Review Papers Claim by Claim

Core claim and evidence chain

The paper asserts that glucose starvation (GS) drives H2Aub — the most abundant monoubiquitinated protein in mammalian cells< Citation 5 — into nucleolar "lysosome-like structures" (LLS) for degradation, releasing glutamine to fuel metabolism when glucose is absent. Three converging pharmacological arguments are offered: MG132 fails to inhibit H2Aub loss (and even exacerbates it), chloroquine (CQ) blocks it, and LAMP1 accumulates as nucleolar foci only under combined GS + CQ. A K119R point mutation abolishes GS-induced H2Aub degradation, supporting residue-specific dependence. Midnolin is proposed as the nucleolar receptor via overexpression knockdown, and its depletion causes S-phase arrest. The companion preprint identifies FBL as an additional candidate carrier< Citation DOI 10.1101/2025.06.30.662258 — raising receptor redundancy ambiguity the authors do not resolve.

What is missing

• No electron microscopy confirms membrane-bound lysosomal organelles inside nucleoli; LAMP1 IF alone cannot distinguish true lysosomes from LAMP1-positive membrane vesicles, condensates, or artifacts of overexpression.
• No isotope tracing (e.g., ¹³C-histone pulse) demonstrates that degraded H2Aub carbon enters TCA cycle or ATP pools. The glutamine-supply calculation is modeled (abundance × −log₂FC × Q-fraction) but not experimentally validated by metabolite flux.
• Single cell line (293T); no primary cells, no in vivo, no cancer-cell-line panel.
• n = 2 biological replicates for proteomics — inadequate for discovering subtle abundance shifts among 7912 proteins.
• Midnolin's canonical role is proteasomal< Citation 37>, not lysosomal. The mechanistic bridge (ubiquitin-binding protein forcing lysosomal routing) is untested — no endogenous co-IP of Midnolin–LAMP1, no substrate binding assays, no resolution of which Midnolin isoform acts in the nucleolus.
• K119R confound: the authors themselves observe that K119R triggers ectopic K120 polyubiquitination, which contaminates interpretation of the "rescue" logic.
• AI-assisted screening for receptor identification is not described with sufficient detail to evaluate blind spots or false-positive risk.
• Nucleolar LLS were hard to visualize: authors admit endogenous LAMP1 signal is low and only overexpression reveals nucleolar LLS — raising antibody-specificity/signal-to-noise concerns.

Falsification pathway

The central claim would be falsified if: (i) endogenous EM shows no membrane-bound lysosomes in nucleoli, (ii) isotopic tracing shows H2Aub degradation products do not enter metabolic carbon pools, (iii) Midnolin's loss can be rescued by a proteasome-targeted rather than lysosome-targeted construct, or (iv) CQ's H2Aub protection is reproduced by a non-lysosomal pH-buffering mechanism. None of these strong falsification tests were performed.

Model visualization

Solid arrows indicate observations directly supported by the paper; dashed arrows mark author- Gianluca interpretations that lack direct experimental confirmation. The "untested flux" step is the biggest mechanistic gap.

Author reviews