Scientists are great at growing E. coli in the lab. They know exactly under which conditions various strains thrive. Unfortunately, there is only so much that can be learned from the bacteria’s behavior in an ideal, isolated and ultimately unrealistic environment. That is why a group of researchers at the University of California, Riverside, decided to study how Escherichia coli O157:H7—a pathogen that can cause hemorrhagic diarrhea and kidney failure—is transmitted to groundwater after a human eats contaminated food.

“The novelty comes from that fact that we were studying bacteria once it goes through the human colon and a septic tank, right up until it reaches groundwater,” says study co-author Ian Marcus, an environmental microbiologist. “And we were doing it without isolating the E. coli strain from other microorganisms—a much more realistic approach than previous experiments.”

But in order for this study, published in Applied and Environmental Microbiology, to be successful Marcus and his colleagues had to build a replica of the human colon first—and “feed” it three times a day. “We were giving it things that the human body isn’t able to digest up until that point, like fiber, and certain sugars and proteins, all of which are based on a Western-style diet,” Marcus says.

By analyzing the faux colon’s contents, the septic tank and the resulting groundwater, the researchers discovered that when E. coli interacts with other microorganisms, like fungi and protozoa, it is prone to forming biofilms. Biofilms are microbial communities composed of many types of microorganisms, including bacteria, which latch onto one another and to a surface. Whereas biofilms aren’t necessarily harmful, once formed they can be very difficult to remove, which is a problem when they contain pathogens. “In our study the E. coli was often protected within a biofilm,” Marcus says, “so it was stronger and persisted much longer in the environment than previously thought.” When isolated strains of E. coli are released into groundwater, they tend to travel much longer distances because they aren’t attached to anything, “but they also die much more quickly,” Marcus notes.

The scientists had considered using human subjects in the experiment, but opted for a colon model because they wanted to make sure every variable was consistent. “We couldn’t let the experiment be affected by whether someone decided to have a candy bar for lunch,” Marcus says.

But because of this experimental simplification, where the food mixture and the colonic environment were unvaried, the researchers can’t definitively say that E. coli cells grown in such systems are more likely to form biofilms. “What we can say is that moving towards a systems-based approach instead of a reductionist approach in environmental biology yields vastly different results,” Marcus explains. “And these results are probably much closer to the real thing.”