Learning What Matters: Neural Mechanisms of Flexible Navigation
Abstract:
Goal-directed navigation in a dynamic world requires quickly identifying important locations and adapting behavioral plans to new information. In this talk I will describe neural circuit mechanisms of rapid spatial learning and of adapting to new information to guide navigation. Identifying crucial locations in a new environment depends on neural computations that rapidly represent locations and connect location information to key outcomes like food, however the mechanisms to trigger these computations at behaviorally relevant locations is not well understood. We find that inhibitory interneurons in hippocampal CA3 play a causal role in identifying and exploiting new food locations. Inhibitory interneurons in CA3 drastically reduce firing on approach to and in goal locations. Sparse optogenetic stimulation to prevent goal-related decreases in interneuron firing impaired learning of goal locations and disrupted neural representations of goal locations. These results reveal that goal-selective decreases in inhibitory activity enable learning important locations. Navigation also requires rapidly updating choices in the face of new information. In hippocampus and prefrontal cortex, neural activity representing future goals is theorized to support navigation planning. Yet how prospective goal representations incorporate new, pivotal information is unknown. Using virtual reality, we precisely introduced new crucial information during navigation and recorded neural activity as mice flexibly adapted their planned destinations. We found that new information triggered increased prospective representations and reorganization to rapidly shift to the new choice. This prospective code updating depended on the degree of behavioral adaptation needed. These studies reveal new mechanisms by which animals rapidly learn crucial new locations and adapt to new information that requires updating navigation plans.

Bio:
Dr. Annabelle Singer is the McCamish Foundation Early Career Professor in the Coulter Department of Biomedical Engineering at Georgia Tech and Emory University. Her research seeks to understand how neural activity produces memories and regulates brain immune function, with the goal of developing new therapies for brain disease. Dr. Singer’s work has shown that coordinated electrical activity across hippocampal neurons encodes memories and fails in models of Alzheimer’s disease. She discovered that driving specific patterns of neural activity, such as gamma oscillations, reduces Alzheimer’s pathology and alters brain immune function. Using non-invasive sensory stimulation, she is translating these discoveries from rodents to humans to pioneer radically new treatments for disease.

Dr. Singer is a Packard Fellow, Kavli Fellow, and recipient of the National Academy of Engineering’s Gilbreth Lectureship, the Society for Neuroscience’s Janett Rosenberg Trubatch Career Development Award, and the American Neurological Association’s Derek Denny-Brown Young Neurological Scholar Award. Her discoveries have inspired more than 20 clinical trials of brain stimulation across multiple diseases and have been featured on PBS, Nature News, Quanta Magazine, The New York Times, Radiolab, and multiple documentaries. Dr. Singer trained as a postdoctoral fellow in Ed Boyden’s Synthetic Neurobiology Group at MIT and earned her Ph.D. in Neuroscience at UCSF.