UC Berkeley

Learning to distinguish ‘good’ from ‘bad’ is essential to an animal’s survival. Taste compounds have intrinsic value to animal survival and serve as innate rewards and punishments. From humans to flies, nutritious substances like sugar are intrinsically positive and promote acceptance behavior, whereas potentially toxic bitter substances like quinine, for example, are met with disgust and aversion. Yet, even the strongest instinctual behavioral drives such as feeding can be modified by prior associations. How does this learning happen? Using the model organism Drosophila Melanogaster, I investigated how the inputs and outputs of a learning center can modulate feeding behavior.

In the first part of this dissertation, I investigated the modulation of proboscis extension by inputs and outputs of the learning center (mushroom body, MB) in the fly brain. I identified 10 split-Gal4 lines that cover 7 different cell types of MB output neurons (MBONs) that decrease the probability of proboscis extension when activated. Silencing these neurons had modest effects on proboscis extension. Additionally, I found 3 split-Gal4 lines labeling dopaminergic neurons (DANs) that decrease proboscis extension to sucrose, upon activation. I also found that the MBONs which suppress proboscis extension upon activation do not respond to sucrose. Lastly, I describe areas for putative neurons that are downstream of MBONs.

In the second part of this dissertation, I outline the methods I developed for a large-scale calcium imaging experiment to track neuronal changes during an aversive taste conditioning paradigm. While this set of experiments was ultimately inconclusive because we lacked a good positive control and behavioral readout to determine successful learning in our in vivo prep, we believe that outlining the experimental methods and analysis pipeline may be useful and applicable for more general analysis of large-scale calcium imaging endeavors to answer questions about neuronal changes during a behavior.

Overall, my thesis research investigated how the activity of MB inputs and outputs impinges on feeding behavior by characterizing mushroom body neurons that antagonize feeding behavior and examining how they alter activity in feeding circuits. The work described in this thesis provides insight into how the mushroom body flexibly alters the response to taste compounds and modifies feeding decisions.