Back

Multisensory pathogen detection drives rapid escape and shapes host-microbe interactions in Drosophila larvae

ZUGASTI, O.; Royet, J.

2025-12-19 animal behavior and cognition
10.64898/2025.12.18.695079 bioRxiv
Show abstract

Larvae of many insects develop immersed in decomposing substrates densely populated with microbes, yet how they detect and respond to pathogenic threats during early life remains poorly understood. Here, we show that Drosophila melanogaster larvae exhibit a previously unrecognized rapid escape behavior triggered by food contaminated with metabolically active Erwinia carotovora carotovora 15 (E. cc15), a natural bacterial pathogen of flies and plants. Stationary-phase cells fail to elicit avoidance, demonstrating that larval detection of E. cc15 depends on bacterial metabolic activity rather than on its mere presence. Using targeted genetic manipulations, we identify two chemosensory pathways required for this response: a gustatory input mediated by the aversion receptor Gr33a and an olfactory input mediated by the Or49a-Orco complex. Disrupting either pathway abolishes escape, revealing that larvae integrate complementary contact-dependent and volatile cues to evaluate microbial danger. Functionally, escape enables larvae to reach uncontaminated food and partially mitigates the developmental impact of early pathogen exposure. However, dispersing larvae also transfer viable bacteria to new substrates, indicating that this defensive response can simultaneously promote pathogen spread. Together, these findings establish the first example of a rapid, multisensory escape behavior induced by a natural pathogen in Drosophila larvae and provide a tractable model for dissecting how microbial cues guide defensive decision-making and influence pathogen dissemination.

Published in PLOS Pathogens (predicted rank #17) · training set

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.