- 제목: Dopaminergic computations of reward proximity during foraging competition 

- 연사: 김형구 교수님 (성균관대학교 글로벌바이오메디컬공학과)

- 초청자: 최명환 교수님(서울대학교 생명과학부)

- 일시: 2026년 4월 16일 (목) 오전 11시

- 장소: 서울대학교 관악캠퍼스 자연과학대학 500 동 1층 목암홀

- 링크: https://snu-ac-kr.zoom.us/j/89840360149?pwd=xSoRKdjq47gkLRmAnGzigHD7tOyCWa.1

  회의 ID: 898 4036 0149  암호: 12345

 - Abstract:

Living with multiple potential goals, we take into account our current status and reward statistics when making decisions. We developed a novel task to model this more realistic reward environment using mice. In a square chamber, an LED cue is positioned at the center of each side, with a water port installed beneath each cue. When one of the four visual cues is randomly activated, a drop of water is dispensed from the port below the activated cue. As mice learned the association between LED cue and reward delivery, they gradually approached reward with shorter latency. We observed that the latency increased with the mouse’s distance to the reward and visual angle of the cue ( < 120 deg from the midline). DA responses to cues decreased with distance and angle, suggesting that mice internally estimate time-to-reward (‘predicted latency’) based on spatial parameters and this process is reflected in dopaminergic responses. Next, we examined how the competitive context influences DA responses by adding another animal. To estimate which animal may win the competition, we computed the predicted latency for each mouse based on both its own geometric parameters and those of the competitor. This predicted latency reflects each animal’s estimated time to reward. Interestingly, when the mouse is closer than the competitor, the DA activity reliably decreases as the predicted latency increases. These results suggest that mice flexibly incorporate both their own and others’ physical parameters to compute reward proximity, reflected in adaptive dopaminergic activity. To test whether the mPFC contributes to computing angle and distance, and consequently influences competitive decision-making, we injected hM4Di into the mPFC and chemogenetically inhibited it using Deschloroclozapine(DCZ). Latency increased following DCZ treatment but stayed constant with saline. CueDA was markedly weakened suggesting that mPFC inhibition disrupts the computation of cue-based reward proximity.