Andy Murray scratches an itch.Carlo Allegri / Reuters

Your nose is a battleground. Just like your mouth or gut, it’s full of microbes. But while those other organs are regularly  flooded with food, the nose is a wasteland. Resources are scarce there, and competition is fierce, so nasal microbes have evolved many ways of outclassing and killing each other. And by raiding their arsenals, we could gain new weapons for our own use.  

Alexander Zipperer and Martin Konnerth from the University of Tübingen have found one such weapon—a chemical called lugdunin. It has all the makings of a good antibiotic. It’s produced by a bacterium that already lives in our noses, which suggests  that it’s safe. It kills another microbe, Staphylococcus aureus, a common and typically harmless inhabitant of the nose. It even kills S. aureus when it’s dressed in its antibiotic-resistant alter-ego, MRSA. And it’s chemically unrelated to existing antibiotics, which opens the door to other new drugs. “It is the founding member of a new class of antimicrobial compounds,” says Andreas Peschel, who led the study.

It will take years to see if this new chemical will work in the clinic, and many similar promising leads have fizzled out. But right now, we need every lead we can get. For decades, pharmaceutical companies have failed to take any new classes of antibiotics to market. Meanwhile, many disease-causing bacteria have evolved to resist our existing weapons, and these impervious strains are predicted to kill 10 million people every year by 2050. To avert an antibiotic apocalypse, we need new drugs.

Even if lugdunin itself doesn’t pan out, that doesn’t matter. The true value of Zipperer and Konnerth’s work lies not in just one chemical, but in highlighting an approach for finding more. They show that the human microbiome—the microbial menagerie that shares our bodies—is a reservoir of potential antibiotics. We just have to tap into it.

We can start by thinking ecologically. Microbes dominate life on Earth, and very few of them cause us harm. Their main foes are each other, and antibiotic chemicals are their means of doing battle. For decades, researchers have stolen these natural weapons to make our own, largely from soil bacteria. But the looming antibiotic crisis has forced them to look further afield, to sources as diverse as the deep ocean, frog skin, panda blood—and now, the human microbiome.

Each of our body parts is its own ecosystem, as distinct in its climate and resident species as a grassland is from a forest. And since the nose is especially nutrient-poor and competitive, Zipperer figured that it would be a good place to look for powerful antibiotics. He was right: After isolating hundreds of strains, he found several species of Staphylococcus that blocked the growth of their close relative, S. aureus. One of these rivals—a strain of Staphylococcus lugdunensis—was especially promising. Drop it into a plate full of S. aureus and it carves a zone of dying microbes around it.

From that strain, Konnerth purified lugdunin. The substance has an unusual chemical structure and it’s unclear how it kills—but it clearly does. At relatively low concentrations, it completely destroys drug-resistant strains of both S. aureus and Enterococcus, as well as a number of other disease-causing bacteria. It even killed S. aureus that had colonized the skin of infected mice, sometimes eradicating it completely, and without any toxic side-effects. Its apparent safety makes sense: The microbes in our bodies are likely to deploy chemicals that have already been optimized for use in a mammalian host.

Of course, any new antibiotic faces the same fate as all the old ones: It will eventually be de-fanged by resistant strains. But Peschel’s team thinks that, for some reason, S. aureus has a tough time evolving resistance to lugdunin. When they exposed the bacterium to the drug over many generations, resistant strains never emerged.  

Karen Bush from Indiana University is skeptical. She notes that the team used low concentrations of bacteria, which would have minimized the odds of finding a resistant colony. She also notes that a few of the mice that were treated with lugdunin still had very high levels of S. aureus, which suggests that resistant populations may have arisen. Still, she says that the results are “respectable as a starting point for a new antibiotic,” and certainly worthy of follow-up work.

Meanwhile, Peschel’s team has already sniffed out other promising chemicals, which they haven’t revealed yet. They’re also considering using S. lugdunensis as a probiotic, dosing people with the entire bacterium rather than just the chemical it makes.

There’s reason to think that might work. When the team swabbed the noses of 187 hospital patients, they found that where S. lugdunensis exists, S. aureus fears to tread. The latter colonizes the noses of 35 percent of people without S. lugdunensis, but just 6 percent of those who have it. “It’s a stunning finding,” says Peschel. “It confirms that lugdunensis is eradicating aureus from its natural habitat.”

But even though S. lugdunensis is less aggressive than S. aureus, it can still cause infections in rare cases. Any probiotic strain would have to be modified, and “that’s science-fiction at the moment,” says Peschel. Alternatively he’s trying to find more benign microbes. After all, S. lugdunensis is only found in 9 percent of people, and 70 percent of people don't carry S.aureus. There must be other species keeping it at bay, and they might lead us to the antibiotics (or probiotics) of the future.

“It sets a great example,” says Cindy Liu from George Washington University. It shows that we could find new ways of fighting infections by exploiting the natural competitions between our resident bacteria. “That’s been one of my hopes for what can be accomplished through human microbiome research.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.