunofluorescent experiments were performed. Colocalization of tau with the immunohistochemical markers used to identify specific populations of cells (see above) was performed by mixing mouse and rabbit primary antibodies prior to incubating with host-specific secondary fluorescent antibodies (donkey anti-rabbit fluorescein from Amersham Pharmacia Biotech, Sydney, Australia; diluted 1 : 50; donkey anti-mouse rhodamine red from Jackson Immunoresearch Laboratories, Medical Dynamics, Sydney, Australia; diluted 1 : 100). Double-labelled sections were analysed using an Olympus BX51 fluorescence microscope fitted with specific filter systems to view fluorescein (filter U-MNIBA2), rhodamine red (filter U-MWIG2) and a wide-band ultraviolet filter (U-MWU2) for non-specific fluorescence.
Glia quantitation
The density of GFAP-positive astrocytic and HLA-DR-positive microglia cell bodies contained within the 50 µm thick sections was estimated by sampling three cortical strips perpendicular to the pial surface (400 µm wide) and four white matter samples (area of 400 x 400 µm) of the superior frontal gyrus from each case at x200 magnification using an Olympus BH-2 microscope. We have shown previously that such sampling methods are adequate for the quantitation of cortical neuronal populations (Kril et al., 1997). The number of glia cells within each sample was counted using inclusion (upper and left) and exclusion (lower and right) borders. The proportion of glia double-labelled with tau-2 and MSR antibodies was determined in five cortical and five white matter samples (400 x 400 µm) of the superior frontal gyrus containing dense cellular labelling from each case at x400 magnification using an Olympus BX51 fluorescence microscope fitted with specific filter systems (see above). Density counts (C.K., J.J.K.) and the proportion of tau-positive glia (E.S., C.E.S.) were verified by two independent researchers blind to case details. Variability in glial counts and the proportions containing tau were <2% for each investigator, with a correlation between investigators of 0.98.
Evaluation of cell loss for correlations
Because of the significant differences in the overall amount and rate of atrophy observed in different FTD cases (Chan et al., 2001), neuronal density measures alone are insufficient to determine the entire extent of cell loss when significant volume changes are occurring (Oorschot, 1994; Benes and Lange, 2001) and the major focus of the glial reaction is the white matter (Cooper et al., 1996; Kitagaki et al., 1997; Nichol et al., 2001). For these reasons, two strategies were adopted to determine the stage of disease regarding the degree of neuronal loss.
(i) For each case, a relative severity rating was given based on the degree of gross brain atrophy, incorporating both grey and white matter changes. This was achieved by analysing three standard coronal photographs approximately equidistant through the brain, using methods similar to those devised for the staging of Huntington’s disease (Vonsattel et al., 1985). The anterior and middle coronal sections incorporated features and structures that are specifically vulnerable in FTD (Fig. 2). Case variability was assessed, common patterns were identified and a scheme of disease severity was developed (E.S., C.E.S.). Cases fell into four characteristic, progressively more severe, stages of global atrophy (stage 1–4) (Fig. 2). The validity of the scheme was tested using 
statistics for multiple raters (Fleiss, 1971). Five independent researchers (C.K., J.J.K., G.M.H., H.McCann, H.Cartwright) blind to case details determined the stage for each case with near perfect agreement ( = 0.92; Landis and Kach, 1977). The data show that this simple scheme was reliable and reproducible for the cases examined.
|
(ii) The loss of large pyramidal neurons was estimated in upper and lower cortical layers. The density of upper and lower layer pyramidal neurons was estimated in five random samples (400 x 400 µm) within the upper and lower superior frontal gyrus from each case at x200 magnification using a Zeiss Axioskop microscope. The number of large Nissl-stained pyramidal neurons was counted (Fig. 3) and the counts were verified by three independent researchers blind to case details (E.S., C.E.S., G.M.H.). Large pyramidal neurons were identified by their size (larger than the macroglia), shape (pyramidal) and staining characteristics (less Nissl substance in their cytoplasm compared with glia), and the number of large pyramidal neurons within each sample was counted using inclusion (upper and left) and exclusion (lower and right) borders. Using this method, very small pyramidal neurons and very atrophic, dysfunctional pyramidal neurons were not counted. Variability was <5% for each investigator, with a correlation between investigators of >0.9. To correct for atrophy, the volume of the superior frontal gyrus was calculated for each case using our standard published method (Cullen et al., 1997
; Harasty et al., 1999). Each coronal photograph was overlaid randomly with a grid of 3848 points (area 286 x 196 mm), and the number of points falling on the superior frontal gyrus and the total number of points within each brain slice were counted. The volume of the superior frontal gyrus was calculated by multiplying the sum of the points falling on this structure by the volume represented by each point (volume/point = number of cerebrum points/cerebrum volume, average 0.05 cm3, mean number of points counted for each cerebrum 23 185 ± 741). Repeated measures gave a 1% maximum error with an inter-rater error of <1% (J.K., G.H., H.McCann, H.Cartwright). The density of large pyramidal cells was corrected for the volume of the superior frontal gyrus and expressed as a percentage of mean control values.
|
Statistics
Statistical analysis was performed using Statview (Abacus, Berkeley, CA, USA). Means and standard deviations are given for all variables and a P value of < 0.05 was taken as the level of significance. Non-parametric Kruskal–Wallis tests were used to determine any significant group differences between demographic and cellular variables, and post hoc Mann–Whitney U tests were used to determine any group differences. Linear regressions were used to correlate disease stage to demographic variables and cellular quantitation, and to determine any linear relationship between the different cellular parameters measured.
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Disease variability
As expected, there was a significant difference between the mean brain weights of the groups (H = 6.9, P = 0.03), those of the FTLD and PiD cases being significantly smaller than those of controls (control, 1262 ± 169 g, FTLD, 954 ± 145 g, P = 0.04; PiD, 940 ± 136 g, P = 0.01). However, the degree of atrophy did not differ between the FTLD and PiD cases (post hoc P = 0.77), nor did their stage of disease (Table 2). For PiD, the stage of gross atrophy correlated linearly with disease duration (3–5 years between stages, R2 = 0.78, P = 0.004), whereas for the FTLD cases the stage of gross atrophy was more severe the earlier the age of onset (onset differing by 5–7 years between stages, R2 = 0.62, P = 0.11).
|
The degree of pyramidal cell loss did not differ significantly between FTLD and PiD (H = 11.2, P = 0.004, post hoc P = 0.12). However, the degree of pyramidal cell loss varied considerably across the stages in FTLD and PiD (Fig. 3). In controls, the density of large pyramidal neurons did not differ between layers 2, 3 and 5. In all FTD and PiD cases, there was significantly more pyramidal cell loss in cortical layer 2 (average 72% of control densities) compared with layers 3 and 5 (average 35–45% of control densities), with a strong relationship between atrophy-corrected pyramidal cell loss and the stage of disease severity (R2 = 0.75, P < 0.0001) (Fig. 3F). The average reduction in large frontal pyramidal neurons was 45 ± 6% in stage 1 cases (H = 13.6, P = 0.009, post hoc P = 0.03), with a further reduction of 32 ± 8% between stage 1 and the later stages (post hoc P values <0.05). The loss of cortical pyramidal neurons appeared nearly complete by stage 2 as no further reduction in pyramidal cell number occurred from stage 2 onwards (average 77 ± 8% cell loss, post hoc P > 0.15). At this stage the cortex was virtually devoid of neurons, but was packed with glial nuclei (Fig. 3C–E). The data show similar patterns of cell loss between cases with different underlying histopathology, and that the degree of cell loss and therefore disease severity can be predicted using a simple staging scheme.
Astroglia changes and relationship to disease stage
As described previously (Dickson, 1998; Mann, 1998; Komori, 1999; Arnold et al., 2000; Martin et al., 2001), astrogliosis was substantial in both FTLD and PiD cases (Fig. 4G–I), with higher average densities in PiD than in FTLD (PiD, 2536 ± 918/mm3; FTLD, 1512 ± 342/mm3; control, 208 ± 66/mm3). GFAP-positive astrocyte density differed significantly between PiD and FTLD cases (post hoc P = 0.46) but did not differ significantly between these cases and controls (H = 8.8, P = 0.01). As described previously (Dickson, 1998; Komori, 1999), many astroglia contained AT8 and tau-2 immunoreactivity in PiD (Fig. 4E, M–O), in contrast to the absence of tau-immunoreactive astroglia in FTLD (Fig. 4F, P–R). Both AT8 and tau-2 antibodies stained filamentous inclusions within the astrocytes (Fig. 4B, E), often displacing the GFAP-positive fibrils within the cytoplasm (Fig. 4M–O). In consecutive sections, the distribution of these tau-positive astroglia was similar to that of GFAP-positive astroglia in the same cases (Fig. 4B, E, H). Double-labelling experiments revealed that 23 ± 8% of cortical GFAP-positive astroglia sampled were also tau-2-positive in PiD (Fig. 4M–O). The density and cortical distribution of both AT8- and tau-2-positive astroglia varied similarly between cases, being largely absent at early stages and maximal at stage 3 (Fig. 5A–H, Table 2).
|
|
Microglia changes and relationship to disease stage
As described previously (Paulus et al., 1993
; Arnold et al., 2000), there was considerable microglial upregulation in the frontal cortex compared with controls (Figs 4J–L and 6D–F). Many upregulated HLA-DR- and MSR-positive microglia had distinctive phagocytic morphology (Figs 4K, L and 6E, F, M, P). The amount of cortical microglial activation was similar in FTLD and PiD (HLA-DR-positive microglia density, PiD, 2616 ± 1468/mm3; FTLD, 2303 ± 1532/mm3; control, 27 ± 52/mm3; H = 6.7, P = 0.04, post hoc P = 0.83). This was consistent for the white matter also, with microglial upregulation similar between FTLD and PiD (HLA-DR-positive microglia density, PiD, 4668 ± 1731/mm3; FTLD, 5156 ± 1175/mm3; control, 0 ± 0/mm3; H = 8.1, P = 0.02, post hoc P = 0.92). However, twice as many activated microglia were found in the white matter underlying the frontal cortex in both disorders (Figs 4K, L and 6E, F, M, P), indicating a preference for this response to concentrate in the white matter, as described previously (Paulus et al., 1993; Mann, 1998). Whereas the FTLD cases were largely devoid of neuronal tau deposition, widespread tau-2-immunoreactivity was observed in white matter microglia in most cases (Table 2). In contrast, no AT8-positive microglia or other cellular structures were observed in consecutively stained sections (Fig. 6L). Also, no tau-positive microglia were observed in PiD, despite high densities of HLA-DR-positive microglia (Figs 4B, E, N and 6H, K, N). In consecutive sections from the FTLD cases, the morphology and distribution of these microglia was similar to that of HLA-DR/MSR-positive microglia in the same cases (Fig. 6F, I). The cellular staining pattern of tau-2 was diffuse throughout the cytoplasm of the microglia and did not appear fibrillar (Fig. 6I, Q). Double-labelling experiments revealed that 35 ± 14% of white matter MSR-positive microglia sampled were also tau-2-positive in FTLD (Fig. 6P–R). The tau-2-positive microglia concentrated in the white matter of FTLD cases at stages 1 and 2 (Fig. 5I–N, Table 2), indicating early involvement in the disease process.
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
This report provides the first direct comparison of the amount of neuronal cell loss and gliosis between samples of cases with PiD and FTLD. Our data indicate that the patterns of gross atrophy (see also Mann and South, 1993
) and the amount of neuronal cell loss and microglial activation are virtually identical between these two types of FTD, and that these degenerative changes occur in unison. These data suggest that microglial activation and neuronal loss are linked in the pathogenesis of both types of FTD. These histopathological changes must be the primary factors underlying the consistent patterns of frontotemporal atrophy and hypometabolism characterizing FTD (Hodges and Miller, 2001; McKhann et al., 2001).
As highlighted in recent reviews (Hodges and Miller, 2001; McKhann et al., 2001), the different underlying histopathologies identified in patients with FTD cannot be predicted in life, and longitudinal studies show considerable heterogeneity in the rate of degeneration between cases (Chan et al., 2001). A standard progression over a relatively predictable time course is yet to be determined. In the absence of such data, we used the methods developed for staging Huntington’s disease (Vonsattel et al., 1985), a disorder with similarities regarding disease variability in onset, progression and stage. By using reliable patterns of gross atrophy, the amount of tissue damage can be staged (stage correlated with the amount of pyramidal cell loss) and early versus late events inferred. In the limited number of cases sampled, the stage of disease was related to the duration of the clinical syndrome for PiD, validating the concept that the stages are progressive and suggesting that this FTD phenotype may have a more predictable disease course. The amount of neuronal degeneration correlated with disease stage and the degree of microglial activation in both PiD and FTLD, linking these phenomena to the time course of disease.
Few previous studies have quantified the amount of neuronal loss in either PiD (Hansen et al., 1988) or FTLD (Arnold et al., 2000). In eight PiD cases, the frontal cortex was thinned by an average of 42% compared with controls and showed a 56% reduction in the density of medium to large neurons, although great variability between cases was found (Hansen et al., 1988). Neuronal density was reduced on average by 45% in five FTLD cases (Arnold et al., 2000). These data are consistent with the results of the present study comparing PiD and FTLD using atrophy-corrected cell counts. As might be expected, our data indicate that neuronal loss is more substantial if tissue loss is taken into account. However, we also show considerable variability in neuronal loss, this variability relating directly to the stage of the disease process.
In addition to the similarities noted above, differences other than those required for diagnostic differentiation were found between the cases of PiD and FTLD. In particular, the glial responses differentiated these two types of FTD. Although not statistically significant, more astrocytes were observed in PiD compared with FTLD, and in PiD a large proportion of these astrocytes contained abnormal phosphorylated tau depositions, as described previously (Komori, 1999; Berry et al., 2001). Tau-immunoreactive astrocytes were not found in any of the FTLD cases. The lower average number of astrocytes seen in FTLD may be due to their active degeneration (Su et al., 2000; Martin et al., 2001; Nichol et al., 2001), although levels of GFAP in the CSF are not changed (Wallin et al., 1996). Increased CSF levels of the astrocytic calcium-binding protein s100b in FTD (Green et al., 1997) support astroglia upregulation in this disorder. In PiD, astrocytic tau differs significantly from the neuronal tau found in Pick bodies, as it is made of four- rather than three-repeat isoforms (Arai et al., 2001). Four-repeat tau isoforms also accumulate in astrocytes in areas undergoing degeneration in familial FTD with tau gene mutations and in sporadic corticobasal degeneration and progressive supranuclear palsy (Komori, 1999; Arai et al., 2001; Berry et al., 2001). This suggests a link between the accumulation of astrocytic tau and neuronal tau, but highlights the fact that significantly different pathogenic processes must occur in neurons in PiD for three- rather than four-repeat tau to accumulate (Arai et al., 2001).
Microglial activation was identical in both PiD and FTLD. Significant microglial activation has been described previously (Paulus et al., 1993; Cooper et al., 1996; Hollister et al., 1997; Arnold et al., 2000) and both neurons and astrocytes have significant complement immunostaining (Yasuhara et al., 1994; Singhrao et al., 1996), suggesting a classical immune response. However, there is no evidence of lymphocytic infiltration using routine (Dickson, 1998; Mann, 1998) or cell-specific (Singhrao et al., 1996; Hollister et al., 1997) stains. It is well known that activation of microglia occurs in response to non-immunological stimuli, including cerebral ischaemia and neuronal cell death (Kreutzberg, 1996). Of course, both frontotemporal atrophy and hypoperfusion are diagnostic for FTD (Talbot et al., 1995; Miller and Gearhart, 1999; Charpentier et al., 2000; McKhann et al., 2001; Santens et al., 2001) and may be the main initiators of the microglia response observed. However, as noted previously (Cooper et al., 1996; Nichol et al., 2001), the region containing the highest density of activated microglia was the white matter underlying the cortex rather than the cortical grey matter. The concentration of activated microglia in the white matter suggests that cell structures in this site are preferentially undergoing damage. There is in vivo evidence for prominent T2- and proton density-weighted magnetic resonance change in the frontal white matter in FTD (Kitagaki et al., 1997; Larsson et al., 2000), with the observation that the increased signal intensity at this site is a sensitive diagnostic marker (Kitagaki et al., 1997). Our data suggest that this observed change is likely to reflect activation of white matter microglia. Such an immune response within the white matter is reminiscent of neuro degenerative demyelinating disorders.
A major difference in the microglial response between cases of PiD and FTLD was that tau-2-immunoreactive, upregulated microglia were observed in FTLD but not PiD. The density of such microglia in FTLD was high (Figs 5I–K and 6I, Q). It has been suggested that microglial tau is of a particular conformational form not present in the pathology of other glial types (Berry et al., 2001), and possibly could be due to Fc binding of tau (Komori, 1999). It has been shown that tau associated with glycolipids is an extremely potent antibody-independent activator of the classical complement pathway [better than IgG itself (Shen et al., 2001)] and that activated microglia accumulate around complement-immunoreactive tau-aggregates (Imamura et al., 2001). Activated tau-2-positive microglia have been reported previously in small numbers in subcortical grey matter structures in Alzheimer’s and Lewy body diseases (Odawara et al., 1995), progressive supranuclear palsy and corticobasal degeneration (Berry et al., 2001). In Alzheimer’s disease, large numbers of mainly resting cortical microglia concentrate tau in an unusual conformational form (Ghoshal et al., 2001), although the relationship of this type of tau to the expression of an immune response in microglia is unknown, particularly as many of these cells appear quiescent (Ghoshal et al., 2001). In contrast, the tau-2 immunoreactivity we have documented in FTLD occurred in a proportion of activated microglia largely within the white matter, and was not observed in PiD. This strongly suggests an association between this non-fibrillar tau and a white matter immune response in FTLD. Confirmation of this finding in more cases is now required. In experimental allergic encephalomyelitis and multiple sclerosis, activated microglia are thought to play a key role in terminating T-cell responses to self antigens (Perry, 1998). In this situation the microglia present antigen in a manner that does not induce lymphocyte proliferation, but actively induces a new state in the surveillance T-cells and results in apoptosis (Sedgwick, 1995; Perry, 1998). The activation of microglia in the FTD cases examined may be similar but associated with different self antigens.
Tau protein concentrates in axons (Binder et al., 1985) and several studies suggest that an impairment in axonal transport may play a role in both PiD (Nakamura et al., 1994; Probst et al., 1996) and FTLD (Zhou et al., 1998). In addition to complement (Yasuhara et al., 1994; Singhrao et al., 1996), non-phosphorylated and hyperphosphorylated tau, phosphorylated neurofilament and kinesin aggregate in Pick bodies, some neurons and axons in PiD (Nakamura et al., 1994; Probst et al., 1996; Dickson, 1998; King et al., 2001), whereas in FTLD non-phosphorylated tau and neurofilaments accumulate in axonal spheroids in the neuropil (Zhou et al., 1998). Tau regulates the attachment of vesicles and organelles to microtubules and their detachment from microtubules by interacting with microtubule-binding ATPases (Ebneth et al., 1998; Trinczek et al., 1999; Almenar-Queralt and Goldstein, 2001) in a phosphorylation-dependent manner (Flaherty et al., 2000). The local arrest of axoplasmic transport is associated with microtubule disruption and eventually results in local breakdown of the axolemma and the formation of end-bulbs (Povlishock and Jenkins, 1995). Disruption of the dynamics between tau and microtubules has been proposed as the major mechanism of neurodegeneration in familial FTD cases with tau gene mutations (Spillantini et al., 1998; Lee et al., 2001), and the location and type of microglia response observed in both PiD and FTLD would suggest that axonal disruption is also important in these types of FTD.
The present study provides evidence linking the major histopathological forms of FTD to abnormalities associated with tau deposition in regions undergoing neurodegeneration. The present data are consistent with different cellular mechanisms of axonal tau dysfunction leading to neurodegeneration. Hyperphosphorylated neuronal and macroglial tau deposition underlies the degeneration in PiD, while tau-2-positive microglia appear related to the primary structural damage in FTLD. This suggests that a proportion of FTD cases may have a dominant inflammatory response to tau. If confirmed in larger cohorts, this finding has significant therapeutic implications.
|
Acknowledgements |
|---|
We wish to thank Heather McCann and Heidi Cartwright for laboratory assistance, preparation of the figures and independently reproducing data for reliability testing. The project was funded by the National Health and Medical Research Council of Australia. C.E.S. is a University of Sydney Rolf Edgar Lake Fellow and G.M.H is a NHMRC Principal Research Fellow.
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Almenar-Queralt A, Goldstein LSB. Linkers, packages and pathways: new concepts in axonal transport. Curr Opin Neurobiol 2001; 11: 550–7.[CrossRef][ISI][Medline]
Arai T, Ikeda K, Akiyama H, Shikamoto Y, Tsuchiya K, Yagishita S, et al. Distinct isoforms of tau aggregated in neurons and glial cells in brains of patients with Pick’s disease, corticobasal degeneration and progressive supranuclear palsy. Acta Neuropathol (Berl) 2001; 101: 167–73.[Medline]
Arnold SE, Han L-Y, Clark CM, Grossman M, Trojanowski JQ. Quantitative neurohistological features of frontotemporal degeneration. Neuromuscul Disord 2000; 21: 913–9.
Ashworth A, Lloyd S, Brown J, Gydesen S, Sorensen SA, Brun A, et al. Molecular genetic characterisation of frontotemporal dementia on chromosome 3. Dement Geriatr Cogn Disord 1999; 10 Suppl 1: 93–101.[Medline]
Benes FM Lange N. Two-dimensional versus three-dimensional cell counting: a practical perspective. Trends Neurosci 2001; 24: 11–7.[CrossRef][ISI][Medline]
Bergeron C, Davis A, Lang AE. Corticobasal ganglionic degeneration and progressive supranuclear palsy presenting with cognitive decline. Brain Pathol 1998; 8: 355–65.[ISI][Medline]
Berry RW, Quinn B, Johnson N, Cochran EJ, Ghoshal N, Binder LI. Pathological glial tau accumulations in neurodegenerative disease: review and case report. Neurochem Int 2001; 39: 469–79.[CrossRef][ISI][Medline]
Binder LI, Frankfurter A, Rebhun LI. The distribution of tau in the mammalian central nervous system. J Cell Biol 1985; 101: 1371–8.
Chan D, Fox NC, Jenkins R, Scahill RI, Crum WR, Rossor MN. Rates of global and regional cerebral atrophy in AD and frontotemporal dementia. Neurology 2001; 57: 1753–63.
Charpentier P, Lavenu I, Defebvre L, Duhamel A, Lecouffe P, Pasquier F, et al. Alzheimer’s disease and frontotemporal dementia are differentiated by discriminant analysis applied to (99m)Tc HmPAO SPECT data. J Neurol Neurosurg Psychiatry 2000; 69: 661–3.
Cooper PN, Siddons CA, Mann DM. Patterns of glial cell activity in fronto-temporal dementia (lobar atrophy). Neuropathol Appl Neurobiol 1996; 22: 17–22.[CrossRef][ISI][Medline]
Cullen KM Halliday GM Double KL Brooks WS Creasey H, Broe GA. Cell loss in the nucleus basalis is related to regional cortical atrophy in Alzheimer’s disease. Neuroscience 1997; 78: 641–52.[CrossRef][ISI][Medline]
Dickson DW. Pick’s disease: a modern approach. Brain Pathol 1998; 8: 339–54.[ISI][Medline]
Ebneth A, Godemann R, Stamer K, Illenberger S, Trinczek B, Mandelkow E-M. Overexpression of tau protein inhibits kinesin-dependent trafficking of vesicles, mitochondria, and endoplasmic reticulum: implications for Alzheimer’s disease. J Cell Biol 1998; 143: 777–94.
Flaherty DB, Soria JP, Tomasiewicz HG, Wood JG. Phosphorylation of human tau protein by microtubule-associated kinases: GSK3ß and cdk5 are key participants. J Neurosci Res 2000; 62: 463–72.[CrossRef][ISI][Medline]
Fleiss JL. Measuring nominal scale agreement among many raters. Psychol Bull 1971; 76: 378–82.[CrossRef]
Ghoshal N, García-Sierra F, Fu Y, Beckett LA, Mufson EJ, Kuret J, et al. Tau-66: evidence for a novel tau conformation in Alzheimer’s disease. J Neurochem 2001; 77: 1372–85.[CrossRef][ISI][Medline]
Green AJ, Harvey RJ, Thompson EJ, Rossor MN. Increased S100beta in the cerebrospinal fluid of patients with frontotemporal dementia. Neurosci Lett 1997; 235: 5–8.[CrossRef][ISI][Medline]
Hansen LA, Deteresa R, Tobias H, Alford M, Terry RD. Neocortical morphometry and cholinergic neurochemistry in Pick’s disease. Am J Pathol 1988; 131: 507–18.[Abstract]
Harasty JA, Halliday GM, Kril JJ, Code C. Specific temporoparietal gyral atrophy reflects the pattern of language dissolution in Alzheimer’s disease. Brain 1999; 122: 675–86.
Hodges JR, Miller B. The classification, genetics and neuropathology of frontotemporal dementia. Introduction to the special topic papers: Part 1. Neurocase 2001; 7: 31–5.[CrossRef][ISI][Medline]
Hollister RD, Xia M, McNamara MJ, Hyman BT. Neuronal expression of class II major histocompatibility complex (HLA-DR) in 2 cases of Pick disease. Arch Neurol 1997; 54: 243–8.[Abstract]
Hosler BA, Siddique T, Sapp PC, Sailor W, Huang MC, Hossain A, et al. Linkage of familial amyotrophic lateral sclerosis with frontotemporal dementia to chromosome 9q21-q22. JAMA 2000; 284: 1664–9.
Imamura K, Sawada M, Ozaki N, Naito H, Iwata N, Ishihara R, et al. Activation mechanism of brain microglia in patients with diffuse neurofibrillary tangles with calcification: a comparison with Alzheimer disease. Alzheimer Dis Assoc Disord 2001; 15: 45–50.[CrossRef][ISI][Medline]
Ishizawa K, Dickson DW. Microglial activation parallels system degeneration in progressive supranuclear palsy and corticobasal degeneration. J Neuropathol Exp Neurol 2001; 60: 647–57.[ISI][Medline]
Jackson M, Lennox G, Lowe J. Motor neurone disease-inclusion dementia. Neurodegeneration 1996; 5: 339–50.[CrossRef][ISI][Medline]
King ME, Ghoshal N, Wall JS, Binder LI, Ksiezak-Reding H. Structural analysis of Pick’s disease-derived and in vitro-assembled tau filaments. Am J Pathol 2001; 158: 1481–90.
Kitagaki H, Mori E, Hirono N, Ikejiri Y, Ishii K, Imamura T, et al. Alteration of white matter MR signal intensity in frontotemporal dementia. AJNR Am J Neuroradiol 1997; 18: 367–78.[Abstract]
Komori T. Tau-positive glial inclusions in progressive supranuclear palsy, corticobasal degeneration and Pick’s disease. Brain Pathol 1999; 9: 663–79.[ISI][Medline]
Kreutzberg GW. Microglia: a sensor for pathological events in the CNS. Trends Neurosci 1996; 19: 312–8.[CrossRef][ISI][Medline]
Kril JJ, Halliday GM, Svoboda MD, Cartwright H. The cerebral cortex is damaged in chronic alcoholics. Neuroscience 1997; 79: 983–98.[CrossRef][ISI][Medline]
Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–74.[CrossRef][ISI][Medline]
Larsson E, Passant U, Sundgren PC, Englund E, Brun A, Lindgren A, et al. Magnetic resonance imaging and histopathology in dementia, clinically of frontotemporal type. Dement Geriatr Cogn Disord 2000; 11: 123–34.[CrossRef][ISI][Medline]
Lee VM-Y, Goedert M, Trojanowski JQ. Neurodegenerative tauopathies. Annu Rev Neurosci 2001; 24: 1121–59.[CrossRef][ISI][Medline]
Mann DMA. Dementia of frontal type and dementias with subcortical gliosis. Brain Pathol 1998; 8: 325–38.[ISI][Medline]
Mann DMA, South PW. The topographic distribution of brain atrophy in frontal lobe dementia. Acta Neuropathol (Berl) 1993; 85: 334–40.[Medline]
Martin JA, Craft DK, Su JH, Kim RC, Cotman CW. Astrocytes degenerate in frontotemporal dementia: possible relation to hypoperfusion. Neurobiol Aging 2001; 22: 195–207.[CrossRef][ISI][Medline]
McKhann GM, Albert MS, Grossman M, Miller B, Dickson D, Trojanowski JQ. Clinical and pathological diagnosis of frontotemporal dementia. Report of the Work Group on Frontotemporal Dementia and Pick’s Disease. Arch Neurol 2001; 58: 1803–9.
Miller BL, Gearhart R. Neuroimaging in the diagnosis of frontotemporal dementia. Dement Geriatr Cogn Disord 1999; 10 Suppl 1: 71–4.[CrossRef][ISI][Medline]
Morris HR, Khan MN, Janssen JC, Brown JM, Perez-Tur J, Baker M, et al. The genetic and pathological classification of familial frontotemporal dementia. Arch Neurol 2001; 58: 1813–6.
Nakamura Y, Takeda M, Yoshimi K, Hattori H, Hariguchi S, Hashimoto S, et al. Involvement of clathrin light chains in the pathology of Pick’s disease; implication for impairment of axonal transport. Neurosci Lett 1994; 180: 25–8.[CrossRef][ISI][Medline]