Mapping neural correlates of visuo-olfactory integration using light field microscopy
1 Universitaet Klinik Bonn, Physiology II, Bonn, Germany
Varying modes of locomotion, be it walking, running, or flying, exert different temporal demands on sensory processing across animal species. The human eye must resolve texts on a sign much more quickly on a moving train than when the train is stopped at the station. A locomoting animal must be able to process sensory stimuli over a dynamic temporal range to effectively interpret its environment. Experimental studies have shown that locomotion increases the gain of sensory processing neurons across vertebrates and invertebrates, and the extent of this modulation scales with the animal’s locomotion mode. However, the complete neural mechanisms underlying these sensorimotor transformations are poorly understood. Whereas separate studies using electrophysiology and in vivo two-photon calcium imaging have described how specific sensory neurons are modulated by locomotion, these findings provide only slivers of the remaining brain wide activity, akin to finely painting the details of several pieces without knowing how they fit with the larger puzzle. To obtain a clearer view of this “big picture”, we employ light field microscopy to image brain wide activity in actively behaving Drosophila. Briefly, in this technique, a microlens array is fitted to a conventional wide-field fluorescence microscope to capture light fields emitted by a biological sample and generate a 3D render of the entire specimen volume, achieving temporal resolutions up to 100Hz and allowing in vivo imaging at rates on par with neural kinetics. Previous work from our lab identified several neural correlates of walk behavior independent of whether the fly actively decided to or were forced to walk, suggesting encoding of motor and proprioceptive information. Building upon this work, here we study how visual and olfactory processing in the fruit fly are modulated by walk behavior. We image pan-neuronal calcium activity in head-fixed, walking or quiescent Drosophila melanogaster while the fly experiences separate or simultaneous presentations of visual wide-field gratings and an appetitive odor. We use principal component analysis and independent component analysis to generate activity maps associated with each of these combinations of locomotor state and sensory stimulus. Furthermore, we align these activity maps to established anatomical templates to assign neuronal activities to anatomical regions. Then, by leveraging connectomics databases and computational techniques, we attempt to extract informational flow within brain regions to propose neural circuits relevant for the locomotor modulation of visuo-olfactory processing.