One Protein, Three Prolines: Where Cells' Power Plants Snag

Researchers have imaged a protein called TACO1 sitting on the protein-making machinery of human mitochondria, and the picture suggests how it keeps that machinery from stalling.

The study, led by Vivek Singh and colleagues, was published in Nature Communications on October 21, 2026. The journal describes it as peer-reviewed, accepted research shared early, before final copy edits ^[2]. It is basic biology, not a treatment trial. But it puts a face on a mechanism that earlier experiments had only inferred.

A small factory with a hard job

Mitochondria are the compartments in our cells that produce most of the cell's usable energy. They carry a little DNA of their own and make a short list of proteins from it, using their own ribosomes, the molecular machines that read genetic instructions and string amino acids into protein. Those few proteins are essential parts of the apparatus that lets mitochondria generate energy ^[1].

Some stretches of instruction are harder to read than others. Ribosomes are known to pause at runs of consecutive prolines, a particular amino acid, and COX1, a key mitochondrial protein, contains three prolines in a row ^[3][4]. It is like a typist who has to hit the same awkward key three times running, and who loses the rhythm at the third.

What the structure shows

The team used cryo-electron microscopy, which flash-freezes molecules and images them with electrons, to determine a structure of human TACO1 bound to a mitochondrial ribosome ^[2]. In that structure, TACO1 sits near the ribosome's A site, the slot where each new amino acid arrives. There it contacts the transfer RNA, or tRNA, the molecule that ferries the amino acid in ^[2]. The study says TACO1 also touches two parts of the ribosome, the sarcin-ricin loop and the L7/L12 stalk ^[2].

The researchers then ran AlphaFold analysis and molecular-dynamics simulations, which are computer models of how atoms move, to ask how TACO1 might behave ^[2]. Their results suggest that its flexible N-terminal domain, one end of the protein, can shift to engage the incoming tRNA and help stabilize it ^[2]. The study also reports structural features that human TACO1 shares with its bacterial relative, YebC2 ^[2].

The movement is a proposal drawn from modeling, not something the imaging recorded directly. The central evidence of the paper is the frozen structure plus computation ^[2].

Where the earlier experiments left off

The structure lands on top of work published in 2024. In that study, in Nucleic Acids Research, researchers used metabolic labeling, puromycin-release and mitoribosome-profiling experiments to test what TACO1 does ^[4]. Puromycin-release is a way of measuring how readily a stalled ribosome lets go of its half-built protein. Mitoribosome-profiling maps where ribosomes sit on their instructions.

They found that TACO1 helps mitochondrial ribosomes translate the proline-rich proteins COX1 and COX3 efficiently, and that the need for it is negligible for the other proteins encoded by mitochondrial DNA ^[4]. When functional TACO1 is missing, ribosomes find it harder to resume synthesis after pausing at consecutive prolines ^[3][4].

That is a narrow job, and a revealing one. The cell has not given every mitochondrial protein its own rescue crew. It has kept one for the sequences that trip the machine.

Why a rare disease cares

TACO1 mutations are associated with cytochrome c oxidase deficiency, a shortfall in a central enzyme of mitochondrial energy production, and with a late-onset form of Leigh syndrome, a severe neurological disease ^[1][3]. COX1 and COX3 are both parts of that enzyme, which fits the pattern. If the rescue protein fails, the proline-rich proteins suffer first.

The new structure offers a possible molecular explanation for how TACO1 dysfunction can impair mitochondrial protein synthesis ^[1][2]. It does not test a treatment, and it does not show that altering TACO1 or its movements could prevent or treat illness ^[1][2].

What comes next

The obvious next experiments are the ones the paper cannot yet supply. Researchers will want to test whether the proposed N-terminal movement and its tRNA-steadying action can be shown in biochemical experiments, rather than in simulation.

They will also want to know how particular disease-associated TACO1 changes alter the way it grips the ribosome or the tRNA. Whether the mechanism helps with other difficult mitochondrial sequences beyond COX1 and COX3 is another open question.

The early-access version of the paper may also change, since the journal says it was shared before final copy edits and the final version of record has yet to appear ^[2]. Specialists checking how robust the structural model is, including how many structures and particle classes underpin it, will need the paper's supplementary material ^[2].