Hypothesis

Aging shifts the composition and activity of AP-1 transcription factors downstream of sustained JNK signaling from plasticity‑permissive heterodimers (c‑Jun/JunB) to consolidation‑locked complexes (c‑Jun/JunD) that preferentially stabilize existing synaptic ensembles. This molecular switch converts adaptive memory consolidation into maladaptive over‑consolidation, producing preserved reference memory alongside impaired reversal learning and reduced tolerance to novelty.

Mechanistic Model

1. Stress‑induced JNK activation – With age, low‑grade inflammatory cues and mitochondrial dysfunction maintain basal MKK4/7 activity, leading to persistent JNK1/2 phosphorylation (JNK functions as a stress-activated kinase).
2. Isoform‑specific AP-1 remodeling – Chronic JNK favors c‑Jun phosphorylation and promotes JunD expression while suppressing JunB via feedback inhibition (chronically activated JNK suppresses transcriptional activity). JunD‑containing AP-1 dimers bind distinct promoter elements that drive genes encoding synaptic adhesion molecules (e.g., NrCAM, SynCAM1) and actin‑crosslinking proteins (e.g., α‑actinin), reinforcing postsynaptic density stability.
3. Synaptic ensemble locking – Elevated JunD‑AP-1 increases transcription of stabilization factors and reduces expression of plasticity‑related genes (Arc, Homer1a, BDNF). The net effect is a higher threshold for long‑term potentiation (LTP) induction and a lower threshold for long‑term depression (LTD) of weak synapses, biasing network dynamics toward preservation of strongly co‑active ensembles.
4. Network‑level phenotype – Over‑consolidated ensembles resist updating, manifesting as intact spatial reference memory but deficient reversal learning and reduced exploratory behavior—exactly the dissociation predicted by the over‑consolidation seed idea.

Novel Mechanistic Insight

We propose that JunD acts as a "molecular brake" on AP-1 transcriptional versatility. Unlike c‑Jun, which can form transient heterodimers with JunB or ATF2 to drive immediate‑early gene bursts, JunD preferentially recruits co‑repressors (HDAC3, NCoR) to AP-1 sites, converting stress‑responsive promoters into constitutively repressed states. This brake is engaged only after prolonged JNK activity, providing a molecular timer that translates cumulative stress into synaptic rigidity.

Testable Predictions

• Prediction 1: In aged mice, nuclear c‑Jun/JunD heterodimers will be enriched in hippocampal excitatory neurons compared with young adults, while c‑Jun/JunB complexes decline.
• Prediction 2: Pharmacological inhibition of JNK (SP600125) or genetic knockdown of JunD in aged animals will increase c‑Jun/JunB binding at plasticity‑gene promoters and rescue reversal learning without affecting reference memory.
• Prediction 3: Viral overexpression of JunD in young mice will recapitulate the aged over‑consolidation phenotype: normal reference memory, impaired reversal learning, and increased synaptic stability markers (PSD‑95, GluA2 surface expression).
• Prediction 4: Single‑cell ATAC‑seq will reveal reduced chromatin accessibility at Arc and BDNF promoters specifically in JunD‑high neuronal clusters of aged brains.

Experimental Approach

1. Biochemical – Perform co‑immunoprecipitation followed by Western blot for c‑Jun, JunB, JunD in hippocampal lysates from young (3 mo) and aged (24 mo) mice; quantify heterodimer ratios.
2. Pharmacological/Genetic – Treat aged mice with JNK inhibitor or inject AAV‑shJunD into CA1; assess behavior in Morris water maze (reference memory) and reversal version of the task.
3. Molecular – Conduct ChIP‑seq for c‑Jun and JunD to map genome‑wide binding; integrate with RNA‑seq to identify deregulated plasticity genes.
4. Electrophysiology – Measure LTP/LTD thresholds in hippocampal slices from each condition to link molecular changes to synaptic plasticity properties.

Falsifiability

If JNK inhibition or JunD knockdown fails to improve reversal learning while leaving reference memory unchanged, or if JunD overexpression in young mice does not induce rigidity phenotypes, the hypothesis that chronic JNK/AP-1 rewiring drives over‑consolidation would be refuted. Conversely, confirmation of the predicted molecular and behavioral shifts would support the over‑consolidation model and suggest a target for cognitive flexibility interventions.