Exploring Non-Visual Cues-Driven Spatial Learning in Drosophila

Abstract

The ability to navigate to the memorized goal locations such as food sources and nests is critical for animals to survive in nature. It is well documented that animals can learn to use visual cues as landmarks to infer the goal location. However, whether animals can use non-visual cues for spatial learning and the underlying neural circuit and molecular mechanisms that support non-visual cues-driven spatial learning (NVSL) are not fully understood. In addition to visual cues, animals receive various non-visual inputs during navigation, such as self-motion, olfactory, auditory, and texture cues. Here, using the high-throughput spatial learning task our lab has developed, I show that Drosophila melanogaster can use both olfactory cues and self-movements (i.e., egocentric cues) to learn a spatial goal. In addition, I characterize the underlying circuit mechanisms and begin to elucidate some of its sex-specific differences and potential molecular basis. Specifically, first, when the environment is featureless, flies can deposit odors near the goal location and associate them with the reward at the goal location as well as using egocentric cues to potentially gauge how far – and in which direction - they have moved from the goal location. Such learning is enabled by mushroom bodies (MBs), an olfactory learning center known to associate odors with reinforcers, as well as neurons that signal egocentric translational velocity (i.e., PFN neurons), respectively. Second, when the environment is enriched with salient non-visual landmarks, flies can significantly improve their learning performance. Interestingly, while such improvement still requires the olfactory learning circuit, the reliance on self-motion sensitive neurons decreases. Third, sexual dimorphism and natural variance exist in both the performance level and the reliance on the different behavioral strategies for NVSL. Lastly, transcriptome analysis of flies trained to perform NVSL identified multiple components of the Toll and Imd signaling pathways whose expression in the brain correlates with high NVSL performance. My research is among the pioneering efforts to demonstrate that Drosophila is capable of combining self-generated and external olfactory markers with egocentric signals to master a spatial goal location, showcasing the adaptability and possible molecular basis of their spatial learning capabilities.