Hypothesis

Aging brains shift from surprise‑driven synaptic remodeling to a state of over‑consolidation because the neuromodulatory signal that tags unexpected events—phasic norepinephrine release from the locus coeruleus (LC)—declines with age. When surprise signaling drops, the cost‑benefit calculation for updating synaptic models tilts toward stability, slowing protein turnover and locking circuits into rigid configurations. This mechanism explains why cognitive flexibility deteriorates while basic memory encoding remains intact.

Mechanistic reasoning

1. Surprise detection → LC‑NE burst – Unexpected sensory outcomes trigger a phasic LC‑NE burst that activates β‑adrenergic receptors on pyramidal neurons, boosting cAMP‑PKA signaling and ubiquitin‑proteasome activity at synapses (1).
2. NE‑dependent proteostasis – β‑adrenergic signaling phosphorylates synaptic scaffolding proteins, marking them for degradation and preventing the accumulation of over‑stabilized complexes (2).
3. Age‑related LC decline – Rodent and human studies show reduced LC neuron firing and lower NE levels in aging (3).
4. Resulting shift – With weaker NE bursts, synaptic proteins acquire longer half‑lives, leading to the proteostasis imbalance described in the seed paper (over 1,700 proteins accumulate at synapses) (4).
5. Behavioral read‑out – The network becomes less able to switch between default‑mode and executive control, manifesting as reduced cognitive flexibility while hippocampal‑dependent memory formation stays relatively preserved (5).

Novel prediction

If the over‑consolidation state is driven by low surprise signaling, then artificially increasing phasic LC‑NE activity in aged animals should restore synaptic protein turnover, reduce the accumulation of stable synaptic complexes, and rescue behavioral flexibility without altering overall neuron counts.

Experimental test (falsifiable)

• Subjects: Young (3‑mo) and aged (24‑mo) mice.
• Groups: Aged mice receive either (a) chemogenetic activation of LC neurons (hM3Dq + CNO) to evoke phasic NE bursts during a surprise‑based learning task, (b) vehicle control, (c) aged mice with LC inhibition (hM4Di + CNO) as a negative control; young mice receive vehicle as baseline.
• Read‑outs:
  • In vivo microdialysis to confirm increased NE peaks during task.
  • Synaptic protein half‑life measured via pulsed SILAC in hippocampal and PFC tissue after 2 weeks of treatment.
  • Behavioral flexibility assessed with reversal learning and attentional set‑shifting tasks.
  • Memory encoding tested with object‑location memory (should remain unchanged across groups).

Expected outcome: Aged mice with LC activation show NE burst amplitudes comparable to young, shortened synaptic protein half‑lives (approaching young values), improved reversal learning performance, and no change in object‑location memory. Vehicle‑treated aged mice retain long protein half‑lives and poor flexibility. LC‑inhibited aged mice perform worse than vehicle controls.

Falsification: If LC activation fails to normalize synaptic protein turnover or does not improve flexibility despite verified NE elevation, the hypothesis that surprise‑driven NE signaling gates synaptic over‑consolidation would be refuted.