Hypothesis

METTL3/14 decline in aged neurons causes locus‑specific loss of m6A on a subset of autophagy‑related mRNAs, impairing their translation and autophagic flux, while simultaneously increasing m6A‑dependent translation of ATF4, leading to sustained PERK‑eIF2α signaling and ER stress. Restoring METTL3/14 specifically in neurons rescues m6A on autophagy transcripts, normalizes autophagy, and attenuates PERK signaling, thereby ameliorating senescence phenotypes.

Mechanistic Model

• Selective m6A loss on autophagy transcripts: METTL3/14 preferentially methylates 5′UTR regions of mRNAs such as MAP1LC3B, BECN1, and ATG5. m6A in these contexts promotes ribosome scanning and initiation. When METTL3/14 levels fall, these transcripts become hypomethylated, reducing their translation efficiency despite stable mRNA levels.
• Compensatory increase in ATF4 translation: In contrast, the 5′UTR of ATF4 contains a regulatory upstream open reading frame (uORF) that is sensitive to m6A. Loss of m6A at this uORF enhances re‑initiation at the main ATF4 ORF, increasing ATF4 protein synthesis. Elevated ATF4 drives chronic PERK‑eIF2α‑ATF4‑CHOP signaling, exacerbating ER stress.
• Feedback to autophagy: Persistent PERK activation phosphorylates eIF2α, globally suppressing translation, which further diminishes autophagy‑related protein synthesis, creating a vicious cycle of impaired proteostasis.
• Cell‑type specificity: Neurons exhibit high basal autophagy demand; thus, neuron‑restricted METTL3/14 loss disproportionately affects autophagic flux compared with proliferative cells where global m6A decline correlates with senescence.

Testable Predictions

1. m6A mapping: MeRIP‑seq of young versus aged human iPSC‑derived neurons will show significant m6A reduction at the 5′UTRs of MAP1LC3B, BECN1, and ATG5, but increased m6A at the ATF4 5′UTR uORF.
2. Translation assay: Riboprofiling will reveal decreased ribosome occupancy on autophagy transcripts and increased occupancy on ATF4 in aged neurons, reversible by neuronal METTL3/14 overexpression.
3. Autophagy flux: LC3‑II turnover assays (with bafilomycin A1) will demonstrate reduced autophagic flux in aged neurons; rescuing METTL3/14 restores flux to youthful levels.
4. PERK signaling: Phospho‑PERK and phospho‑eIF2α levels will be elevated in aged neurons; pharmacological PERK inhibition (GSK2606414) will normalize autophagy metrics without altering METTL3/14 expression.
5. Phenotypic rescue: Neuronal METTL3/14 overexpression in aged mice will reduce senescence markers (p16^INK4a^, SA‑β‑gal), improve motor performance, and extend lifespan in SOD1^G93A^ models, whereas ATF4 knockdown will mimic these effects.

Experimental Approach

• Generate a Cre‑dependent METTL3/14 AAV vector targeting Synapsin‑positive neurons.
• Inject into young (3 mo) and aged (18 mo) wild‑type mice; harvest cortical neurons after 4 weeks.
• Perform MeRIP‑seq, riboprofiling, and western blot for LC3‑II, phospho‑PERK, ATF4.
• Assess autophagy flux using mCherry‑GFP‑LC3 reporter and lysosomal inhibition.
• Behavioral assays (rotarod, grip strength) and senescence staining (p16, SA‑β‑gal) to correlate molecular changes with phenotype.
• Include controls: AAV‑GFP, PERK inhibitor, ATF4 shRNA.

This hypothesis is falsifiable: if METTL3/14 restoration fails to rescue m6A on autophagy transcripts, autophagic flux, or PERK signaling, or if manipulating ATF4 does not phenocopy METTL3/14 loss effects, the model would be refuted.