Skip to contentSkip to filters

We use cookies for analytics.

Does the fly’s brain center for vector navigation know that the world is 3D?

Neuroscience ePoster by Angel Stanoev, Hannah Haberkern, Brad Hulse, Sandro Romani and Vivek Jayaraman

Part of COSYNE 2023 posters, HHMI Janelia Research Campus; Janelia Research Campus, Howard Hughes Medical Institute

Saturday

COSYNE 2023
Montreal, Canada

Abstract

Across species, many navigational spatial representations, including those of head direction and traveling direction, require the integration of changes in sensory information induced by self-motion. In Drosophila, connectomic analysis points to a variety of so-called LNO neurons as key inputs for this information to the central complex (CX), a navigational center of insect brains. Recent studies have reported that the responses of several of these neurons correlate with self-motion and with optic flow resulting from rotations and translations in 2D. However, flying animals, like bats and insects, navigate in 3D. We used the fly’s connectome and genetic tools to identify, target and systematically characterize the optic-flow responses of the near-complete set of LNO neurons using two-photon calcium imaging in tethered and fully-restrained flies. We presented different wide-field optic flow visual stimuli on a cylindrical LED arena, thereby deriving the fly’s encoding of visual motion inputs in isolation. The flies were virtually embedded in a featureless flow field corresponding to rotational or translational motion around a given axis, sampled systematically from the set of 3D vectors on a unit sphere around the fly’s position. We found that all LNO neuron types exhibited consistent tuning profiles centered around a preferred axis and direction to rotational and translational motion. We verified that our results for previously characterized LNO types were in agreement with recent studies. Further, our analysis of the complete set of LNO types showed that appropriately weighted linear combinations of their responses by putative downstream neurons could produce responses tuned to rotations around any 3D vector. These results provide concrete predictions about the possible responses of downstream neurons to 3D optic flow that can be tested with further connectomic analysis and functional measurements and paves the way for the search for 3D navigational representations in the fly CX.

Topics

Details

Session
Poster Session II