Hypothesis

Age‑associated shifts in the gut microbiota increase production of phenylacetic acid (PAA) and other microbial metabolites that peripheral‑ly activate CD38 in endothelial and immune cells, raising systemic NAD+ consumption. Simultaneously, PAA‑induced mitochondrial ROS impairs vagal afferent signaling, reducing cholinergic tone that normally sustains neuronal NAMPT activity. The combined effect is a brain‑specific NAD+ deficit that drives microglial senescence and cognitive decline.

Mechanistic Model

• Microbial trigger: Aged microbiota overexpress PAA‑producing pathways (e.g., phenylacetate‑CoA ligase) [1]. Elevated PAA crosses the compromised gut barrier and reaches the circulation.
• Peripheral NAD+ sink: PAA stimulates CD38 expression in gut‑associated macrophages and endothelial cells via NF‑κB signaling, increasing NAD+ glycohydrolase activity and lowering plasma NAD+ [3].
• Vagal dampening: PAA‑induced mitochondrial H2O2 in vagal ganglion neurons reduces action potential firing, decreasing afferent input to the dorsal motor nucleus of the vagus [2]. Lower vagal output lessens acetylcholine release in the brainstem, which normally activates α7‑nicotinic receptors on microglia to boost NAMPT transcription.
• Central NAD+ failure: Reduced neuronal NAMPT limits NAD+ salvage, while accumulated NADH from impaired oxidative phosphorylation (due to lactate overload) further skews the NAD+/NADH ratio. Lactate, elevated because PAA‑exacerbated microglial glycolysis shifts to anaerobic metabolism, inhibits SIRT1 deacetylase activity, exacerbating inflammaging.
• Outcome: NAD+‑dependent sirtuin and PARP activity fall, promoting microglial senescence, neuroinflammation, and memory loss.

Testable Predictions

1. Metabolite correlation: Plasma PAA levels will positively correlate with CD38 activity (measured by NAD+ metabolomics) and negatively with brain NAD+ in aged mice.
2. Vagal rescue: Chemogenetic activation of vagal afferents in old mice will restore brain NAMPT expression and NAD+ levels despite high PAA.
3. CD38 blockade: Treating aged mice with a CD38 inhibitor (e.g., 78c) will normalize plasma NAD+ and improve cognition without altering microbiota composition.
4. Lactate link: Intraventricular lactate infusion will blunt the NAD+‑raising effect of vagal stimulation, confirming lactate’s inhibitory role on SIRT1.
5. FMT specificity: Fecal microbiota transplant from young donors engineered to lack phenylacetate‑CoA ligase will fail to improve brain NAD+, indicating PAA as a necessary mediator.

Experimental Approach

• Mouse cohorts: Young (3 mo), aged (20 mo), aged + PAA‑producing bacteria colonization, aged + CD38 inhibitor, aged + vagal DREADD activation.
• Measurements: (a) fecal PAA via GC‑MS; (b) plasma NAD+, NADH, CD38 activity via enzymatic assay; (c) brain NAD+/NADH by LC‑MS/MS; (d) vagal firing in nodose ganglia using ex‑vivo electrophysiology; (e) microglial senescence markers (p16, SASP) and cognitive performance (Morris water maze).
• Interventions: Antibiotics to deplete PAA producers, AAV‑mediated NAMPT overexpression in neurons as a positive control.

If PAA‑driven CD38 activation and vagal‑mediated NAD+ salvage suppression are central, manipulating either node should decouple gut dysbiosis from brain NAD+ loss and rescue cognition. Conversely, if brain NAD+ remains low despite these interventions, the hypothesis is falsified, prompting search for alternative gut‑derived NAD+ consumptive pathways.