Inner Workings: Microbiota munch on medications, causing big effects on drug activity

Millions of patients with Parkinson’s disease rely on the drug Levodopa for relief from tremors, slowed movement, and other motor symptoms. But many patients experience side effects such as cardiac arrhythmias, nausea, and gastrointestinal problems. Levodopa’s side effects and benefits vary widely among patients. Those puzzling disparities, it turns out, have a lot to do with the microbes in their guts.

Researchers have found that many side effects of the Parkinson’s drug Levodopa were the result of a bacterial decarboxylase enzyme, produced by the commensal gut microbe E. faecalis (pictured in colored-scanning electron micrograph). Image credit: ScienceSource/Dennis Kunkel Microscopy.

Earlier this year, chemist Emily Balskus and her colleagues at Harvard University in Cambridge, MA, found that many side effects were the result of a bacterial decarboxylase enzyme, produced by the gut microbe Enterococcus faecalis. Levodopa is an inactive form of the neurotransmitter dopamine and must be activated by a human decarboxylase enzyme to work. Activate the medicine too soon—before it crosses the blood–brain barrier—and side effects will occur. To block this premature activation, drug makers have long added an enzyme inhibitor known as carbodopa to Levodopa. But Balskus and her colleagues found that although carbodopa works on human enzymes, it does not inhibit bacterial decarboxylase. In fact, the bacterial enzyme acts on the drug in the intestines before it crosses the blood–brain barrier, triggering problematic symptoms (1).

Balskus and others are learning that the interactions among our microbes and medications are far more complex than previously assumed, potentially causing toxic side effects or altering drugs’ activity. Medications left unabsorbed in the body are usually marked for removal in the liver and then transported to the gut. Although human cells no longer recognize these excretory products, intestinal bacteria can act on inactivated drug molecules. This stage could be labeled the “fourth phase of drug metabolism,” says chemist Matthew Redinbo of the University of North Carolina at Chapel Hill. “Bacteria perform incredibly sophisticated chemical reactions that no human systems are able to do.”

Using a combination of chemistry and genomics, researchers are now beginning to identify these mechanisms. As they do, they’re uncovering ways to inhibit the microbial enzymes that cause distressing side effects. The end result could be drugs that are less toxic, as well as better predictions about how patients respond to medications.

Potent Pathways

Microbes’ drug-altering abilities have been known for nearly a century. In the preantibiotic era, German researchers discovered that the antimicrobial drug prontosil—once it was digested by enzymes in the liver and kidney as well as by gut bacteria—became a potent sulfanilamide, effective against many gram-positive bacteria (2).

Then there’s the case of the antiviral drug sorivudine. The drug turned deadly in 18 cancer patients, triggering its removal just two months after FDA approval in the 1990s. Mouse studies later suggested that intestinal microbes had likely digested sorivudine into a product that blocked the liver enzymes needed to metabolize the common cancer drug 5-fluorouracil (5-FU)—leading to a lethal buildup of 5-FU (3).

Only recently have researchers begun to systematically dissect the pathways involved in such unexpected interactions. Microbiologist Andrew Goodman of the Yale School of Medicine was studying the links between dietary chemicals and gut bacteria when he discovered that certain species depended on specific vitamins for their survival. Remove the chemical, and the commensal bacteria die out. “It was really in the process of trying to understand that conversation between commensals and the host diet that we started thinking about other small molecules that gut microbes might be recognizing and then transforming for their own purposes,” Goodman says. “That’s what led us to think about medical drugs.”

In a recent study, Goodman and his colleagues measured how 76 different species of human gut bacteria digested 271 common drugs that are taken orally. They found that even when 80% of a medication dose quickly entered circulation, microbial enzymes could act on the remaining 20% to produce toxic metabolites. Although the team chose drugs that were chemically very different from one another, they found that at least two-thirds of the molecules were metabolized by at least one gut microbe studied (4). “The capacity of these microbes to metabolize these drugs was much broader than we had expected,” Goodman says.

It wasn’t easy to predict precisely which medications would be metabolized. “Drugs with chemical structures that look like perfect bait for microbes were untouched for some reason,” Goodman says. “And drugs that didn’t have anything that looked like a microbe might recognize it were very efficiently degraded.”

In experiments with human microbiota samples, the researchers have begun to spot the enzymes that can help predict which bacterial communities are likely to metabolize drugs. The work is “just the beginning of understanding the degree to which almost everything in our bodies is subject to transformation by microbes,” Redinbo says.

Redinbo and others want to understand the microbial enzymes that act on drug molecules. Typically, the unabsorbed remnants of orally consumed drugs are transported to the liver, where enzymes inactivate them and add tags such as glucuronic acid to mark the molecules for excretion. Then they are transported to the intestines via bile acids. The chemical tags make inactivated drugs unusable by human cells.

But gut bacteria see the molecules as food sources, and they carry enzymes known as β-glucuronidases that can chomp up the glucuronic acid for energy and toss the drug molecule—now reactivated—back into the intestines. Redinbo’s studies began with the anticancer drug irinotecan, which can cause intense, delayed diarrhea as a result of this bacterial activity. They also found that a molecule they named SBX-1 inhibited these microbial enzymes and thus could block these toxic effects (5). “That was really the first demonstration that the microbiota contained druggable targets, and that they could be targeted for a clinical outcome,” Redinbo says.

A person’s microbiome can determine whether they will experience a drug's benefits, side effects, toxicity, or some combination. Image credit: From ref. 7. Reprinted with permission from AAAS.

Targeting Bacterial Enzymes

The team is also working to identify analogs of SBX-1 that might prove effective against other bacterial enzymes. Approximately 25% of clinical drugs, including several nonsteroidal inflammatory drugs, are targeted by bacterial glucuronidases. So enzyme inhibitors that block this activity could potentially reduce the toxic side effects of various other drugs.

In a similar vein, Stanley Hazen of the Cleveland Clinic in Ohio and his team are working on drugs that can block gut microbes from synthesizing trimethylamine N-oxide (TMAO), a potentially undesirable molecule that’s produced from fatty foods such as egg yolks, meat, and dairy. TMAO accelerates the buildup of plaque on artery walls and can lead to cardiovascular diseases. In animal studies, the researchers have found that blocking microbial TMAO synthesis reduced fatty deposits and could serve as a route to treating cardiac and metabolic disorders (6).

Thanks to recent genomic advances, drug developers are finding ways to tailor molecules to specific variants in the human genome. These so-called pharmacogenomic approaches allow patients to choose from amongst a few different drugs based on the person’s genetic profile. But if no good options exist, it may soon be feasible to tweak the patient’s microbiome instead, Goodman says. “We have the opportunity to alter a person’s microbiome in a way that we really wouldn’t think about altering their genome to improve their response to a drug.”

Matching Meds and Microbes

To perfect drug-modulating microbiome interventions, researchers will need to better characterize drug function and better gauge the effects of diet and gut community on microbial enzymes, says microbiologist Peter Turnbaugh of the University of California, San Francisco.

“The idea that the microbiome would be used in preclinical drug development is still sort of controversial, and definitely not the standard for drugs on the market.”

–Peter Turnbaugh

“We don’t have a simple rule to say this drug won’t be metabolized and that one will,” Turnbaugh says. “And even assuming your microbiome has some enzyme that affects a compound, we really don't understand what determines whether or not those enzymes are active.”

Elucidating these data could also boost the process of drug development. Many potential drugs are discarded early in development because of toxic side effects. But if those effects are the result of microbial meddling—and if researchers can identify the sources of such interference—the rewards could be significant: older drugs resurrected, and perhaps new ones tailored to patients based on their gut microbiota. “The idea that the microbiome would be used in preclinical drug development is still sort of controversial, and definitely not the standard for drugs on the market,” Turnbaugh says.

Nonetheless, given the mounting evidence for the microbiota’s drug-regulating effects, he suggests that microbial activity should be factored into drug design and clinical trials. “I'm not trying to argue that the microbiome is more important than the human genome,” Turnbaugh says. “But just as people look at specific human genes or mutations that might matter to a drug’s activity, we need to be at the same point with the microbiome.”