Our research goal is to understand, on the whole-brain but single-cell level, how entire neural circuits generate adaptable behaviors and how plasticity reorganizes the functional properties of these circuits to implement learned changes in behavior. We are particularly interested in adaptive motor control, a behavior that relies on circuits in the cerebellum, inferior olive, and other brain areas, and is as important to fish as it is to humans.
We use advances in microscopy, genetics, and a new virtual reality setup for zebrafish to systematically study the neural basis of such adaptive behaviors.
In my previous work I developed a paradigm for recording neural activity, at the single-cell level, in the entire brain of larval zebrafish, as they behave in a virtual environment. I showed that in this virtual setting, these tiny animals are able to perform short-term motor learning, and using whole-brain imaging, I identified the cerebellum and the inferior olive as brain areas with activity particularly correlated with the behavior. I then showed that lesioning the inferior olive removes the ability of the fish to adapt its motor programs to changes in visual feedback, opening the door to studying cerebellar-driven motor control in this tractable brain.
Imaging whole-brain neuronal activity while the fish navigate through a virtual environment is made possible by electrical recordings from their motor neuron axons, from which their intended swimming speed and direction is decoded in real-time and fed into a "swim simulator" that alters a visual display as if the fish were actually swimming. This work was recently published in Nature and has been likened to "The Matrix" (Nature.com Blogs)
Are neural circuits hard-wired to generate a fixed pattern of behavior in response to a stimulus? In most cases, the answer is noanimals continuously adapt their behavior to changing environments. This includes changes in the reward structure as well as changes in the physics of the body and the environment, such as when humans step onto a slippery floor or a fish swims into more viscous water, or after injury. This adaptability is the key to the successful function of the central nervous system in driving behavior. How is this continuous learning and adaptation implemented on the circuit level? How does the function of neural networks change when an animal is confronted with a change in the environment? With the virtual environment described above, we aim to tackle these questions, in particular in the case of cerebellar-dependent short-term motor learning, using a combination of imaging, and perturbing, neuronal activity during fictive virtual-reality behavior.
As of August 22, 2012