Slime City: Where Germs Talk to Each Other and Execute Precise attacks

From the July-August special issue, published online July 17, 2009

For 300 years, scientists thought of bacteria as individual killers, like a bunch of piranhas. Recently, we've found that's almost entirely wrong.

Perhaps you notice it after a visit to the dentist.You pass your tongue across the front of your teeth and they feel slick and squeaky-clean. Four hours later, although you might not yet be able to tell the difference, the beginning of a rough fuzz is growing. These are streptococci, the first bacterial settlers in the film that saliva deposits on your teeth. Another four hours and the bridge germs, the fusobacteria, have climbed on board. They are the ones that make it possible for the really bad actors, like Porphyromonas gingivalis, to grab on and start building colonies.

By the next morning, if you still have not brushed your teeth, a definite fuzzy scum is starting to form. If you could look at that fuzz under a microscope without disturbing its structure, you would see towers or entire communities of bacteria, each building upon others. Some of those microbes are dangerous indeed. P. gingivalis not only grows in the pockets of your gums, helping to loosen your teeth from your jaws, but also causes the release of inflammatory chemicals that get into your circulation, complicating diabetes treatment and possibly increasing the risk of heart disease. Traces of the germ have also been found in arterial plaque.

If you have ever been admitted to a hospital, it is very likely you have experienced another, related kind of scary bacterial growth—and in this case you almost certainly did not notice it. Hospitalized patients are routinely hooked up to urinary catheters that enable doctors and nurses to measure urine flow (not incidentally, the catheters also liberate health-care workers from having to take patients to the bathroom). Swiftly coated by a conditioning film made of proteins in the urine, the catheters are then inexorably covered by layers of interacting bacteria, which alter the chemistry of their surface and can cause crystals to form. Within a week, an infection is growing on the catheters of 10 percent to 50 percent of catheterized patients. Within a month the infection has reached virtually everyone.

These slimy bacterial colonies, known as biofilms, add a remarkable new dimension to our understanding of the microbial world. Ever since Louis Pasteur first grew bacteria in flasks, biologists have pictured bacteria as individual invaders floating or swimming in a liquid sea, moving through our blood and lymph like a school of piranhas down the Amazon.

But in recent years, scientists have come to understand that much, and perhaps most, of bacterial life is collective: 99 percent of bacteria live in biofilms. They vary widely in behavior. Sometimes these collectives are fixed, like a cluster of barnacles on a ship’s hull; other times they move, or swarm, like miniature slime molds. Bacteria may segregate into single-species biofilms, or they may, as in the case of dental bacteria, join together in groups that function like miniature ecological communities, competing and cooperating with each other.

The unifying factor in all these biofilms—the thing that makes them so strange and wonderful and dangerous—is that their cooperation is, in a sense, verbal. Using streams of chemicals that they pump outside their cell walls and membranes, they “talk” incessantly, among their own clones and species and even to unrelated bacteria dwelling nearby. Understanding that chatter could be vital for gaining the upper hand in the endless battle against infectious disease.

Biofilms were first discovered in 1978 in the clear waters of a frigid mountain stream in British Columbia. Microbiologist William Costerton, now of Allegheny General Hospital in Pittsburgh, and his team of scientists wondered why there were so few bacteria in the water, while billions upon billions of the germs nestled in the crevices of the streambed’s rocks. “We were finding 9 bacteria per milliliter in the water, but there must have been 100 million in a square centimeter when we took a rock out of a stream and brought it down to the lab,” Costerton says.

The bacteria were not just sitting idly on the rocks, he found. They were forming complicated structures, cities of germs encased in a slippery substance the bacteria exude called an exopolysaccharide matrix. This slime protects them from grazing amoebas and provides them with food that is excreted by bacteria within the biofilm or even bits of DNA released when other germs die.

When Costerton published his results, he coined the term biofilm and introduced a whole new understanding of how bacteria behave. “We reasoned one stubborn fact,” he recalls. “Bacteria have no idea of where they are. They are just programmed to do their thing.” In other words, they are always going to form biofilms —whether they are living on a rock or in the human body.
Two years later Tom Marrie, a young doctor working in Halifax, Nova Scotia, examined a feverish homeless man who had wandered off the street and into his emergency room.

The man had a raging staph infection and, on his chest, a lump the size and shape of a cigarette pack. It was an infected pacemaker, Marrie reasoned. For three weeks the man was given huge doses of antibiotics but did not get better, so Marrie and his team decided to operate. They invited Costerton to sit in. “If there were ever going to be a biofilm infection in a human being, it was going to be on the end of that pacemaker,” Costerton says. “We took out the pacemaker and there was our first medical biofilm. It was a great big thick layer of bacteria and slime, just caked on.”

Biofilms on implants are now recognized as a serious and growing health problem. Bacterial infections hit 2 percent to 4 percent of all implants. Of the 2 million hip and knee replacements performed worldwide each year, 40,000 become infected. More than a third of these infections lead to amputation, and not with very successful results: Most of those people die. “Implant operations have a 98 percent success rate, so people don’t want to talk about the infections,” Costerton says. “They’re a bit of a disgrace, really.”

Biofilm infections are not limited to implants. They can be found in the bodies of the young and the healthy. Many children suffer from undiagnosed biofilm infections in their ears, which require months of oral antibiotic therapy while the underlying infection smolders untouched. Millions of others live with chronic biofilms: urinary tract infections in women that last for years; prostatitis that no antibiotics permanently cure; bone infections (osteomyelitis) that cripple and immobilize people for the rest of their lives. Each year roughly 500,000 people in the United States die of biofilm-associated infections, nearly as many as those who die of cancer.

As Marrie’s experience shows, biofilms repel antibiotics, although scientists do not fully understand how. Some drugs cannot fully penetrate the biofilm’s protective matrix. In other cases, even though most of the germs die, enough remain alive to regroup and develop another biofilm. The matrix also keeps its resident germs under cover, hiding the chemical receptors on the bacteria so that drugs cannot latch onto them and kill the germs.

The study of this newly discovered behavior is rooted in the basic and ancient biology of bacteria. Geneticist Bonnie Bassler of Princeton University thinks group-living bacteria may give us a window onto the origins of multicellular life. “Bacteria grow best when each one does its own thing…together,” she says. “Bacteriologists had it wrong for the past 300 years—bacteria don’t live alone.”

As these social bacteria talk to each other, we can now listen in. Bassler and other scientists are learning how to eavesdrop on the chemical language of bacteria, seeking ways to scramble or block those messages. Disrupting the formation of films could be a powerful way to neutralize harmful infections.

Originally trained as a biochemist at Johns Hopkins University, the blue-eyed, athletic Bassler walked into a lecture hall on a whim in the late 1980s to listen to a talk by geneticist Michael Silverman of the Agouron Institute in La Jolla, California. It was one of only a handful of talks that the notoriously reserved Silverman had given in 10 years. Bassler was riveted by what she heard. Silverman talked about how bacteria make light inside the inch-long luminescent squid that live in the shallow waters off the Hawaiian coast.

Infant squid cannot glow until they excrete a mucuslike net to entrap the ubiquitous luminescent bacteria floating in the water. The squid draw captured bacteria into their “light pouches,” where the bacteria are bathed in nutrients —a diet richer than what they can find outside in the sea. In return, the bacteria (Vibrio fischeri, a close relative of the cholera germ) produce a dim blue-green light that is directed downward through small reflective organs in the squid to shine into the water below. When the squid swim at the ocean surface at night, hunting for shrimp, they are invisible to predators below because they look like moonlight on the water. Both squid and bacteria benefit. “The host wants the light, the bacteria get fed,” Bassler says.

The glow of V. fischeri provides an instructive glimpse into the communal behavior of bacteria. Autoinducers (chemical signaling molecules that produce more of themselves inside the cell) control the switch that turns the light genes off and on. Each bacterium secretes a bit of this light-evoking substance into the environment. When a crowd of bacteria and their autoinducers become dense enough, the lights in all the bacteria switch on at once. “This counting of heads is called quorum sensing,” Bassler explains. More broadly, this is how bacteria coordinate their actions in large groups: When the local concentration of autoinducers gets high enough, the bacteria know a crowd is present, and they flip over from solitary mode to group behavior.

The autoinducer molecule that triggers bacterial glow is made by a protein called LuxI, which has a very focused effect. “The molecule that the LuxI protein makes is acylated homoserine lactone, or AHL,” Bassler says. “Each LuxI protein and the molecule it produces is species-specific. There are two kinds of bacteria, and each talks in a different language. Gram-negative bacteria [which have a thin cell wall surrounded by an outer membrane] use the AHLs as autoinducers, while gram-positives [which have a thick cell wall] use peptides. This is a very ancient split.” When the V. fischeri make enough AHL autoinducer—called AI-1 for short—the cells wink on. But that is far from the only autoinducer.

Working with a related bacterium, Vibrio harveyi, in the early 1990s, Bassler discovered another kind of chemical signal that a wide range of bacteria emit. In many species this chemical, called autoinducer 2 (AI-2) has properties of a waste product, says molecular biologist Stephen Winans of Cornell University. AI-2 is the by-product of a complex process of metabolism in these species. Not all bacteria create AI-2, however. According to Winans, eons ago one line of early bacteria began to break down waste products along a pathway leading to the excretion of AI-2; another line did not. The latter are the bacteria that eventually gave rise to eukaryotic organisms, including humans. “That’s why you don’t excrete ?AI-2,” Winans says.

But Bassler found that AI-2 is much more than a waste product. “This little leftover molecule,” she says, got pressed into service as another bacterial language, one that can carry messages between different kinds of germs. Most forms of quorum sensing, including V. fischeri’s luminescence circuit, act as a private language—that is, each germ speaks only to others of its own kind. But AI-2 is a kind of bacterial Esperanto, Bassler determined. After she and her team purified the small AI-2 molecule and its protein receptor, they were able to show that the two form a lock-and-key structure, the telltale sign of a chemical signaling mechanism.

The big question was, what are different germs saying when they talk to each other? Bassler says that in some instances—such as in dental biofilms, in which some 600 species may be growing at a time—AI-2 is necessary for collective or cooperative behavior. First, though, the bacteria must be right next to each other to receive the signal, especially in a dynamic system like the mouth, where saliva is constantly washing across the teeth. The earliest colonists on freshly cleaned teeth, the streptococci, produce only low levels of AI-2; the fusobacteria produce moderate levels. The appallingly destructive germs love a very high level of AI-2, which sends them into overdrive. “They grow like gangbusters,” says Paul Kolenbrander of the National Institute of Dental and Craniofacial Research of the National Institutes of Health.

Quorum-sensing molecules also play an important part in bacterial virulence, or deadliness. If a lethal germ released toxic chemicals immediately after entering the host’s body, the immune system would quickly sense the toxin and go after the invader. So it pays for bacteria to wait, stealthily multiplying until the unwitting host is full of them. Then they can release their toxins all at once, overwhelming immunity and sickening or killing the host.

In their more recent work, Bassler and her colleagues are searching for ways to scramble the quorum-sensing signals of cholera germs. The researchers have demonstrated that in test tubes a particular chemical, called CAI-1, can induce deadly cholera cells to turn off their virulence genes.

Building on our understanding of how germs communicate, Naomi Balaban, a molecular biologist at Tufts University, has spent 17 years studying Staphylococcus aureus, a strain of bacterium that is the main cause of hospital-acquired infections.

Antibiotic-resistant forms of S. aureus, known collectively as methicillin-resistant Staphylococcus aureus, or MRSA, have spread widely in hospitals throughout the world, forming long chains of infection. There are 19,000 MRSA-associated deaths in the United States alone each year.

Other forms of MRSA have begun to spread outside of hospitals; one strain, known as USA300, is especially deadly. It has infected and killed children and athletes, and no one knows where it came from or exactly how it spreads, though athletic locker rooms have been implicated in some cases. Like other forms of staph, USA300 can form invisible biofilms outside the body, making it almost impossible to eradicate. It is difficult to judge the actual prevalence of MRSA, since many staph infections do not get much more serious than a small pimple.

Some cases do progress, though, and they may cause debilitating and almost untreatable soft-tissue infections like cellulitis and folliculitis, pneumonia, and often-fatal heart infections, or endocarditis. Another form of staph, Staphylococcus epidermidis, grows commonly in sheets of invisible biofilm on our skin, where it is normally benign. But if it is introduced into the body during a medical procedure—especially if a joint implant, catheter, or pacemaker is contaminated during insertion—both S. epidermidis and S. aureus can form dangerous biofilms that often cannot be treated without removal of the infected implant.

Balaban has discovered that all forms of staph, whether in a free-floating state or in a biofilm, have a complex form of chemical communication that can activate the agr (accessory gene regulator) system, producing a number of toxins. Somewhat controversially, Balaban also claims to have discovered another system that controls the agr system. The second system involves two proteins known as RNAIII activating protein (RAP) and TRAP, which Balaban calls “the most beautiful protein in the world.” TRAP is RAP’s target protein, Balaban says. It is found both on and within the staph cell. S. aureus secretes RAP into the environment, where the chemical collects and binds to the TRAP molecules on the cells.

When enough RAP molecules adhere to enough target molecules, staph bacteria switch on their cell-to-cell communication and stress-response systems and begin producing the toxin that makes them so lethal. S. aureus bacteria, depending on their strain, can produce 40 or more different toxins. The toxins break down the cells in the host—which could very well be you—in order to release nutrients to the germs. That is why staph infections can be so destructive. When there are enough staph germs present, the host’s immune system is overwhelmed, and tissues are destroyed at a frightening rate, leading sometimes to shock and death.

Balaban reasoned that if she could find a way to block RAP from reaching its target molecule, she could break down the signaling system that allows the release of staph’s devastating toxins. She discovered a chemical she calls RIP (RNAIII inhibiting peptide), which blocks RAP from linking to its target. It is as if an outfielder were standing ready to catch a fly ball heading his way, but he already has a grapefruit in his mitt, preventing the ball from going in. If RAP does not reach its target molecule, the whole communication process breaks down, toxins do not get made, and human immune cells converge on the now-helpless staph germs, ready to mop them up. Balaban claims that RIP can have this effect on free-floating and biofilm-embedded staph alike.

Bacteriologists had it wrong for the past 300 years—bacteria don’t live ?alone. They grow best when each one does its own thing...together.

Some researchers remain unpersuaded by Balaban’s work, however. Richard Novick of the NYU Langone Medical Center, a well-respected staph expert who was also Balaban’s postdoctoral adviser, insists that the TRAP protein does not have any known role in staph biology. In a series of letters to the journal The Scientist, he argues that only one quorum-sensing system has been discovered in Staphylococcus: the agr system. Neither Novick nor any other scientist has been able to reproduce Balaban’s RAP/TRAP experiments in the laboratory. Novick does acknowledge, though, that RIP works. “I don’t question that it has activity.

But whatever it’s doing, it’s not inhibiting agr,” he says. “I would guess it could work by interfering with assembly of a biofilm. It should not have any effect on planktonic Staphylococcus. If it did, I would have to revise my view.”

Despite these questions, RIP—which Balaban discovered in Novick’s laboratory—is in the first stages of preclinical testing as a new kind of antibiotic. It costs millions of dollars to develop drugs and get them tested in animals before they can ever be used in clinical trials for safety and efficacy in humans. Fortunately, Balaban has found a naturally occurring chemical equivalent to RIP: hamamelitannin, an extract of witch hazel bark. She has shown that this old-fashioned household remedy, long used by Native Americans, also serves to knock the ball from the outfielder’s mitt. In her tests, hamamelitannin has the same chemical effect as synthetically produced RIP.