THE BOTTLE LOOKED PERFECTLY ORDINARY. It contained a greenish, translucent liquid—jasmine tea. I had left it by my rowing machine a few days earlier. But when I opened it and took a swig, I nearly choked. The liquid had congealed into a jellylike ooze that slipped down my throat like an oyster, with shocking, nauseating ease. I had just swallowed a biofilm.

Most of us, including many scientists and physicians, think of bacteria as free-floating planktonic germs awash in a sea of solution—water, blood, pus, growth medium—or drifting in the air or buried in the soil. When researchers (or high school biology students) grow solutions of bacteria on dry or liquid media, they see generally unattached colonies, and studying germs in this way has probably conditioned us to think we are observing them in something like their natural state.

But during the past 10 years, bacteriologists have discovered that more than 99% of bacteria live enmeshed in a substance they produce, an extracellular polymeric matrix consisting of complex sugars known as exopolysaccharides, which may form as much as 85% of the biofilm’s entire volume. This matrix creates structures more akin to a gelatin of frog eggs than to a wash of plankton. A biofilm is actually a tiny ecosystem, composed sometimes of one species but often of many living together. In either case, bacteria in the biofilm both take from and contribute to their environment, as do the residents of any ecosystem.

In the 1980s, the pioneering researcher J. William Costerton, director of the Center for Biofilms at the University of Southern California in Los Angeles, dubbed these ecosystems biofilms, but you could also call them slime—and it is in the form of slime, as in that innocuous-looking bottle of tea, that we usually encounter them. The most complex examples, comprising many kinds of bacteria, form by adhering to a surface and then aggregating, piling up upon one another and communicating through chemicals called autoinducers. Biofilms may even contain pillars and channels through which water flows, bringing nutrients to microbes clinging to the pillars.

Biofilm architecture—“Biofilms: City of Microbes” was the title of a recent review article that Boston researchers Paula Watnick of Children’s Hospital and Roberto Kolter of Harvard Medical School published in the Journal of Bacteriology—is fascinating in itself. But as Watnick, among others, has shown, biofilms also play a significant—and until recently, largely unappreciated—role in human disease.

The germs that cause certain deadly infectious diseases, notably plague and cholera, form biofilms, an important phase in the germ’s life cycle. And in a particularly pernicious development, some bacteria have evolved the ability to occupy a recently created ecological niche, forming on hospital implants—ranging from simple urinary catheters to artificial hips to heart-valve implants—biofilms that are virtually impervious to antibiotics.

Learning how to treat diseases that, until recently, no one knew were caused by biofilms is daunting. Only by understanding how these structures work can scientists find ways to break through the exopolysaccharide matrix and assault the more vulnerable bacteria within biofilms.

BIOFILMS FALL INTO TWO BASIC CATEGORIES—those essentially composed of a single species and those that are more complex ecosystems, often incorporating not only bacteria but also protozoa and fungi. According to Bonnie Bassler, professor of molecular biology at Princeton University, some bacterial species are happiest growing in films by themselves, while others, such as the 600-odd species that together form the biofilm known as tooth plaque, seem to need one another’s presence to survive. (Only recently have scientists found the genetic footprints of all these dental inhabitants, and culturing them singly is almost impossible because they may not be able to grow alone. What all 600 species are doing on your teeth, how they affect your health and how they all cooperate, no one yet knows.)

The plague germ Yersinia pestis forms biofilms of the single-species variety—in the gut of carrier fleas. When an uninfected flea bites an infected mammal, the flea swallows a few plague germs that lodge in its midgut and then creep back into the proventriculus (a section of the foregut that leads to the midgut). As the bacteria multiply, according to plague expert Joseph Hinnebusch of the National Institutes of Health Rocky Mountain Laboratories in Hamilton, Mont., they exude an exopolysaccharide, forming a sticky matrix. The resultant biofilm seals off the proventriculus, trapping the blood the flea has ingested in the foregut. In short, its foregut is bulging, but the flea is starving. So it bites another mammal—this time an uninfected one—and part of the biofilm is injected into that host, where the mammal’s higher body temperature halts production of the sugar matrix. New bacterial cells, no longer confined to the matrix, are free to travel into the mammal’s deep tissue, causing another deadly infection.

While plague forms simple structures, multiple-species biofilms such as dental plaque are actually microbial cities, as Watnick and Kolter note. According to Watnick, these metropolises arise from a single layer of cells adhering to a surface. Once this monolayer has established itself, other bacteria are drawn to nutrients that the early arrivals release. After several layers of interacting bacteria have formed, says Watnick, you have a biofilm, whether it’s composed of a single clone or of many interacting species.

Some bacteria prefer to live attached to the substrate, which can be metal, plastic or cellular, while others live adjacent to channels of water that move nutrients through the biofilm. Bacteria don’t typically eat one another or any other intact organism—they acquire nutrients from their environment, often in the form of particles of DNA or other chemicals such as polysaccharides. Sometimes, though, bacteria within a biofilm kill and degrade other bacterial cells, yielding nutrients the predatory bacteria need. Some members of a biofilm community may be overtly predatory—for example, protozoa may graze on fields of biofilm like cows in a pasture.

On occasion, biofilms, especially those found in polluted water (“pond scum”) or in the lungs of cystic fibrosis patients, may become thick. But based on her research, Watnick suspects that healthy biofilms tend to be thin—like a film. “A thick biofilm may be a sick biofilm,” she says, “quite different from the balanced ecosystem existing in an unpolluted environment.” In a thick biofilm, cells in the middle die for want of nutrients. That means not all biofilm development is beneficial to all denizens of the microbe city.

FURTHER DEMONSTRATING THE EXTRAORDINARY COMPLEXITY of biofilms—and adding to the difficulty in treating diseases linked to them—some bacteria can exist both as environmental biofilms and as pathogens. For example, scientists have discovered that Vibrio cholerae, the germ that causes cholera, can float for years in both fresh water and seawater enmeshed in a biofilm, possibly associated with tiny crustaceans or blue-green algae. Cholera infects only humans, but its ability to survive outside the body and longer than any epidemic outbreak makes the disease almost impossible to eradicate.

According to Watnick’s experiments, when nutrients are plentiful, cholera germs form a biofilm by coating themselves with an exopolysaccharide matrix. Watnick thinks this matrix may serve as an extracellular nutrient-storage system, like fat cells in mammals. When environmental nutrients run out, the cells draw upon the matrix, and eventually the matrix dissolves. It’s possible that cholera germs then take refuge elsewhere. Watnick thinks V. cholerae may inhabit the intestinal tracts of flies or other arthropods, perhaps forming a biofilm as plague germs do in fleas. This raises the intriguing possibility that cholera may be spread by flies as well as through contaminated water or fecal-to-oral contact, long thought to be the only routes of transmission. It also suggests that cholera may have a complex life cycle—in biofilms, flies and people—that would make it all the more difficult, if not impossible, to banish from the environment.

Given evidence that cholera might form biofilms in humans, Watnick speculates that biofilms might fuel cholera’s rapid spread of infection by concentrating its toxin, which invades the cells lining human intestines and prompts ion channels in those cells to open wide. Fluid pours out into the intestines, causing victims to die of dehydration. The cholera germ is shed into the environment in feces, and may be picked up by other humans or returned to seawater to form another biofilm.

Other, less exotic human pathogens also form biofilms. Staphylococcus epidermidis, which lives as a harmless commensal, or fellow traveler, on human skin, shows an unnerving capacity to form low-grade—but stubborn and dangerous—biofilm infections on artificial implants. According to Paul Fey, associate professor at the University of Nebraska Medical Center in Omaha, S. epidermidisappears to be able to bind to any surgical material—stainless steel, zirconium, ultra-high-molecular-weight polyethylene—though when it comes to infection, not all S. epidermidis strains are created equal.

The strains of this normally harmless germ that cause infection all seem to contain a special packet of genes, or an operon, that codes for the formation of S. epidermidis biofilms, while cultures grown from S. epidermidis that has been taken from human skin generally lack it. Though no one is sure where the biofilm-forming genes came from, bacterial strains with the operon have clearly evolved to thrive in the hospital environment. In fact, Fey believes many S. epidermidis germs infecting catheters worldwide may be members of a single clone.

In patients these staph biofilms (with the operon) produce chronic infections that are almost impossible to cure. This is because biofilms have several ways to fight off attacks by antibiotics and antiseptics, says Philip Stewart, director of the Center for Biofilm Engineering at Montana State University in Bozeman. For example, the usually potent antiseptic chlorine bleach (which has been tried against biofilms that form in pipes and cooling systems) loses its killing power as it filters through a biofilm’s dense network of organic matter. In the human body, white blood cells normally kill bacterial invaders with tiny bursts of oxygen in the form of superoxide, peroxide and hydrochloric acid. But this mechanism too seems powerless against biofilms. It may deactivate surface layers, but it can’t get to all of the cells embedded in the matrix.

Antibiotics, effective against free-moving planktonic bacteria, also seem unable to destroy those that reside in biofilms. It appears that the exopolysaccharide matrix does not block the drug from reaching the bacterial cells, so there must be some other mechanism by which the biofilm neutralizes the antibiotic’s effect.

Part of the problem, says Stewart, is that antibiotics typically target only growing, metabolically active cells—for example, by preventing the cells from reproducing. But many of the bacterial inhabitants of biofilms exist in a dormant or inactive state—and thus are unaffected by an onslaught of antibiotics. As a result, even if a drug kills almost all of the bacteria in a biofilm, a few dormant cells may live to infect another day, re-forming a pernicious biofilm on an implant.

With conventional therapies thwarted, doctors are left with only one radical choice: to cut out the infected implant. For a patient with an infected artificial hip, that means having the entire implant removed, with a spacer left behind that must stay in place until the infection has been completely eradicated. Only then can a new implant be installed. This extreme approach is a throwback to the days before antibiotics, when surgeons had to cut out pockets of infection, and adds up to a lot of surgery—with all of the risks and pain that come with it.

Searching for less drastic therapies, Stewart and his colleagues are studying how biofilms function in the chronic foot ulcers of diabetes patients. They hope that if they find a topical therapy that works against these surface infections, which are easier to experiment with than are internal biofilms, it might also effectively combat biofilm infections on implants.

OTHER SCIENTISTS ARE EXPLORING WAYS to prevent biofilms from forming on implants, rather than trying to kill them after they develop. One pioneering technique, developed at the Centers for Disease Control and Prevention (CDC), uses bacteria-destroying viruses known as phages. Phages are strain-specific—in other words, each strain of S. epidermidis would need to be attacked by its own phage. So implants would have to be coated with several phages to fight the multiple kinds of staph that may form a biofilm, explains Rodney Donlan of the CDC. There is always the danger too that bacteria will develop resistance to their phages, and scientists don’t know how the human immune system will react to phage-coated implants. Phages are, after all, viruses and could prompt an immune response that might carry unknown risks. Still, Donlan is optimistic, pointing out that in the countries of the former Soviet Union, phage-based antibacterial therapy has been used successfully for decades.

Coating implants with antimicrobial agents may reduce but not entirely prevent the formation of biofilms and would be particularly ineffective against those caused by antimicrobial-resistant strains.

Some of the same issues affecting doctors trying to treat infected implants also bedevil clinicians treating run-of-the-mill infections. Biofilms have been implicated in middle-ear infections and in persistent urinary tract and prostate infections. Finding a way to break up the biofilm to expose the bacteria within or developing the means to penetrate the protective matrix could help in treating these infections. But researchers are still a long way from finding safe, practical treatments.

Nonetheless, our growing understanding of the natural history of bacteria—and their propensity to live enmeshed within a physical and even a social structure—has far-ranging implications. Researchers’ discoveries about biofilms during the past two decades are already challenging many assumptions and may eventually alter the ways we think about, and defend ourselves against, these strange and ubiquitous forms of microbial life.