Periodic Reporting for period 1 - INTENDeD (INTegration of EmotioNs During Decision-Making)
Reporting period: 2021-03-01 to 2023-02-28
Emotions are often contrasted with reason. However, the integration of emotional information into the decision-making processes can advantageously bias adaptive behavior selection. Miss-regulation of emotional information integration in cortical processing is a hallmark of psychiatric disorders with affect dysregulation. Our knowledge of the etiology and pathophysiology of these disorders is strongly limited by our understanding of the non-pathological neuronal circuits underlying emotional processing and makes targeted intervention methods scarce.
The project addressed how valence and value representation is encoded in the BLA using in vivo two-photon calcium imaging during a newly established unpredicted gustatory reward delivery paradigm in mice. This allowed us to characterize gustatory value representations in detail. Leading to the description of how neuronal value representations are maintained and modified across different stimulus contrasts, physiological needs, and affective states. We find dynamically updated value signals, which could be used to 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.
The results obtained here will inform our continued characterization of whether the BLA transmits differential valence and value information to specific PFC subregions and how BLA information is integrated in the local PFC circuit. The long-term basic research goal is to achieve a circuit- and cellular-level mechanistic understanding of the influence of BLA activity on PFC processing and decision-making. Characterizing the BLA – PFC network in the healthy brain will serve as a reference for the characterization and optogenetic modulation of abnormalities in disease model mice, ultimately providing a roadmap for a translational path, like optogenetically-inspired deep-brain stimulation in human patients.
The project addressed how valence and value representation is encoded in the BLA using in vivo two-photon calcium imaging during a newly established unpredicted gustatory reward delivery paradigm in mice. This allowed us to characterize gustatory value representations in detail. Leading to the description of how neuronal value representations are maintained and modified across different stimulus contrasts, physiological needs, and affective states. We find dynamically updated value signals, which could be used to 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.
The results obtained here will inform our continued characterization of whether the BLA transmits differential valence and value information to specific PFC subregions and how BLA information is integrated in the local PFC circuit. The long-term basic research goal is to achieve a circuit- and cellular-level mechanistic understanding of the influence of BLA activity on PFC processing and decision-making. Characterizing the BLA – PFC network in the healthy brain will serve as a reference for the characterization and optogenetic modulation of abnormalities in disease model mice, ultimately providing a roadmap for a translational path, like optogenetically-inspired deep-brain stimulation in human patients.
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.
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.
Miss-regulation of emotional information integration in cortical processing is a hallmark of psychiatric disorders spanning anxiety disorders, addiction, schizophrenia, and autism spectrum disorders. The neurobiological mechanisms underlying these disorders' etiology and pathophysiology are poorly understood. At the same time, they represent a major burden on the affected individuals and our economy and society at large. Brain disorders affect at least one in three people during their lifetime – currently 165 million people in Europe, with an estimated cost of 800 billion euros per year. This accounts for 35% of the burden of all diseases in Europe. Due to our limited understanding of the underlying causes leading to disease manifestation, targeted intervention methods are scarce. This holds especially true at the neuronal circuit level. Emotions are multifaceted, but a key aspect of emotion involves the evaluation of external sensory stimuli. The progress made on the state- and contrast-dependent value processing in the amygdala during this project, provides a solid foundation for future research aimed at investigating information transfer between the BLA and the PFC, both hub regions of emotional processing. The current research project is at the level of basic research; however, the improved understanding of the healthy state is a needed reference for the characterization of abnormalities in disease model mice.