Abstract
Drosophila melanogaster is a pest of stored fruits as well as the model of insect olfaction. Despite recent breakthroughs unravelling its olfactory signalling cascade, it is unknown what makes natural blends (‘banana’) attractive, i.e., what olfactory code underpins ‘attraction’. Yet, the fly affords a unique ‘backdoor approach’ on studying this very aspect of olfaction. I propose to generate a computer model of the fly based on published sets of all olfactory receptor neuron types (ORNs) and the synthetic odors to which they respond. Based on the ORNs’ responses to ‘banana’ as input, the computer model will compute synthetic blends that generate patterns similar to ‘banana,’ which can be tested behaviorally. Interestingly, the blend odors need not at all be similar to the ‘banana’ odor, but do mimic ‘banana’ in the fly brain. To increase the model’s power, I will physiologically characterize ORNs with additional odor sets, and perform a sensitivity analysis of ‘attraction’ through scoring the effect of genetic exclusion of selected ORNs on the fly’s behavior. This entirely novel approach bypasses the tedious and often ineffective route of research on attractants, and will provide an extremely valuable tool for attractant research on other pest insects, such as mosquitoes or moths. In addition, the project will cross-fertilize basic research by providing a wealth of information about the fly’s perception of attractants or repellents, an unresolved piece of the basic research
Sweden