For many inherited and age-related diseases of the nervous system, current molecular and gene therapeutic approaches hold great promise for restoring function. However, injuries leading to irreversible damage of the delicate neural tissue at the earliest stage of sensory processing (e.g., in the retina at the back of the eye), require other methods to rescue function. Action potentials leaving the retina reach the brain via a series of unreliable synapses at each successive stage of the visual processing hierarchy. The rate-limiting step in finding a cure for many forms of retinal blindness is to determine exactly how to electrically stimulate the retinal axons projecting to the brain in a manner that recreates some sense of visual perception. With new ultra-high resolution molecular imaging methods, we can now track the neural activity in thousands of neurons in a functional column of the visual cortex with single-cell resolution while presenting natural or artificial stimuli to the retina.
We hypothesize that being able to monitor the neural activity at this unprecedented level of resolution in the visual cortex, while delivering natural visual or artificial electrical stimuli to the retina, will allow us to determine the exact spatio-temporal patterns of retinal ganglion cell activity that are needed to drive individual neurons in the visual cortex.
Our Specific Aims are to:
1. Determine the mechanism by which a single retinal ganglion cell's activity is encoded in a three dimensional (3D) population of neurons in the visual cortex,
2. Determine the exclusive sets of spatio-temporal retinal ganglion cell firing patterns that are most effective in driving a population of neurons in the visual cortex.
Successful determination of the rules of retino-cortical activation will necessitate that natural visual and artificial electrical stimulation of retinal ganglion cells produce the same activation dynamics within the imaged cortical region.
To accomplish these aims, we propose to use an in vivo animal model wherein an array of recording and stimulating electrodes can be placed on the retina while simultaneously imaging the visual cortex with a two-photon microscope. We will load many cortical neurons with a cell-permeant (AM-ester) form of calcium indicator. This method of loading results in a dye concentration in neuronal cell bodies which produces activity-dependent peak amplitude fluorescence changes that correlate linearly with action potential activity. Thus, calcium imaging in neuronal cell bodies will serve as an excellent surrogate for spiking activity in visual cortex.
Fifty years of neurophysiology research has provided us a reasonable idea of the computations performed in isolation at each stage of sensory processing. Our imaging proposal may lead to a new approach to studying sensory systems, in which one determines the exact firing events from one stage of processing that are actually used by neurons at subsequent stages. Thus, it may provide a more efficient assay for determining the neural code for early visual processing than simply measuring the output of retinal ganglion cells in isolation.