Finding. In a dish that pairs human iPSC retinal pigment epithelium with a porcine neuroretina, the smoke oxidant hydroquinone injures both sides. Photoreceptors apoptose, run ERAD, lose outer-nuclear-layer rows, and shorten outer segments. Blocking valosin-containing protein with ML240 does not save the RPE. It does save photoreceptor outer-segment length and cone density. The stressed neuroretina’s proteome then shows less ERAD, more antioxidant proteins, and cytochrome c that stays enriched in inner segments, which the authors read as better mitochondrial integrity.
Why this paper matters
AMD models often stop at RPE. This co-culture asks how RPE stress becomes photoreceptor death, then tests a proteostasis drug that was motivated by HQ’s earlier proteostasis hit. The split result is the story: same oxidant, RPE still dies, photoreceptors keep structure if VCP is inhibited.
The mitochondrial sentence is cytochrome c remaining where inner-segment mitochondria live. That is a localization-as-integrity claim. It is better than nothing and not a bioenergetic proof.
What they actually measured
HQ on iPSC-RPE/porcine neuroretina: oxidative stress, apoptosis/caspase, ERAD, ONL rows, outer-segment length. ML240: no RPE apoptosis rescue; photoreceptor OS and cone density preserved. Proteomics of HQ-exposed neuroretina ± ML240.
How to read the score
Low seventies. Platform plus a cell-type-selective protectant plus a mitochondrial-integrity marker. Confidence is medium. Score 72.
What to do with it
If you model AMD, RPE–photoreceptor coupling, or VCP in neurodegeneration, steal the co-culture and the inner-segment cytochrome c readout. Do not treat ML240 as an AMD drug. The directional implication is that photoreceptor mitochondria and outer segments can be spared by VCP inhibition even when RPE apoptosis proceeds.