The hallmarks of aging may trace back to a single epigenetic switch: age‑dependent loss of the m6A writer complex METTL3/METTL14 in cortical neurons. We propose that this loss initiates a cascade that simultaneously disrupts ER calcium homeostasis, autophagic flux, and mitochondrial proteostasis, thereby driving multiple aging phenotypes. Specifically, reduced METTL3/14 activity lowers m6A modification on transcripts encoding the ER Ca2+ ATPase Atp2a3 and the mitochondrial protease LONP1, leading to their destabilization. The resulting ER calcium leak activates the unfolded protein response (UPR) via PERK‑eIF2α signaling, which in turn phosphorylates eIF2α and reduces global translation. Concurrently, diminished m6A on autophagy regulators such as ATG5, ATG7, and ULK1 impairs their translation, weakening autophagic clearance of damaged organelles and protein aggregates. Mitochondrial dysfunction stems from decreased LONP1, causing accumulation of oxidized proteins and ROS, which further exacerbates ER stress. This triad of UPR activation, autophagy insufficiency, and mitochondrial proteostasis failure creates a feed‑forward loop that accelerates cellular senescence and neuroinflammation.

Key predictions of this model are testable: (1) In normally aged human cortex, METTL3 and METTL14 protein levels will show a gradual decline correlated with donor age, independent of pathology. (2) Neuron‑specific overexpression of METTL3/14 in aged mice will rescue Atp2a3 and LONP1 mRNA stability, attenuate UPR markers (BiP, CHOP), restore LC3‑II conversion and p62 clearance, and improve mitochondrial membrane potential. (3) Pharmacological inhibition of PERK will ameliorate autophagy deficits in METTL3/14‑deficient neurons, indicating that UPR activation lies upstream of autophagy failure. (4) Conversely, CRISPR‑mediated disruption of the m6A consensus site on Atp2a3 or LONP1 will phenocopy the proteostasis collapse seen with METTL3/14 loss, confirming that transcript‑specific m6A loss is sufficient.

Falsifiable outcomes include: absence of age‑related METTL3/14 decline in healthy cortical neurons, or failure of METTL3/14 restoration to normalize UPR, autophagy, and mitochondrial readouts despite verified m6A rescue. Such results would redirect the search for an upstream controller toward other nuclear‑cytoplasmic regulators (e.g., NAD+‑SIRT1 signaling or chromatin remodelers). By integrating m6A writer activity with the three major proteostasis arms, this hypothesis offers a mechanistic framework that links a single epigenetic lesion to the multidimensional decline observed in brain aging.