Effects of Exposure Time and Biological State on Acquisition and Accumulation of Erwinia amylovora by Drosophila melanogaster

Abstract

Fire blight, caused by the bacterium Erwinia amylovora, is a disease devastating the production of rosaceous crops, primarily apple and pear, with worldwide distribution. Fire blight begins in the spring when primary inoculum is produced as ooze, which consists of plant sap, E. amylovora, and exopolysaccharides. Ooze is believed to be transferred to healthy tissues by wind, rain, and insects. However, the mechanisms by which insects locate and transmit ooze are largely undocumented. The goals of this study were to investigate the biological factors affecting acquisition of E. amylovora from ooze by a model dipteran, Drosophila melanogaster, and to determine whether flies are able to mechanically transfer this bacterium after acquisition. We found that the percentage of positive flies increased as exposure time increased, but nutritional state, mating status, and sex did not significantly alter the number of positive individuals. Bacterial abundance was highly variable at all exposure times, suggesting that other biological factors play a role in acquisition. Nutritional state had a significant effect on E. amylovora abundance, and food-deprived flies had higher E. amylovora counts than satiated flies. We also demonstrated that D. melanogaster transmits E. amylovora to a selective medium surface and hypothesize that the same is possible for plant surfaces, where bacteria can persist until an opportunity to colonize the host arises. Collectively, these data suggest a more significant role for flies than previously thought in transmission of fire blight and contribute to a shift in our understanding of the E. amylovora disease cycle.IMPORTANCE A recent hypothesis proposed that dissemination of Erwinia amylovora from ooze by flies to native rosaceous trees was likely key to the life cycle of the bacterium during its evolution. Our study validates an important component of this hypothesis by showing that flies are capable of acquiring and transmitting this bacterium from ooze under various biotic conditions. Understanding how dipterans interact with ooze advances our current knowledge of its epidemiological function and provides strong evidence for an underappreciated role of flies in the disease cycle. These findings may be especially important as they pertain to shoot blight, because this stage of the disease is poorly understood and may involve a significant amount of insect activity. Broadly, this study underscores a need to consider the depth, breadth, and origin of interactions between flies and E. amylovora to better understand its epidemiology.

Keywords: Diptera; fire blight; insect-microbe interactions; plant disease epidemiology.

FIG 1

Marginal effect of exposure time on overall percentage of flies positive for E. amylovora (black line with ±95% confidence interval [CI]) as predicted by a generalized mixed-effects model with exposure as a fixed effect and date tested as a random effect with a binomial distribution. Line shows the predicted probability of infection as exposure time increases based on the model (χ² = 24.08, P < 0.001). Blue dots represent raw overall percentage of positive flies at each exposure time (3, 6, 12, and 24 h).

FIG 2

E. amylovora log₁₀ CFU/insect at 3, 6, 12, and 24 h of exposure. Red dot and error bars represent mean log₁₀ CFU/insect ± 95% CI. Data are presented as discrete here to better visualize the variability at each time point but were analyzed as a continuous variable. There was a negative effect of exposure time on abundance (χ² = 3.92, P = 0.048). Black dots represent raw log-transformed CFU of individual D. melanogaster flies.

FIG 3

E. amylovora log₁₀ CFU/insect for food-deprived and satiated D. melanogaster flies. Red dot and error bars represent mean log₁₀ CFU/insect ± 95% CI. Food-deprived flies had significantly higher bacterial abundances than satiated flies (χ² = 5.36, P = 0.02). Black dots represent log-transformed CFUs of individual D. melanogaster flies. *, P < 0.05.

FIG 4

E. amylovora log₁₀ CFU/insect for mated and unmated D. melanogaster flies. Red dot and error bars represent mean log₁₀ CFU/insect ± 95% CI. Unmated flies had marginally higher bacterial abundances than mated flies (χ² = 3.68, P = 0.055). Black dots represent log-transformed CFUs of individual D. melanogaster flies. Dot above line, P = 0.055.