For more than half a century, the standard approach to antibiotic discovery has been straightforward: find a microbe that kills other microbes, isolate the molecule, and purify it. But bacteria themselves have been playing a different game, one that is far more sophisticated than the single-molecule, single-target strategy that has dominated medicine. A team at McMaster University has now uncovered a striking example of nature’s combinatorial approach: a “megacluster” of bacterial DNA that encodes four different antibiotic classes and a biotin-stealing protein, all working together to starve and kill competing bacteria.
The discovery, published June 24 in *Nature* by first author Rodion Gordzevich and senior authors Gerry Wright and Eric Brown, reveals a stretch of DNA roughly 66,000 base pairs long in the soil bacterium *Streptomyces* sp. WAC05950. Within this single locus sit four separate biosynthetic gene clusters, a “cluster of clusters” that Brown called “unheard of” in bacterial genomics.
A multi-pronged attack on a single target
What makes the megacluster remarkable is not just its size, but its coordination. All four antibiotic families it produces, the stravidins, acidomycin, the newly discovered dapamycins, and alpha-methyl-KAPA, attack different steps of the same essential metabolic pathway: biotin synthesis.
“What this bacterium has done is evolve a complete chemical arsenal that hits one vulnerable point in its competitors from every angle,” Wright told *Nature News*.
The megacluster also encodes streptavidin, a protein that binds free biotin with extraordinary affinity, further starving nearby bacteria of the vitamin they need to survive. Lab tests confirmed that the combination is synergistic, stravidin S2 and alpha-methyl-KAPA together were more effective against multidrug-resistant *E. coli* in mice than either compound alone.
The four antibiotic families work as follows:
– Stravidins inhibit BioA, an enzyme in the early biotin pathway, while co-produced streptavidin sequesters free biotin from the environment
– Acidomycin inhibits BioB, a later step in biotin synthesis, and is selective against mycobacteria
– Dapamycins, an entirely new class of natural product, inhibit BioD via covalent cofactor mimicry
– Alpha-methyl-KAPA, a molecule known to science but never before recognized as an antibiotic, acts as a prodrug that inhibits BioA after metabolic activation
The blind spot in conventional screening
Perhaps the most significant implication of the discovery is what it says about the way antibiotics have been discovered historically. Standard laboratory screening uses nutrient-rich media that contain abundant biotin. Bacteria that rely on attacking biotin metabolism are invisible under those conditions, their weapons simply don’t work when the target vitamin is freely available.
“There is no reasonable way to know whether molecules that target these nutrient acquisition and synthesis systems are actually working when the nutrients themselves are so overwhelmingly abundant in lab conditions,” Brown said. This suggests that an entire class of naturally occurring antibiotics, those that attack essential nutrient pathways, may have been systematically missed by conventional screening for decades.
The megacluster itself is not rare. The team found that it is even more widespread across *Streptomyces* genomes than the gene cluster responsible for producing streptomycin, one of the most classic antibiotics discovered in the 1940s. It is highly conserved and ubiquitous among *Streptomyces* species.
“We’ve already catalogued dozens of other nutrient-targeting natural products,” Gordzevich noted. The team plans to publish a comprehensive review of these hidden antibiotics in a forthcoming paper.
From soil to clinic
While the discovery is exciting, clinical translation faces significant hurdles. Efficacy has only been demonstrated in a mouse model of multidrug-resistant *E. coli* infection. The natural compounds have unknown pharmacokinetic profiles, metabolic stability, and toxicity in humans. And biotin is essential in mammals too, the mouse model required pre-treatment with streptavidin to sequester the host’s biotin, adding complexity to any future formulation.
Nevertheless, the multi-target strategy offers a theoretical advantage: by attacking multiple points in a single essential pathway simultaneously, it could make it much harder for bacteria to evolve resistance.
“This is nature’s own combination therapy,” Wright said. “The question now is whether we can learn from it and apply the same logic to our own drug design.”
The researchers note that the finding also opens the door to targeted genome mining for other megaclusters, computational searches for similarly large, coordinated biosynthetic loci that may have been overlooked by conventional analysis tools.
—
Source: Gordzevich, R., Xu, M., Wang, W., et al. (2026). A *Streptomyces* megacluster encodes synergistic biotin-targeting antibiotics. *Nature*. https://doi.org/10.1038/s41586-026-10647-9