Hypothesis

In cognitively normal aging, a global rise in m6A modification preferentially enhances YTHDF1‑dependent translation of proteasome subunit mRNAs, thereby sustaining proteostasis. In Alzheimer’s disease, loss of this age‑associated m6A shift reduces YTHDF1 recruitment and increases YTHDF2‑mediated decay of the same transcripts, compromising proteasome assembly and activity. The switch is mediated by age‑regulated expression of a neuronal lncRNA (lnc‑ProtAS) that masks YTHDF2 binding sites on proteasome transcripts, tipping the balance toward YTHDF1.

Mechanistic Rationale

1. m6A as a translational rheostat – METTL3/14‑catalyzed m6A creates a binding platform for both YTHDF1 (promotes translation) and YTHDF2 (promotes decay). The outcome depends on competing RNA‑binding proteins and local RNA structure.
2. lnc‑ProtAS as an age‑dependent modulator – We propose that lnc‑ProtAS, whose expression rises during healthy aging, binds near YTHDF2 consensus motifs on proteasome subunit transcripts (e.g., PSMB5, PSMC1), sterically hindering YTHDF2 recruitment while leaving YTHDF1 sites accessible.
3. Disease‑specific erosion – In AD, lnc‑ProtAS levels fall (or are disrupted by hyperphosphorylated Tau), exposing YTHDF2 sites. Consequently, m6A‑modified proteasome mRNAs are earmarked for decay, reducing proteasome biogenesis despite elevated global m6A.
4. Proteostatic consequence – Reduced proteasome subunit synthesis impairs 26S proteasome assembly, leading to accumulation of ubiquitinated proteins—a hallmark of AD neuropathology.

Testable Predictions

• Prediction 1: In human prefrontal cortex, m6A levels on proteasome subunit transcripts will positively correlate with age in cognitively normal individuals but negatively correlate with Braak stage in AD cases.
• Prediction 2: lnc‑ProtAS expression will increase with age in control neurons and decrease in AD‑derived neurons; its knockdown in young neurons will mimic the AD‑like loss of m6A‑dependent translation.
• Prediction 3: Overexpression of lnc‑ProtAS in AD mouse models will restore YTHDF1 binding to proteasome mRNAs, increase proteasome activity, and ameliorate cognitive deficits.
• Prediction 4: Disrupting the lnc‑ProtAS‑RNA duplex (using antisense oligonucleotides) will shift m6A‑modified proteasome transcripts toward YTHDF2 binding, decreasing their half‑life and proteasome function even in young mice.

Experimental Approach

• Transcriptome‑wide m6A mapping (MeRIP‑seq) of laser‑captured pyramidal neurons from young, aged, and AD human brains; focus on proteasome subunit genes.
• RNA immunoprecipitation for YTHDF1 and YTHDF2 followed by qPCR to assess isoform‑specific binding under varying lnc‑ProtAS levels.
• Luciferase reporters containing wild‑type or mutant lnc‑ProtAS binding sites fused to proteasome subunit 3’UTRs to test translational repression/activation.
• AAV‑mediated lnc‑ProtAS overexpression in APP/PS1 mice; measure proteasome activity (fluorogenic substrates), ubiquitin‑positive inclusions, and behavior (Morris water maze).
• Antisense oligonucleotide knockdown of lnc‑ProtAS in wild‑type aged mice to see if it precipitates AD‑like proteostatic decline.

Falsifiability

If any of the following observations are found, the hypothesis is weakened or refuted:

• No age‑dependent increase in m6A on proteasome transcripts in control brains.
• lnc‑ProtAS expression does not change with age or disease status.
• Manipulating lnc‑ProtAS fails to alter YTHDF1/YTHDF2 binding or proteasome activity as predicted.
• Restoring m6A writer activity (METTL3/14) without altering lnc‑ProtAS does not rescue proteasome deficits in AD models.

By integrating the observed age‑related m6A rise with a novel lncRNA‑mediated reader competition mechanism, this hypothesis provides a concrete, testable framework to explain why increased m6A is neuroprotective in normal aging yet lost in Alzheimer’s disease, and it points toward a targeted therapeutic strategy to reestablish proteostasis.