Hypothesis

Active suppression of autophagy in old age is driven by aging‑associated extracellular vesicles (EVs) that deliver specific microRNAs to target cells, thereby dampening VPS34‑complex activity and reinforcing mTORC1 signaling as a compensatory mechanism to limit proteotoxic overload.

Rationale

The review shows that autophagy inhibition in aging is an active process: Rubicon (RUBCN) rises, mTORC1 stays active, and post‑translational modifications inhibit VPS34 and BECN1 [1] [2]. Young plasma can reverse this block, indicating that circulating factors convey the suppressive signal [3]. Senescent cells accumulate when autophagy is blocked, and clearing them improves function, suggesting a feedback loop [4].

We propose that the circulating factor is not a generic protein but a subset of EVs enriched in miRNAs that directly target autophagy‑promoting genes. In aged tissues, EV secretion increases, and their cargo shifts toward miR‑34a, miR‑146a, and miR‑21—microRNAs known to suppress ULK1, Beclin‑1, and VPS34 translation [1]. By reducing the core autophagic machinery, these EVs lower the risk of catastrophic self‑digestion when the proteostasis network is already strained by accumulated damage. Simultaneously, EV‑mediated inhibition of autophagy sustains mTORC1 activity (via reduced AMPK activation), which further suppresses TFEB nuclear translocation, creating a stable, low‑autophagy state.

Testable Predictions

• EV isolation: EVs harvested from old mouse plasma will show elevated miR‑34a/miR‑146a/miR‑21 levels compared with young EVs. Transfer of old EVs to young hepatocytes will decrease LC3‑II/I ratio and increase p62, mimicking the aged autophagy block [1].
• miRNA inhibition: Antagomirs against miR‑34a, miR‑146a, or miR‑21 administered to aged mice will restore VPS34 activity (measured by PI3P production) and increase autophagic flux, while concurrently reducing mTORC1‑pS6K signaling.
• Proteostasis trade‑off: Restoring autophagy via EV‑miRNA inhibition will improve clearance of damaged mitochondria but will also increase detectable ubiquitinated protein aggregates in the short term, reflecting the cell’s shift from a suppression‑to‑a‑clearance mode.
• Senescence link: Mice treated with EV‑miRNA antagomirs will exhibit reduced senescent‑cell markers (p16^INK4a^, SA‑β‑gal) in heart and liver, consistent with the autophagy‑senescence feedback loop described [4].

Experimental Approach

1. EV profiling: Use ultracentrifugation or size‑exclusion chromatography to isolate EVs from plasma of 3‑month vs. 24‑month mice. Perform small‑RNA sequencing to quantify miRNA enrichment.
2. Gain‑of‑function: Inject old‑derived EVs into young mice; monitor autophagy markers (LC3‑II, p62) and mTORC1 activity in liver and muscle after 48 h.
3. Loss‑of‑function: Treat aged mice with locked‑nucleic‑acid antagomirs targeting the candidate miRNAs (or neutral control) via tail‑vein injection twice weekly for 4 weeks. Assess autophagic flux (chloroquine‑based LC3‑II accumulation), mTORC1 signaling (p‑S6K), and senescence (p16, SASP cytokines).
4. Proteotoxicity assay: Filter‑trap assay for ubiquitinated aggregates and mitochondrial membrane potential (JC‑1 staining) to capture the predicted trade‑off.

Falsifiability

If old‑derived EVs do not carry elevated miR‑34a/miR‑146a/miR‑21, or if their transfer fails to suppress autophagy in young recipients, the EV‑miRNA mechanism is refuted. Likewise, if antagomir treatment restores autophagy without affecting senescence markers or aggregate load, the proposed protective trade‑off loses support.

Broader Implication

This hypothesis reframes age‑related autophagy decline not as a mere breakdown but as a programmed, EV‑mediated safeguard that temporarily limits autophagic activity to avoid overwhelming a compromised proteostasis network. Therapeutically, timing the inhibition of these suppressive signals—perhaps after clearing senescent cells or reducing aggregate burden—could allow a controlled “reset” without triggering catastrophic self‑digestion.