Microglial activation is a determining factor that allows amyloid-beta accumulation to drive reactive astrogliosis, amplifying inflammatory responses in the brain. This was the conclusion of a study conducted by researchers in Brazil, Canada, and the US.
When microglia are not activated, amyloid stops correlating with elevated plasma levels of glial fibrillary acidic protein (GFAP), indicating that the inflammatory process triggers the cascade of Alzheimer’s disease (AD).
The authors further demonstrated that the microglial-astroglial pathway amplifies tau phosphorylation and contributes to the accumulation of neurofibrillary tangles. These findings reinforce the notion that neuroinflammation is critical for disease progression.
The Study
AD is defined by amyloid-beta and tau deposits; however, there is a growing consensus that these proteinopathies alone do not explain clinical heterogeneity. Individuals with similar amyloid burdens can follow radically different trajectories. Recent studies suggest that microglia and astroglia play key roles in this divergence.
In experimental models, activated microglia release cytokines such as interleukin-1 alpha, TNF, and C1q, which can induce neurotoxic reactive astrogliosis. Reactive astroglia also organize around plaques, follow the topography of amyloid-beta, and amplify tau phosphorylation. However, it remains unclear whether microglia modulate the extent to which amyloids drive astroglial activation in humans. To date, this hypothesis has been supported only by experimental or neuropathological evidence.
To address this question, the researchers evaluated 101 participants aged ≥ 50 years from the Canadian Translational Biomarkers of Aging and Dementia (TRIAD) cohort. All patients underwent plasma analyses of p-tau217 and GFAP, as well as PET imaging of amyloid-beta, translocator protein (TSPO), and tau. The sample included individuals with normal cognition, mild cognitive impairment, and dementia associated with AD.
A second sample of 251 participants from the Wisconsin Registry for Alzheimer’s Prevention and TRIAD cohorts were used to validate the findings using soluble TREM2 (sTREM2), a biomarker of microglial activation in the cerebrospinal fluid. All data were analyzed using regression analysis adjusted for age, sex, and cognitive status, as well as interaction models, regional analyses, and structural equation modeling.
The investigators also incorporated postmortem transcriptome data from the Allen Human Brain Atlas, allowing the evaluation of whether the anatomical distribution of TSPO expression correlates with in vivo amyloid-beta glial interactions.
The central finding was that amyloid-beta was associated with increased GFAP only when microglia were activated. Among individuals with microglial activation, a higher amyloid load was strongly correlated with plasma GFAP levels, whereas in those without activation, this relationship disappeared completely. The interaction between amyloid-beta and TSPO outperformed any model without interaction, and the result was replicated using sTREM2. Amyloid alone is insufficient to induce reactive astrogliosis, and microglial activation determines this outcome.
Brain region analyses showed that the association between amyloid-beta and GFAP in individuals with microglial activation was concentrated in the frontal, parietal, temporal, and cingulate cortices, which are the brain regions most vulnerable to AD. Notably, the areas where amyloid-beta and microglia interacted to increase GFAP levels corresponded to the cortical regions with the highest physiological TSPO expression in the Allen Human Brain Atlas. This topographic correspondence suggests that the basal molecular architecture of the brain helps define the most pronounced amyloid-beta glial interface.
Researchers have also demonstrated that microglial-astroglial signals are directly related to pathological tau. GFAP was associated with p-tau217 and PET detected tau only when microglial activation was present, indicating that neuroinflammation is not merely a late response but an intermediate step in the transition from soluble to aggregated tau.
Structural modeling showed that the amyloid-beta → GFAP → p-tau217 → tau PET → cognition pathway accounted for 76% of the cognitive variability but only in individuals with activated microglia. In the absence of microglial activation, the chain was incomplete, and amyloid-beta did not exert a meaningful effect on cognition.
Microglia and Astroglia
These findings clearly reposition neuroinflammation within the AD cascade.
A 2015 study described close interactions between microglia, astrocytes, and neurons in patients with AD. By placing microglia and astroglia at the center of clinical progression rather than at the margins, this study provides a proof-of-concept in humans.
The inflammatory axis, defined by activated microglia, modulates the timing and location of amyloid deposition, which is biologically relevant for the induction of astrogliosis. These results reinforce the evidence that astrogliosis arises early and may function as a precursor to tauopathy, and that the amyloid-beta → astroglia → tau pathway critically depends on microglial activation.
From a biomarker perspective, this study strengthens the role of plasma GFAP as a practical indicator of amyloid-associated astrogliosis.
In a cross-sectional analysis of more than 300 individuals across the AD continuum, researchers showed that plasma GFAP increases in both preclinical and symptomatic phases, shows greater alterations than CSF GFAP, and better discriminates amyloid-beta-positive from amyloid-beta-negative individuals, including those in the early stages of the disease.
By demonstrating that the association between amyloid-beta and GFAP depends on microglial activation, the current study helps clarify these findings: Plasma GFAP is not simply a marker of nonspecific damage but also a sensitive indicator of amyloid-linked glial neuroinflammation, with direct implications for trial enrolment and monitoring.
Conceptually, these findings support the expansion of the framework to include an inflammatory glial component, given that the chronology and intensity of neuroinflammation modulate the risk for clinical progression in amyloid-beta-positive individuals. Glial markers included YKL 40, MCP 1, GFAP, and sTREM2.
These results provide empirical evidence that amyloid-beta microglial astroglia interactions explain a meaningful proportion of the cognitive variance, supporting the use of these biomarkers for risk stratification and therapeutic design.
Interventions aimed at reducing the formation of A1 astrocytes or restoring synaptic support have shown neuroprotective benefits in tauopathy models.
Therapeutic and Practical Implications
In this context, the results suggest that anti-amyloid-beta treatments may be more effective when combined with strategies that modulate the glial response, particularly in individuals showing evidence of hyperreactive microglia and astroglia.
Selecting individuals based on a high-risk glial profile, defined, for example, by elevated plasma GFAP levels together with markers of microglial activation, may increase the likelihood of detecting clinical effects in phase 3 trials.
In summary, the AD cascade resembles a less rigid amyloid-beta → tau → cognitive impairment sequence and a network in which microglia and astroglia function as obligatory amplifiers between amyloid deposition and clinical manifestations.
This scenario has two clinical implications: First, it reinforces the importance of incorporating glial biomarkers into research and, potentially, clinical practice. It also opens up a therapeutic pathway beyond plaque removal, targeting neuron-glial interactions to slow disease progression.
Future trials should evaluate microglial modulators, cytokine-specific blockers, and interventions aimed at reactive astrocytes. Longitudinal studies are essential to define the temporality of microglial-astroglial tau interactions and identify therapeutic windows.
Daniela Barros is a journalist with postgraduate training in social journalism from the Pontifical Catholic University of São Paulo, São Paulo, Brazil. She is a special student in the Department of Social Medicine at Ribeirão Preto Medical School, University of São Paulo, São Paulo.
This story was translated from Medscape’s Portuguese edition.
Ju Luo†
Yang Yang
