Presynaptic computation of reward intensities through the dual autoreceptor system

To make optimal decisions, animals must accurately differentiate reward intensities, yet synaptic mechanisms for this computation remain poorly understood. Here, we show presynaptic gain control of reward signals in the dopaminergic neurons (DANs) of Drosophila melanogaster, mediated by two opposing dopamine autoreceptors, Dop1R1 and Dop2R. Cell-type-specific endogenous protein tagging and functional imaging of the reward-signaling DANs revealed the localization of both receptors at active zones and the regulation of presynaptic calcium in response to distinct reward intensities. Reward learning with cell-type-specific silencing of these receptors uncovered the role of Dop2R in attenuating reward signals specifically at high concentrations of sugar and alcohol, in contrast to selective amplification of low-intensity rewards by Dop1R1. This dose-specific and bidirectional regulation may extend the dynamic range of perceived reward intensity, enabling the selection of options that predict better outcomes.