Hypothesis

In cognitively normal aging, neuronal METTL3/METTL14 activity increases, leading to elevated m6A deposition preferentially in the 3'UTRs of synaptic and proteostasis genes. This modification recruits IGF2BP2, which not only stabilizes transcripts but also interferes with upstream nonsense-mediated decay (NMD) by blocking exon junction complex (EJC) assembly downstream of stop codons. Consequently, m6A-modified proteostasis mRNAs (e.g., HSP70, BAG3, ATG5, PSMB5) evade NMD-mediated degradation, maintaining protein quality control despite age-related stress. In Alzheimer's disease, the age-associated m6A increase fails, resulting in loss of IGF2BP2 binding, restored NMD activity, and selective depletion of proteostasis transcripts, thereby linking m6A loss to proteolytic dysfunction. Premature aging models show decreased METTL3 and global m6A loss, which similarly elevates NMD of proteostasis mRNAs, explaining their early-onset aggregation phenotypes.

Testable Predictions

1. 'm6A‑NMD competition': In young neurons, inhibiting IGF2BP2 (siRNA or CRISPRi) will increase NMD susceptibility of m6A‑modified proteostasis transcripts, measurable by increased upstream open reading frame (uORF)‑derived isoforms and reduced mRNA half‑life, without affecting total m6A levels.
2. 'Rescue of proteostasis': Overexpressing a catalytically dead METTL3 that retains RNA‑binding but cannot methylate will not rescue NMD suppression, indicating that the methyl group itself is required for IGF2BP2‑mediated NMD inhibition.
3. 'Disease specificity': AD‑derived iPSC neurons will show heightened NMD activity on HSP70 and ATG7 transcripts compared to age‑matched controls, correlating with reduced m6A peaks at their 3'UTRs (MeRIP‑seq).
4. 'Proteostasis readout': Restoring IGF2BP2 binding via a tethered IGF2BP2‑MS2 system to the 3'UTR of a reporter proteostasis gene will decrease NMD reporter activity and increase protein levels, even in AD neurons.

Experimental Approach

• Perform MeRIP‑seq and parallel RNA‑seq on sorted neurons from young, aged, and AD brains to map m6A sites on proteostasis transcripts.
• Use transcriptome-wide NMD inhibition (e.g., UPF1 knockdown) to identify NMD‑sensitive isoforms; compare changes upon IGF2BP2 depletion.
• Measure mRNA half‑life with 4sU labeling in conditions: control, IGF2BP2 KD, METTL3 overexpression, and AD pathology.
• Assess proteostasis via autophagy flux (LC3-II turnover), proteasome activity (fluorogenic substrates), and aggregation (filter trap assay for ubiquitin‑positive species).

Potential Impact

If validated, this hypothesis reframes m6A not merely as a stability mark but as a direct antagonist of NMD, providing a mechanistic link between epitranscriptomic regulation and proteostasis collapse in neurodegeneration. It also suggests that targeting the IGF2BP2‑m6A‑NMD axis could restore proteostasis in AD without altering global methylation levels.