Predatory Bacteria Attenuate Klebsiella pneumoniae Burden in Rat Lungs

Abstract

Bdellovibrio bacteriovorus and Micavibrio aeruginosavorus are predatory bacteria that naturally-and obligately-prey on other Gram-negative bacteria, and their use has been proposed as a potential new approach to control microbial infection. The ability of predatory bacteria to prey on Gram-negative human pathogens in vitro is well documented; however, the in vivo safety and efficacy of predatory bacteria have yet to be fully assessed. In this study, we examined whether predatory bacteria can reduce bacterial burden in the lungs in an in vivo mammalian system. Initial safety studies were performed by intranasal inoculation of rats with predatory bacteria. No adverse effects or lung pathology were observed in rats exposed to high concentrations of predatory bacteria at up to 10 days postinoculation. Enzyme-linked immunosorbent assay (ELISA) of the immune response revealed a slight increase in inflammatory cytokine levels at 1 h postinoculation that was not sustained by 48 h. Additionally, dissemination experiments showed that predators were efficiently cleared from the host by 10 days postinoculation. To measure the ability of predatory bacteria to reduce microbial burden in vivo, we introduced sublethal concentrations of Klebsiella pneumoniae into the lungs of rats via intranasal inoculation and followed with multiple doses of predatory bacteria over 24 h. Predatory bacteria were able to reduce K. pneumoniae bacterial burden, on average, by more than 3.0 log₁₀ in the lungs of most rats as measured by CFU plating. The work presented here provides further support for the idea of developing predatory bacteria as a novel biocontrol agent.

Importance: A widely held notion is that antibiotics are the greatest medical advance of the last 50 years. However, the rise of multidrug-resistant (MDR) bacterial infections has become a global health crisis over the last decade. As we enter the postantibiotic era, it is crucial that we begin to develop new strategies to combat bacterial infection. Here, we report one such new approach: the use of predatory bacteria (Bdellovibrio bacteriovorus and Micavibrio aeruginosavorus) that naturally-and obligately-prey on other Gram-negative bacteria. To our knowledge, this is the first study that demonstrated the ability of predatory bacteria to attenuate the bacterial burden of a key human pathogen in an in vivo mammalian system. As the prevalence of MDR infections continues to rise each year, our results may represent a shift in how we approach treating microbial infections in the future.

FIG 1

Histological examination of rat lungs after intranasal introduction of predatory bacteria. B. bacteriovorus 109J or HD100 bacteria or M. aeruginosavorus ARL-13 (MICA) bacteria were introduced into the lungs of SD rats via intranasal inoculation. Histological examination of rat lungs revealed no pathological abnormalities compared to rats inoculated with the control (PBS). All images are representative micrographs taken at 24, 48 h, and 10 days postintranasal inoculation and at ×40 total magnification.

FIG 2

Inflammatory protein profile within rat lungs in response to intranasal inoculation of predatory bacteria. ELISA analysis of responses of IL-1β, IL-4, IL-5, IL-8, IL-10, IL-13, CXCL-1/KC, gamma interferon (IFN-γ), and TNF to intranasal inoculation of predatory bacteria relative to those seen with a PBS control was performed. B. bacteriovorus 109J or HD100 bacteria or M. aeruginosavorus ARL-13 bacteria (and also K. pneumoniae [Kp] bacteria as a control) were introduced into the lungs of SD rats via intranasal inoculation. Inflammatory proteins were assessed within the lungs at (A) 1, (B) 24, and (C) 48 h postinoculation. Twelve rats per treatment group (eight for K. pneumoniae) were used at each time point. Data are combined from two independent experiments. Data represent means ± standard errors of the means. Significant differences between treatment groups and respective control were determined using ANOVA (*, P < 0.05; ****, P < 0.0001).

FIG 3

Predatory bacterial dissemination within hosts. qPCR detection of predatory bacteria within the host was performed. The (A) lungs, (B) kidneys, (C) livers, and (D) spleens were probed for B. bacteriovorus 109J and HD100, M. aeruginosavorus (MICA), and K. pneumoniae at 1, 24, and 48 h and 10 days (d) (lung only) postinoculation. Twelve rats per treatment group (eight for K. pneumoniae) were analyzed at each time point for the lungs, with the exception of the M. aeruginosavorus 1-h treatment group, which had six rats. Only six rats per treatment group were analyzed for the kidney, liver, and spleen. Each data point represents a single rat’s respective bacterial load. A line represents the mean of the results from each treatment set. Data are combined from the results of two independent experiments.

FIG 4

Histological examination of rat lungs after treatment of K. pneumoniae inoculation with predatory bacteria. K. pneumoniae (or PBS for control groups) was initially introduced into the lungs of rats via intranasal inoculation. Animals were then treated with PBS, B. bacteriovorus 109J, or M. aeruginosavorus (MICA) at 30 min and 6, 12, and 18 h postinoculation. All images are representative micrographs taken at 24 h postinoculation and at ×40 total magnification.

FIG 5

K. pneumoniae bacterial burden within lungs of rats after treatment with predatory bacteria. K. pneumoniae (or PBS for control groups) was initially introduced into the lungs of rats via intranasal inoculation. Animals were then treated with PBS, B. bacteriovorus 109J, or M. aeruginosavorus (MICA) at 30 min and 6, 12, and 18 h postinoculation. At 24 h, lungs were harvested, homogenized, and plated on MacConkey agar plates to recover K. pneumoniae CFU. Twelve rats per treatment group were used at each time point. Each data point represents a single rat’s respective bacterial load. Horizontal lines represent the median of the results from each treatment set. Data are combined from the results from two independent experiments. Analyses of significant differences between treatment groups and respective controls were performed using the Mann-Whitney test (*, P < 0.05).