If your cholesterol levels are high, your doctor might prescribe you a statin, a drug that blocks one of the enzymes involved in creating cholesterol. But in the future, she might also prescribe a second drug that technically doesn’t target your body at all. Instead, it would manipulate the microbes in your gut.

Each of us is home to trillions of bacteria and other microbes—a teeming mass collectively known as the microbiome. These microbes affect our risk of several health problems, from obesity to bowel cancer. But they also offer some opportunities for controlling these problems.

Over the last five years, Stanley Hazen from the Cleveland Clinic has been amassing strong evidence that the microbes in our guts can affect our risk of heart disease. Certain gut bacteria can transform dietary nutrients like choline into a chemical called TMAO, which slows the body’s breakdown of cholesterol. As TMAO levels build, so do fatty deposits in our arteries, leading to atherosclerosis—a hardening of arterial walls—and other heart problems. Hazen and his team have tracked this process in mice, and they’ve also shown that higher TMAO levels are linked to heart-disease risk in people.

Now, Hazen’s colleague Zeneng Wang has found a chemical that stops this process at the beginning, at least in mice: It prevents microbes from turning choline into TMA, which in turn reduces the animals’ risk of atherosclerosis.

The specific chemical isn’t the important part of this discovery. “There have been many compounds that block atherosclerosis in rodents. So, in and of itself, that’s not super exciting,” says Hazen. What matters here is the team’s approach: If there’s a chemical that can prevent heart disease in mice by targeting their microbes, it may be possible to make a drug for humans that works the same way.

To be clear, the researchers aren’t trying to kill the microbes. Their substance isn’t an antibiotic; it just nudges the microbes’ behavior away from certain actions that negatively affect our health. “It’s a new approach to treating not just the individual Homo sapiens but also the microbes that live with us, and collectively contribute to disease,” says Hazen.

It takes two to TMAO: Bacteria first transform choline into TMA, before an enzyme from the host animal changes TMA into TMAO. At first, Hazen’s team tried to prevent the second part of this chain by blocking the animal enzyme. They succeeded, lowering TMAO levels in mice and making them resistant to  atherosclerosis. But there was just one problem: Disabling the enzyme leads to a build-up of TMA, which doesn’t harm the heart but does smell of rotting fish.

So Wang went after the microbes instead. He identified a substance called DMB that looks a little like choline, and acts as a gobstopper. It gums up the enzymes that the bacteria normally use to digest choline, which prevents them from producing TMA.

When Wang laced his lab rodents’ drinking water with DMB, he found that the mice produced far less TMAO, even when they ate choline-rich food. And their arteries benefited: Even though these mice were genetically bred to be prone to atherosclerosis, they developed fewer signs of the disease.

“This study is of potentially groundbreaking significance,” says Marius Troseid from Oslo University Hospital, who also studies TMAO. “So far, interventions targeting gut microbes, including probiotics, antibiotics, and fecal transplantaion, have been non-specific,” taking a broader approach to manipulating the microbiome. By contrast, Wang and Hazen’s technique seems to target one particular bacterial process.

The usual caveat applies: The team have only looked at mice, so it’s not clear if DMB would work in the same way in humans. But Hazen argues that there’s good reason to think it would. His work on TMAO started in people—he was searching for chemicals that predicted the risk of heart disease. Since then, he has bounced back and forth between mice and humans, checking that the same microbial processes are at work in both species. And they are.

If DMB, or some other chemical like it, works in humans, it would carry several advantages. It seems safe—it’s naturally found in foods like balsamic vinegar, extra-virgin olive oil, and some red wine.  It might also help to prevent diseases others than atherosclerosis, since TMAO has been linked to chronic kidney disease.

Perhaps most importantly, DMB doesn’t harm gut microbes or slow their growth. That’s because the gut is a nutrient-rich environment with plenty of molecules for the microbes to feast upon. Choline may be off the menu, but it’s a rich menu—the bacteria can just pick a different option. Hazen hopes that this will stop them from evolving resistance to DMB, as they would eventually do against a drug made to kill them.

But in the meantime, there are other variables that present their own challenges. Even though it wasn’t what the researchers intended, the chemical did change the makeup of the mouse microbiome, reducing the numbers of bacteria that produce TMA. This suggests that DMB might have been disadvantageous to those particular species in some unknown way. “We still don’t understand very much about what controls the whole microbiome,” says Hazen.

Troseid also notes that the gut microbiome varies a lot from person to person, which could make it difficult to develop a drug that works across the board. “Targeting the enzymatic activity of certain bacteria might be beneficial in some individuals, but it may not work in others,” he says.