Massive-Scale Multi-Area Single Neuron Recordings to Reveal Circuits Underlying Short-Term Memory

Short-term memory is a crucial component of cognitive function and pervades nearly all aspects of our mental lives. Previous research has shown that short-term memory involves multiple cognitive components and diverse brain regions. However, it is not mechanistically understood what regions are involved when, what neuronal subsets are recruited within these regions, or how they interact to represent information relevant to behavior. This proposal aims to elucidate the role of visual, association, and motor cortex in mice performing a visually-cued short-term memory task. This will be accomplished using massive-scale two-photon calcium imaging in behaving mice to measure activity of thousands of neurons simultaneously across these multiple brain regions. Subsequently, optogenetic manipulation of brain regions and of computationally identified neuronal assemblies will be used to determine their causal role in behavior. These technologies and results will have wide impact on understanding neural circuits underlying behavior and cognition. New approaches will be introduced for massive-scale mapping of single neuron activity in relation to a quantifiable behavior. New ways to determine circuit connectivity, and novel combination computational and optogenetic technologies to manipulate critical circuit components, will be introduced. These large data sets will be made widely and freely available, enabling other research groups to avail of these data for novel analyses.

The goal of this proposal is to develop novel tools and provide unprecedented information on neuronal activity patterns and circuits in order to understand the role of multiple cortical areas during short-term memory in mice. Classical electrophysiological recordings are limited to relatively small numbers of neurons with unknown identity. In addition, while microstimulation or pharmacological manipulations can be used to activate or inhibit all the neurons within a local area, it is not possible to selectively excite or inhibit specific neuronal subpopulations that are known to play a role in the behavior. The proposal addresses these issues by developing novel tools to study mice performing a visually-cued memory-guided discrimination task. First, methods for massive scale imaging (up to ten thousand neurons simultaneously) of multiple cortical regions spanning several millimeters in the mouse cortex will be developed. Second, mice will be trained on a visually cued short-term memory task with suitable behavioral richness, including separate sensory, memory and response epochs, so that activity in distributed cortical regions (such as visual, parietal, and frontal motor cortices) can be imaged and the role of individual areas in each epoch can be ascertained. Third, targeted inactivation of specific brain areas will be performed to determine their role in the behavior. Finally, computationally identified neuronal subsets in specific areas will be stimulated in order to determine if they are sufficient for altering behavior. Together, these will be the first studies in the field to link behavior, extremely large-scale multiple-area recordings, and causal manipulations of areas and identified neuronal assemblies. By introducing tools for a radically different approach from previous analyses of memory and memory-guided functions, it is expected that the project will have a significant impact on the field.