Discerning the complexity of community interactions using a Drosophila model of polymicrobial infections

Abstract

A number of human infections are characterized by the presence of more than one bacterial species and are defined as polymicrobial diseases. Methods for the analysis of the complex biological interactions in mixed infections with a large number of microorganisms are limited and do not effectively determine the contribution of each bacterial species to the pathogenesis of the polymicrobial community. We have developed a novel Drosophila melanogaster infection model to study microbe-microbe interactions and polymicrobe-host interactions. Using this infection model, we examined the interaction of 40 oropharyngeal isolates with Pseudomonas aeruginosa. We observe three classes of microorganisms, one of which acts synergistically with the principal pathogen, while being avirulent or even beneficial on its own. This synergy involves microbe-microbe interactions that result in the modulation of P. aeruginosa virulence factor gene expression within infected Drosophila. The host innate immune response to these natural-route polymicrobial infections is complex and characterized by additive, suppressive, and synergistic transcriptional activation of antimicrobial peptide genes. The polymicrobial infection model was used to differentiate the bacterial flora in cystic fibrosis (CF) sputum, revealing that a large proportion of the organisms in CF airways has the ability to influence the outcome of an infection when in combination with the principal CF pathogen P. aeruginosa.

Figure 1. Characterization of the P. aeruginosa Drosophila infection.

(A) The CFU/fly during the first four days of feeding on PA01. (B) Survival curves of Drosophila chronically infected with PA01: open boxes, flies fed 5% sucrose; open circles, flies continuously exposed to PA01; open triangles, flies exposed to PA01 for 24 hours and then transferred to sterile food.

Figure 2. Fluorescence microscopy and direct measurement of P. aeruginosa infected Drosophila.

(A) Red auto-fluorescence from an uninfected Drosophila crop (10×). (B) mCherry fluorescence from P. aeruginosa in a Drosophila infected crop (10×). (C,E) Yellow auto-fluorescence from an uninfected crop, 10× and 40× respectively. (D,F) Yellow auto-fluorescence from an infected crop, 10× and 40× respectively. Crops were harvested from twenty infected flies 24 hours post-infection and bacterial load was measured. The PAO1 in the crops and in whole flies were determined by plating (G). The entire gut (the foregut, the midgut and the hindgut), the crop, and remaining fly body were removed from twenty flies infected with a PA01 strain expressing luciferase from the lasI promoter (24 hours post-infection) and gene expression measured for each individually (H).

Figure 3. Cluster representation of the three OF infection classes.

The survival curve of each OF infection alone was compared to the survival curve of flies feeding on 5% sucrose. The survival curves of the OF PA01 co-infections were compared to the survival curve of PA01 alone. Green boxes indicate time points where fewer flies were alive as compared to controls; red indicates time points where more flies were alive as compared to controls. The data was collected from six independent infections with a minimum of 25 flies per infection. The * indicate infections that were further characterized.

Figure 4. Kaplan-Meier survival curves and quantitative bacteriology of two representative OF strains for each infection class.

Open triangles, survival curve of flies feeding on 5% sucrose; filled triangles, survival of flies infected with PA01; filled circles, survival of flies infected with the OF strain alone; open circles, survival of flies co-infected with PA01 and the OF strain. The black bar shows the OF CFU/fly 24 post-infection; black bars with white hatches, OF CFU/fly 48 post-infection; open white bars, PA01 CFU/fly 24 hours post-infection; white bars with black hatches, PA01 CFU/fly 48 hours post-infection. The C102 bacterial load was unable to determined in monoinfections due to PA01 overgrowth. (A–D) are Class I organisms; (E–H) are Class II organism; (I–L) are Class III organisms. Log-rank analysis (Mantel-Cox) was used to compare OF infections with the sucrose control and mixed infections with the PA01 control; statistical significance between survival curves is shown with * P<0.05, ** P<0.005 and *** P<0.0005; ns = not significant.

Figure 5. In vivo modulation of P. aeruginosa virulence factor gene expression by OF.

(A) The level of gene expression of 24 P. aeruginosa virulence factors in Drosophila prior to death during infection with P. aeruginosa alone (circles) or co-infection with C90 (squares) or C87 (diamonds). The average CPS/Fly before death was calculated from the average CPS/Fly during the last two hours of life from eight flies per condition. Red and blue symbols indicate promoters activated or repressed in the presence of the OF respectively. (B) Temporal expression of lasI and lasB in Drosophila infected with P. aeruginosa alone (black circles) and in the presence of C90 (open squares) or C87 (open diamonds). The * indicate those profiles considered to be statistically significant (Student's t test, p<0.05) compared to infections with P. aeruginosa alone.

Figure 6. Drosophila AMP gene expression 24 hours post-infection.

Cecropins have a broad spectrum of activity against Gram-positive, Gram-negative, protozoan parasites, and fungi and are induced in the intestine during gut infections . Real-time PCR was used to calculate the fold transcriptional activation above uninfected flies: (A) ditericin expression; (B) cecropin A1 expression; (C) drosomycin expression. Solid black circles indicate the response to PA01 alone; blue circles, examples of suppressed AMP expression during co-infection; red circles, AMP expression that can be explained by an additive effect; green circles, synergistic activation of AMP expression.