Hypothesis

Age-related decline in brain aerobic glycolysis triggers increased deposition of perineuronal nets (PNNs) around excitatory neurons, which stabilizes existing synaptic ensembles and reduces the brain’s ability to incorporate surprising information. This mechanism shifts the rigidity observed in aging from a passive loss of plasticity to an active, metabolism‑driven stabilization of neural maps.

Mechanistic Basis

• Normal aging shows a progressive loss of aerobic glycolysis, approaching zero by ~60 years [Loss of brain aerobic glycolysis characterizes normal aging].
• Reduced glycolysis lowers ATP availability for enzymes that degrade extracellular matrix components, favoring net synthesis.
• PNNs envelop parvalbumin‑positive interneurons and pyramidal cells, restricting spine turnover and long‑term potentiation [Neural ageing is characterized by impaired synaptic plasticity].
• Stabilized networks increase the signal‑to‑noise ratio for familiar patterns, making the system less tolerant of prediction error—a computational signature of over‑consolidation.

Testable Predictions

1. Older individuals with higher CSF lactate (a proxy for glycolysis) will show thinner PNN layers in posterior cortex, measured via PET‑ligand binding or post‑mortem histology.
2. Pharmacological enhancement of glycolysis (e.g., with dichloroacetate) in aged mice will reduce PNN intensity, increase spine turnover, and improve performance on surprise‑based learning tasks (e.g., reversal learning with unexpected outcome changes).
3. Genetic knockdown of brevican, a core PNN component, will rescue plasticity deficits in aged animals without altering neuronal death markers.
4. In humans, individual differences in aerobic glycolysis (assessed by FDG‑PET) will predict the magnitude of the mismatch negativity response to auditory oddball stimuli, linking metabolism to surprise detection.

Experimental Approach

• Human cohort: Recruit adults aged 20‑80, acquire FDG‑PET to quantify cortical aerobic glycolysis, MR spectroscopy for lactate, and EEG mismatch negativity. Correlate glycolysis metrics with PNN‑sensitive MRI markers (e.g., magnetization transfer ratio) and behavioral surprise tolerance.
• Mouse model: Use 24‑month‑old C57BL/6 mice. Treat one group with dichloroacetate (500 mg/kg i.p. daily for 2 weeks) to boost glycolysis, another with control saline. Assess PNN density via Wisteria floribunda agglutinin staining, spine density via two‑photon imaging, and reversal learning in a tactile discrimination task where the reward contingency is reversed unexpectedly.
• Intervention validation: Include a group receiving chondroitinase ABC to digest PNNs, confirming that any glycolysis‑dependent effects are mediated through the extracellular matrix.

Potential Limitations and Confounds

• Age‑related vascular changes could influence FDG‑PET signals independent of neuronal glycolysis; we will control for cerebrovascular reactivity using breath‑hold CO₂ challenges.
• Systemic effects of dichloroacetate (e.g., on peripheral metabolism) might affect behavior; we will monitor blood glucose and lactate and include a pair‑fed control group.
• PNN labeling may vary across brain regions; we will focus on sensorimotor and prefrontal cortices where surprise‑dependent plasticity has been documented.

By linking a specific metabolic deficit to a structural mechanism that actively locks neural configurations, this hypothesis reframes age‑related rigidity as a maladaptive over‑stabilization rather than a simple wear‑and‑tear process, offering a clear, falsifiable pathway for therapeutic intervention.