In 1928, the British chemist Alexander Fleming returned from a vacation in the countryside to find that his lab was a frightful mess. There was, for example, a pile of Petri dishes in his sink, each of which contained a carpet of Staphylococcus aureus—a bacterium that can cause severe skin infections. On one such carpet, Fleming noticed that a bit of mold had landed, and carved out a kill-zone of slaughtered bacteria. From that mold, Fleming isolated a chemical called penicillin, and kicked off the modern antibiotic era.

Like penicillin, all our antibiotics were created naturally by microbes to suppress or kill other microbes. We then found and exploited these weapons. Fleming himself was always clear about giving due credit: “I did not invent penicillin. Nature did that. I only discovered it by accident.” His successors were more deliberate, searching for new antibiotics among microbes that live in soil, oceans, and other habitats. Now, many are screening the human microbiome—the communities of microbes that share our bodies. And their searches are already paying off.

Teruaki Nakatsuji and Richard Gallo from the University of California, San Diego, have discovered that some bacteria which naturally live on human skin produce chemicals that kill S. aureus—the same species that Fleming was studying. But rather than harvesting said chemicals, the duo went after the bacteria themselves—isolating them from people with a skin disease called atopic dermatitis (eczema), growing them, and adding them to a cream. The result: a personalized ointment for killing S. aureus—and hopefully treating eczema—using bacteria that come from a person’s own skin.

S. aureus is a common part of the skin microbiome. But it’s particularly abundant on people who suffer from eczema, and especially on the dry, itchy, inflamed patches that characterize the disease. The lines of cause and effect aren’t entirely clear, but Gallo and others have suggested that S. aureus could at least partly drive the symptoms of eczema, by causing inflammation and triggering allergic reactions.

By contrast, other species of Staphylococcus (staph) could have the opposite effect—reducing inflammation and secreting antibiotics that keep their more unruly cousins in check. Perhaps, Gallo reasoned, the balance of these related microbes influences a person’s risk of eczema. And perhaps by amplifying the beneficial strains, he could keep the disease at bay.

He and Nakatsuji swabbed the forearms of several volunteers, and identified two species—S. epidermidis and S. hominis—that could suppress the growth of S. aureus. Not every bacterium within these species was the same; only some strains had the suppressive ability. The team focused on one—a strain of S. hominis called A9. When grown in a Petri dish, it carved out a zone of S. aureus, just as Fleming’s mold had done a century earlier.

Strain A9 produces several new antibiotics that seem to specifically suppress the growth of S. aureus, including the drug-resistant versions that we know as MRSA. By contrast, none of these chemicals harmed A9 itself, nor any of our other common skin bacteria. And these antibiotics seem to work together with chemicals that are naturally produced by the human skin—a case of host and microbiome cooperating to defend against disease.

Protective staph strains like A9 dominate the skins of healthy people, but they’re much rarer on those with eczema. There, “they accounted for maybe one out of every 300 bacteria,” says Gallo. “There’s a dramatic difference in this part of the microbiome between healthy people and those with the disease.” He had no idea why. A person’s genes could all influence which strains rise to the fore, as could their diet, environment, or skincare regimens. That’s a question for another day. For now, what matters is that even in people with eczema, the protective strains aren’t totally absent. They’re still there. So what happens if you give them a boost?

To find out, the team recruited five volunteers with eczema, and searched for those rare protective strains of S. epidermidis or S. hominis on their skin. They isolated these strains, sequenced their genomes, checked that they had antibiotic-producing genes, confirmed that they could suppress S. aureus, grew them in a lab, and then added them to a creamy lotion. The volunteers applied these probiotic creams to one of their arms, and refrained from washing for a day. As expected, the levels of S. aureus fell by more than 90 percent. In two cases, the troublesome microbe disappeared entirely.

“It’s a big step towards using microbial therapies to treat skin disease,” says Shruti Naik, from Rockerfeller University. “It will be interesting to take it a step further, and test if the beneficial microbes can dampen or cure eczema.” Gallo is keeping quiet on that for now, and doesn’t want to say how his five volunteers fared. That would be anecdotal and misleading, he says. His small study was only designed to see if the personalized skin probiotics could kill S. aureus. He’s now running larger clinical trials, in which up to 60 patients use the creams for weeks or months, to see if they’re safe, and to test if they can actually treat eczema.

Eczema has many potential causes, says Gallo. “Many patients have genetic problems in their skin, or underlying difficulties with their immune system. This won’t alleviate any of that. But we’ve shown that there’s a microbiome defect, and we can repair that part. In some patients, that may be enough.” And he argues that his approach beats slathering the skin with broad-spectrum antibiotics. “Those not only target S. aureus but also kill beneficial bacteria,” he says. “Our approach identified antimicrobials that have evolved to kill S. aureus while leaving the good bacteria alone.”

In a similar way, last June, German scientists identified a microbe called S. lugdunensis that lives in our noses and that also produces a chemical that specifically kills S. aureus. They called it lugdunin. That study, and Gallo’s latest one, suggest that we may be at the start of a new antibiotic era—one in which we look to our own microbial companions to defend our health.

These studies also herald a smarter approach to developing probiotics—products containing supposedly beneficial microbes. Current probiotics are medically underwhelming because they contain D-list microbes that aren’t well-adapted to life in the human body. By contrast, Gallo and others are now creating probiotics using species and strains that already common in the human microbiome. They recognize that only certain strains have the right abilities; A9 might suppress S. aureus, but other strains within the same species might not. And they’re personalizing their treatments to individual people, rather than going for a one-size-fits-all approach.

“This is the most compelling study to date showing the potential of skin probiotics in remediating atopic dermatitis,” says Julia Oh, from the Jackson Laboratory. And although the particular microbes and antibiotics that the team discovered may be specific to S. aureus, the techniques they developed can be used again and again to find other strains and molecules that protect against different threats.