Laboratory of Neurobiology
Charles D. Gilbert
Arthur and Janet Ross Professor
Our aim is to understand mechanisms of sensory processing in the cortex by working at the interface of microcircuitry, receptive field properties, and functional architecture, and to establish the relationship between the operation of neuronal ensembles and perceptual performance. We study the mechanisms of neural integration by exploring how the various sources of inputs to a cell interact to generate its functional properties. A key element in our studies of cortical microcircuitry was the discovery of a plexus of long-range horizontal connections in the cortex, which we now find mediate spatial integration and dynamic properties of cortex. We have studied the functional implications of these connections, exploring how cells' responses are modulated by the context within which a stimulus is presented. These results fit with psychophysical studies demonstrating the dependence of perception on context, and lead to a novel view of cortical processing: Adult neurons, rather than having fixed functional properties, are dynamically tuned, changing their specificities with varying sensory experience. We have used these findings as a basis for making computational models relating the properties of neuronal ensembles to perceptual performance. These models enable us to understand how neurons, acting as an interactive ensemble, determine visual attributes in a manner that depends on the conjoint activity within the ensemble.
The functional role of intrinsic cortical connections depends on their registration with cortical functional architecture. We have investigated the nature of the functional architecture and connectivity within the primary and secondary visual areas using a combination of optical and single cell recording, by cross-correlation analysis and by labeling of cortical connections.
Additional insight into the dynamic nature of cortical processing in the adult visual system has come from experiments in which we make permanent alterations of visual input by means of retinal lesions. This procedure produces changes in the receptive field properties of cells in the cortex immediately following the lesion, and alterations in the visual topography of the cortex over a period of weeks. Once again, it appears that receptive fields are mutable, changing in size and position as a consequence of alterations of sensory input. We have shown that the long-term changes involve axonal sprouting and synaptogenesis, and are now studying the underlying molecular mechanisms.
We have observed shorter term changes in the functional properties of cortical cells, occurring within a time span of minutes, that result from changes in visual experience and the context within which a stimulus is placed. We are determining the pathway along which these changes take place, characterizing the synaptic modifications involved. The remapping of visual space on the cortex and the mutability of receptive field properties represent useful models in several respects: as tools for determining the role of horizontal connections, as a means for exploring changes in neuronal assemblies that result from visual experience (such as those associated with long-term memory), and as a way of exploring the mechanisms underlying functional recovery following lesions of the central nervous system. We are extending our findings of dynamic changes in receptive field properties and cortical functional architecture, occurring over a wide range of time scales, to explore the mechanisms of spatial learning and perceptual learning.
In meeting these objectives we are using a number of tools. We combine anatomy at the light and ultrastructural levels with physiology, based both on single-cell and optical recording. We are integrating molecular approaches with the systems level analysis to understand the mechanisms of adult cortical plasticity. This multidisciplinary approach provides insight into the mechanisms, at the level of circuitry, of the dynamic changes in functional properties of cortical cells. In addition, we do electrophysiological recordings to study the more complex properties of cortical cells, their dependence on state of alertness or attention, and on early perceptual learning. In parallel with the physiological studies we are doing human psychophysics to explore the perceptual consequences of dynamic changes in cortical properties. Taken together, the approaches outlined above should greatly enhance our ability to understand cortical sensory mechanisms.