A person’s bacteria, yeast and other tiny companions have, in recent decades, gotten an image boost. A thriving microbiome is now understood as a sign of good health, and the market for “probiotics” is thriving, though concerns have been raised about the safety and effectiveness of many over-the-counter products. But what if science could add bacteria to the body that was guaranteed to have a therapeutic effect?

Enter the idea of engineering the human microbiome. New strains of microbes can be, according to some researchers, molded to specific therapeutic ends: to calm or stimulate the immune system, to help with metabolic disorders or to produce a constant supply of some critical medication. The early explorers of this idea believe the microbiome has the potential to become a portable pharmacy in the gut, though they caution patience; the microbiotic frontier is still, for the most part, untamed.

One immediate use for engineered microbes might be to hitch a ride to tricky places in the gastrointestinal system, such as the large intestine. According to Nathan Crook, an assistant professor of chemical and biomolecular engineering at North Carolina State University, drugs aimed at the large intestine conventionally come either as an injection, which can have off-target effects, or “oral medications that have to pass through the stomach and small intestine, where enzymes and low pH can break the drug down and make it less effective,” Crook says.

In contrast, certain hardy microbes—species of so-called lactic acid bacteria and the yeast Saccharomyces boulardii, for instancecan survive these harsh environments and deliver a therapeutic payload.

Crook’s lab is working on a treatment for infections with Clostridium difficile, a drug-resistant bacterium that can cause severe diarrhea and dehydration. C. diff leads to thousands of deaths per year. One arm of his team’s research is developing the yeast Saccharomyces boulardii to be a microbial delivery system. In a study published in ACS Synthetic Biology, the lab showed it could produce small molecules in a mouse gut. The other arm is trying to create the right molecules that would bind to C. diff’s toxins and prevent the damage caused by an infection.

Different engineered microbes could target other parts of the body. Christy Carter, associate professor of gerontology at the University of Alabama at Birmingham, is working with a bacterium engineered to produce a peptide called angiotensin (1-7), or Ang (1-7). This peptide is one of many players in the renin-angiotensin system, which is important for controlling blood pressure, but also in regulating pathways related to aging and disease, says Carter.

In a 2020 study, Carter’s team showed that aged rats who ingested these engineered bacteria had reduced proinflammatory gene expression in their brains. Brain inflammation is a hallmark of many forms of dementia, and this engineered microbe might serve as a therapy for Alzheimer’s disease.

Cancer treatment may be another frontier for engineered bacteria. In work recently published in Science Translational Medicine, the lab of Tal Danino, associate professor of biomedical engineering at Columbia University, showed that an engineered strain of E. coli produced antibody fragments with effects that were similar to those of checkpoint inhibitors, a promising class of immunotherapy drugs. A single injection of these bacteria into the tumors of mice caused many to shrink, and those that got the shots lived longer than other mice that received a single dose of traditional antibody therapies. The researchers hope that this more localized production of the drug can reduce immune-related adverse reactions elsewhere in the body.

Some engineered probiotics are already being tried in humans, including as a treatment for Type 1 diabetes, in which the immune system destroys the body’s insulin-producing cells. One clinical trial is testing bacteria that have been engineered to produce two substances: the autoantigen proinsulin—a trigger for the damaging immune response—and interleukin-10, an anti-inflammatory cytokine. The hope is that the combination can help the immune system “learn” to tolerate insulin-producing cells, halting their destruction. In an initial analysis of results, more than half of adult patients showed stable or increased production of insulin, a sign that the disease may have stopped progressing.

Another clinical trial is enrolling patients with phenylketonuria, a genetic disorder in which the body is unable to process a certain amino acid. Those who have the disease must adhere to an extremely limited diet—avoiding milk, eggs, cheese, meat, fish, nuts and beans—or risk serious neurological complications. The engineered bacteria in the study do what the body can’t, producing enzymes that break down the amino acid. In a previous trial, the microbe was shown to be safe and it metabolized the amino acid. Now researchers are measuring how well it reduces levels of this amino acid in patients’ blood.

Although these advances are promising, engineering a living microbe and setting it loose in the gut brings many uncertainties. Unlike a traditional drug, these so-called “biotherapeutics” aren’t static: They can grow, evolve and die in the body. Science doesn’t yet fully understand the ecology of the human gut and how different microbes find their niche. If an introduced microbe dies too quickly, it won’t be able to produce enough of the drug to be effective, but if it colonizes the gut too effectively, it could produce toxically high levels of the drug.

Nathan Crook at North Carolina State says that differences between peoples’ microbiomes—determined by genetics, diet and behavior—could also be a major hurdle. “Probiotics are going to have to be either robust enough to function in different hosts, or you’re going to have to develop different probiotics for different people,” he says.

Even if an engineered microbe successfully colonized the gut, it wouldn’t necessarily stay the same over many generations. Because producing a drug costs the microbe energy and likely provides no survival benefit, this trait may eventually be lost. That could require repeated dosing of an engineered microbe not intended to survive long-term, says Crook.

“Horizontal gene transfer” is another concern. Bacteria occasionally swap genetic material with other bacterial species. So, any gene that scientists put in an engineered bacterium —a gene to promote survival in the human gut, for example—could potentially end up in the DNA of another species. That might improve survival for a pathogen, too.

Despite these open questions, and the challenges they pose to scientists and regulators, researchers in the field are hopeful that engineered microbes can serve as a new way to introduce chemical changes that can be either broad or narrowly targeted, depending on the need. In this way, the body’s small companions can carry a great deal more than their own weight.