CAREER: Neural circuit mechanisms of spatial target selection in the mammalian midbrain

Johns Hopkins University, Baltimore MD

Investigators

Abstract

When faced with multiple competing stimuli in the natural world, animals display a remarkable ability to process selectively the most salient target location to guide behavior. Such stimulus selection is central to behavioral and cognitive functions such as spatial attention and decision-making, as well as their impairments in psychiatric conditions. However, how the brain achieves spatial target selection remains poorly understood. This proposal combines cutting-edge technologies and analytical methods to discover neural mechanisms underlying spatial target selection in mammals. Specifically, it focuses on the mouse midbrain and unravels the biological implementation of key aspects of ‘winner-take-all’ selection. In parallel with this research effort that links experimental neuroscience with mathematical descriptions, the educational, training and outreach goals, here, also couple neuroscience and mathematics training. This proposal supports the expansion and improvement of a recently created course for graduate students and undergraduate seniors called ‘Quantitative Methods for Brain Sciences’, trains high school students to perform ‘physical’ simulations of our neural circuit models in robot vehicles, engages in STEM outreach with local elementary schools, and creates an annual ‘Neurons in robots’ showcase for the public. In sum, the proposed efforts will advance understanding of how the brain implements selection in its neural hardware, and create a foundation for addressing the dysfunction of stimulus selection found in psychiatric disorders. Additionally, deconstruction of the biological implementation of the winner-take-all operation will inform the design of efficient artificial intelligence systems for target selection in cluttered environments (during navigation, scene analysis, etc.). The proposed research employs a combination of electrophysiological, optogenetic, and calcium imaging methods in mice to dissect the mechanistic logic of spatial target selection. The research hypotheses tested here are motivated by insights from recent work in barn owls. Specifically, they focus on the evolutionarily conserved midbrain sensorimotor hub, the superior colliculus (SC), and a group of satellite inhibitory neurons, called the periparabigeminal lateral tegmental nucleus (pLTN). Aim 1 will investigate the neural correlates of spatial target selection in the intermediate and deep layers of SC (SCid) using electrophysiology. The hypothesis is that SCid signals the most salient stimulus categorically (i.e., in a winner-take-all manner). Aim 2 uses optogenetic manipulations to investigate inhibitory mechanisms underlying this SCid selection signal. The hypothesis is that a donut-like pattern of inhibition from pLTN controls spatial target selection in SC. Aim 3 uses endoscopic calcium imaging to investigate the mechanisms that allow selection in SCid to operate at all pairs of stimulus locations. The hypothesis is that a combinatorial inhibitory strategy enacted by sparse pLTN neurons that tile space in an unusual manner achieves such selection. Uncovering neural mechanisms of spatial target selection in mice will not only address long-standing open questions pertinent to spatial attention, decision-making, and perceptual categorization, but also illuminate whether neural circuit principles underlying selection may be conserved across vertebrate species. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.