Lethality and developmental delay in Drosophila melanogaster larvae after ingestion of selected Pseudomonas fluorescens strains

Abstract

Background: The fruit fly, Drosophila melanogaster, is a well-established model organism for probing the molecular and cellular basis of physiological and immune system responses of adults or late stage larvae to bacterial challenge. However, very little is known about the consequences of bacterial infections that occur in earlier stages of development. We have infected mid-second instar larvae with strains of Pseudomonas fluorescens to determine how infection alters the ability of larvae to survive and complete development.

Methodology/principal findings: We mimicked natural routes of infection using a non-invasive feeding procedure to study the toxicity of the three sequenced P. fluorescens strains (Pf0-1, SBW25, and Pf-5) to Drosophila melanogaster. Larvae fed with the three strains of P. fluorescens showed distinct differences in developmental trajectory and survival. Treatment with SBW25 caused a subset of insects to die concomitant with a systemic melanization reaction at larval, pupal or adult stages. Larvae fed with Pf-5 died in a dose-dependent manner with adult survivors showing eye and wing morphological defects. In addition, larvae in the Pf-5 treatment groups showed a dose-dependent delay in the onset of metamorphosis relative to control-, Pf0-1-, and SBW25-treated larvae. A functional gacA gene is required for the toxic properties of wild-type Pf-5 bacteria.

Conclusions/significance: These experiments are the first to demonstrate that ingestion of P. fluorescens bacteria by D. melanogaster larvae causes both lethal and non-lethal phenotypes, including delay in the onset of metamorphosis and morphological defects in surviving adult flies, which can be decoupled.

Figure 1. Timeline of the experimental protocol.

Above the daily timeline are cartoons of the developmental stage that most closely replicates the control-treatment group trajectory (see Material and Methods for details). On Day −2, eggs were transferred from the egglaying plates to non-nutritive agar plates. The split image on Day 4 indicates the transition period between wandering larvae and prepupal stages around the onset of metamorphosis. Treatments distributed on the agar surface on Day 0 (open arrow) were composed of a yeast suspension, which served as a food source for the larvae, alone (control) or amended with bacterial inoculum. An additional yeast supplement (food) was added on Day 2 to all plates (solid arrow) and every 2 days as long as feeding larvae were visible (dashed arrow). Using this method, larvae were never directly handled and the only disturbance was the application of treatments or food to the surface of the plate. The plates were housed in an incubator on a 12L:12D cycle initially at 25°C; on Day 2, the temperature was shifted to 22°C to slow down larval development.

Figure 2. Developmental time course after bacterial inoculation.

The percentage of larvae that pupariated, counted as prepupae and/or pupae (solid lines), or emerged as adults (dashed lines) were determined for two different experiments (A and B) in which second instar larvae were fed with a yeast suspension having no bacteria (black) or amended with bacterial strains Pf0-1 (blue), SBW25 (green, Fig. 2A), Pf-5 (red), or a Pf-5 gacA mutant (open red circle, Fig. 2B). On the timeline in this and Figure 3, the egg-laying period is denoted by the egg cartoon and the treatment by an open arrow at Day 0. Bacterial treatment as cfu/plate: (A) Pf0-1, ; SBW25, ; Pf-5, , or (B) Pf0-1, ; Pf-5 gacA mutant, ; Pf-5 (squares), ; Pf-5 (triangles), .

Figure 3. Developmental time course after inoculation with different doses of Pf-5.

The percentage of larvae that pupariated, counted as prepupae and/or pupae (solid lines), or emerged as adults (dashed lines) after inoculation with Pf-5 (filled red symbols or star), killed Pf-5 (killed cfu/plate, gray circles) or control (black circles ) treatment were determined for each time point. The Pf-5 cell densities as cfu/plate were (circles), (squares), (diamonds), (triangles), (stars) cfu/plate.

Figure 4. Dose-dependent effects of P. fluorescens strains on survival and developmental delay.

For each experimental treatment, the percent pupal or adult survival is normalized to that of the appropriate control, which has been set to 100%, and the relative delay to pupariation is normalized to that of the appropriate control, which has been set at 0. In this figure CS-A larval treatment groups are represented using circles and OR larval treatment groups by diamonds. The size of each data point is scaled to the inoculation dose (highest dose/largest circle is cfu/plate; smallest dose/smallest circle is cfu/plate; highest dose/largest diamond is ; lowest dose/smallest diamond is ); Pf0-1 (blue), SBW25 (green), Pf-5 (red), killed Pf-5 (gray) and a Pf-5 gacA mutant (open red circle). (A) The relative developmental delay (see Materials and Methods) during the larval stage (X-axis) was plotted against the percentage of larvae that survived to pupariate (Y-axis) for six different experiments. (B) The relative developmental delay during the pupal stage (X-axis) was plotted against the percentage of pupae that survived to adult survivors (Y-axis). (C) The relative developmental delay during larval and pupal stages (X-axis) was plotted against the percentage of larvae that survived to adult (Y-axis).

Figure 5. Effects of P. fluorescens strains on the morphology of third instar larvae.

Third instar larvae on the surface of the treatment plates were photographed using a digital camera mounted on a dissecting scope. Bar is 1 mm with 0.1 mm divisions. (A) 96 hour post control treatment. (B) 72 hour post Pf0-1 treatment ( cfu/plate). (C) 72 hour post Pf-5 gacA mutant treatment ( cfu/plate). (D) 48 hour post Pf-5 treatment ( cfu/plate), showing small body size and convoluted longitudinal tracheal trunks. (E) 72 hour post Pf-5 treatment ( cfu/plate), showing small body size and convoluted tracheal trunks. (F) 72 hour post Pf-5 treatment ( cfu/plate), showing loss of fat body opacity and a melanotic nodule. (G) 72 hour post SBW25 treatment (2.9×10⁷ cfu/plate), with no melanization within the body cavity. (H) 72 hour post SBW25 treatment ( cfu/plate) showing melanization within the posterior third of the body. (I) 72 hour post SBW25 treatment ( cfu/plate) showing complete melanization reaction.

Figure 6. Morphological defects in adult survivors after infection with Pf-5.

(A) Eye images from control and bacterial treatment groups. Top panel from left to right: control; cfu/plate, Pf0-1; cfu/plate Pf-5 gacA mutant; cfu/plate killed Pf-5. Bottom panel from left to right: cfu/plate Pf-5; cfu/plate Pf-5; cfu/plate Pf-5; cfu/plate SBW25. Magnification bar = 500 µm. (B) Images of wing blades from adult flies after control (upper left) or Pf-5 at cfu/plate (bottom left, right panel) treatments. Developmental defects visible in these wings include loss of cells at the distal margin (upper right) and/or the presence of ectopic wing veins (bottom left and right). Magnification bar = 100 µm.

Figure 7. Population sizes of P. fluorescens in larvae of D. melanogaster.

Second instar larvae were inoculated with strains Pf0-1 (blue), SBW25 (green), or Pf-5 (red) (A and B), or the Pf-5 gacA mutant (open circle) or wildtype Pf-5 (filled circle) (C and D). Each day following inoculation, larvae were individually surface sterilized, homogenized and bacterial numbers were estimated by culturing. Each point represents the average of five larvae. Bacterial population sizes in the inocula were 10⁴ cfu/plate (A and C), or 10⁷ cfu/plate (B and D). The dotted line denotes the limit of detection. On Day 4, larvae inoculated with Pf0-1 at 10⁷ cfu/plate had pupariated and were not sampled. Error bars denote standard errors.