The assumption that calming neuroinflammation would protect Alzheimer's-affected brain circuits turns out to be far more complicated — and a new set of findings from transgenic mouse research challenges a popular therapeutic strategy in ways that matter for anyone tracking the microglia-targeting drug pipeline.
Using acute brain slices from APP/PS1 mice at 9–11 months of age — a model that recapitulates amyloid pathology — investigators mapped the electrophysiological consequences of microglial suppression via the CSF1R inhibitor GW2580. In untreated APP/PS1 animals, gamma oscillations (30–80 Hz) were significantly reduced in amplitude, beta-band power was elevated, and epileptiform activity was heightened — a constellation consistent with the oscillatory slowing recognized as a neurophysiological hallmark of Alzheimer's disease. Immunohistochemical staining confirmed losses in parvalbumin- and somatostatin-positive inhibitory interneurons, two populations critical for sustaining gamma rhythms. Gabazine challenge further exposed a compromised inhibitory tone via GABA-A receptor antagonism sensitivity. Crucially, when microglial proliferation was curtailed by GW2580, oscillatory disruptions were exacerbated rather than rescued, and network hyperexcitability increased.
This finding carries significant weight in the broader context of microglia-targeting therapeutics. CSF1R inhibitors have attracted pharmaceutical interest as anti-neuroinflammatory candidates, partly informed by studies showing amyloid plaque reductions with microglial depletion. Yet microglia are not uniformly pathological in Alzheimer's disease — they also perform protective synaptic surveillance and support inhibitory circuit maintenance. This study implies that reactive microglia may be partially compensating for interneuron loss or otherwise stabilizing network excitability, such that their removal tips the system toward seizure-like dynamics. The work is preclinical and restricted to a single amyloid overexpression model, which does not fully replicate the complexity of human Alzheimer's disease. Nonetheless, the demonstration that a seemingly logical anti-inflammatory intervention can worsen network physiology is an important cautionary signal for translational programs. It argues for circuit-level electrophysiological readouts as essential biomarkers in any microglia-targeting trial.