Researchers at UCL, King's College London and the UK Health Security Agency, working with a collaborator at Mount Holyoke College in the United States, have described a new ring-shaped peptide from Bacillus smithii, a heat-loving bacterium used in industry. They named it smithiosin. The study was published online in ACS Chemical Biology on 5 November 2026.
This is early chemistry and genome work. The peptide turned out to be only weakly active against bacteria, and the authors say so themselves.
How it was found
The team mined the genome of Bacillus smithii DSM 4216, a strain first isolated from cheese, which grows at 55 degrees Celsius. They were looking for a sactipeptide, a peptide made by the cell's ordinary protein machinery and then closed into a ring by sulphur bridges called sactionines.
Smithiosin is a close relative of subtilosin A, a sactipeptide from Bacillus subtilis. High-resolution mass spectrometry showed three sactionine bridges in smithiosin, and the authors also used circular dichroism to look at its shape. They recovered about 0.4 milligrams of peptide per litre of culture, and they report that growing B. smithii was inconsistent from one run to the next.
An unexpected result in a lab host
The researchers also put the gene for smithiosin's precursor into Bacillus subtilis 168 without removing that bacterium's own subtilosin gene. The host made both peptides: subtilosin A at significantly reduced levels, and fully formed smithiosin, which normal B. subtilis does not make.
Smithiosin output was about a thousand times lower than subtilosin A, and the engineered strain was somewhat unstable when grown repeatedly. Even so, the authors write that the foreign precursor can compete with the native one and outcompete it for the host's own enzymes, including a radical SAM enzyme called AlbA. Their abstract describes this as shown for the first time, and they say they know of no earlier precedent. The two precursor peptides share 69.4% of their sequence.
Does it kill bacteria?
Not well. In tests run at the UK Health Security Agency, smithiosin stopped the growth of four of six Gram-positive strains of Staphylococcus aureus and Enterococcus faecalis at 128 to 256 micrograms per millilitre, while subtilosin A did so in none at the highest dose of 256. Neither peptide showed any effect on seven Gram-negative strains, which included Klebsiella, Acinetobacter, Pseudomonas and E. coli.
Standard antibiotics tested alongside worked at far lower doses in most strains. The authors say the difference between smithiosin and subtilosin A is not significant given how different the sequences are, and that smithiosin is currently insufficiently active to be used as an antimicrobial.
Why it still matters
The authors say that describing smithiosin as a natural product of B. smithii is valuable in itself. They also expect that, with more optimisation, expressing foreign precursor peptides in Bacillus subtilis could become a useful way to screen new sactipeptides. Nothing here has been tested in animals.