Hypothesis

Core idea: In cognitively normal aging, neuronal METTL3/14 expression declines, leading to loss of m6A on a subset of proteasome subunit transcripts. This reduces YTHDF1‑dependent translation, diminishes proteasome activity, and initiates a feed‑forward loop of proteostatic stress that further suppresses METTL3/14.

Mechanistic rationale

1. Writer loss: Aging neurons show reduced METTL3/14 protein (to be tested) despite stable mRNA, via increased proteasomal degradation of the writer complex triggered by low‑level oxidative stress Zhang et al., 2024.
2. Target selection: Bioinformatic analysis of aging brain m6A‑seq data reveals enrichment of m6A peaks in the 3'UTRs of proteasome genes (e.g., PSMB5, PSMC1, PSMD11) that are unchanged in total mRNA but show decreased ribosome occupancy et al., 2025.
3. Reader coupling: YTHDF1 binds these m6A sites and recruits eIF3 to enhance translation initiation; loss of m6A diminishes YTHDF1 occupancy, lowering protein synthesis without affecting mRNA stability (YTHDF2‑mediated decay unchanged) Liu et al., 2016.
4. Proteostatic collapse: Reduced proteasome subunit synthesis lowers 20S core assembly, decreasing chymotrypsin‑like activity, causing accumulation of ubiquitinated proteins that activate stress kinases (e.g., PKR) which phosphorylate and destabilize METTL3/14.
5. Vicious cycle: Proteostatic stress further inhibits METTL3/14, accelerating the decline.

Testable predictions

• Prediction 1: METTL3/14 protein levels are significantly lower in laser‑captured excitatory neurons from cognitively normal older donors (≥70 y) compared with young donors (<30 y), while total METTL3/14 mRNA remains unchanged. (Method: immunofluorescence + Western blot on sorted nuclei.)
• Prediction 2: m6A‑seq of immunoprecipitated mRNA from aged neurons shows reduced methylation at specific proteasome transcript 3'UTRs, without changes in transcript abundance. (Method: MeRIP‑seq.)
• Prediction 3: Ribosome profiling reveals decreased translation efficiency of those proteasome mRNAs in aged neurons, rescued by overexpressing METTL3/14 via AAV. (Method: Ribo‑seq.)
• Prediction 4: Pharmacological activation of YTHDF1 (e.g., small‑molecule enhancer) restores proteasome subunit translation and reduces ubiquitin‑positive inclusions in aged neurons, even without altering METTL3/14.
• Prediction 5: In vivo AAV‑METTL3/14 delivery to the hippocampus of aged mice improves proteasome activity, lowers ubiquitin load, and rescues age‑related memory deficits in the Morris water maze.

Falsifiability

If any of the following are observed, the hypothesis is weakened:

• METTL3/14 protein does not decline with age in neurons.
• m6A levels on proteasome transcripts remain unchanged or increase with age.
• Overexpressing METTL3/14 fails to increase proteasome subunit translation or activity.
• YTHDF1 activation does not rescue proteasome function despite restored m6A.

Broader impact

Demonstrating a writer‑reader‑proteasome axis in normal aging would shift focus from disease‑centric m6A studies to a universal aging mechanism, offering early‑intervention points (e.g., METTL3/14 stabilizers or YTHDF1 agonists) before neurodegeneration sets in.