Four Individually Identified Paired Dopamine Neurons Signal Taste Punishment in Larval Drosophila

Abstract

Dopaminergic neurons (DANs) carry out multiple tasks in the brain, including the transmission of information related to rewards and punishments across various animal species. They are responsible for evaluating sensory input, storing resulting associations as memory, and continuously updating them based on their relevance and reliability. Accurate comprehension of the dopaminergic systems operation necessitates an understanding of the specific functions mediated by individual DANs. To this end, our research employs Drosophila larvae, which possess approximately 12,000 neurons in their brains, of which only around 1% (approximately 120) are DANs. The presynaptic projections to the mushroom body (MB) - a brain region pivotal for associative olfactory learning in insects - are limited to only eight larval dopaminergic neurons. These DANs are further subdivided into two clusters: the primary protocerebral anterior medial cluster (pPAM) comprises four cells, and the dorsolateral 1 cluster (DL1) comprises the remaining four cells. Our findings confirm previous research that demonstrates that the pPAM DANs innervating the MBs medial lobe encode for a gustatory sugar reward signal. Furthermore, we have identified four DANs in the DL1 cluster - DAN-c1, DAN-d1, DAN-f1, and DAN-g1 - each of which innervates distinct compartments of the MB peduncle, lateral appendix, and vertical lobe. Optogenetic activation of DAN-f1 and DAN-g1 alone suffices to substitute for punishment. Furthermore, optogenetic inhibition, calcium imaging results and electron microscopy-based reconstruction of all sensory input circuits to the four DL1 DANs demonstrate that each DAN encodes a different aspect of punishment, with DAN-g1 being of central importance for the salt dependent teaching signal. To summarize, our investigation has revealed the existence of a cellular division of labor among larval DANs concerning the transmission of dopaminergic reward (pPAM cluster) and punishment signals (DL1 cluster). Individual DANs in each cluster encode for distinct but partially overlapping aspects of the teaching signal. The striking resemblance in the organizing principle of larval DANs with that of its adult counterpart and the mammalian basal ganglion suggests that there may be a limited number of efficient neural circuit solutions available to address more complex cognitive challenges in nature.