Now a new study makes the interplay between the microbiome and general health more explicit: bacteria in the gut, it suggests, could cause a brain disorder.

The paper, which was published in Nature, traces a connection between the formation of blood-filled bubbles in the brain, called cerebral cavernous malformations, and the presence of a particular strain of bacteria.

Dr. Mark Kahn, one of the study’s authors, has studied this disorder, which can lead to serious conditions including seizures and stroke, for nearly a decade. It was only recently, however, that he make the connection between the disease and gut bacteria populations.

Below, he explains how his team of researchers stumbled on the link between cerebral cavernous malformations and the microbiome, and how this revelation could shape our understanding of how we treat a range of disorders. (Spoiler alert: It could involve fecal transplants via poop pills.)

This interview has been edited for clarity and length.

How common are cerebral cavernous malformations?

Pretty common. They occur in up to one in 100 people, although most of this group isn’t born with a genetic predisposition. In most cases, there is typically a sudden, single, solitary lesion.

But about 20% of patients have what is called a familial form of the disease, which is much more severe. Some present as infants, and experience stroke or seizures. Instead of having one lesion, they can have tens or even hundreds of lesions.

You’ve previously linked the familial form of the disorder to genetic mutations. How did you discover the microbiome might also be involved in the presentation of the disease?

When we deleted genes in mice, they developed cerebral cavernous malformations [in humans, this gene is mutated in many CCM patients]. At least at first.

But then our lab moved from one building at the University of Pennsylvania to a different building. When we moved the mice to the new building, they stopped showing lesions. Everything was being done exactly the same as in the building a block away. It was pretty mysterious and it pointed to an environmental influence. We hadn’t changed the genetics of these mice; all we had done was put them in a different facility.

This was a huge problem for us, until we came across an important clue. [To delete the relevant gene] we would inject the mice at birth. By putting a needle into an animal’s abdomen, occasionally the needle might go through part of the bowel, and cause a bacterial abscess. It’s actually pretty common in people too. If bacteria escapes from your gut and into your belly, it can end up causing a localized infection, which is like a large pimple, but in your abdomen.

Although this was a rare event, the animals that developed these abscesses were also the only animals developing lesions.

The bacterial abscesses were somehow causing the lesions?

Yes. We started to give them infections deliberately and when we did, we found they drove the development of massive numbers of cerebral cavernous malformations.

What did this suggest, and how did it relate to the relocation of your lab?

That a component of bacteria, when it gets into the blood, is what was driving the formation of lesions. This was totally unexpected! Why would a brain vascular malformation come from bacteria?

Once we knew that, we went back to the move in lab locations. If we had animals with the disease in facility A, and not in facility B, maybe that’s because when you move to a new place, the bacteria changed.

So a change in location can alter the microbiome?

It happens all the time in science and in human life. I’m in Italy now. If I stayed here for the next five years, I’d probably have a different microbiome than if I stayed in Philadelphia.

When we changed the mice’s environment, the bacteria in their guts changed. It wasn’t a deliberate experiment, it was serendipitous. But it had a profound effect. When we sequenced the types of bacteria in their gut, we discovered a change in one type of bacteria. They still had the deleted genes, but they weren’t developing lesions.

How does this apply to human patients?

In the familial form of the disease, there is an enormous variation in clinical course. There are patients who have the exact same mutation in the same gene, but some of them never develop lesions, and some of them do, as infants.

This is very unusual for a genetic disease. If everyone has same genetic problem, they should all present the same.

So it’s possible that in humans, as well as mice, gut bacteria could determine whether patients with the same genetic mutation develop lesions?

It’s a huge amount of work to look at the human microbiome. We haven’t done it yet. But what we are starting to do is look at several hundred patients in New Mexico who all have the same exact mutation on the CCM1 gene.

By collecting fecal samples from these patients, then, you’re hoping to make the same connection in humans that you did in mice?

We haven’t proven it. But [previous research in mice suggests] maybe there is a biome that is protective, and a biome that is not.

What we want to do is measure this bacteria [in humans], which requires getting samples and careful analysis. That’s exactly the next step.

If the microbiome and the development of the disorder is linked in humans, how could this lead to new treatments?

It’s not that hard to change someone’s biome. You give them antibiotics, which kills off their bacteria. And then you perform a fecal transplant, in which you literally giving them feces from someone else.

In theory, this will seed the gut with new bacteria. If we can prove the biome is as influential in people as it is in mice, we can actually treat people by giving them biomes we believe are protective. This is all still very hypothetical, but I do think we will see some diseases treated this way in the next 10 to 20 years.

How big of a deal is this?

The implications are pretty big. No drugs are without side effects, which makes the microbiome very attractive. If we are right, [i.e. gut bacteria impacts the formation of lesions in people with a familial form of the disorder], by changing a patient’s’ biome, he can be protected for the duration of his life without taking drugs. It could offer more elegant therapies.

What are the potential drawbacks?

We are just starting to learn about the microbiome. There are almost no diseases right now for which altering the microbiome is an accepted therapy. This is all very new.

There’s been a lot of research linking the microbiome to a host of disorders, including obesity, diabetes, and ADHD. Is it possible fecal transplants could inadvertently cause new problems, even as they solve old ones?

You know, there’s no free lunch: whenever you alter something, you maybe get some good, but you are probably going to get some bad. We’re at the very beginning of this era of medicine and biology, and we don’t know what the answers are yet.

Are there any general takeaways you think your study reveals?

We’ve worked on this disease for almost a decade. We never imagined in our wildest flights of idea that it could be affected by bacteria in the gut. The gut is in no obvious way connected to the brain.

What this really tells you is that the biology in the body is complicated—there are so many things going on that we are not aware of yet.