Age‑Dependent METTL3 Auto‑Phosphorylation Drives Proteostatic m6A Enhancement that Fails in Alzheimer’s Due to Oxidative Inhibition
Hypothesis
In healthy aging, neuronal activity‑dependent CaMKII phosphorylates METTL3 at a specific serine residue, boosting its catalytic activity and leading to a global increase in m6A on transcripts encoding proteasome subunits, chaperones, and autophagy receptors. This m6A‑mediated boost enhances translation of proteostasis machinery, counteracting age‑related protein damage. In Alzheimer’s disease, chronic oxidative stress activates phosphatases and/or directly oxidizes the METTL3 phosphorylation site, abolishing the age‑related increase in METTL3 activity. Consequently, m6A on proteostasis transcripts falls, proteasome and chaperone synthesis decline, and aggregates accumulate.
Mechanistic Basis
- CaMKII‑METTL3 link: Prior work shows CaMKII can phosphorylate RNA‑binding proteins in response to calcium spikes that rise with synaptic activity during aging (https://doi.org/10.1101/2025.05.02.651974). We propose METTL3 contains a conserved CaMKII motif that, when phosphorylated, increases its affinity for SAM and RNA substrates.
- Oxidative inhibition: Oxidative modifications (e.g., S‑glutathionylation) of the same serine block phosphorylation and reduce METTL3 activity, a mechanism observed for other enzymes in AD models (https://www.pnas.org/doi/10.1073/pnas.2204933120).
- Cell‑type specificity: Endothelial METTL14 upregulation with age (https://academic.oup.com/eurheartj/article/46/45/4953/8221892) contrasts with neuronal METTL3 regulation, underscoring that the proposed mechanism is neuron‑restricted.
Testable Predictions
- Phospho‑METTL3 rises with age in wild‑type mouse prefrontal cortex neurons and is reduced in APP/PS1 or tauopathy models. Detectable by phospho‑specific Western blot or mass spectrometry.
- Phospho‑mimetic METTL3 (S→D) expressed in aged neurons restores m6A levels on proteasome subunit transcripts (e.g., Psmc1, Psmc5) and improves chymotrypsin‑like proteasome activity, whereas phospho‑dead (S→A) fails to do so.
- Antioxidant treatment (N‑acetylcysteine) in aged AD‑model neurons prevents loss of METTL3 phosphorylation and rescues m6A on chaperone transcripts (Hsp90aa1, Hspa5).
- Cell‑type‑specific RNA immunoprecipitation shows increased m6A on 3′UTRs of proteasome genes in young adult neurons, a further increase in middle‑aged neurons, and a drop in late‑aged AD‑model neurons.
Potential Caveats
- Off‑target effects of CaMKII inhibitors could confound interpretation; use kinase‑dead CaMKII as control.
- Oxidative stress may affect m6A demethylases (FTO/ALKBH5) as well; measure their activity to isolate writer‑specific effects.
If these predictions hold, the hypothesis would explain why the compensatory m6A rise seen in normal aging is absent in AD and would point to METTL3 phosphorylation as a druggable node to bolster proteostasis in neurodegenerative disease.