Finding. Mitochondrial Complex III is an obligate dimer of 20-odd subunits. For years the assembly story was: fold each piece, add them in order, dimerize, then finish both protomers together. Kazmi, Vercellino and colleagues’ cryo-EM intermediates say two of those clauses are false. Cytochrome c1 can be incorporated before it is folded. After the dimer exists, the two protomers do not mature in parallel; the intermembrane-space domain is the example. In non-vertebrates, supercomplex CIII2CIV is born on that same path, which is what a cooperative assembly model predicted and a late-docking model did not. AlphaFold did not draw these steps.
Why this paper matters
Complex III sits in the middle of the chain and in the middle of a metabolic-disease gene list. If assembly is asymmetric after dimerization, patient alleles and assembly-factor knockouts will pile intermediates that look “half-done” on one protomer. Biochemical add-a-subunit schemes will mis-order those states.
The supercomplex claim is the second headline. If CIII2CIV is intertwined with CIII2 maturation, you do not first finish Complex III and Complex IV and then hope they collide. You grow the supercomplex while III is still being built. That is a different genetic expectation for COX and UQCR assembly factors.
How to read the score
Low nineties. Central OXPHOS machine, cryo-EM, a model that changes textbooks, an explicit AlphaFold miss. Confidence is high for the observed intermediates, medium for “this is the general eukaryotic path,” because the supercomplex half is non-vertebrate.
Caveats
Isolation bias. Species. Not every assembly factor is reassigned here. Do not remap a patient’s CIII deficiency onto one figure from this brief.
What to do with it
If you write Complex III assembly, replace parallel-protomer cartoons. If you score supercomplex papers, add cooperative CIII2CIV as the non-vertebrate default until someone shows otherwise. Pull the cytochrome c1 and IMS-domain states first.