Microglial activation occurs late during preclinical Alzheimer's disease

Abstract

Sporadic Alzheimer's disease (AD) is marked by a lengthy preclinical phase during which patients are nonsymptomatic but show pathology in variable manifestations. Whether or not neuroinflammation occurs in such nondemented individuals is unknown. We evaluated the medial temporal lobe of 66 nondemented subjects, aged 42–93, in terms of tau pathology, Aβ deposition, and microglial activation. We show that 100% of subjects had neurofibrillary degeneration (NFD), 35% had Aβ deposits, and 8% revealed microglial activation in individuals where early amyloid formation was apparent by Congo Red staining. Amyloid‐induced neuroinflammatory clusters of Iba1, CD68, and ferritin‐positive microglia were evident in the immediate vicinity of aggregated Aβ. Microglia in the adjacent neuropil were nonactivated. Thus, neuroinflammation in AD represents a highly localized phagocyte reaction, essentially a foreign body response, geared toward removal of insoluble Aβ. Because clustered microglia in some amyloid plaques were dystrophic and ferritin‐positive, we hypothesize that these cells were exhausted by their attempts to remove the aggregated, insoluble Aβ. Our findings show that the sequence of pathologic events in AD begins with tau pathology, followed by Aβ deposition, and then by microglial activation. Because only 8% of our subjects revealed all three hallmark pathologic features, we propose that these nondemented individuals were near the threshold of transitioning from nonsymptomatic to symptomatic disease. The onset of neuroinflammation in AD may thus represent a tipping point in AD pathogenesis. Our study suggests that the role of microglia in AD pathogenesis entails primarily the attempted removal of potentially toxic, extracellular material.

Dystrophic microglia in late‐onset Alzheimer's disease

“…The early symptomatic stages of motor neuron disease in this model are characterized by massive microglial activation in the spinal cord ventral grey matter, but this neuroinflammatory response wanes during end stage disease when microglia transition from activated to dystrophic cells and the spinal cord grey matter becomes inundated with fragmented microglia (Fendrick et al, 2007;). Thus, there is a parallel situation to what we have observed in LOAD, namely, that microglial activation occurring in response to amyloid eventually converts to microglial dystrophy (Streit et al, 2018), illustrating how immune activation can lead to immune exhaustion in pathological situations. The fact that this transition from activation to dystrophy occurs in rats bearing a Cu/Zn superoxide dismutase mutation strongly suggests that genetic defects in antioxidant enzymes contribute to microglial dystrophy, and that dystrophy is therefore likely caused by oxidative damage.…”

“…We also know that treatment with anti‐inflammatory drugs in clinical trials does not halt the aging‐related progression of neurodegeneration and cognitive decline, and thus it is clear that NFD in AD has been frequently misrepresented as a condition characterized by an inflammatory etiology (Streit, ). Although there is an inflammatory component to AD, neuroinflammation in LOAD represents a very focal and targeted response of microglia to abnormal extracellular amyloid deposits that occurs late during preclinical LOAD when NFD is already well established (Braak et al, ; Streit et al, ). Moreover, microglial activation (neuroinflammation) triggered by amyloid formation may not be as chronic as commonly thought, because our work shows that even amyloid plaque‐associated, activated microglia undergo dystrophic degeneration (Streit et al, , ).…”

“…Although there is an inflammatory component to AD, neuroinflammation in LOAD represents a very focal and targeted response of microglia to abnormal extracellular amyloid deposits that occurs late during preclinical LOAD when NFD is already well established (Braak et al, ; Streit et al, ). Moreover, microglial activation (neuroinflammation) triggered by amyloid formation may not be as chronic as commonly thought, because our work shows that even amyloid plaque‐associated, activated microglia undergo dystrophic degeneration (Streit et al, , ). These observations suggest that immune exhaustion rather than chronic inflammation is more consequential in terms of LOAD pathogenesis, and thus future therapeutic strategies should perhaps be guided by this insight.…”

“…What most researchers failed to realize and take into consideration is the extraordinary specificity of the microglial reaction to amyloid, namely, that inflammation is spatially confined only to the immediate vicinity of amyloid. Microglia located in the neuropil between amyloid plaques are nonactivated and ramified (Prokop, Miller, & Heppner, ; Streit et al, ). This kind of highly localized neuroinflammation is consistent with a targeted cellular response to a foreign body (amyloid) by showing microglia acting as macrophages wanting to clear out an undesirable, insoluble substance.…”

“…Increased brain iron likely contributes to increasing levels of oxidative stress and to aggregation of Aβ into amyloid (R. J. Castellani, Moreira, Perry, & Zhu, ), which triggers microglial activation. At some point all of the various pathological components (NFD, dystrophy, amyloid, neuroinflammation, iron accumulation) converge to where preclinical LOAD turns into clinical disease thus marking a tipping point in pathogenesis (Streit et al, ).…”