Berberine‑AMPK Signaling Shifts Age‑Related Neuronal Pruning From Passive Loss to Active Metabolic Quality Control

Fatima Moreno (PhD)· Mar 23, 2026

Hypothesis

Berberine, via AMPK activation and PCSK9 inhibition, modulates the brain’s energy‑sensing machinery to convert age‑related neuronal loss from a passive degenerative process into an active, metabolically driven eviction of inefficient neurons.

Mechanistic Rationale

AMPK activation signals low cellular ATP, a cue that in developing tissue triggers complement‑mediated synapse pruning and microglial phagocytosis ([1][2]).
In aging neurons, sustained AMPK activation could raise the threshold for metabolic competence, marking cells with low oxidative phosphorylation or high ROS as targets for removal.
Berberine’s inhibition of PCSK9 raises hepatic LDL‑R activity, lowering circulating LDL and potentially reducing lipid‑laden microglia that otherwise impair phagocytic capacity ([3][4]).
Combined, these actions may shift the balance toward selective removal of high‑energy‑cost, low‑output neurons, thereby increasing ATP yield per surviving cell.

Testable Predictions

Marker shift – Chronic low‑dose berberine in 20‑month‑old mice will decrease neuronal death markers unrelated to activity (e.g., cleaved caspase‑3 in densely connected hippocampal CA1) while increasing markers of microglial phagocytosis (Iba1⁺/CD68⁺ contacts with NeuN⁺ cells lacking synaptic vesicle protein SV2A).
Metabolic read‑out – Isolated neurons from berberine‑treated aged brains will show higher ATP/ADP ratios and lower lactate production compared with vehicle controls, indicating improved metabolic efficiency per cell.
Network efficiency – In vivo two‑photon calcium imaging will reveal reduced pairwise correlation variance and increased information transmission (mutual entropy) in local circuits despite a ~15% reduction in total neuron number.
Lipid mediator link – Berberine treatment will lower hippocampal PCSK9 protein and raise LDL‑R expression, correlating with the degree of neuronal eviction (Spearman r > 0.6).

Experimental Design (Falsifiable)

Groups (n=10 per group): young (3 mo) vehicle, aged (20 mo) vehicle, aged berberine (50 mg/kg/day, oral, 12 weeks), aged metformin (positive control), aged berberine + AMPK inhibitor (Compound C).
Readouts at endpoint: immunostaining for C1q, cleaved caspase‑3, Iba1/CD68, NeuN, SV2A; lipidomics for LDL‑R/PCSK9; Seahorse ATP assay on acutely isolated hippocampi; in vivo calcium imaging during whisker‑stimulation task.
Falsification: If berberine does not alter the ratio of phagocytic markers to apoptotic markers, or if ATP/ADP ratios remain unchanged while neuron number stays constant, the hypothesis that berberine drives active metabolic eviction is unsupported. Conversely, a selective increase in phagocytic clearance of low‑activity neurons without rise in apoptotic death would support it.

Implications

Confirming that berberine can harness the brain’s intrinsic quality‑control system would reframe age‑related neuronal loss as a potentially tunable, energy‑sensing process rather than irreversible damage, opening metabolic‑based strategies to preserve circuit efficiency.

Mechanistic Rationale

AMPK activation signals low cellular ATP, a cue that in developing tissue triggers complement‑mediated synapse pruning and microglial phagocytosis ([1][2]).

In aging neurons, sustained AMPK activation could raise the threshold for metabolic competence, marking cells with low oxidative phosphorylation or high ROS as targets for removal.

Berberine’s inhibition of PCSK9 raises hepatic LDL‑R activity, lowering circulating LDL and potentially reducing lipid‑laden microglia that otherwise impair phagocytic capacity ([3][4]).

Combined, these actions may shift the balance toward selective removal of high‑energy‑cost, low‑output neurons, thereby increasing ATP yield per surviving cell.

Testable Predictions

Marker shift – Chronic low‑dose berberine in 20‑month‑old mice will decrease neuronal death markers unrelated to activity (e.g., cleaved caspase‑3 in densely connected hippocampal CA1) while increasing markers of microglial phagocytosis (Iba1⁺/CD68⁺ contacts with NeuN⁺ cells lacking synaptic vesicle protein SV2A).

Metabolic read‑out – Isolated neurons from berberine‑treated aged brains will show higher ATP/ADP ratios and lower lactate production compared with vehicle controls, indicating improved metabolic efficiency per cell.

Network efficiency – In vivo two‑photon calcium imaging will reveal reduced pairwise correlation variance and increased information transmission (mutual entropy) in local circuits despite a ~15% reduction in total neuron number.

Lipid mediator link – Berberine treatment will lower hippocampal PCSK9 protein and raise LDL‑R expression, correlating with the degree of neuronal eviction (Spearman r > 0.6).

Experimental Design (Falsifiable)

Groups (n=10 per group): young (3 mo) vehicle, aged (20 mo) vehicle, aged berberine (50 mg/kg/day, oral, 12 weeks), aged metformin (positive control), aged berberine + AMPK inhibitor (Compound C).

Readouts at endpoint: immunostaining for C1q, cleaved caspase‑3, Iba1/CD68, NeuN, SV2A; lipidomics for LDL‑R/PCSK9; Seahorse ATP assay on acutely isolated hippocampi; in vivo calcium imaging during whisker‑stimulation task.

Falsification: If berberine does not alter the ratio of phagocytic markers to apoptotic markers, or if ATP/ADP ratios remain unchanged while neuron number stays constant, the hypothesis that berberine drives active metabolic eviction is unsupported. Conversely, a selective increase in phagocytic clearance of low‑activity neurons without rise in apoptotic death would support it.