Overview of the results:
To allow for adaptive decision-making, an organism must integrate the value of stimuli or expected outcomes. This value needs to be adjusted to maximize survival, resulting in a perceived value that allows dynamically assessing the currently possible actions. Previous studies have mainly investigated the association of a stimulus with a predictor, a process called conditioning. However, how neuronal value representations are maintained and modified across different stimulus contrasts, as well as physiological needs, and affective states remained poorly understood. Here, we utilized an unpredicted reward delivery paradigm in mice, to characterize gustatory value representations in the Basolateral amygdala (BLA). We show that responsive cells in the BLA, a region known for their role in associative learning, scale their response magnitude with caloric content, in line with a single-cell-based value code. However, the reward-responsive population shows sensory tuning, leading to the scaling response being encoded by different sets of neurons dependent on reward nutrient composition, making integrating all reward-responsive neurons necessary to achieve a stimulus-invariant value representation. Furthermore, by introducing a novel large reward, we found that changes in stimulus contrast led to an updated relative value representation within minutes, pointing to a relative value scale. Finally, we find that the internal state plays a dominant role, as reward representations are sensitive to thirst and aversive experiences, leading to an increased response to rewards addressing a current physiological need and a diminished response to rewards in an aversive context. Such a dynamically updated value signal could determine the reinforcing properties of a stimulus, while at the same time allowing for sensory specific associations, thereby providing a flexible basis for adaptive learning and decision-making.
Dissemination of the results:
We are preparing a manuscript describing the state- and contrast-dependent value processing in the amygdala, which we will publish on a freely accessible preprint server (e.g. BioRxiv). This will allow other scientists to profit from novel information as soon as possible. The project data will be made openly accessible along with analysis scripts as soon as our manuscript has undergone the standard peer-review system. So far, I have presented a poster describing our discoveries at the last Society for Neuroscience annual meeting (Neuroscience 2022), as well as during an invited lecture (Bernstein Seminar Series Freiburg). Attending this conference and seminar allowed the presentation and discussion of our research and led to the initiation of new collaborations.
Exploitation of the results:
The current research project is at the level of basic research. Results will mainly be exploited as the basis for further research. This will be done by the scientific community at large, through the integration of the results we publish, as well as by myself in my continued research career. Based on the results of the current project I was able to attain a Swiss National Science Foundation transition to independence fellowship (Ambizione), within which I will continue the current line of research together with a PhD student. Furthermore, I will use the established methodologies and insights to compare wild-type to disease model mice in my independent research career.
