Hypothesis

Chronic low‑grade inflammasome activation in aged neurons triggers caspase‑1–mediated cleavage of METTL3 and METTL14, producing a loss‑of‑function shift that converts the compensatory rise in m6A seen during normal aging into the pathogenic m6A deficit observed in Alzheimer’s disease. In contrast, vascular endothelial cells experience sustained laminar shear stress that activates KLF2, which transcriptionally upregulates METTL14 and drives an m6A‑dependent senescence‑associated secretory phenotype. This dichotomy explains why METTL14 shows opposite trends in neurodegeneration versus vascular aging.

Mechanistic rationale

1. Inflammasome link – Aging microglia and astrocytes release IL‑1β and ATP, priming the NLRP3 inflammasome in neurons. Caspase‑1, besides processing IL‑1β, cleaves METTL3 at D395 and METTL14 at D210 (predicted cleavage sites based on conserved caspase motifs). Cleavage removes the catalytic MT‑A domain of METTL3 and the RNA‑binding region of METTL14, abolishing m6A deposition without altering total protein levels detectable by standard western blots.
2. Selective transcript impact – Cleaved METTL3/14 preferentially lose activity on m6A sites located in the 3’UTRs of autophagy‑lysosome genes (e.g., BECN1, SQSTM1/p62, HSPA8) and ubiquitin‑proteasome components (UCHL1). Loss of m6A reduces YTHDF1‑mediated translation efficiency and increases susceptibility to endonucleolytic decay, lowering protein synthesis of key proteostasis factors.
3. Proteostasis collapse – Reduced translation of autophagy and chaperone proteins impairs clearance of aggregation‑prone substrates such as TDP‑43 and α‑synuclein, creating a feed‑forward loop where aggregates further activate NLRP3 via lysosomal destabilization.
4. Vascular divergence – In endothelial cells, atheroprotective shear stress phosphorylates KLF2, which binds the METTL14 promoter and enhances its transcription. Elevated METTL14 increases m6A on inflammatory mRNAs (e.g., IL6, CXCL8), stabilizing them via YTHDC1 and amplifying the senescence‑associated secretory phenotype. This pathway is absent in neurons due to low KLF2 expression and different chromatin accessibility.

Testable predictions

• Prediction 1: In human post‑mortem tissue from cognitively normal aged donors, neuronal METTL3 and METTL14 will show increased caspase‑1 cleavage fragments compared with young adults, while total METTL3/14 levels remain unchanged.
• Prediction 2: Pharmacological inhibition of NLRP3 (MCC950) or caspase‑1 (VX‑765) in aged human iPSC‑derived neurons will preserve METTL3/14 full‑length protein, restore m6A levels on BECN1 and HSPA8 3’UTRs, and increase autophagy flux (LC3‑II/I ratio) without altering global m6A.
• Prediction 3: Expression of cleavage‑resistant METTL3(D395A) and METTL14(D210A) mutants in AD‑model neurons will rescue m6A deposition on proteostasis transcripts, reduce TDP‑43 aggregation, and improve survival, whereas wild‑type METTL3/14 overexpression will not.
• Prediction 4: In human aortic endothelial cells exposed to laminar shear stress, KLF2 knockdown will abolish shear‑induced METTL14 upregulation, decrease m6A on IL6 CXCL8 transcripts, and attenuate senescence markers (SA‑β‑gal, p16INK4a).

Experimental approach

1. Biochemical detection – Use neo‑epitope antibodies specific for the caspase‑1 cleavage neo‑termini of METTL3 (after D395) and METTL14 (after D210) in western blots and immunohistochemistry across young, aged, and AD brain cohorts.
2. m6A mapping – Perform miCLIP‑seq on sorted neurons from the same cohorts to quantify m6A changes at predicted sites on BECN1, HSPA8, UCHL1, and IL6; compare conditions with and without caspase‑1 inhibition.
3. Functional assays – Measure autophagic flux (mRFP‑GFP‑LC3), chaperone activity (luciferase refolding), and aggregate load (filter‑trap, immunofluorescence) in iPSC‑neurons treated with NLRP3/caspase‑1 modulators or expressing cleavage‑resistant mutants.
4. Vascular validation – Subject human aortic endothelial cells to parallel‑plate flow shear stress (± KLF2 siRNA) and assess METTL14 mRNA/protein, m6A‑seq of inflammatory transcripts, and senescence readouts.

If these experiments confirm that inflammasome‑driven cleavage of METTL3/14 converts a protective m6A increase into a deleterious decrease specifically in neurons, while shear stress–KLF2 upregulates METTL14 in endothelium, the hypothesis will provide a mechanistic bridge between aging, neurodegeneration, and vascular aging, and suggest caspase‑1 inhibition or cleavage‑resistant METTL3/14 as therapeutic strategies to preserve epitranscriptomic proteostasis.