Protein swaps reveal why a photosynthetic bacterium needs TrxA

Why similar proteins are not interchangeable

Photosynthetic cells must coordinate the chemical reactions that capture carbon, build proteins and respond to oxidative stress. Thioredoxins help regulate these processes by changing chemical bonds involving sulfur-containing cysteine residues in other proteins. Although different thioredoxins share a basic molecular structure, they do not necessarily regulate the same partners equally well.

In the cyanobacterium Synechocystis sp. PCC 6803, TrxA is essential. It is also much more abundant than the organism’s other thioredoxins. That creates a straightforward question: do cells need TrxA because they need a large supply of thioredoxin activity, or because this particular protein has properties its relatives lack?

Researchers addressed that question through genetic replacement experiments, biochemical measurements and computational structural predictions. Their study was posted on bioRxiv on November 9, 2026, and has not been certified by peer review.

More protein was not enough

The team engineered Synechocystis strains to produce extra amounts of TrxA or the related proteins TrxB, TrxQ and TrxC. An arsenite-responsive genetic control allowed them to drive expression experimentally.

Protein measurements showed that TrxB rose to roughly seven times its normal level and TrxQ to about 35 times its normal level. These increases brought both proteins into a range comparable to the amount of TrxA normally present in unmodified cells.

The researchers then attempted to disrupt the original trxA gene. With an additional TrxA supply, they could obtain strains in which the disrupted version fully replaced the intact gene at its original location. With extra TrxB, TrxQ or TrxC, they could not: intact trxA remained detectable. Similar results in a second genetic background supported the finding.

The team also introduced NTRC, a different redox-regulating protein from another cyanobacterium, Anabaena. Its expression did not improve growth in a strain with reduced TrxA levels. Under the tested conditions, neither increasing related thioredoxins nor introducing this alternative system overcame the requirement for TrxA.

Swapping sections reveals a critical region

To investigate what makes TrxA distinctive, the researchers constructed hybrid proteins combining sections of TrxA and TrxB. Initial swaps exchanged regions on either side of the catalytic site, where the thioredoxin’s chemical reaction occurs.

The direction of the swap mattered. A hybrid with the beginning of TrxB and the later, C-terminal region of TrxA supported complete disruption of the original trxA gene. The reverse combination did not.

Further swaps narrowed the relevant region. One successful hybrid retained TrxA residues 49–106, encompassing several folded structural elements beyond the catalytic site. Hybrids retaining smaller portions of this region did not support complete replacement. This implicates a substantial region of TrxA rather than only the residues immediately adjacent to its catalytic machinery.

Successful replacement did not mean that the hybrids reproduced every aspect of normal function. One strain grew slightly more slowly even when expression was induced. Removing the inducer caused growth retardation, especially in the strain carrying the initial successful hybrid, and both successful hybrid strains showed reduced chlorophyll and phycobilin pigments.

Testing interactions with metabolic enzymes

The researchers next examined whether the hybrids could activate purified enzymes involved in photosynthetic carbon metabolism. These experiments were performed outside cells, with a chemical reducing agent included in the reaction mixtures.

For the cyanobacterial enzyme FBP/SBPase, TrxA and the two hybrids that supported genetic replacement produced similar activation. TrxB had a weaker effect, while the reverse hybrid and two more extensively substituted variants failed to activate the enzyme.

Tests with FBPase from pea chloroplasts showed a broadly similar pattern. However, one hybrid that failed to activate the cyanobacterial enzyme retained reduced activation of the pea enzyme. The differences underline that performance depends on the target protein being tested.

A surface-recognition explanation, with limits

Computational models suggested that the proteins differ in the distribution of electrical charge and in surface features surrounding their catalytic sites. These features could affect how a thioredoxin recognizes and positions a partner protein. The successful hybrids retained important TrxA-like regions that may help create productive contacts.

Those molecular contacts remain predictions, not experimentally determined structures of the interacting complexes. The work also does not establish which individual residues explain essentiality, or which combination of cellular targets makes TrxA indispensable. Activating one purified enzyme cannot account for all its functions inside a cell.

The strongest conclusion is narrower: abundance alone does not explain TrxA’s essential role under the conditions tested. Its C-terminal region carries molecular features that allow substantial TrxA-like function when transferred into a related protein.