BRAIN project (Histed): Readout and Control of Spatiotemporal Neuronal Codes of Behavior

National Institute Of Mental Health

Investigators

Linked publications & trials

Abstract

In this project period, we have developed means of stimulating single neurons in the cortex. To make cortical neurons accessible for imaging activity and stimulation, we have developed an approach to express soma-targeted ChrimsonR (stChrimsonR) combined with expression of GCaMP variants. This allows high efficiency all-optical stimulation at 1030 nm (laser repetition rate: 500 kHz) and imaging at 920 nm (laser repetition rate: 80 MHz). We have established a GCaMP8s-ChRimson opsin expression system. With these tools we can express both GCaMP8 and a soma-targeted ChRimson stably over many months in the mouse cortex. Neurons are two-photon excitable and neural activity can be recorded with high temporal resolution. We have disseminated these tools to other laboratories inside the consortium as well as some laboratories in the broader scientific community. Both the plasmid and the bicistronic virus are available from Addgene, and a manuscript on the work was published in the journal eNeuro. We have obtained results about how the visual cortical network changes in response to learning, to amplify fixed inputs. This work shows that the recurrent cortical network can adapt over a few days to produce larger responses when those benefit behavioral perception. The amplification is dependent on behavioral learning. A matched control with stimulation applied with similar statistics, but without learning, produces no change. Background: Cerebral cortex supports hierarchical representations of the world in patterns of neural activity, used by the brain to make decisions and guide behavior. Key computations in sensory learning might occur in areas downstream of the primary sensory cortex, with cortical representations specified by genetics and refined during development undergoing only small changes in the adult. Alternatively, local cortical circuitry may have substantial capacity to change to accommodate new processing. To examine this, we trained mice to recognize entirely novel, non-sensory patterns of cortical activity created by direct optogenetic stimulation in the primary visual cortex (V1). As animals learned to use these off-manifold patterns, we found large increases in cortical responses to fixed optogenetic input, and animals detection performance improved in concert with the increased cortical amplification. The neural changes were dependent on animals using the new patterns for behavior, and the increased responses for novel optogenetic inputs had little effect on existing visual sensory responses. Increased amplification would seem to be optimal for improving decision-making in the detection task we used. In sum, we find adult cortical plasticity can act over a few days to optimize processing of novel inputs. A report on the above work was published this year in Current Biology, and a Dispatch short review by James Marshel was also published by Current Biology on our results. In ongoing work, we are studying the role in brain function played by the dense excitatory-excitatory recurrent connections in the cerebral cortex. We have developed a hypothesis that these connections filter and amplify certain sequences of input in sensory cortex. This allows the brain to be sensitive to the dynamic and changing aspects of the world, from understanding visual movies we see to processing speech that we hear.