Hypothesis

Aging microglia become sensitized to NLRP3 inflammasome activation, leading to chronic, low‑grade IL‑1β release that shifts the BCM‑like threshold for synaptic modification in hippocampal CA1 neurons toward long‑term potentiation (LTP) and away from long‑term depression (LTD). This creates a state of over‑consolidation: synapses are potentiated more easily and resist depotentiation, manifesting behaviorally as reduced memory flexibility (impaired reversal learning, heightened perseveration) while baseline acquisition appears intact. We predict that acute, transient NLRP3 inhibition combined with mild environmental novelty will reset the modification threshold, restoring reversal learning without compromising initial memory formation, whereas chronic NLRP3 blockade will impair acquisition by excessively lowering the threshold.

Mechanistic Rationale

1. Inflammasome‑IL‑1β signaling modulates calcium‑dependent kinases/phosphatases – IL‑1β activates PKA and CaMKII via downstream NF‑κB and MAPK pathways, increasing phosphorylation of GluA1 (Ser845) and promoting AMPAR insertion (1). Simultaneously, IL‑1β suppresses protein phosphatase‑1 (PP1) activity, which is required for LTD‑associated dephosphorylation of GluA1 (2). The net effect is a lowered LTP threshold and raised LTD threshold, consistent with a BCM‑like shift.
2. Microglial priming amplifies the signal – Aged microglia exhibit exaggerated NLRP3 activation upon secondary stimuli (4, 5). This creates a spatially heterogeneous but persistent IL‑1β milieu that preferentially affects synapses with high activity histories, reinforcing existing traces.
3. Behavioral signature – Over‑consolidation predicts preserved initial learning (LTP‑dependent) but deficits in tasks requiring memory updating: reversal learning in the Morris water maze, extinction of conditioned fear, and probabilistic reversal learning where perseverative errors increase.

Experimental Design (Testable & Falsifiable)

Subjects: 20‑month‑old C57BL/6J mice (n=12/group).

Groups:

• Vehicle + home‑cage (control)
• MCC950 (NLRP3 inhibitor, 10 mg/kg i.p., 30 min before testing) + home‑cage
• Vehicle + novelty exposure (5 min exposure to a novel object arena 24 h before testing)
• MCC950 + novelty exposure (same timing as above)

Procedures:

1. Baseline acquisition – Standard Morris water maze (hidden platform) for 5 days; measure escape latency and path length.
2. Reversal learning – Platform moved to opposite quadrant on day 6; measure trials to criterion and perseverative entries into former location.
3. Fear extinction – Auditory fear conditioning followed by extinction trials; assess freezing across sessions.
4. Ex vivo slice physiology – After behavioral testing, prepare hippocampal slices; record LTP (theta‑burst) and LTD (low‑frequency stimulation) in CA1; quantify GluA1‑Ser845 phosphorylation via Western blot.
5. Neuroinflammatory read‑out – ELISA for IL‑1β and IL‑18 in hippocampal lysates.

Predictions:

• Vehicle controls will show normal acquisition but significantly increased perseverative errors and slower reversal vs. young adults (baseline).
• MCC950 alone will normalize reversal performance without altering acquisition latency.
• Novelty exposure alone will produce a modest improvement in reversal (due to endogenous neuromodulatory arousal) but not reach control levels.
• The combined MCC950 + novelty group will show the fullest restoration of reversal learning, approaching young‑adult performance, while acquisition remains unchanged.
• Electrophysiology will reveal restored LTD magnitude and normalized GluA1‑Ser845 phosphorylation only in the combined group.
• IL‑1β levels will be reduced proportionally to behavioral improvement.

Falsifiability: If MCC950 fails to improve reversal learning, or if improvement occurs without novelty and is accompanied by impaired acquisition, the hypothesis that NLRP3‑driven IL‑1β specifically creates an over‑consolidation state (rather than a global plasticity deficit) would be refuted. Likewise, if LTD is not rescued despite behavioral improvement, the proposed synaptic threshold mechanism would need revision.