There were three major components to the study. The first established that isolates of two antibiotic-resistant ocular pathogens, Pseudomonas aeruginosa and Serratia marcescens, were all susceptible to being attacked and killed by at least one of two other bacteria , Micavibrio aeruginosavorus and Bdellovibrio baceriovorus, which act as predators against the pathogens but are believed to be "good," or non-infectious, bacteria when they exist within the human body.

In the second phase, human corneal-limbic epithelial cells that are native to the eye were exposed in vitro to M. aeruginosavorus and B. baceriovorus to test whether those "good" predator bacteria would cause either toxicity or inflammation in those cells. They did not.

In the third phase, the two "good" predator bacteria were injected into live worms from the species Galleria mellonella, which is well established as a suitable model to test the toxicity of various microbes as well as a live organism's innate immunity to those microbes. Where injection the pathogenic bacterium P. aeruginosa as a positive control was one hundred percent fatal to the worms, other worms injected with the two "good" predator bacteria had 11-day survival rates between 93.3 and 100 percent, a strong sign that the "good" bacteria were not toxic to the worms. In addition a lack of change in larval pigmentation following injection suggested that the "good" bacteria also did not provoke an aggressive innate immune response in the worms.

"Taken together, our findings leave us confident that, in isolation, pathogenic bacteria are susceptible to successful attack by predator bacteria, predator bacteria do not appear inherently harmful to ocular cells when applied topically, and a live organism can tolerate the predator bacteria well," says Kadouri. "The time to test all three phenomena simultaneously in the eye tissue of a live organism may now be at hand."

The current study builds on another recent paper published in PLoS ONE, which also described research led by Kadouri. That study used the predatory bacteria Bdellovibrio baceriovorus 109J, B. bacteriovorus HD100, and Micavibrio aeruginosavorus strain ARL-13, in targeting 14 strains of dangerous bacteria that are known to be multidrug resistant (MDR). Species targeted in that earlier research included Acinetobacter baumannii, Escherichia coli, Klebsiella pneumoniae, and Pseudomas spp., all of which are commonly encountered in health care settings.

After predator bacteria and MDR strains were co-cultured in the laboratory, the researchers found that cell viability had diminished to varying degrees in all 14 MDR strains, suggesting that while the MDR strains are strongly resistant to current antibiotics, they may have no innate defenses that would protect them against one or more of the predator bacteria.

This research presents hope that a wide variety of MDR bacteria may be susceptible to attack by predatory bacteria in settings where antibiotic therapy has increasingly failed. It remains unknown whether predatory bacteria can have the same effect on MDR bacteria when used systemically within the human body that they do in the lab, given the immune system's potential ability to neutralize the predators before they can do their beneficial work. However Kadouri's promising research in the lab appears to strengthen the case for further study.