The long-term goal of our lab is to understand how animals learn from experience and how they subsequently use this knowledge to guide behavior. One form of learning, commonly observed in humans and animals, relies on the reinforcing effect of rewarding or aversive experience. As a result of such learning, sensory cues, behavioral responses or specific actions become associated with positive or negative values, which is critical for seeking resources and avoiding danger.
In a dynamically changing environment it is also important to identify the boundaries of current knowledge, in order to initiate new learning whenever necessary. Consistent with this idea, humans and animals are capable of rapidly detecting novelty, taking as little as 85ms in humans. Novel stimuli trigger distinct orienting and exploratory behaviors, which habituate after only a few exposures, suggesting a very rapid form of memory formation. Moreover, animal learning theories postulate that the rate of learning depends on the degree to which rewards or punishments are novel, thus also highlighting the importance of novelty for learning.
Research in our lab focuses on the neural mechanisms underlying the two problems described above:
- What circuit computations are involved in learning from reward and punishment?
- What are the mechanisms for novelty detection and behavioral novelty responses?
Previous studies have identified key brain areas and neuromodulatory systems involved in reinforcement learning and novelty processing. However, a mechanistic description of how different neurons, interacting dynamically within local and larger scale circuits, mediate learning and generate behavioral responses is currently lacking. We approach this problem using a combination of multielectrode recording and optogenetic techniques in awake, behaving mice. By genetically tagging specific neuronal populations for electrophysiological identification and manipulation, this approach allows for the precise characterization of the firing properties of defined neuron types and for testing the causal contribution of their firing for specific aspects of behavior. By complementing this approach with anatomical circuit mapping using viral tracing tools, we aim at revealing some of the neural computations involved in learning and decision making.
Disturbances in reinforcement learning and novelty processing have been linked to human pathological conditions including schizophrenia, depression and autism, as well as human personality traits associated with maladaptive behaviors such as addiction. Therefore, our research program also has implications for understanding these mental diseases and may ultimately support the development of novel interventional strategies.
> video on basic research on nanotech meets biotech - Sebastian Haesler - ©VIB, 2015