Finding. Mild mitochondrial electron-transport-chain (ETC) lesions extend Caenorhabditis elegans life, and textbooks have started to call that one word: mitohormesis. Chiamkunakorn, Braeckman, Suthammarak and colleagues knock the idea apart by complex. RNA interference against the Complex I subunit nuo-6 and the Complex IV subunit cco-1 both turn on the mitochondrial unfolded protein response (UPRmt). That is where the similarity ends. Complex I longevity leans hard on ATFS-1, the UPRmt transcription factor, and does not need AAK-2, the worm AMP-activated protein kinase (AMPK). Complex IV longevity ignores ATFS-1 and requires AAK-2-driven metabolic reprogramming.
Metformin, which is both a Complex I inhibitor and an AMPK activator, reads the two genotypes differently. In nuo-6;aak-2 animals it markedly suppresses the longevity phenotype. In cco-1;aak-2 animals it trends toward lifespan extension that still looks AAK-2-independent. Same drug, opposite manners, because the lesions were never the same signal.
Why this paper matters A lot of mitochondrial aging papers treat 'less ETC, more life' as a shared retrograde circuit: UPRmt, AMPK, maybe both. If that were true, a Complex I knockdown and a Complex IV knockdown should fail the same mutants and take metformin the same way. They do not. The practical consequence is that you cannot use hsp-6 reporters or an AMPK blot as a universal mitohormesis badge. You have to name the complex.
The metformin split is the translational tease. People already argue whether metformin is a Complex I drug, an AMPK drug, or a gut-microbiome story. Here the pharmacology is a diagnostic: it hurts a Complex I longevity genotype that has already lost AMPK, and it does not behave that way on a Complex IV genotype. That is a tool for pathway assignment, not a reason to dose worms or humans differently this week.
How to read the score High eighties. Core organelle biology, a clean genetic double dissociation, UPRmt versus AMPK, and a named drug that respects the split. Confidence is high for the dependency architecture as stated, medium for the metformin Complex IV trend. Heuristic copy would have said 'mitochondrial stress extends lifespan via UPRmt and AMPK.' The paper says pick one, and which one depends on which complex you broke.
Caveats RNAi of one subunit is not a null, and worm life is not human healthspan. The Complex IV plus metformin result is a trend. Both lesions induce UPRmt, so a glowing reporter will not tell you which path you are on. No abstract numbers for respiration, ATP, or metabolites.
What to do with it If you curate mitohormesis, aging, or metformin-mitochondria papers, split your tags by complex and by ATFS-1 versus AAK-2. If you build screens, use the two RNAi backgrounds as a filter: hits that copy both are the rare shared node. Do not write that metformin is 'an AMPK longevity drug' from this brief. Write that metformin interrogates which retrograde path a lesion chose.