Microbes make metabolic mischief by targeting drugs

Tests of whether a range of gut bacteria can metabolize a diverse group of drugs has revealed that all the microbes metabolized some drugs and that more than half of the drugs were metabolized.
Kim Lewis is in the Department of Biology, Northeastern University, Boston, Massachusetts 02115, USA.

Search for this author in:

Philip Strandwitz is at Holobiome, Cambridge, Massachusetts 02139, USA.

Search for this author in:

All humans are different and, unsurprisingly, also differ in their response to drug treatments. It is usually thought that this variation is due mainly to differences in liver enzymes that specialize in detoxifying ingested molecules. Such enzymes can metabolize drugs, with consequences that include reducing or eliminating drug potency or making them toxic. Understanding how an individual will respond to a given drug is important in developing treatment plans. Yet our knowledge of drug fate in the body is still rudimentary, despite a long history of studies in this area. Writing in Nature, Zimmermann et al.1 put human gut bacteria in the spotlight in the quest to understand how drugs are naturally metabolized.

A handful of previous examples have revealed that the community of microorganisms residing in the gut, termed the gut microbiota, can affect drugs. A classic example is the case of prontosil, the first widely used antibiotic. In the 1930s, the microbiologist Gerhard Domagk found that prontosil could tackle infection by the bacterium Streptococcus pyogenes in mice2. It was later established that prontosil is metabolized by gut bacteria to generate the molecule sulfanilamide, which is the active form of the drug3. Interestingly, had prontosil been tested for activity against S. pyogenes in a test tube, as we do today, its capacity to generate an antibiotic would have been missed.

Other examples of gut bacteria affecting drugs include the microbial inactivation of digoxin, which is used for heart conditions4, and the bacterial modification of the chemotherapeutic agent irinotecan, which causes toxic side effects5. Zimmermann and colleagues devised a large-scale approach to tackle the open question of how widespread drug metabolism by the microbiota is.

The authors conducted in vitro tests to assess the ability of 76 bacterial strains from the human gut, representing 68 species from the main bacterial taxonomic groupings, to metabolize 271 drugs (Fig. 1). These drugs were chosen to provide a diverse group in terms of factors such as molecular structure or effect on the body. Zimmermann and colleagues report that 176 of the drugs tested underwent a substantial metabolic change, caused by least one bacterial strain, that resulted in a reduction in the level of the active drug molecule in bacteria. Each bacterial strain tested metabolized some of the drugs, with the numbers ranging from 11 to 95 drugs per strain. Given that the authors tested a broadly representative panel of drugs, the scale of these results is remarkable because it raises the possibility that most drugs are modified by the microbiota. This type of testing could also be a useful way of singling out drugs that would probably be deactivated by the microbiota.

Figure 1 | Studying drug metabolism by gut bacteria. a, To assess how commonly drugs are metabolized by bacteria in the human gut, Zimmermann et al.1 tested the ability of 76 bacterial strains (representing 68 species across the main bacterial taxonomic groupings) to metabolize 271 drugs that have diverse structures and functions. This revealed that 65% of the drugs were metabolized — an unexpectedly high number. Some drugs were metabolized into more than one molecular form, and all the bacteria metabolized some of the drugs tested. b, To identify some of the bacterial enzymes responsible for drug metabolism, the authors focused on the gut bacterium Bacteroides thetaiotaomicron, which metabolized numerous drugs. Zimmermann and colleagues isolated sections of the B. thetaiotaomicron genome and inserted them into pieces of circular DNA called plasmids. Plasmids were inserted into the bacterium Escherichia coli, which expressed the proteins, such as enzymes, encoded by the B. thetaiotaomicron DNA. When these E. coli bacteria were exposed to one of the drugs tested, diltiazem, some of the bacteria did not metabolize the drug, but those that did helped to identify the B. thetaiotaomicron enzymes responsible for its metabolism.

Zimmermann and colleagues analysed the products of the 176 metabolized drugs using mass spectrometry. This revealed that 868 molecules are derived from these drugs. The numbers indicate that more than one metabolite can be produced from the metabolism of some drugs by gut bacteria. The mass-spectrometry analysis revealed the types of drug modification that occurred, which covered a wide range of chemical alterations, including oxidation, reduction and acetylation (the addition of a C2H3O group). The implications of this unexpectedly high diversity of drug alterations will no doubt take researchers a while to address. In the meantime, Zimmermann et al. report a few cases of drug metabolism that they examined in detail.

To identify some bacterial enzymes responsible for drug metabolism, the authors chose to profile the gut bacterium Bacteroides thetaiotaomicron. This species was a prolific drug metabolizer in their study, modifying 46 of the drugs tested. Zimmermann and colleagues studied how B. thetaiotaomicron metabolizes diltiazem, which is used to treat hypertension. The authors engineered Escherichia coli bacteria to express sequences from the genome of B. thetaiotaomicron, and tested whether the engineered bacteria could metabolize diltiazem. They found that the B. thetaiotaomicron gene bt4096 is required to metabolize the drug.

To validate their finding, Zimmermann et al. engineered a strain of B. thetaiotaomicron that lacked bt4096, gave germ-free mice either this strain or wild-type B. thetaiotaomicron, and then gave all the animals diltiazem. This confirmed that bt4096 encodes an enzyme that metabolizes diltiazem. Taking a similar approach, the authors identified genes that are needed to metabolize 18 of the drugs that B. thetaiotaomicron can modify.

This type of general strategy should enable the identification of the enzymes in gut bacteria that can metabolize any given clinically used drug or therapeutic molecule in development. Such information would also be useful when testing candidate therapeutics in clinical trials, to try to determine whether a person has gut bacteria that are particularly good at inactivating a particular drug.

Zimmermann and colleagues’ study offers a remarkable advance in our understanding of drug dynamics in the body, and will serve as a blueprint for other studies in the fledgling field that seeks to track the effect of microbes on drug metabolism. Yet despite the impressive scope and depth of this analysis, many questions remain, inviting an impatient reader to speculate in the meantime. One issue to consider is that, rather than being taken orally, many drugs are delivered by injection and thus would not be expected to encounter gut bacteria (although some drugs that are delivered by injection can reach the gut and re-emerge in the bloodstream). However, there is a general trend in drug delivery towards oral administration, and advanced methods to facilitate this are in development6,7. Over time, there might be a large-scale transition from the use of injected drugs for therapy to more widespread oral delivery. If so, the need to understand the microbiota’s role in drug metabolism will become even more urgent.

Drug metabolism by gut bacteria adds to the growing list of ways in which the microbiota can affect the human body. The considerable variation in the microbiota from individual to individual probably also results in variation in drug metabolism. In addition, diet can have a major effect on the composition of the microbiota8. Does diet affect the efficiency of drugs by affecting the microbiota? Such issues highlight the complexity of considering a person’s microbiota when trying to take a personalized-medicine approach. Adjusting the microbiota to suit our needs, including achieving individually tailored approaches to tackling drug metabolism, is probably where this field is heading.

Nature 570, 453-454 (2019)


  1. 1.

    Zimmermann, M., Zimmermann-Kogadeeva, M., Wegmann, R. & Goodman, A. L. Nature 570, 462–467 (2019).

  2. 2.

    Domagk, G. Dtsch. Med. Wochenschr. 61, 250–253 (1935).

  3. 3.

    Tréfouël, J., Tréfouël, T., Nitti, F. & Bovet, D. C. R. Soc. Biol. 120, 756–758 (1935).

  4. 4.

    Haiser, H. J. et al. Science 341, 295–298 (2013).

  5. 5.

    Guthrie, L., Gupta, S., Daily, J. & Kelly, L. NPJ Biofilms Microbiomes 3, 27 (2017).

  6. 6.

    Abramson, A. et al. Science 363, 611–615 (2019).

  7. 7.

    Banerjee, A. et al. Proc. Natl Acad. Sci. USA 115, 7296–7301 (2018).

  8. 8.

    David, L. A. et al. Nature 505, 559–563 (2014).

Download references

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.