Conclusions

This interaction-based secondary reanalysis of GSE85718 identified tissue-dependent patterns consistent with NMN-associated attenuation of selected age-related transcriptional changes. It did not produce FDR-significant gene-level discoveries, and it should not be presented as definitive evidence that any single gene mediates NMN's effects. Its value is in prioritization. Under relaxed, hypothesis-generating criteria, NMN-rescue candidates were identified in skeletal muscle, liver, and WAT, but overlap across tissues was modest, and no gene was rescued in all three tissues. Within that tissue-specific landscape, RAB11A and CPT2 stand out. RAB11A is the strongest cross-tissue trafficking candidate and points toward endosomal recycling, vesicular movement, and membrane homeostasis as under-appreciated parts of the NMN response. Its repeated GSEA leading-edge representation supports prioritization but does not independently establish a change in endocytic recycling. CPT2 is the strongest fatty-acid oxidation candidate, linking the transcriptional findings to substrate utilization in skeletal muscle and WAT. The available mitochondrial analysis does not support broad mitochondrial restoration; instead, CPT2 suggests a narrower carnitine-shuttle and fatty-acid oxidation hypothesis. Liver adds the clearest pathway-level result, with suppression of fatty-acyl-CoA and fatty acid elongation programs.

The broader implication is that NAD+ precursor biology should move beyond generic claims about mitochondrial improvement. NMN may act through tissue-specific programs that include cellular logistics, lipid handling, and selected metabolic enzymes. These findings provide a practical roadmap for validation rather than proof of mechanism: confirm RAB11A and CPT2 at RNA and protein levels, test trafficking and fatty-acid oxidation functionally, re-run mitochondrial enrichment with full MitoCarta3.0, and connect tissue-specific signatures to metabolic outcomes.