Hypothesis Aged stromal cells secrete exosomes enriched in miR-21-5p and miR-155-3p that are taken up by nearby T cells and macrophages, where they suppress ULK1 and ATG5 expression and concomitantly activate PI3K/Akt/mTORC1 signaling. This paracrine inhibition of autophagy leads to mitochondrial ROS accumulation, NLRP3 inflammasome priming, and a feed‑forward loop that sustains inflammaging. Blocking EV release or neutralizing these miRNAs restores autophagic flux and reduces inflammatory output even in an aged microenvironment.

Mechanistic rationale

• Immunosenescence is characterized by defective autophagy in lymphocytes and myeloid cells, yet core autophagy proteins remain expressed [1].
• Extracellular vesicles (EVs) from old serum carry a distinct miRNA repertoire that alters recipient cell phenotype [2].
• miR-21-5p targets PTEN, leading to Akt activation and mTORC1 up‑regulation [3].
• miR-155-3p directly binds the 3′‑UTR of ULK1, impairing its translation [4].
• Combined, these EV‑derived miRNAs create a dual hit: translational blockade of autophagy initiators and hyperactive mTORC1, reproducing the active suppression described in aged fibroblasts.

Testable predictions

• EVs from old serum show elevated levels of miR-21-5p and miR-155-3p compared with young serum (quantified by qPCR after ultracentrifugation or size‑exclusion chromatography).
• Incubation of naïve young bone‑marrow‑derived macrophages with old‑mouse EVs decreases the LC3‑II/I ratio and increases phospho‑S6K within 6 h; this effect is blocked by GW4869 (EV release inhibitor) or by pretreating EVs with RNase plus detergent.
• Transfecting young macrophages with antagomiRs against miR-21-5p and miR-155-3p prior to EV exposure restores autophagic flux (measured by mCherry‑GFP‑LC3 reporter) and reduces NLRP3 inflammasome activation (caspase‑1 cleavage, IL‑1β secretion) despite the presence of old EVs.
• Systemic administration of GW4869 to aged mice for 2 weeks lowers exosomal miR-21‑5p/miR‑155‑3p levels in serum, increases splenic T‑cell LC3 puncta, and decreases circulating IL‑1β and IL‑6 compared with vehicle‑treated controls.

Falsifiability

• If old‑mouse EVs are not enriched for miR-21‑5p or miR-155‑3p relative to young EVs, the premise fails.
• If EV uptake does not alter ULK1/ATG5 protein levels or mTORC1 activity in young immune cells, the mechanism is unsupported.
• If pharmacological blockade of EV release does not improve autophagy markers or cytokine profiles in aged animals, the in vivo relevance is questionable.
• If antagomiR transfection does not rescue autophagic flux despite effective miRNA knock‑down, the hypothesized miRNA targets are incorrect.

Broader implication This model shifts the focus from cell‑intrinsic brakes to a tissue‑level communication network where the aged microenvironment actively enforces autophagy suppression via EV‑borne miRNAs. Therapeutically, combining EV‑targeting strategies (e.g., GW4869, miRNA sponges) with intermittent mTOR inhibition could synergistically rejuvenate immune function beyond what either approach achieves alone.