The intensive care unit, cloistered and cosseted though it may seem, is among the riskiest spots in the hospital. The hardiest of bugs take up residence there since only they can withstand the daily onslaught of disinfectants that scrub the ICU walls and floors. Moreover, the damp tangle of tubing and catheters that enmesh patients provide a breeding ground for a range of resistant strains. Indeed, the main thing stopping these invaders from flourishing may be each other.
That microbes attack one another is well known—penicillin is an ingenious weapon released by a species of fungus to kill nearby bacteria. It now appears that a highly resistant strain of bacteria, in an effort to gain the upper hand, turns the tables in a novel way—by preventing its fungal competitor from achieving virulence. Anton Peleg, Eleftherios Mylonakis, and colleagues observed this cross-kingdom battle playing out in a living worm host between Acinetobacter baumannii, a rising and resistant species of bacteria, and the infamous fungus Candida albicans.
“Their ecological niche within the hospital is very intertwined,” said Peleg, an HMS research fellow in medicine at Massachusetts General Hospital. “They infect similar patients in a similar setting.”
What makes the interaction so attention-getting is its timing: A. baumannii targets C. albicans at its most noxious stage, when it is putting out invasive filaments and developing into a biofilm. C. albicans launches a counterattack, but only in its most mature biofilm form.
Peleg, Mylonakis, an HMS assistant professor of medicine at MGH, and colleagues also found that when the dust settled the worm hosts had benefited from the conflict. Those infected with both A. baumannii and C. albicans lived longer than worms infected with C. albicans alone, though eventually even the doubly-infected worms died. The findings appeared online Sept. 15 in Proceedings of the National Academy of Sciences.
Until now, most researchers have taken a monomicrobial approach, focusing on the struggle between host and individual pathogen. But the observation that microbes may enter the body in dueling pairs—and possible triads and even more complex arrangements—opens the door to richer scenario-making and possibly to more inventive therapeutic approaches. One might exploit their antagonism, turning the microbial antagonists into human allies. “If we can understand the molecular mechanisms of how A. baumannii kills C. albicans, maybe we could exploit those pathways to develop novel antimicrobials,” Peleg said.
Creative thinking is much needed. Drug-resistant strains of all manner of microbe are on the rise, not just in ICUs but also in burn units. “We have patients on the ward where sometimes we are now running out of effective antibiotics or antifungals and other antimicrobials,” said Peleg. The issue has become especially acute in field hospitals housing newly injured soldiers from Afghanistan and Iraq, where broad-spectrum antibiotics are required.
It was while working in Australia as an infectious disease doctor that Peleg first encountered A. baumannii. “The hospitals had major problems with Acinetobacter baumannii—it was very resistant to drugs. And very little was known about it,” he said. Sensing an open niche, he set out to discover how the bacterium launches its infection, but he needed to find a good animal model. He heard that Mylonakis, who has a long-standing interest in C. albicans, had created an elegant model using the soil-dwelling Caenorhabditis elegans.
A scientific cross-kingdom partnership was cemented when the pair met. “I always go with the person first and the project second,” said Mylonakis.
Until then, Mylonakis and his colleagues had mostly been occupied with understanding how C. albicans launches its infection in the worm. They observed the yeasts making their way into the gut and then exiting through tiny tears. Once in the bloodstream, the bugs underwent a metamorphosis—putting out tiny filaments, which they then used to bore through organs. Often, the microbes teamed up, weaving their processes together into a fine mesh, or biofilm, which in many cases killed the worm.
Through conversation with Peleg, a new question arose: what would happen if A. baumannii was added to this model system? The researchers infected the worms with C. albicans, put them into a liquid medium, and then added A. baumannii. They monitored the worms for several days. Compared to worms infected only with Candida, the twice-infected ones appeared to be much better off: fewer died. And the candida seemed to be having trouble putting out filaments (center image below). “It was a green light to run forward,” said Peleg.
A. baumannii was attacking filamentous C. albicans, but it was not clear how. The researchers suspected that either it was targeting the filaments directly, cell-to-cell, or it was secreting some kind of factor. To test the first hypothesis, they cultured A. baumannii with two strains of C. albicans—one that forms filaments constitutively and a mutant incapable of forming filaments altogether. The results were striking. A. baumannii appeared to flock to the filamentous form while avoiding the filament-free mutant. Under the microscope, the researchers could even see the bacteria adhering to filaments.
But that was not all the microbes were doing. Peleg, Mylonakis, and colleagues tested the second hypothesis, that the bacteria were secreting a toxic substance, by growing the bacteria in liquid media for varying amounts of time, removing the bacteria, and exposing C. albicans to the liquid. The liquid alone was enough to inhibit the formation of filaments. And the longer the bacteria spent in the liquid, the greater its filament-inhibiting powers.
To keep themselves from overgrowing, microbes often secrete special quorum-sensing molecules. Peleg and Mylonakis suspect that the bacteria might be secreting a quorum-sensing signal that also acts to suppress filament growth, though they have not yet found one.
Intriguingly, it appears Candida, in its own defense, might be using this tactic. Peleg, Mylonakis, and colleagues grew Candida biofilms in liquid for varying lengths of time, siphoned off the liquid, and mixed it with A. baumannii. Bacterial growth was inhibited, but only when exposed to the liquid of mature 24-hour biofilms.
Again the researchers suspected a quorum-sensing molecule and, in this case, they were able to home in on one, the molecule farnesol—ironically, a substance used in the well-known perfume Chanel No. 5. “The same thing that is used for candida to attract and communicate with one another is used for humans to attract and communicate,” Mylonakis said.
Might humans use the substance to subdue the growing A. baumannii menace? With their newly developed polymicrobial worm model, Peleg and Mylonakis plan to explore this and other questions. What excites them is the possibility of developing drugs that disarm resistant microbes rather than kill them outright.
“Until now, we have concentrated on finding antimicrobials that kill bacteria or fungi,” said Peleg, “but we must expand our ideas and think about antimicrobials that alter the virulence potential of the pathogen.”
Conflict Disclosure: The authors declare no conflicts of interest.
Funding Sources: The National Institutes of Health, the Ellison Medical Foundation, and the University of Queensland
