Finding. Mitochondrial import stress sends hydrophobic oxidative phosphorylation proteins to the endoplasmic reticulum. Once there they do not share one fate. Some sit stably in the ER. Others are cleared by ER-associated degradation, with the ubiquitin ligase MARCHF6 as a central, partly redundant actor. The ER is a quality-control station for mistargeted mitochondrial proteins, not a bystander of organelle identity.
Why this paper matters
Every textbook says organelle identity is protein composition. Targeting errors should therefore be lethal to identity, yet the detection-and-disposal map for mitochondrial proteins that miss the TOM gate is still thin. This brief fills a specific hole: the hydrophobic OXPHOS subunits that are worst at living in the cytosol do not only clog chaperones. They go to the ER.
That recasts import failure as a two-organelle problem. Mitochondria lose clients. The ER gains a toxic, hydrophobic census and must decide what to keep and what to extract. If you only score mitochondrial proteomes after an import hit, you will miss the ER holding pen.
What they actually measured
The authors combine proximity proteomics, a split-fluorescence reporter, and a genome-wide CRISPR screen. The proteomic claim is enrichment of hydrophobic OXPHOS components among ER-rerouted mitochondrial proteins. The reporter and the screen split those clients into classes: stable ER residence versus rapid ERAD. Clearance uses partially redundant ERAD branches; MARCHF6 is the named central ligase.
Keep the scope tight. This is about mistargeted mitochondrial proteins during import stress, not about canonical ERAD of secretory clients. The redundancy clause matters: a single-ligase knockout will leak. The class split matters: not every stranded OXPHOS subunit is a degradation substrate.
The abstract does not tell you the cell type, the import-stress reagent, or the size of the ER-resident versus ERAD cohorts. Those are PDF problems, not reasons to ignore the geometry.
How to read the score
High seventies. Method mix is serious, the client class (hydrophobic OXPHOS) is the right mitochondrial object, and a named ligase gives you something to pull. Confidence is medium because the teaser is mechanism-rich and number-poor. It is not a clinical paper.
What to do with it
If you model import stress, mitoprotein toxicity, or ER–mitochondria contact proteomes, add MARCHF6 and ERAD branches to the parts list. Pull the proximity-proteomic ER census and the CRISPR hit table. Do not call MARCHF6 a therapeutic target from this brief. The directional implication is that ER quality control is part of how cells survive mitochondrial targeting failure.