Hypothesis

Age‑related loss of cholinergic neurons is not passive vulnerability but an active, gut‑mediated culling process in which circulating indole‑3‑propionic acid (IPA) activates neuronal pregnane X receptor (PXR), shifting metabolism toward glycolysis and triggering AMPK‑dependent synaptic pruning that culminates in soma death.

Mechanistic Rationale

The gut‑brain axis provides a conduit for microbial metabolites to influence neuronal homeostasis. IPA, produced by commensal Clostridia, enhances intestinal barrier integrity via epithelial PXR activation [4]. When the barrier deteriorates with age, microbial products such as CML leak into the circulation and impair microglial mitochondria [3]. We propose that, in parallel, increased systemic IPA reaches the brain and binds neuronal PXR, which is expressed at low levels in cholinergic populations [5]. Neuronal PXR activation upregulates cytochrome P450 enzymes and shifts cellular metabolism from oxidative phosphorylation to aerobic glycolysis, lowering ATP yield. This energetic deficit activates AMPK, which phosphorylates downstream targets (e.g., ULK1) to initiate autophagy of synaptic vesicles and postsynaptic densities. Chronically weakened synapses trigger retrograde degeneration signals that lead to soma loss, preferentially affecting high‑firing, metabolically expensive cholinergic neurons that rely on mitochondrial ATP for acetylcholine synthesis.

Testable Predictions

1. Elevating circulating IPA in young mice will recapitulate age‑like cholinergic neuron loss in the basal forebrain and distal colon enteric plexus.
2. Neuron‑specific knockout of PXR (ChAT‑Cre;Pxr^fl/fl) will protect cholinergic neurons from loss despite induced gut barrier permeability (e.g., DSS colitis).
3. Pharmacological inhibition of AMPK (Compound C) will rescue synaptic markers and prevent soma death in IPA‑treated wild‑type mice.
4. Metabolomic profiling of aged brain interstitial fluid will show elevated IPA and glycolytic intermediates alongside reduced TCA‑cycle metabolites in cholinergic regions.

Experimental Approach

• Metabolite manipulation: Administer IPA (10 mg/kg/day) via osmotic pump to 3‑month‑old mice for 8 weeks; assess cholinergic neuron numbers (ChAT immunostaining) in medial septum and myenteric plexus.
• Genetic targeting: Cross ChAT‑Cre mice with Pxr^fl/fl; induce barrier leak with 2 % DSS for 7 days; quantify neuron loss and microglial activation (Iba1).
• Metabolic read‑outs: Perform Seahorse analysis on FACS‑sorted basal forebrain neurons to measure OCR/ECAR; Western blot for p‑AMPK, LC3‑II, and synaptic proteins (Synaptophysin, PSD‑95).
• Rescue experiments: Treat IPA‑exposed mice with Compound C (10 mg/kg IP) or overexpress a dominant‑negative AMPK isoform via AAV‑ChAT vector.

Falsifiability

If neuronal loss proceeds unchanged in PXR‑deficient cholinergic neurons despite elevated IPA and barrier breach, or if AMPK inhibition fails to preserve synapses and soma, the hypothesis would be refuted. Conversely, demonstration that IPA alone is sufficient to induce metabolic switching and synaptic autophagy specifically in cholinergic neurons would support the model.