Changes in glial cell phenotypes precede overt neurofibrillary tangle formation, correlate with markers of cortical cell damage, and predict cognitive status of individuals at Braak III-IV stages

Abstract

Clinico-pathological correlation studies show that some otherwise healthy elderly individuals who never developed cognitive impairment harbor a burden of Alzheimer's disease lesions (plaques and tangles) that would be expected to result in dementia. In the absence of comorbidities explaining such discrepancies, there is a need to identify other brain changes that meaningfully contribute to the cognitive status of an individual in the face of such burdens of plaques and tangles. Glial inflammatory responses, a universal phenomenon in symptomatic AD, show robust association with degree of cognitive impairment, but their significance in early tau pathology stages and contribution to the trajectory of cognitive decline at an individual level remain widely unexplored. We studied 55 brains from individuals at intermediate stages of tau tangle pathology (Braak III-IV) with diverging antemortem cognition (demented vs. non-demented, here termed `resilient'), and age-matched cognitively normal controls (Braak 0-II). We conducted quantitative assessments of amyloid and tau lesions, cellular vulnerability markers, and glial phenotypes in temporal pole (Braak III-IV region) and visual cortex (Braak V-VI region) using artificial-intelligence based semiautomated quantifications. We found distinct glial responses with increased proinflammatory and decreased homeostatic markers, both in regions with tau tangles (temporal pole) and without overt tau deposits (visual cortex) in demented but not in resilient. These changes were significantly associated with markers of cortical cell damage. Similar phenotypic glial changes were detected in the white matter of demented but not resilient and were associated with higher burden of overlying cortical cellular damage in regions with and without tangles. Our data suggest that changes in glial phenotypes in cortical and subcortical regions represent an early phenomenon that precedes overt tau deposition and likely contributes to cell damage and loss of brain function predicting the cognitive status of individuals at intermediate stages of tau aggregate burden (Braak III-IV).

Keywords: AD; Cognition; Cortical cell vulnerability; Glial phenotypes; Neurofibrillary tangles; White matter changes.

Fig. 1

Example of the artificial intelligence based semiautomated quantification method used. Selection of cortical and subcortical regions of interest (a, b), tissue layer detection in green (b, c, f), P2RY12 counter in purple (c, d), IBA1 area detector in blue (f, g) and unannotated images stained with P2RY12 (e) and IBA1 (h)

Fig. 2

Regional Aβ plaque burden (defined as the percentage of cortex occupied by amyloid beta (Aβ) deposits immunostained with 4G8 antibody) (a) and number of neurofibrillary tangles (as reported by immunolabeling with AT8 antibody) (b, c) did not significantly differ in demented compared to resilient. Representative photomicrographs of plaques (a) and tangles (b, c) on sections from control, resilient and demented cases containing temporal pole (a and b) and visual cortex (c). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); *p < 0.05; **p < 0.01. Scale bars 100 μm and 50 μm

Fig. 3

Quantifications for total IBA1 burden (percentage of cortex or white matter covered by IBA1 + microglia), total number of astrocytes (ALDH1L1), and total number of oligodendrocytes (Olig2) in temporal and visual cortex (microglia and astrocytes) and white matter (microglia and oligodendrocytes) (b) did not show statistically significant differences between demented, resilient and control brains. Representative photomicrographs of IBA1 + microglia, ALDH1L1 + astrocytes and OLIG2 + oligodendrocytes on sections from control, resilient and demented cases containing temporal pole are displayed on the top (a). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV). Scale bar 50 μm

Fig. 4

Number of reactive astrocytes as labelled by GFAP antibody were significantly higher in the temporal pole and visual cortex of demented but not resilient when compared to control brains (a). Representative photomicrographs of GFAP + astrocytes on sections from control, resilient and demented cases containing temporal pole are displayed on the right (b). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); **p < 0.01; ****p < 0.0001. Scale bar 50 μm

Fig. 5

Number of CD68 + microglia was significantly higher in the temporal pole and visual cortex of demented but not in resilient when compared to control brains (a). A similar trend that did not reach statistical significance was observed in the number of HLA-DR + microglia (a). Representative photomicrographs of CD68 + and HLD-DR + microglia on sections from control, resilient and demented cases containing temporal pole are displayed on the right (b). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); *p < 0.05; **p < 0.01; ****p < 0.0001. Scale bars 50 μm

Fig. 6

Homeostatic microglia stained with TMEM119 and P2RY12 antibodies in temporal pole and visual cortex was significantly decreased in the temporal pole and visual cortex of demented but not in resilient when compared to control brains (a). Representative photomicrographs of TMEM119 + and P2RY12 + microglia on sections from control, resilient and demented cases containing visual cortex are displayed on the right (b). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); *p < 0.05; **p < 0.01; ****p < 0.0001. Scale bars 50 μm

Fig. 7

Number of γH2AX + cells was significantly increased in temporal pole and visual cortex of demented but not in resilient when compared to control brains (a). Representative photomicrographs of γH2AX + cells on sections from control, resilient and demented cases containing visual cortex (b). Representative immunofluorescence images of colocalization of γH2AX + cells (brown) with neurons (red) (7c 1–3, showing a progressively zoomed-in brain region), GFAP + astrocytes (green) (c 4–6, showing GFAP alone, 7c4, γH2AX alone, 7c5, both GFAP-γH2AX, 7c6), and IBA1 + microglia (green) (7c 7–9, showing IBA1 alone, 7c7, γH2AX alone, 7c8, both IBA1-γH2AX, 7c9). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); **p < 0.01; ***p < 0.001. Scale bars 1 mm, 50 μm, and 20 μm. White arrows indicate colocalization between γH2AX+ cells and neurons, astrocytes, and microglia, respectively

Fig. 8

Density of reactive microglia (CD68+ , HLA-DR +) in temporal and occipital white matter (WM) was significantly increased in demented but not in resilient when compared to control brains (a). Representative photomicrographs of CD68+ and HLA-DR+ microglia on sections from control, resilient and demented cases containing temporal WM are displayed on the right (b). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); **p < 0.01; ****p < 0.0001. Scale bar 50 μm

Fig. 9

Density of homeostatic microglia (TMEM119 + , P2RY12 +) in temporal and occipital white matter (WM) was significantly decreased in demented but not in resilient when compared to control brains (a). Representative photomicrographs of TMEM119 + and P2RY12 + microglia on sections from control, resilient and demented cases containing temporal WM are displayed on the right (b). C Control (Braak 0-II); R Resilient (Braak III/IV); D Demented (Braak III/IV); *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Scale bar 50 μm