Hypothesis

In aged neurons, METTL3 undergoes activity‑dependent phosphorylation at serine‑43 (S43) that reduces its catalytic efficiency toward a subset of 3′UTR sites in autophagy‑related mRNAs, producing a compensatory rise in m6A that sustains proteostasis. In Alzheimer’s disease, aberrant phosphatase activity dephosphorylates METTL3‑S43, restoring high writer activity and causing excessive methylation that disrupts autophagic flux, thereby converting a protective modification into a pathogenic one.

Mechanistic Rationale

• Neuronal firing triggers CaMKII‑mediated phosphorylation of METTL3‑S43, a site conserved across mammals and previously shown to modulate protein‑protein interactions without altering METTL3 stability (see METTL3 regulation in non‑neuronal senescence models)【2】.
• Phospho‑METTL3 exhibits preferential binding to the RNA‑binding protein YTHDF1 over YTHDF2, shifting the fate of methylated transcripts from decay to translation initiation. This bias favors synthesis of autophagy receptors (e.g., p62/SQSTM1, NBR1) and ubiquitin‑proteasome subunits, reinforcing clearance pathways【3】.
• Global m6A levels rise because the majority of transcripts remain unaffected, but site‑specific loss of methylation at autophagy‑related 3′UTRs reduces YTHDF2‑driven decay, increasing mRNA stability and protein output—a compensatory mechanism that matches the observed increase in m6A at ubiquitin‑mediated proteolysis genes during normal cognitive aging【1】.
• In AD, chronic oxidative stress activates PP2A phosphatases that dephosphorylate METTL3‑S43. The dephosphorylated writer regains high catalytic activity across all substrates, leading to hypermethylation of the same autophagy transcripts. Excess m6A now recruits YTHDF2, accelerating transcript decay and lowering p62/NBR1 levels, which impairs autophagosome formation and aggregates clearance, mirroring the maladaptive excessive m6A phenotype seen in ALS models【3】.

Experimental Plan

1. Phospho‑specific profiling – Generate a polyclonal antibody against METTL3‑pS43. Quantify its signal in human post‑mortem prefrontal cortex across Braak stages (0–VI) and correlate with global m6A (LC‑MS) and autophagy‑gene m6A peaks (MeRIP‑seq).
2. Kinase/phosphatase manipulation – Treat primary cortical neurons from aged mice with CaMKII inhibitor (KN‑93) or PP2A activator (FTY720). Measure METTL3‑pS43 levels, site‑specific m6A at p62 and NBR1 3′UTRs (targeted MeRIP‑qPCR), and autophagic flux (LC3‑II/I, p62 turnover).
3. Rescue experiments – Express phospho‑deficient (S43A) and phospho‑mimetic (S43D) METTL3 in METTL3‑KO neurons derived from aged human iPSCs. Assess whether S43D restores autophagy‑gene expression and reduces phosphorylated tau accumulation, whereas S43A exacerbates pathology.
4. Behavioral readout – Deliver AAV‑shRNA targeting PP2A subunits to the hippocampus of APP/PS1 mice and evaluate memory performance (Morris water maze) alongside METTL3‑pS43, m6A patterns, and neurodegeneration markers.

Expected Outcomes

• Normal aging will show elevated METTL3‑pS43, increased m6A at autophagy‑gene 3′UTRs, and heightened autophagy flux; AD brains will display reduced METTL3‑pS43, hypermethylation of the same sites, and diminished autophagic activity.
• Pharmacological or genetic restoration of METTL3‑pS43 in AD models should normalize autophagy‑gene expression and improve cognitive phenotypes, falsifying the hypothesis if no rescue occurs.
• Conversely, forcing METTL3 dephosphorylation in young neurons should recapitulate AD‑like hypomethylation of autophagy transcripts and impair proteostasis, providing a causal link.

This hypothesis integrates the observed age‑related m6A increase【1】 with the known regulation of METTL3 via proteasomal degradation【2】 and the opposing effects of writer activity on autophagy【3】, offering a precise, testable mechanism that explains why global m6A elevations are protective in normal aging yet detrimental in neurodegeneration.